Kin17 regulates proper cortical localization of Miranda in Drosophila neuroblasts by regulating Flfl expression

Abstract

During asymmetric division of Drosophila larval neuroblasts, the fate determinant Prospero (Pros) and its adaptor Miranda (Mira) are segregated to the basal cortex through aPKC phosphorylation of Mira and displacement from the apical cortex, but Mira localization after aPKC phosphorylation is not well understood. We identify Kin17, a DNA replication and repair protein, as a regulator of Mira localization during asymmetric cell division. Loss of Kin17 leads to aberrant localization of Mira and Pros to the centrosome, cytoplasm, and nucleus. We provide evidence to show that the mislocalization of Mira and Pros is likely due to reduced expression of Falafel (Flfl), a component of protein phosphatase 4 (PP4), and defects in dephosphorylation of Serine-96 of Mira. Our work reveals that Mira is likely dephosphorylated by PP4 at the centrosome to ensure proper basal localization of Mira after aPKC phosphorylation and that Kin17 regulates PP4 activity by regulating Flfl expression.

Graphical Abstract

Summary:

Connell et al. identify a regulator of Miranda localization, Kin17, which promotes the proper localization of Mira by regulating the expression of PP4 component Falafel. They further show that proper segregation of the cell fate determinant Mira likely requires PP4-mediated dephosphorylation of Serine-96 of Mira at the centrosome and subsequent dissociation of Mira from the centrosome/spindle in Drosophila neuroblasts, a critical step that has been overlooked previously.

Introduction

The generation of diverse cell types from a single population of stem cells is essential for proper development. In the brain, neural stem cells (NSCs) divide to produce neurons and glia. Asymmetric cell division allows for the generation of cells of different fates from one parental cell by producing daughters of different protein content, fate, niche, and/or size. This process must be tightly regulated to ensure the proper balance of stem and differentiating cells, and defects in this process can lead to overproliferation or premature differentiation of the stem cell population, both of which can have detrimental effects on the organism, such as cancer or developmental disorders.¹

Drosophila NSCs, termed neuroblasts (NBs), provide a useful model for the study of asymmetric cell division as many of the key components are conserved in mammalian systems.² In developing Drosophila brains, most NBs (type I NBs) divide asymmetrically to generate a self-renewing NB and a differentiating ganglion mother cell (GMC).² GMCs divide a single time to produce neurons and glia. The identity of the GMC is established by the specific segregation of the fate determinants, Prospero (Pros)^3,4, Numb4, and Brain tumor (Brat)^5–7 to the GMC during mitosis. In order to deliver these factors to the GMC, neuroblasts establish a cell polarity axis during mitosis through the recruitment of the Par protein complex to the apical pole of the cell.^8,9 This apical complex establishes cell polarity by excluding the localization of the fate determinants from the apical pole, leading to basal localization of these factors.

One of the primary factors that must be excluded from the apical pole is the adaptor protein, Miranda (Mira), which localizes pros mRNA¹⁰, Pros^11,12, Staufen¹³, and Brat⁶ to the basal pole and ensures their proper segregation to the GMC. Prior to the onset of mitosis, Mira localizes to the entire cell cortex through an interaction between its Basic and Hydrophobic (BH) motif and the cell membrane¹⁴. At the onset of mitosis, Inscuteable and Bazooka (Baz, also known as Par-3) are recruited to the apical cortex, where Cdc42 acts downstream of Baz to recruit the Par-6/aPKC (atypical Protein Kinase C) complex to the cortex through a direct interaction between the CRIB domain of Par-6 and Cdc42.^15–19 This interaction between Par-6 and Cdc42 relieves the Par-6-mediated suppression of the serine/threonine kinase activity of aPKC¹⁹, and subsequently, aPKC phosphorylates Mira within the BH motif breaking the interaction between Mira and the apical cell cortex, leading to basal localization.^14,20 Although aPKC phosphorylates multiple residues, a phosphomimetic mutation of S96, Mira^S96D, is displaced from the cortex even in the absence of aPKC, indicating that phosphorylation of S96 is sufficient for displacement from the cortex.²⁰ This raises the question of how Mira localizes to the basal cortex after aPKC phosphorylation. A previous study reports that depletion of Falafel (Flfl), the targeting subunit of Protein Phosphatase 4 (PP4), leads to a reduction in Mira localization to the basal domain²¹, implying that Mira needs to be dephosphorylated before interacting with the basal domain. However, while Flfl has been shown to physically interact with Mira²¹, PP4 has only been shown to be required for the dephosphorylation of threonine-591, which is required for cortical localization of Mira prior to its phosphorylation by aPKC.²² Therefore, although how Mira is phosphorylated by aPKC and displaced from the apical cortex has been well studied, the exact mechanism of how Mira localizes to the basal cortex after aPKC phosphorylation is unclear.

Mira has also been shown to localize to the centrosome of wild-type NBs during prophase²³ and to the mitotic spindle in syncytial embryos and various mutants, including Mira^S96D.^{18,19,23–25} The mutants that lead to accumulation of Mira to the centrosome and mitotic spindle also lead to defects in basal localization, suggesting that the localization of Mira to the centrosome in prophase and its clearance may be an essential step in localization to the basal domain downstream of aPKC phosphorylation.

Kin17 is known as a DNA- and RNA-binding protein that is essential for DNA replication and DNA damage repair.^26–29 Kin17 is upregulated in cervical³⁰, breast^31–33, colorectal³⁴, ovarian³⁵, and lung cancers³⁶, where it is associated with proliferation and invasion. Here, we identify Kin17 as a factor regulating the localization of Mira during asymmetric cell division in Drosophila larval NBs. We demonstrate that Kin17 promotes the proper localization of Mira in part by regulating the expression of the PP4 component Flfl. We further show that PP4 likely dephosphorylates Mira at S96 at the centrosome to ensure dissociation of Mira and Pros from the centrosome/spindle and proper cortical localization. Therefore, our work reveals a mechanism that ensures proper localization of Mira during the asymmetric division of NBs.

Results

Kin17 knockdown leads to reduced brain size and a reduced mitotic index in NBs

We identified Kin17 in a screen for factors that affect Drosophila brain development. Kin17 knockdown using the pan-neuroblast (NB) driver, insc-Gal4, led to reduced brain size (Figs. 1A,B). This decrease in brain size was not due to a reduction in the number of type I NBs in the central brain (Figs. 1C,D). However, we observed a significant decrease in the mitotic index of Kin17 RNAi NBs (Figs. 1E,F) and a corresponding reduction in GMCs (Pros+, Elav-) and newly born neurons (Pros⁺, Elav⁺) (Fig. 1G,H). These phenotypes were rescued by expressing Kin17, ruling out off-target effects of the RNAi (Figs. 1A,B,E–H). These results indicate that Kin17 is required for normal brain development and mitosis in NBs.

Figure 1. Kin17 knockdown leads to a reduction in brain size and nuclear Pros.

(A,B) Brain size in control (ctrl), Kin17 RNAi, and Kin17 rescue brains. Scale bar, 50 μm; n ≥ 6 brains. (C,D) Quantification of type I neuroblast (NB) number in Kin17 RNAi brains. Scale bars, 50 μm; n ≥ 10 brains. (E,F) Mitotic index in Kin17 RNAi. Yellow circles, mitotic NBs; White circles, non-mitotic NBs; Scale bars, 20 μm; n=3 biological replicates. (G) Pros+ progeny in Kin17 RNAi brains. Yellow arrowheads, Pros⁺ GMCs; white dashed line, Pros⁺, Elav⁺ neurons; Scale bars, 50 μm. (H) Quantification of GMC number in Kin17 RNAi NB lineages. N ≥ 6 brains with each n representing an average of 5 lineages from 1 brain. (I,J) Pros staining in Kin17 RNAi brains and quantification of the NBs with nuclear Pros. Scale bars, 5 μm; n ≥ 8 brains. (K,L) Pros nuclear localization in Kin17 RNAi; pros^17/+ NBs. Dashed lines outline NBs. Scale bar, 5 μm; n ≥ 6 brains. (M,N) Mitotic index in Kin17 RNAi; pros^17/+ NBs. Yellow circles, mitotic NBs; Grey circles, non-mitotic NBs; Scale bar, 20 μm; n ≥ 8 brains. (O) Pros+ progeny in Kin17 RNAi; pros^17/+ brains. Scale bar, 50 μm. (P) Quantification of GMC number in Kin17 RNAi; pros^17/+ NB lineages. n ≥ 7 brains with each n represents an average of 5 lineages from 1 brain. (Q) Quantification of brain size in Kin17 RNAi; pros^17/+ brains. n ≥ 7 brains. Data are presented as the mean ± 1 Standard deviation (SD). n.s., not significant; **, p<0.01; ***, p<0.001, unpaired student’s t-test for (D) and unpaired student’s t-test with the Bonferroni correction for (B, H, J, L, N, P, Q).

