Regulation of Notch signaling by the chromatin-modeling protein Hat-trick

Notch signaling is an evolutionarily conserved pathway that regulates a variety of developmental processes, including acquisition of specific cell fates, cell proliferation, differentiation, self-renewal and cell death programs (Cabrera, 1990; Egan et al., 1998; Artavanis-Tsakonas et al., 1999; Fehon et al., 2007; Fortini, 2009; Liu et al., 2010; Andersson et al., 2011; Guruharsha et al., 2012). Notch is subjected to tight regulation at the level of receptor and ligand biosynthesis, post-translational modifications, ligand–receptor interaction and trafficking. The intricate regulatory mechanisms of Notch signaling account for the diverse functions of same pathway in numerous cellular contexts during organism development (Baron et al., 2002). The Notch receptor is synthesized as a 300 kDa single polypeptide precursor that is cleaved by furin-like convertase(s) during maturation in the trans-Golgi network. The N-terminal extracellular subunit and a C-terminal transmembrane intracellular subunit join via a noncovalent bond to form a single transmembrane heterodimeric receptor that translocates to the cell membrane (Blaumueller et al., 1997; Logeat et al., 1998). Notch signal transduction is initiated when the Notch receptor interacts with its ligands, Delta and Serrate (Drosophila), from neighboring cells (Rebay et al., 1991), and initiates proteolytic cleavage by the A Disintegrin And Metalloprotease domain (ADAM)/TNFα Converting Enzyme (TACE)/Kuzbanian family of metalloproteases (Brou et al., 2000; Mumm et al., 2000). This cleavage of the extracellular domain from the transmembrane domain generates a membrane-tethered Notch, known as Notch extracellular truncation (NEXT) fragment, that is subsequently cleaved by the γ-secretase complex (De Strooper et al., 1999; Struhl and Greenwald, 1999; Ye et al., 1999). This cleavage results in the release of Notch-intracellular domain (Notch-ICD/NICD), which eventually translocates to the nucleus via Importin-α3 (Sachan et al., 2013). In the nucleus, Notch-ICD binds to, and activates, the transcription factor, Suppressor of Hairless [Su(H)] in Drosophila (CBF1/RBPJ in vertebrates) and displaces the co-repressors and recruits co-activators, including Mastermind (Mam), leading to the activation of Notch target genes (Fortini and Artavanis-Tsakonas, 1994; Struhl and Adachi, 1998; Wu et al., 2000; Cau and Blader, 2009). In Drosophila, most of the Notch target genes identified so far are in the Enhancer of Split complex [E(spl)]. This is a complex locus containing, among others, seven transcription units (m8, m7, m5, m3, mβ, mγ and mδ), which encode basic helix-loop-helix (bHLH) transcriptional repressors (Delidakis and Artavanis-Tsakonas, 1992; de Celis et al., 1996; Bray and Furriols, 2001). This complex of genes is often regulated by Notch signaling and is important for neurogenesis and cell fate decisions.

To identify novel regulators of Notch-ICD, a yeast two-hybrid screen was performed, which identified Drosophila Hat-trick (Htk) as an interacting partner of Notch-ICD. Htk is a DNA-binding protein that harbors the AT-rich interacting domain (ARID), chromatin organization modifier (Chromo) and Tudor domains (Fig. 1A). These domains are present in proteins that are involved in DNA binding and chromatin modeling (Herrscher et al., 1995; Iwahara, and Clubb, 1999; Akhtar et al., 2000; Brehm et al., 2000; Bouazoune et al., 2002; Iwahara et al., 2002; Kim et al., 2006). In addition to the sequence homology with chromatin-remodeling proteins, Htk is also reported to interact with the histone deacetylase, Sin3A (Spain et al., 2010). We recently revealed its role in mediating important chromatin functions during Drosophila oogenesis, such as karyosome structure and double-strand break (DSB) repair (Singh et al., 2018). Htk is presumed to be a chromatin-modeling protein, although its functional characterization is incomplete. Its name derives from its putative role in heterochromatinization and its influence on TDP-43-mediated toxicity (Sreedharan et al., 2015). The htk gene encodes two annotated transcript variants that are translated into two polypeptides of 186 kDa and 259 kDa, respectively. Bioinformatic analysis revealed that Htk can regulate transcription from the RNA polymerase II promoter (Gaudet et al., 2010). The mammalian orthologs of Htk, ARID4A and ARID4B, are members of the chromatin-remodeling complex and function as transcriptional repressors upon recruitment by tumor suppressor RB; thus, they are also known as RB-Binding Protein 1 (RBBP1) and RBBP1-Like Protein 1 (RBBP1L1), respectively (Defeo-Jones et al., 1991; Lai et al., 1999, 2001; Cao et al., 2001; Fleischer et al., 2003).