Kin17 knockdown leads to aberrant nuclear and centrosomal accumulation of Pros

Nuclear Pros promotes cell cycle exit of GMCs³⁷ and quiescence or premature differentiation in NBs.³⁸ Therefore, we examined if the reduction in the mitotic index of Kin17 RNAi NBs was caused by Pros localization to the nucleus. We observed that 37.5± 15.0% of Kin17 RNAi NBs had nuclear Pros (Figs. 1I,J), which was never observed in control NBs. To determine if the aberrant nuclear Pros is responsible for the Kin17 knockdown phenotypes, we examined if reducing Pros levels would alleviate the Kin17 phenotypes. Kin17 knockdown in pros^17/+ NBs leads to a signification reduction in nuclear Pros localization (Figs. 1K,L), the mitotic index was partially rescued to 20.2 ± 2.8% compared to 34.3% in control and 13.4% in Kin17 RNAi (Figs. 1M,N), and the reduction in Pros+ neurons, GMCs, and brain volume was also partially rescued (Fig. 1O–Q), suggesting that the Kin17 phenotypes are due to aberrant nuclear localization of Pros.

In addition to the nuclear localization, we also observed significantly increased centrosomal localization of Pros at interphase and metaphase (particularly at the apical centrosome) in Kin17 RNAi ; pros^17/+ NBs (Figs 2A,B) as indicated by its colocalization with the centrosomal protein, Pericentrin-like protein (Plp).³⁹ In control NBs, centrosomal Pros appeared to be cell cycle-dependent, with a peak occurring in prophase (Figs.2A,B). However, Pros basal crescents were observed at similar rates in control and Kin17 RNAi; pros^17/+ NBs at metaphase (Fig. S1A,B). Therefore, Kin17 is required to prevent aberrant localization of Pros to the nucleus and centrosome.

Figure 2. Kin17 knockdown leads to centrosomal localization of Pros and Mira.

(A,B) Pros localization to the centrosome (marked by Pericentrin-like protein (Plp)) in ctrl and Kin17 RNAi; pros^17/+ NBs. Inset, Pros localization at apical (yellow boxes) and basal (grey boxes) centrosomes; Scale bars, 5 μm; n ≥ 16 NBs. (C,D) Mira localization in ctrl, Kin17 RNAi, pros^17/+, and Kin17 RNAi pros^17/+ NBs throughout the cell cycle. Inset, Mira localization at apical (yellow boxes) and basal (grey boxes) centrosomes; Arrow, Mira localization on spindle; Scale bar, 5 μm; n ≥ 18 NBs. (E,F) Cortical localization of Mira in ctrl, pros^17/+, and Kin17 RNAi pros^17/+ NBs. Metaphase depicts normal (left) and weak (right) basal cortical localization of Mira. Scale bars, 5 μm; n ≥ 18 NBs. (G,H) Colocalization of Mira and Pros at the centrosome in Kin17 RNAi NBs. Insets, Grey boxes, Pros; Purple boxes, Mira; Scale bar, 5 μm; r, Spearman coefficient; n = 50 NBs. (I,J) Neuroblast numbers in Kin17 and Eiger single knockdown or double knockdown brains. Scale bar, 5 μm; n ≥ 6 brains. Data are presented as the mean ± 1 SD. n.s., not significant; **, p<0.001; ***, p<0.001, unpaired student’s t-test for (B), unpaired student’s t-test with the Bonferroni correction for (D, J), and Spearmann’s rank correlation coefficient for (H). See also Figures S1 and S2.

Kin17 is required for proper Mira localization

As Mira is the adaptor protein of Pros¹¹, we asked if Mira localization was altered in Kin17 RNAi NBs. In control NBs, Mira localized to the cortex during interphase as previously described (Figs. 2C). During prophase, Mira showed increased localization to the apical centrosome, which decreases through the rest of cell cycle until it returns to interphase levels in telophase (Figs. 2C,D).²³ Similar to Pros, we found Mira also aberrantly localized to the centrosome in Kin17 RNAi NBs in interphase (Fig. 2C,D). Since Kin17 RNAi NBs have a low mitotic index and, we performed Kin17 RNAi in the pros^17/+ background to systematically quantify Mira localization at different stages of the cell cycle. Pros is not required for proper Mira localization¹² and Mira localization in pros^17/+ NBs was similar to that in the control throughout the cell cycle (Fig. 2C,D). In Kin17 RNAi; pros^17/+ NBs, there was a significant increase in Mira centrosomal localization in interphase and telophase and at the basal centrosome in prophase and metaphase (Figs. 2C,D). Besides the centrosomal localization, about 25% of Kin17 knockdown NBs also exhibited Mira localization to the mitotic spindle at anaphase/telophase, which was only observed in 3–6% of control or pros^17/+ NBs (Fig. 2C). Furthermore, we found that Mira accumulation at the centrosome correlated with defects in cortical localization, with 24.8% and 9.1% of Kin17 RNAi; pros^17/+ NBs having cytoplasmic Mira at interphase and metaphase, respectively, as well as reduced basal cortical localization in 20–30% of prophase and metaphase NBs (Figs. 2E,F). Mira centrosomal localization was significantly correlated with Pros centrosomal localization in interphase Kin17 RNAi NBs. Among NBs with centrosomal localization of Mira, about 80% of them exhibited a centrosomal Pros to cytoplasmic Pros ratio of greater than 1.5 (Figs. 2G,H); whereas among the NBs with centrosomal localization of Pros, 93% of them also showed a centrosomal Mira to cytoplasmic Mira ratio of greater than 1.5 (Fig. S1C). In addition, we observed nuclear localization of Mira that was also correlated with Pros nuclear localization in Kin17 knockdown NBs at the interphase (Figs. S1D,E). Taken together, our data suggest that Kin17 regulates Mira localization in a cell cycle-dependent manner and that aberrant centrosomal and nuclear Mira localization correlates with of Pros localization to the centrosome and nucleus. This regulation appears to be specific as we did not observe mislocalization of the basal component Numb in Kin17 knockdown NBs (Figs. S2A,B).

Kin17 RNAi can potentially lead to defects in asymmetric cell division of NBs

Although we observed defects in Mira localization at metaphase during Kin17 knockdown, we rarely saw defects in basal cortical localization at telophase. The lack of basal cortical localization defects of Mira at the telophase could be due to correction of the Mira localization defects in a process known as “telophase rescue”⁴⁰ and may account for the fact we did not observe changes in the number of type I NBs during Kin17 RNAi. Telophase rescue involves Eiger and Traf4.^40,41 In order to examine if Kin17 RNAi could affect asymmetric division of NBs, we thus inhibited telophase rescue by Eiger RNAi. Eiger RNAi alone did not change the basal cortical localization of Mira or the number of type I NBs (Figs. 2I,J, S2C,D). Interestingly, Kin17 and Eiger double knockdown led to a loss of basal cortical localization of Mira in about 80% of telophase NBs and about a 5% increase in the number of type I NBs (Figs. 2I,J, S2C,D), indicating that Kin17 knockdown can lead to defects in asymmetric cell division when there is no telophase rescue. The increase in the number of type I NBs was small probably because most Kin17 RNAi NBs were not actively dividing due to the aberrant nuclear localization of Pros.

Phosphorylation at Serine-96 leads to centrosomal localization of Mira

As localization of Mira to the centrosome in wild-type NBs occurs in prophase when aPKC is recruited to phosphorylate Mira, we hypothesized that phosphorylation of Mira leads to its centrosomal localization and that defects in dephosphorylation are leading to accumulation of Mira to the centrosome and cortical localization defects in Kin17 RNAi NBs. Serine-96 (S96) is phosphorylated by aPKC and this phosphorylation is required and sufficient for the displacement of Mira from the cell cortex.^14,20,24 However, it is currently unknown if Mira is dephosphorylated prior to localization at the basal cortex. To directly assess if phosphorylation of S96 regulates the localization of Mira to the centrosome, we utilized phosphomutant alleles of mira. We observed that Mira^WT localizes similarly to endogenous Mira, except that in anaphase/telophase, Mira^WT was enriched at the centrosome and/or mitotic spindle in about 20% of NBs (Figs. 3A,B, S3A, compared to Figs. 2C,D,F). For phosphomimetic Mira^S96D, we observed an increase in cytoplasmic localization at interphase and prophase (Fig. S3A), and increased centrosomal localization through the entire cell cycle (Figs. 3A, B). In contrast, the phosphodead Mira^S96A primarily localized to the cell cortex and rarely to the centrosome at all stages (Figs 3A,B, S3A). The lack of centrosomal localization of Mira^S96A is unlikely due to sequestration at the cortex as Mira^S96A could not localize to the mitotic spindle either as Mira^WT and Mira^S96D did in syncytial embryos, which lack cell membranes (Fig. 3C). Therefore, the mislocalization of Mira to the centrosome observed in Kin17 knockdown NBs is likely due to phosphorylation of Mira on S96.

Figure 3. Phosphorylation of Mira residue S96 is required and sufficient for centrosomal localization.

(A,B) Localization of Mira^WT-mCherry, Mira^S96D-mCherry, and Mira^S96A-mCherry in NBs. Inset, Mira localization at apical (yellow boxes) and basal (grey boxes) centrosomes; Arrow, spindle localization of Mira; Scale bar, 5 μm; n ≥ 30 NBs. (C) Localization of Mira^WT-mCherry, Mira^S96D-mCherry, and Mira^S96A-mCherry in syncytial embryos. Scale bar, 25 μm. (D,E) Nuclear localization of Pros in embryonic NBs homozygous for Mira^WT-mCherry, Mira^S96D-mCherry, and MiraS96A-mCherry. Arrow, centrosome; Arrowhead, nuclear Pros; Scale bar, 5 μm; n ≥ 8 embryos. (F,G) Pros localization to the centrosome in interphase embryonic NBs homozygous for Mira^WT-mCherry, Mira^S96D-mCherry, and MiraS96A-mCherry. Yellow boxes, Pros localization at the centrosome; Scale bar, 5 μm; n ≥ 80 NBs. Data are presented as the mean ± 1 SD. **, p<0.01; ***, p<0.001, unpaired student’s t-test with the Bonferroni correction. See also Figure S3.

Next, we asked if aberrant phosphorylation of Mira is responsible for the localization of Pros to the nucleus and centrosome observed in Kin17 knockdown NBs. We examined Pros localization in embryonic NBs homozygous mutant for the phosphomutant mira alleles. We observed that Pros localized to the nucleus in 20 ± 8.7% of mira^S96D NBs compared to 0% and 5.6 ± 7.9% for mira^WT and mira^S96A NBs, respectively, at the interphase (Figs. 3D,E). Accordingly, we observed a significant increase in Pros localization to the centrosome in interphase mira^S96D NBs compared to mira^WT and mira^S96A NBs (Figs. 3D,F,G). These results suggest that nuclear and centrosomal Pros localization observed in Kin17 RNAi NBs could be due to phosphorylation of S96 of Mira.

PP4 dephosphorylates Mira at Serine-96

Aberrant Mira phosphorylation could be due to an increase in aPKC activity or a decrease in phosphatase activity. We observed that localization of apical Bazooka and aPKC were not altered in Kin17 knockdown during metaphase (Figs. S4A,B), but we observed aPKC localization to the cell cortex was reduced relative to the cytoplasm when compared to Kin17 rescue, suggesting that there is a reduction in aPKC levels (Figs. S4C). Together, this suggests that Kin17 may function through a phosphatase.

Falafel (Flfl), the targeting subunit of Protein Phosphatase 4 (PP4), is required for basal Mira localization and interacts with Mira, and PP4 is required for dephosphorylation of Mira at T591^21,22. However, it has not been shown if PP4 is required for dephosphorylation at S96. To test if PP4 dephosphorylates Mira at S96, we first examined the role of PP4 in Mira centrosomal localization. We found that knockdown of Flfl or the catalytic subunit of PP4, PP4–19c, led to increased Mira localization to the centrosome throughout the cell cycle (Figs. 4A,B). Furthermore, flflⁿ⁴² mutants exhibited nuclear Pros localization in 52.5% of NBs and a significant reduction in the mitotic index (Figs. S4D–G). However, we did not observe increased centrosomal localization of Mira^T591D in NBs compared to Mira^WT or MiraT591A (Figs. S4H,I). These phenotypes suggest that decreased PP4 activity but not defects in dephosphorylation at T591 could account for the aberrant centrosomal localization of Mira in Kin17 knockdown NBs.

Figure 4. PP4 is required for displacement of Mira from the centrosome.

(A,B) Localization of Mira in PP4-19c and Flfl RNAi NBs throughout the cell cycle. Inset, Mira localization at apical (yellow) and basal (grey) centrosomes; Scale bar, 5 μm; n ≥ 27 NBs. (C) IP of PP4-19c (left) and quantification of dephosphorylation of Mira phosho-Serine96 peptide by PP4 precipitates (right). n = 4 biological replicates. (D) Localization of RFP-Flfl in NBs. Arrowheads, mitotic spindle; Arrows, centrosome; Scale bar, 5 μm. (E,F) Localization of PP4-19c in ctrl and PP4-19c RNAi NBs. Inset, PP4-19c localization at apical (yellow) and basal (grey) centrosomes; Scale bar, 5 μm; n ≥ 61 NBs. (G,H) Localization of Mira in prophase NBs expressing V5-PACT or V5-Flfl-PACT. Inset, Mira localization at apical (yellow boxes) and basal (grey boxes) centrosomes; Scale bar, 5 μm; n ≥ 25 NBs. Data are presented as the mean ± 1 SD. n.s., not significant; *, p<0.05; **, p<0.01; ***, p<0.001, paired student’s t-test for (C), unpaired student’s t-test for (H), and unpaired Student’s t-test with the Bonferroni correction for (B, F). See also Figure S4.

To further test if PP4 can dephosphorylate Mira at S96, we utilized an in vitro malachite green-based phosphatase assay. We immunoprecipitated the PP4 complex from Drosophila S2 cells (Fig. 4C, left) and incubated this with a synthetic phospho-peptide corresponding to the region surrounding S96 [92-FRTP(pSer)LPQR-100]. We found that incubating PP4 immunoprecipitates with the Mira pS96 peptide led to a 3.6-fold increase in dephosphorylation compared to IgG immunoprecipitates (Fig. 4C, right). These data together with the similar centrosomal localization of Mira observed in Flfl/PP4-19c knockdown NBs and Kin17 knockdown/Mira^S96D mutant NBs strongly suggest that PP4 can dephosphorylates Mira at S96 and reduced PP4 activity and subsequent defects in Mira dephosphorylation at S96 is likely responsible for the aberrant Mira localization in Kin17 knockdown NBs.

The PP4 complex localizes to centrosomes and the mitotic spindle

We next wanted to identify the localization of the PP4 complex in NBs. In NBs, Flfl localizes to the nucleus during interphase.²¹ However, in Drosophila D.mel-2 cells, Flfl localizes to the kinetochores⁴² and PP4 is centrosomal in Drosophila embryos.^43,44 To visualize the localization of Flfl in NBs, we utilized UAS-RFP-Flfl²¹, which localized to the nucleus in interphase and was enriched at the centrosome and mitotic spindle in 67.3% of mitotic NBs (Fig. 4D). Additionally, PP4-19c was significantly enriched at the apical centrosome during mitosis compared to interphase and this enrichment at the apical centrosome diminished after PP4-19c knockdown (Figs. 4E,F). The localization of Flfl and PP4 supports a model that Mira is dephosphorylated by PP4 at the centrosome.

To investigate if Flfl functions at the centrosome to dephosphorylate Mira, we targeted Flfl specifically to the centrosome using the PACT domain of Plp.^45,46 We quantified Mira localization in prophase NBs expressing V5-Flfl-PACT and found that targeting Flfl to the centrosome lead to a significant reduction in Mira localization to the centrosome during prophase (Figs. 4G,H). A similar reduction was not observed when we overexpressed V5-PACT (Figs. 4G,H), indicating that targeting Flfl to the centrosome via the PACT domain may account for this difference and supporting the idea that Flfl/PP4 dephosphorylates Mira at the centrosome.