In the present study, we characterized the functional significance of the Notch–Htk interaction, hypothesizing that Htk is a part of the Notch-Su(H) activation complex. Using molecular and genetic analyses, we found that Htk acts as a positive regulator of Notch signaling by physically interacting with Notch-ICD, a result confirmed by co-immunoprecipitation experiments. htk also showed genetic interactions with components of the Notch signaling pathway. Loss- and gain-of-function studies revealed its pleiotropic role in regulating various developmental events. Several phenotypes generated by loss or gain of htk function mimicked Notch mutant phenotypes, further supporting the role of htk as a modulator of Notch signaling. Overexpression of htk upregulated Notch signaling, as evidenced by the ectopic expression of the Notch target Cut, without a change in the level of Notch-ICD in developing wing imaginal discs. Conversely, the Notch targets Cut and Wingless were downregulated in htk mutant clones, indicating htk to be a positive regulator of Notch signaling. Additionally, the Htk protein colocalized with expressed Notch-ICD in the same nuclear compartment. Furthermore, Htk immunoprecipitated Su(H), a transcription factor involved in Notch signaling. Fragments of the regulatory sequences of E(spl) complex genes, which are Notch targets, were chromatin immunoprecipitated with Htk, confirming Htk as a component of the regulatory complex. Real-time analysis of E(spl) complex gene expression from htk overexpressed tissue confirmed Htk as part of the activation complex. Thus, our functional analyses indicate Htk to be a novel modulator of Notch signaling in Drosophila.

In a yeast two-hybrid screen, we identified Htk as an interacting partner of Notch. In the same screen, multiple positive clones of Su(H), a well-established binding partner of Notch-ICD, were also identified, which validated our approach. The amino terminus of Notch-ICD (amino acids 1765-1895) was used as a bait to screen 6×10⁶ cDNAs from a Drosophila 0-24 h embryonic library. Eleven positive clones (His⁺) were isolated and found to encode overlapping htk cDNAs. Sequence analysis of overlapping domains revealed that the C terminus of Htk (amino acids 2424-2486), which is specific to the protein, was sufficient for binding to Notch-ICD (Fig. S1A).

Co-immunoprecipitation experiments further validated the physical interaction of Notch with Htk. Using extracts from wing discs coexpressing hemagglutinin (HA)-tagged Htk and Notch-ICD, we demonstrated that Htk or Notch could be co-immunoprecipitated with either anti-HA or anti-Notch antibodies (Fig. 1B,C). Moreover, expressed HA-Htk was able to co-immunoprecipitate endogenous Notch (Fig. 1D). Endogenous Htk was also sufficient to immunoprecipitate Notch from larval salivary glands when only Notch-ICD was overexpressed (Fig. 1E).

Immunocytochemical analysis revealed that endogenous Htk (marked by an anti-Htk antibody, generated previously; Singh et al., 2018) and overexpressed Notch-ICD colocalized in the same nuclear compartment (Fig. 1F-Q).

To further analyze the functional implications of the physical interaction between the Htk and Notch proteins, we investigated whether mutations in htk and Notch pathway components displayed genetic interactions in trans-heterozygous combinations. Three htk loss-of-function alleles, htk⁷¹, htk⁴⁹ and htk³⁷, were used for genetic interaction studies. Both Notch and htk are located on I chromosome and their null alleles are hemizygous lethal; therefore, it was not possible to check the genetic interaction between them under trans-heterozygous conditions. A trans-heterozygous combination of a htk allele and a dominant negative allele of Notch (UAS-Notch-DN) enhanced the wing-nicking phenotype, indicating a further decrease in Notch function (Fig. 2A-D). An enhanced wing-nicking phenotype also occurred when different htk alleles combined with the Su(H) gain-of-function allele, Su(H)T4 (Fig. 2E-H, Fig. S1B) and loss-of-function allele, Su(H)1 (Fig. 2I-L, Fig. S1B). Su(H) is a DNA-binding protein component of the Notch signaling pathway; its gain- or loss-of-function should result in reduced Notch signaling activity, as it is a component of both the repression and activation complexes (Furriols and Bray, 2000). Excess Su(H) competes with the normal Su(H) and Notch-ICD-containing activation complex, and free Su(H) alone can bind to DNA, inactivating those downstream Notch target genes that require both Su(H) and Notch-ICD for their activation (Furriols and Bray, 2000). Thus, we hypothesize that, similar to the case of the Su(H) loss-of-function allele, when we reduced the dose of htk along with the heterozygous Su(H) gain-of-function allele, there was a further reduction of Notch signaling, as shown by the increased wing-nicking phenotype in both cases. C96-GAL4-driven expression of dominant negative C-terminal Mam truncation displays a fully penetrant wing-nicking phenotype (Fig. 2M) (Helms et al., 1999; Kankel et al., 2007). Reducing the dose of htk in these individuals elicited an enhanced wing-notching phenotype (Fig. 2M-P). Our genetic interaction screen revealed that, when combined with Notch pathway mutants, different htk loss-of-function mutants further reduced the Notch signaling activity, resulting in more severe phenotypes. We note that htk mutants showed a strong genetic interaction with the transcription factors and co-activators involved in the Notch signaling pathway.