The centrosome is required for dephosphorylation and basal localization of Mira

The localization of the PP4 complex to the centrosome and the reduction in Mira centrosomal localization when Flfl-PACT is over-expressed suggests that dephosphorylation of Mira occurs at the centrosome. However, Mira could accumulate to the centrosome because of defects in dephosphorylation rather than as part of the dephosphorylation mechanism. To test that Mira dephosphorylation occurs at the centrosome, we knocked down Sas-4. Sas-4 is required for centriole replication and centrosomes are completely lost in Sas-4 RNAi NBs at late larval stages, although NBs assemble mitotic spindles and undergo mitosis.⁴⁷ We hypothesized that if dephosphorylation of Mira did not require the centrosome, Mira would still localize to the basal domain during mitosis. We confirmed that Flfl levels and aPKC localization were not altered in Sas-4 RNAi NBs (Figs. 5A–D). Knockdown of Sas-4 led to an increase in cytoplasmic Mira localization in interphase, prophase, and metaphase (Figs. 5E,F), demonstrating that the centrosome is essential for proper cortical localization of Mira. We then combined Sas-4 RNAi with the Mira S96 phosphomutants and found that Sas-4 RNAi with the mira^S96D allele led to a more severe phenotype in metaphase and telophase than mira^S96D alone or Sas-4 RNAi combined with mira^WT (Figs. 5G,H). In contrast, MiraS96A remained consistently cortical in Sas-4 RNAi NBs, indicating that increased cytoplasmic localization of Mira in Sas-4 RNAi NBs is due to enhanced phosphorylation at S96. In both experiments, we observed that endogenous Mira and Mira^WT showed increased cytoplasmic localization during interphase (prior to aPKC phosphorylation). This is likely because Mira could not be dephosphorylated properly in the previous rounds of mitosis due to the lack of the centrosome. Subsequently, the phosphorylated Mira could not be was fully segregated to the daughter GMC and remained in the cytoplasm of newly generated NBs during interphase. Further, the phosphorylated Mira may also bring the newly synthesized unphosphorylated Mira to the cytoplasm by forming dimmers. Together, these data indicate that the centrosome is required for proper cortical localization of Mira and that defects in cortical localization of Mira in NBs lacking the centrosome is likely due to defects in S96 dephosphorylation.

Figure 5. Centrosomes are required for proper localization and dephosphorylation of Mira.

(A,B) Flfl levels in Sas-4 RNAi NBs. Dotted lines, NBs; Scale bar, 10 μm; n = 3 biological replicates. (C,D) aPKC localization in Sas-4 RNAi NBs. Arrows, apical aPKC crescent; Scale bar, 5 μm; n ≥ 116 NBs. (E,F) Cortical localization of Mira in Sas-4 RNAi NBs. Yellow arrows, cortex; Scale bar, 5 μm; n ≥ 29 NBs. (G,H) Localization of Mira in NBs expressing Sas-4 RNAi in combination with Mira^WT, Mira^S96D, and MiraS96A. Yellow arrows, cortex; white arrows, mitotic spindle; Scale bar, 5 μm; n ≥ 25 NBs. (I) Verification of specificity of antibody to recognize phosphorylated S96 of Mira. (J,K) Levels of phosphorylated S96 in ctrl and Sas-4 RNAi brains. n = 4 biological replicates. Data are presented as the mean ± 1 SD. n.s., not significant; *, p<0.05, paired student’s t-test for (B, K) and unpaired student’s t-test for (D).

We then tested if Mira phosphorylation levels at S96 was increased in Sas-4 RNAi brains, utilizing an antibody specific to phosphorylated Phospho-(Ser) 14-3-3 binding motif which has the same sequence as the region surrounding S96 of Mira and was confirmed to specifically recognize Mira phosphorylated at S96 (Fig. 5I). We found that there was an increase in pS96 Mira relative to total Mira in Sas-4 RNAi brains compared to control brains (Fig. 5J,K), indicating that Sas-4 RNAi leads to an increase in Mira phosphorylated at S96. Taken together, these data indicate that dephosphorylation of Mira by PP4 likely occurs at the centrosome and this is required for proper basal cortical localization of Mira.

Kin17 regulates cellular levels of Flfl

We then investigated if Kin17 regulates PP4 activity by examining the expression of Flfl and PP4-19c. We found that Kin17 RNAi led to a reduction in Flfl levels to about half that observed in control interphase NBs (Figs. 6A,B). Additionally, we found a change in PP4-19c localization from nuclear to cytoplasmic at interphase (Figs. 6C). In Dictyostelium, SMEK, a Flfl homolog, is required for PP4C to localize to the nucleus.⁴⁸ Therefore, the cytoplasmic localization of PP4-19c in Kin17 RNAi NBs could be due to the reduction in Flfl expression. Indeed, PP4-19c became cytoplasmic in flflⁿ⁴² mutant NBs (Figs. 6C,D).

Figure 6. Kin17 regulates Flfl levels and functions upstream of Flfl.

(A,B) Relative Flfl levels in ctrl, Kin17 RNAi, and Kin17 rescue NBs. Dotted lines, NBs [identified by size and CD8-GFP expression (not shown)]; Scale bar, 5 μm; n = 3 biological replicates. (C,D) PP4-19c localization in Kin17 RNAi and flflⁿ⁴² NBs. Dashed circles, NBs. Ase staining indicates nucleus. Scale bar, 5 μm; n ≥ 51 NBs. Data are presented as the mean ± 1 SD. *, p<0.05; ***, p<0.001, paired student’s t-test with the Bonferroni correction for (B) and unpaired student’s t-test with the Bonferroni correction for (D).

The minor spliceosome component U6atac regulates Mira localization and Flfl expression

As previous experiments suggest that Kin17 may be involved in splicing^49–52 and Kin17 RNAi led to a reduction in the levels of Flfl, we hypothesized that Kin17 may regulate the splicing of the flfl transcript. To test this hypothesis, we examined how loss of the major or minor spliceosomes would affect Mira localization. Knockdown of the major spliceosome component, U2A, did not lead to Mira mislocalization (Figs. 7A,B). However, there was a significant increase in Mira localization to the centrosome in interphase and telophase and to the basal centrosome in prophase and metaphase in NBs homozygous mutant for u6atac (u6atac^k01105), a minor spliceosome component (Figs. 7A,B). Furthermore, we observed Mira cortical localization defects in u6atac^k01105 mutants (Figs. 7C,D). Consistent with the mislocalization of Mira, Pros localized to the nucleus in 45 ± 19.7% of u6atac^k01105 NBs (Figs. 7E,F). Consequently, the mitotic index of u6atac^k01105 mutant NBs was reduced to 14.4 ± 3.1% from 26.3 ± 5.9% in the control (Figs. 7G,H). Additionally, loss of U6atac led to a reduction in Flfl levels by 56% and localization of PP4-19c to the cytoplasm (Figs. 7I–K). These phenotypes are similar to those observed in Kin17 RNAi NBs, suggesting that Kin17 could function in splicing of the flfl pre-mRNA. Consistent with a role in splicing, Kin17-HA localized to the nucleus when expressed in NBs (Fig. 7L).

Figure 7. Kin17 regulates splicing of the flfl pre-mRNA.

(A,B) Localization of Mira throughout the cell cycle in u6atac^K01105 (u6atac^−/−) mutant and U2A RNAi NBs. Inset, Mira localization at apical (yellow boxes) and basal (grey boxes) centrosomes; Scale bar, 5 μm; n ≥ 19 NBs. (C,D) Cortical localization of Mira in metaphase u6atac^−/− NBs. Arrows, basal cortex. Scale bar, 5 μm; n ≥ 19 NBs. (E,F) Pros nuclear localization in u6atac^−/−mutant NBs. Dashed circles outline NBs. Scale bar, 5 μm; n ≥ 6 brains. (G,H) Mitotic Index in u6atac^−/− mutant brains. Yellow circles, mitotic NBs; Grey circles, non-mitotic NBs; Scale bar, 20 μm; n ≥ 7 brains. (I,J) Flfl staining in u6atac^−/− mutant NBs. Dashed circles, NBs; Scale bar, 10 μm; n = 3 biological replicates. (K) PP4-19c staining in u6atac^−/− mutant NBs. Dashed circles, NBs; Scale bar. (L) Localization of Kin17-HA in NBs during interphase and mitosis. Scale bar, 5 μm. (M) Quantitative RT-PCR results for Flfl transcript. Left panel, position of primers. n = 8 biological replicates (N) RNA immunoprecipitation of flfl mRNA by Kin17-HA in Drosophila embryos. Top panel, IP of Kin17-HA proteins; Bottom panels, flfl RT-PCR products obtained from cDNA libraries generated from RIP immunoprecipitates. Lysate and IP samples were run on the same blot but images were taken from different exposures. (O) aPKC and Pros RT-PCR results obtained from cDNA libraries generated from RIP immunoprecipitates. Data are presented as the mean ± 1 SD. n.s., not significant; *, p<0.05; **, p<0.01; ***, p<0.001, unpaired student’s t-test with Bonferroni correction (B). unpaired Student’s t-test for (F, H), and paired Student’s t-test for (J).

Kin17 interacts with the flfl mRNAs and knockdown of Kin17 leads to a reduction in flfl mRNAs

To assess if Kin17 and U6atac are involved in regulating splicing of flfl pre-mRNAs, we performed quantitative real-time PCR using primers to the mRNA of flfl. We found that in both Kin17 RNAi and u6atac^k01105 larval brains, there was a reduction in the spliced form of the flfl mRNA relative to control and eEF1a1 (Fig. 7M). Furthermore, immunoprecipitation of Kin17-HA from embryos showed that Kin17 interacts with the flfl mRNA (Fig. 7N). Although we were not able to immunoprecipitate flfl pre-mRNAs likely due to low abundance, the reduction in flfl mRNAs in Kin17 RNAi and u6atac^k01105 mutant brains and binding of Kin17 to flfl mRNAs support the hypothesis that Kin17 is required for splicing of the flfl mRNA. However, Kin17 may not just specifically bind to flfl mRNAs. We observed that aPKC mRNAs and pros mRNAs were also immunoprecipitated by Kin17 (Fig. 7O). Given that aPKC expression was also reduced in Kin17 RNAi NBs, it is possible that Kin17 may be involved in pre-mRNA splicing of other genes such as aPKC.

Discussion

We have identified Kin17 as essential for the proper localization of Mira in Drosophila NBs. The localization of Mira during asymmetric division is essential to the proper segregation of the fate determinants to the daughter cell and the regulation of its phosphorylation state via an interplay between kinases and phosphatases is essential to this localization. We demonstrate that PP4 is required for the dephosphorylation of Mira at S96 at the centrosome beginning in prophase and completing dephosphorylation by the end of mitosis. Dephosphorylation of Mira at the centrosome is essential for proper localization of Mira to the basal domain and prevents accumulation of Mira at the centrosome and in the cytoplasm. Knockdown of Kin17 leads to reduction in Flfl expression and PP4 activity, consequently, Mira cannot be fully dephosphorylated at S96 and accumulates in the cytoplasm and the centrosome instead of binding to the basal cortex, which leads to Pros being translocated to the nucleus. Further, similar mislocalization of Mira/Pros and reduction in flfl mRNAs observed in u6atac mutant and Kin17 knockdown NBs and binding of Kin17 to flfl mRNAs suggest that Kin17 is likely involved in flfl splicing.

Kin17 has not been well-studied in Drosophila. In mammals, it is essential for DNA repair and replication. Kin17 has been shown to interact with the spliceosome^49,51,52 and has very recently been identified as a splicing factor in C. elegans, where Kin17 is required to maintain the 5’ splice site identity during spliceosome assembly.⁵⁰ In this system, mutations in Kin17 led to changes in alternative 5’ splice site usage.⁵⁰ Kin17 may play a similar role in humans, as mass spectrometry has identified it to primarily interact with the spliceosome in B stage, which is the first formation of the entire spliceosome.⁴⁹ Here we show that Kin17 binds to flfl mRNAs and knockdown of Kin17 leads to reduction in flfl mRNAs, and that mutants for the minor spliceosome component u6atac phenocopy Kin17 knockdown, supporting that in Drosophila Kin17 functions to regulate splicing of transcripts. In the mass spectrometry analysis, Kin17 was only found in complexes isolated from both HeLa and Drosophila Kc cells using fushi tarazu as bait but not zeste⁵¹, suggesting that Kin17 likely regulates splicing of specific transcripts rather than acting as a general splicing factor. However, Kin17 may not just regulates the splicing of flfl pre-mRNAs in the NBs as Kin17 also binds to aPKC and pros mRNAs and Kin17 knockdown also leads to reduction in aPKC expression. It is possible that Kin17 may also regulate splicing of other transcripts such as aPKC transcripts. Our findings that Kin17 is potentially involved in regulating splicing of particular transcripts could be helpful for investigating underlying mechanisms of pathogenesis of various cancers associated with increased Kin17 expression.^{30,31,34,36,53}

Mira localization to the mitotic spindle/centrosome has been observed in the syncytial embryo²³, and interactions between Mira and the microtubules have also been observed in the anterior pole of oocytes.⁵⁴ Mutants in polarity proteins can lead to localization of Mira to the centrosome and mitotic spindle in NBs, raising the question of why Mira does not localize to the mitotic spindle in NBs, and whether localization to the mitotic spindle during mitosis leads to consequences for asymmetric cell division. Our work provides evidence that localization to the centrosome and potentially the mitotic spindle (based on the S96 alleles) does occur in wild-type NBs but it is likely transient and appears to be cell-cycle dependent. Our work also reveals that localization of Mira to the centrosome/spindle depends on the phosphorylation status of Mira at S96 and that localization of Mira to the centrosome is essential for dephosphorylation of S96 and cortical localization as Sas-4 RNAi leads to defects in cortical localization and an increase in the levels of phosphorylated Mira. Although PP4 has been implicated in dephosphorylation of Mira at T59122, our work identifies S96 as a dephosphorylation site of PP4 and shows that T591 phosphorylation does not contribute to the observed phenotype. In fact, we observed a slight decrease in T591D centrosomal localization. This may be due to the fact that T591 dephosphorylation is required prior to aPKC phosphorylation of Mira to ensure proper localization.²² Thus, Mira removal from the centrosome is specific to dephosphorylating S96. Our finding of dephosphorylation of S96 by PP4 at the centrosome answers several outstanding questions regarding Mira localization after aPKC phosphorylation, including if PP4 dephosphorylates Mira, the function of centrosomal localization of Mira, and if dephosphorylation of S96 occurs prior to localization of Mira to the basal domain.

Interestingly, we observed an increase of Mira localization to the centrosome in PP4 and Flfl RNAi NBs with a minimal increase in cytoplasmic Mira. This suggests that centrosomal/spindle localization of Mira may be more sensitive to changes in the levels of PP4 activity. The phosphorylated Mira may preferentially localize to the centrosome/spindle and bring its cargo protein Pros to the centrosome as well as indicated by the colocalization of Pros and Mira to the centrosome in Kin17 RNAi NBs. Additionally, in Kin17 RNAi NBs, we observed defects in Mira cortical localization in interphase, which is likely related to the localization of Pros to the nucleus. mira null mutants exhibit nuclear Pros localization²⁴, suggesting that Pros must be tethered to something to prevent nuclear localization and that defects in S96 dephosphorylation cause tethering to the cortex to be impaired so Mira/Pros then localizes to the centrosome and nucleus. It is tempting to speculate that when the amount of phosphorylated Mira exceeds the binding capacity of the centrosome/spindle, it may “overflow” to the cytoplasm, leading to untethered Pros being transported into the nucleus.

In summary, our studies identify a factor, Kin17, that regulates the proper localization of Mira during asymmetric cell division potentially through regulation of the splicing of flfl transcripts and provides evidence to demonstrate that dephosphorylation of Mira at S96 by PP4 at the centrosome is essential for proper localization of Mira in NBs during asymmetric cell division.

Limitations of the Study

There are several limitations to the study that prevent making further conclusions. First, while our data suggest that Kin17 functions in the splicing of the flfl transcript, this will have to be confirmed through further experiments that are able to capture the flfl pre-mRNA through sequencing or other methods that will detect the transcript. Second, the presence of redundant mechanisms, such as telophase rescue, may not allow us to see the full effect of the loss of Kin17 in terms of asymmetric cell division and development. While our data suggests that when telophase rescue is also lost, Kin17 does affect cell polarity in telophase indicating Kin17 could potentially lead to defects in cell fate, further studies will have to be carried out to determine the effect this has on development. Thirdly, as we used RNAi, some of the phenotypes may not be fully penetrant, such as Flfl RNAi and PP4-19c RNAi, which may not lead to complete phenotypes and prevent us from determining the full effect on cell polarity that these proteins have.

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, Sijun Zhu (zhus@upstate.edu).

Materials availability

All materials generated in this study are available from the lead contact upon request.

Data and code availability

• All data reported in this paper will be shared by the lead contact upon request.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Experimental model and study participant details

Fly Stocks

insc-Gal4⁶ was used for transgene expression in NBs and UAS-CD8-GFP⁵⁵ was used to mark neuroblast lineages. UAS transgenes for RNAi knockdown or overexpression include: UAS-Kin17 RNAi⁶⁰ (#55692; Bloomington Drosophila Stock Center (BDSC) (IN, USA)), UAS-kin17 (this work), UAS-kin17-HA (this work), UAS-V5-Flfl-PACT (this work), UAS-V5-PACT (this work), UAS-Mira^T591WT-GFP (this work), UAS-Mira^T591D-GFP (this work), UAS-Mira^T591A-GFP (this work), UAS-RFP-flfl²¹ (#66538, BDSC), UAS-numb^ΔCT-GFP⁵⁸, UAS-flfl RNAi²¹ (#66541, BDSC), UAS-U2A-RNAi⁵⁶ (#33671; BDSC), UAS-Sas-4 RNAi⁶⁰ (#35046, BDSC), UAS-EigerIR⁶¹ (#58993; BDSC), and UAS-PP419c RNAi⁶⁰ (#57823; BDSC). The Mira phosphomimetic alleles, mira^WT-HA-mCherry, mira^S96A-HA-mCherry, mira^S96D-HA-mCherry24 were generously provided by Jens Januschke. snRNA:U6atac^{k01105 62} (#10492, BDSC), flfl^{n42 21} (#66534, BDSC), pros^{17 63} (#5458, BDSC) and aPKC^{K06403 62} (#10622, BDSC) mutant alleles were used for phenotypic analysis. Mutant chromosomes were balanced over CyO, weeP or TM6B, Tb for larval analysis and TM3, Tw-Gal4, UAS-2xGFP for embryonic analysis. Lines were raised at 25°C on standard fly food. Larva were dissected at 3rd instar larval stage, except for mutants that did not survive until that stage. Embryos were collected at syncytial stages or stages 9–11.

Cell Culture

Drosophila S2 cells were grown in Schneider’s medium (ThermoFisher Scientific) supplemented with 10% FBS (ThermoFisher Scientific) and 5% streptomycin/penicillin (Gibco) at 37°C.

Method details

RNAi knockdown and transgene analysis

For RNAi knockdown and transgene overexpression, embryos were collected for 24 hours at 25°C, staged at 25°C for an additional 24 hours, and then shifted to 29°C until dissection at third instar larval stages. To visualize RFP-Flfl localization, crosses were collected at 25°C and shifted to 29°C, 24 hrs prior to dissection at third instar larval stages. Crosses involving UAS-numb^ΔCT-GFP were performed at 25°C. For syncytial embryo staining, embryos were collected for 1 hour and staged for 1 hour at 25°C before fixation. For stage 9–11 embryonic NB staining, embryos were collected for 4 hours and staged for 4 hours at 25°C before fixation.

Immunostaining and confocal microscopy

Third instar larval brains were dissected in PBS and fixed within 20 minutes in 4% formaldehyde (Fisher (NH, USA)) in PBS for 20 minutes except for brains stained for Mira which were fixed for 35 minutes. Following fixation and washing, brains were blocked for 30 min in PBS supplemented with 5% normal donkey serum (Jackson Immunoresearch (PA, USA)) and 0.3% Triton X-100 (Fisher). Brains were incubated with primary antibodies overnight at 4°C and secondary antibodies for 2 hours at RT. Embryos were stained as previously described.⁶⁴ Coverslips were mounted using antifade reagent in glycerol (Invitrogen). Images were taken using a Zeiss LSM 780 (Zeiss, Germany) inverted confocal microscope using a 40× 1.4 numerical aperture oil immersion objective. Images were collected using Zen Software (Zeiss) and processed with Fiji.⁶⁵

Construction of plasmids and generation of transgenic lines.

For pUAST-Kin17 and pUAST-Kin17 HA, the open reading frame was amplified from a larval brain cDNA library using CloneAmp HiFi PCR Premix (Takara Bio). For the HA-tagged version, the HA tag was integrated into the primer at the c-terminus of Kin17. Primers used were: Kin17 F and Kin17 R or Kin17 HA F and Kin17 HA R. The PCR products were digested by EcoRI and XbaI or XhoI before cloning into the pUAST vector.⁵⁹ UAS-Kin17 and UAS-Kin17 HA were integrated randomly into the genome of w1118 embryos and positive lines were mapped. Injections were performed by The Best Gene (Chino Hills, CA, USA).

To generate pUAST-Mira T591 mutants, Mira was amplified from genomic DNA using the following primers: Mira F and Mira R. Mira was cloned into pJET using the CloneJet PCR cloning kit (ThermoFisher) and T591A/D mutations were induced using quickchange PCR with the following primers: T591D F and T591D R or T591A F and T591 R. Mira^WT, MiraT591A, and Mira^T591D were digested from pJET using BglII and XhoI and cloned into pUAST-GFP (see above). UAS-Mira^WT-GFP, UAS-Mira^T591D-GFP, and UAS-MiraT591A-GFP were integrated randomly into the genome of w¹¹¹⁸ embryos and positive lines were mapped. Injections were performed by Rainbow Transgenic Flies, Inc. (CA, USA).

To generate pUAST-V5-PACT and pUAST-V5-Flfl-PACT, Flfl was amplified from a Drosophila cDNA library using the Flfl (PACT) F and Flfl (PACT) R primers. The PACT domain of Drosophila Plp was cloned from a cDNA library using the PACT F and PACT R primers. Flfl and PACT were cloned into pUAST-V5 using EcoRI and KpnI. UAS-V5-PACT and UAS-V5-Flfl-PACT were integrated randomly into the genome of w¹¹¹⁸ embryos and positive lines were mapped. Injections were performed by Rainbow Transgenic Flies, Inc. (Camarillo, CA, USA).

Phosphatase assays

S2 cells were lysed in lysis buffer (50 mM Tris-HCl pH 7.4, 50 mM NaCl, 1 mM EDTA, 0.1% NP40) supplemented with HALT protease inhibitor cocktail (Thermo Scientific (MA, USA)). Endogenous PP4 was immunoprecipitated with Protein A/G Plus Agarose beads (Santa Cruz Biotechnology) and 4 μg/mg protein of anti-PP4-19c antibody (Rb, Proteintech (IL, USA)) with anti-IgG (Rb, Invitrogen) as a control. Immunoprecipitation was confirmed by western blotting. The immunoprecipitates were washed three times with TBS and then with Ser/Thr assay buffer (Ser/Thr Phosphatase Assay Kit 1 Millipore Sigma (MA, USA)). We utilized 1 mM phosphoserine peptide generated by (Biomatik, (Canada)) with the sequence, FRTP(pSer)LPQR, corresponding to the S96 region of Mira. The phosphatase assay was carried out using the Ser/Thr Phosphatase Assay Kit 1 following the manufacturer’s protocol, with the incubation time of the phosphatase with the peptide increased to 2 hours.

Immunoprecipitation and Immunoblotting

Brains were homogenized in RIPA buffer supplemented with HALT protease inhibitor cocktail and Halt phosphatase inhibitor cocktail (Thermo Scientific). Mira^WT-HA-mCherry was precipitated with Protein A/G Plus Agarose beads and 1 μg/mg protein of anti-HA antibody (Rb, Invitrogen). The immunoprecipitates were washed three times with lysis buffer and resuspended in SDS sample buffer.

Lysates/Immunoprecipitates were mixed with SDS sample buffer, separated by SDS-PAGE, transferred to an Immersion-PSQ PVDF membrane (Millipore Sigma), and blotted with indicated primary antibodies. Blots were visualized with SuperSignal West Pico or Femto PLUS Chemoluminescent Substrate (ThermoFisher Scientific) on a BioRad Chemidoc MP Imaging System. For phospho-S96 blots, after probing with the phospho-specific antibody, blots were stripped using mild stripping buffer (1.5% glycine, 0.1% SDS, 1% Tween 20, pH 2.2) and reprobed with HA antibody.

PCR and quantitative real-time PCR

RNA for quantitative Real-time PCR was extracted from 3rd instar larval brains using the Qiagen RNeasy kit, digested with DNase I (NEB) following the manufacturer’s protocol, and repurified with the Qiagen RNeasy kit. Primers were designed using Primer3web version 4.1.0 (http://bioinfo.ut.ee/primer3) as recommended by the manufacturer and generated by IDT. Primer used were: eEF1a1 qPCR F, eEF1a1 qPCR R, Flfl qPCR F, and Flfl qPCR R. Optimal concentrations of each primer and primer efficiency were determined, and products were run on an agarose gel to ensure correct products were generated. Real-time PCR reactions were set up using the iTaq Universal SYBR Green One-Step RT-PCR kit (Biorad) according to manufacturer’s directions. Reactions were run in triplicate with a BioRad CFX 384 Real Time PCR system. Results were analyzed using the Delta-Delta Ct method. The expression of each transcript was normalized to eEF1a1.⁶⁶

RNA Immunoprecipitation

Embryos expressing Kin17-HA under the control of insc-Gal4 were collected for 16 hours, dechorionated with 50% bleach, homogenized with a pestle, and lysed in lysis buffer (100 mM KCl, 10 mM Hepes, pH 7.0, 5mM MgCl2, 0.25% Triton, 0.25% NP40, 0.01% SDS, 2x HALT Protease Inhibitor Cocktail (ThermoFisher Scientific), 100 units/ml Ribolock RNase Inhibitor (ThermoFisher Scientific)). One milligram of protein was incubated with 4 μg of antibody (IgG (Rb, Invitrogen) or HA (Rb, Invitrogen)) and 50 μl of Protein A/G Plus agarose beads (Santa Cruz Biotechnology) for 4 hours. Prior to immunoprecipitation, the beads were pre-blocked overnight with single-stranded Salmon Sperm DNA (0.5 mg/ml, Sigma Aldrich). Beads were collected, washed 3 times with wash buffer. RNA was purified using RNAzol RT (Molecular Research Center, Inc. (OH, USA)) following the manufacturer’s protocol. RNA was then treated with DNase I (NEB) following the manufacturer’s protocol, and 250–500 ng of RNA was used to generate a cDNA library with MMLV H- Reverse transcriptase (Promega) and Random primers (Promega) following the manufacturer’s protocol. PCRs were performed using DreamTaq Hot Start Green PCR Master Mix (ThermoFisher Scientific) in 10 μl volume with 1 μl of the RT reaction with the following primers: Flfl RIP F1, Flfl RIP R1, Flfl RIP F2. Flfl RIP R2, aPKC RIP F, aPKC RIP R, Pros RIP F, and Pros RIP R. PCR conditions were as follows: 95°C for 5 min, followed by 30 cycles of 95°C for 30 sec, 54°C for 30 sec, 72°C for 30 sec, and a final extension at 72°C for 5 min. Products were detected on a 1.5% agarose gel by staining with SYBRGold (Invitrogen).

Quantification and statistical analysis

Quantification was performed by pooling neuroblast numbers from multiple brains unless otherwise indicated. Statistical analysis was performed using Microsoft Excel. Pairwise comparisons were made using two-tailed, paired, or unpaired student’s t-tests. Multiple comparisons were performed using two-tailed, paired, or unpaired t-tests with the Bonferroni correction. Data was presented as mean ± standard deviation (SD). P values of <0.05 were considered statistically significant.

To quantify the mitotic index, the ratio of Dpn+, PH3+ NBs to Dpn+ NBs was determined for each brain, with a total of at least 30 NBs counted per brain lobe. The average and standard deviation were calculated from the mitotic index calculated for each brain.

Quantification of nuclear Pros was determined by colocalization of Pros with Deadpan in interphase cells. NBs were determined to be positive for nuclear Pros if the nuclear Pros intensity was greater than 2x than that of the cytoplasmic intensity. The percentage for each brain was determined by quantifying 10 NBs per brain (blinded by turning off the Pros channel during measurement), and the average and standard deviation were calculated from the percentage calculated for each brain. In embryos, quantification was the same except 10 NBs from a single plane were quantified using Mira staining as a guide for nuclear localization (blinded by turning off the Pros channel during measurement).

To quantify colocalization of Mira and Pros at the centrosome in Kin17 RNAi NBs, we first visually identified NBs with a bright spot of Mira staining at the centrosome with the channel of Pros staining turned off, followed by measuring the intensity of Pros and Mira staining at the centrosome and cytoplasm and calculated the ratio. We also did the reverse way by identifying NBs with a bright spot of Pros staining at the centrosome with the channel of Mira staining turned off. The ratio of staining intensity of Pros at the centrosome vs that in the cytoplasm was then plotted against that of Mira or vice versa. The significance of correlation was determined by calculating the spearman coefficient (r).

To quantify the Flfl levels in each NB, the average intensity of the Flfl signal for 10 NBs from each brain was quantified and the average intensity for each brain was determined. For each experiment, an average was calculated from the average intensity of each brain and the intensity for the mutants was normalized to that of the control.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit polyclonal PP4-19c (IP: 4 μg/mg; IF: 1:1000; IB: 1:1000)	Proteintech	#10262-1-AP
Rabbit monoclonal HA (IP: 1μg/mg; IF: 1:500; IB: 1:1000)	Invitrogen	#3724
Guinea Pig Ase (IF: 1:5000)	Y.N. Jan	N/A
Rabbit Dpn (IF: 1:500)	Y.N. Jan1	N/A
Mouse monoclonal Pros (IF: 1:5 (cortical/centrosomal), 1:50 (nuclear))	Developmental Hybridoma Studies Bank	#Prospero (MR1A)
Rabbit polyclonal Plp (IF: 1:500)	Abcam	#4448
Mouse monoclonal PH3 (IF: 1:1000)	Cell Signaling Technology	#9706
Mouse monoclonal Elav (IF: 1:100)	Developmental Hybridoma Studies Bank	#Elav-9F8A9
Mouse monoclonal α-tub (IF: 1:1000)	Sigma	#T199
Mouse monocolonal γ-tub (IF: 1:1000)	Sigma	#T5326
Rabbit Mira (IF: 1:1000)	Y.N. Jan	N/A
Rat Flfl (IF: 1:500)	Z. Lipinszki2	N/A
Guinea Pig Baz (IF: 1:500)	C. Doe3	N/A
Mouse monoclonal aPKC (IF: 1:500)	Santa Cruz Biotechnology	#sc-17781
Chicken polyclonal GFP (IF: 1:500)	Aves Lab	#GFP-1010
Rabbit DsRed (IF: 1:500)	Takara Bio	#632496
Chicken Alexafluor-488 (IF: 1:500)	Invitrogen	#A-11039
Rabbit Alexafluor-647 (IF: 1:500)	Invitrogen	#A-21245
Mouse Alexafluor-647 (IF: 1:500)	Invitrogen	#A-21235
Rabbit Rhodamine Red RRX (IF: 1:500)	Jackson Immunoresearch	#611-295-215
Rat Rhodamine Red RRX (IF: 1:500)	Jackson Immunoresearch	#712-295-153
Mouse Rhodamine Red RRX (IF: 1:500)	Jackson Immunoresearch	#712-295-150
Rabbit IgG (IP: 1–4 μg/mg)	Invitrogen	#02–6102
Rabbit polyclonal Phospho-(Ser) 14-3-3 binding motif (IB: 1:1000)	Cell Signaling Technology	#9601S
Rabbit HRP (IB: 1:2000)	Cell Signaling Technology	#7074
Rabbit HRP (IB 1:2000)	Jackson Immunoresearch	#112-035-003
Bacterial and virus strains
DH5 alpha E. coli
Biological samples
Chemicals, peptides, and recombinant proteins
Normal Donkey Serum	Jackson Immunoresearch	#017-000-121
Triton X-100	Fisher	#AAA16046AE
37% Formaldehyde	Fisher	#F79-1
HALT protease inhibitor cocktail	Thermo Scientific	#78430
HALT Phosphatase inhibitor cocktail	Thermo Scientific	#78420
Fetal Bovine Serum	Thermo Scientific	#26150079
Penicillin/Streptomycin	Thermo Scientific	#15140122
FRTP(pSer)LPQR	Biomatik	N/A
Critical commercial assays
Ser/Thr Phosphatase Assay Kit 1	Millipore Sigma	#17-127
iTaq Universal Sybr Green One-Step RT PCR kit	Biorad	#1725120
Deposited data
Experimental models: Cell lines
D. melanogaster S2 cells	F. Pignoni	N/A
Experimental models: Organisms/strains
D. melanogaster insc-Gal4	J. Knoblich4	N/A
D. melanogaster UAS-CD8-GFP	L. Luo5	N/A
D. melanogaster UAS-Kin17 RNAi (y1, sc*, v1, sev21; P{TRIP.HMC03906}attP40)	Bloomington Drosophila Stock Center	#55692; http://flybase.org/reports/FBal0294464
D. melanogaster UAS-Kin17	This work	N/A
D. melanogaster UAS-Kin17-HA	This work	N/A
D. melanogaster UAS-V5-flfl-PACT	This work	N/A
D. melanogaster UAS-V5-PACT	This work	N/A
D. melanogaster UAS-MiraT591WT-GFP	This work	N/A
D. melanogaster UAS-MiraT591D-GFP	This work	N/A
D. melanogaster UAS-MiraT591A-GFP	This work	N/A
D. melanogaster UAS-RFP-Flfl	Bloomington Drosophila Stock Center	#66538; http://flybase.org/reports/FBal0322866
D. melanogaster UAS-Flfl RNAi	Bloomington Drosophila Stock Center	#66541; http://flybase.org/reports/FBal0322868
D. melanogaster UAS-numbΔCT-GFP	B. Lu6	http://flybase.org/reports/FBtp0073064
D. melanogaster UAS-U2A RNAi (y1, sc*, v1, sev21; P{TRIP.HMS00535}attP2/TM3, Sb1)	Bloomington Drosophila Stock Center	#33671; http://flybase.org/reports/FBal0257223
D. melanogaster UAS-Sas4 RNAi (y1, sc*, v1, sev21; P{TRIP.HMS01463}attP2)	Bloomington Drosophila Stock Center	#35049; http://flybase.org/reports/FBal0263371
D. melanogaster UAS-EigerIR	Bloomington Drosophila Stock Center	#58993; http://flybase.org/reports/FBti0164915
D. melanogaster UAS-PP4-19c RNAi (y 1 , v 1 ; P{TRIP.HMJ21831}attP40)	Bloomington Drosophila Stock Center	#57823; http://flybase.org/reports/FBal0300358
D. melanogaster miraWT-HA-mcherry	J. Januschke7	http://flybase.org/reports/FBal0361098
D. melanogaster miraS96D-HA-mcherry	J. Januschke7	http://flybase.org/reports/FBal0361230
D. melanogaster miraS96A-HA-mcherry	J. Januschke7	http://flybase.org/reports/FBal0361228
D. melanogaster snRNA:U6atac k01105	Bloomington Drosophila Stock Center	#10492; http://flybase.org/reports/FBal0064601
D. melanogaster flfl n42	Bloomington Drosophila Stock Center	#66534; http://flybase.org/reports/FBal0241927
D. melanogaster pros 17	Bloomington Drosophila Stock Center	#5458; http://flybase.org/reports/FBal0032479
D. melanogaster aPKC k06403	Bloomington Drosophila Stock Center	#10622; http://flybase.org/reports/FBal0064438
Oligonucleotides
Random primers	Invitrogen	#481900111
Kin17 F: 5’GGAATTCATGGGTCGCGCCGAGGTA-3’	This work, IDT	N/A
Kin17 R: 5’-GCTCTAGACTAGGCGCCATGTAGTTTAGATAT-3’	This work, IDT	N/A
Kin17 HA F: 5’-ATCGATGAATTCATGGGTCGCGCCGAGGTAGGT-3’	This work, IDT	N/A
Kin17 HA R: 5’-ATCGATCTCGAGTTCCTAAGCGTAATCTGGAACATCGTAAGGGTAGGCGCCATGTAGTTTAGATAT-3’	This work, IDT	N/A
Mira F: 5’-TTCAGTAGATCTATGTCTTTCTCCAAGGCCAAG-3’	This work, IDT	N/A
Mira R: 5’-TTCAGTCTCGAGGATGTTGCGCGCCTTGAGCAC-3’	This work, IDT	N/A
T591D F: 5’-CTGAGATCCTCCTCCCAGGATCTGCAGAGCGAGGTATCG-3’	This work, IDT	N/A
T591D R: 5’-CGATACCTCGCTCTGCAGATCCTGGGAGGAGGATCTCAG-3’	This work, IDT	N/A
T591A F: 5’-CTGAGATCCTCCTCCCAGGCCCTGCAGAGCGAGGTATCG-3’),	This work, IDT	N/A
T591A R: 5’-CGATACCTCGCTCTGCAGGGCCTGGGAGGAGGATCTCAG-3’	This work, IDT	N/A
Flfl (PACT) F: 5’- GAATTCATGACGACTGACACCCGCCGAC-3’	This work, IDT	N/A
Flfl (PACT) R: 5’- GAATTCTGCCTGACGCGCGCGCTTTTGTGC-3’	This work, IDT	N/A
PACT F: 5’- GGTACCTTATTGCTCTGCAGAAGAAATGCG-3’	This work, IDT	N/A
PACT R: 5’- GGTACCATGATGCCGCGCATGCGCTC-3’	This work, IDT	N/A
eEF1a1 qPCR F: 5’-CTACAAGTGCGGTGGTATCG	This work, IDT	N/A
eEF1a1 qPCR R: 5’-TTATCCAAAACCCAGGCGTA-3’	This work, IDT	N/A
Flfl qPCR F: 5’-CTCCAGCTGTGTCCAGTCC-3’	This work, IDT	N/A
Flfl qPCR R: 5’-TCGTAGTCATCTTCGCCAGA-3’	This work, IDT	N/A
Flfl RIP F1: 5’-CCACGTGTCATCCACCTATG-3’	This work, IDT	N/A
Flfl RIP R1: 5’-CTCAGGGCCAGATCAAAGTTAT-3’	This work, IDT	N/A
FlflI RIP F2: 5’-TAGAGTTCTCGCCTCTGGTAG-3’	This work, IDT	N/A
FlflI RIP R2: 5’-CCTTCTCGGTGAGCATGTTT-3’	This work, IDT	N/A
aPKC RIP F: 5’-CAGCATCTATCGACGCGGTGC-3’	This work, IDT	N/A
aPKC RIP R: 5’-CTGACGTCCCAAACCCCAGAT-3’	This work, IDT	N/A
Pros RIP F: 5’-GCACGACAAGCTGTCACCGA-3’	This work, IDT	N/A
Pros RIP R: 5’-CTCGTGTCTTTGCCACCCT-3’	This work, IDT	N/A
Recombinant DNA
pUAST vector	N. Perrimon8	N/A
Software and algorithms
ImageJ (Fiji)	http://fiji.sc	N/A
Zen software	Zeiss	N/A
Other
Antifade Reagent in Glycerol	Invitrogen	#S36936
Zeiss LSM 780 inverted confocal microscope	Zeiss	N/A
Protein A/G Plus Agarose Beads	Santa Cruz Biotechnology	#sc-2003
Immersion-PSQ PVDF membrane	Millipore Sigma	#ISEQ00010
SuperSignal West Pico PLUS Chemoluminescent Substrate	ThermoFisher Scientific	#34579
SuperSignal West Femto Chemiluminescent Substrate	ThermoFisher Scientific	#34094
CloneAmp HiFi PCR Premix	Takara Bio	#639298
Schneiders Medium	ThermoFisher Scientific	#21720024
RNeasy Kit	Qiagen	#74104
RNAzol RT	Molecular Research Center Inc	#RN 190
Ribolock RNase Inhibitor	ThermoFisher Scientific	#EO0381
Salmon Sperm DNA	Sigma Aldrich	#7656
Proteinase K	Sigma Aldrich	#124568
DNase I	New England Biolabs	#M0303
MMLV H- Reverse transcriptase	Promega	#M5301
DreamTaq Hot Start Green PCR Master Mix	ThermoFisher Scientific	#K9021
EcoRI	ThermoFisher Scientific	#ER0271
XhoI	ThermoFisher Scientific	#ER0692
XbaI	ThermoFisher Scientific	#ER0682
BglII	ThermoFisher Scientific	#ER0082
KpnI	ThermoFisher Scientific	#ER0521
CloneJet PCR Cloning Kit	ThermoFisher Scientific	#K1231
SYBR Gold	Invitrogen	#S11494