We generated htk mutant somatic clones in larval wing imaginal discs using different htk loss-of-function mutant alleles and the FLP-FRT system (Xu and Rubin, 1993). In htk⁷¹ mutant clones, no significant change in the Notch protein level was observed compared with the surrounding wild-type cells (Fig. 3A-A‴). We also checked the status of Notch signaling activity by investigating the expression level of downstream targets of Notch, Cut and Wingless. Notch induces Cut and wingless expression at the dorsoventral (DV) boundary of the developing wing disc (Neumann and Cohen, 1996). In the current study, protein levels of both Cut and Wingless were significantly reduced in htk⁷¹ mutant clones compared with wild-type sister cells at the DV boundary (Fig. 3B-D‴). In addition, expression of a Notch signaling reporter, Notch responsive element Green Fluorescent Protein (NRE-GFP), was significantly reduced in htk loss-of-function clones (Fig. 3E-E‴). Adults carrying htk clones displayed Notch loss-of-function phenotypes, such as wing notching and increased scutellar bristle phenotypes (Fig. S5G,H). Thus, loss of Htk resulted in reduced Notch signaling activity without affecting the level of the Notch receptor.

We further examined the influence of ectopic expression of htk on Notch expression and its localization. HA-tagged htk was overexpressed in the anterior-posterior (AP) margin of the wing disc using the ptc-GAL4 driver; however, expression of HA-Htk did not affect the level or localization of the endogenous Notch protein (Fig. 4A-C). We then investigated the level of Notch signaling activity by checking the status of Cut, which expresses under the influence of Notch signaling at the DV boundary of the wing disc (Neumann and Cohen, 1996). Overexpression of HA-htk at the AP boundary of the wing discs resulted in ectopic expression of Cut along the AP junction of the pouch region within the wing disc (Fig. 4D-I).

Using tissue-specific GAL4 drivers (Brand and Perrimon, 1993), we also observed the downregulatory and gain-of-function effects of htk in adult tissues, as also seen in imaginal discs (Fig. S1C,D). htk-RNAi was used to investigate the tissue-specific loss-of-function effects of htk. However, this RNAi stock caused weak RNA interference and, as a result, variability in expressivity and incomplete penetrance were observed. Downregulation of htk using apterous-GAL4 in the dorsal region of wing discs, which form the adult thorax and wing, resulted in increased scutellar bristles in comparison with wild type (Fig. S2C,D) and wings that were bent out and upwards, probably because of thorax muscle defects (Fig. S2A,B). These wings also had extra rows of bristles (Fig. S2G). Downregulation of htk using the patched-GAL4 driver in the AP boundary reduced the intervein distance between the L3 and L4 veins and also reduced the first cross-vein (Fig. S2F). dpp-GAL4-driven downregulation of htk also resulted in extra vein material and, occasionally, an extra cross-vein at the AP boundary (Fig. S2I). Furthermore, loss of htk at the wing margin using C96-GAL4 resulted in extra vein material, and areas with thinner cuticle at, or near, the wing margin (Fig. S2J). MS1096-GAL4-driven htk-RNAi also resulted in similar phenotypes (data not shown). engrailed-GAL4-driven htk-RNAi resulted in thinner wing blades in the posterior region of the wings (Fig. S2K). Thus, loss of htk in different regions of wings resulted in extra vein material, extra bristles and small patches of disorganized tissue (Fig. S2H-H″). Loss of htk in the eye using eyeless-GAL4 resulted in loss of ommatidia and reduced eye size (Fig. S2L). Most of these htk-downregulated phenotypes also corresponded to Notch loss-of-function phenotypes. It is well established that Notch is needed for wing margin, vein and sensory bristle development, and its loss also results in reduced intervein distance and eye size (Hartenstein and Posakony, 1990; Go and Artavanis-Tsakonas, 1998; Casso et al., 2010).

Additionally, using various tissue-specific GAL4 drivers, HA-tagged htk was overexpressed in different regions of the wing discs and eye discs. Gain of function of HA-Htk in the dorsal region of wing discs, using an apterous-GAL4 driver, resulted in the loss of scutellar bristles, reduced scutellum size and severe wing blisters (Fig. S3B-D). Overexpression in the posterior region of wing discs using engrailed-GAL4 resulted in bending of the third wing vein, a thinner wing blade and an incomplete fifth vein (Fig. S3H). HA-htk gain of function in the eye using GMR-GAL4 and eyeless-GAL4 drivers resulted in eye roughening (Fig. S3J,L) and loss of ommatidia (Fig. S3F), respectively. A few of these observed phenotypes were similar to Notch gain-of-function phenotypes, such as the loss of bristles and vein material and an increase in eye roughening (Hartenstein and Posakony, 1990; Go and Artavanis-Tsakonas, 1998; Casso et al., 2010).

To determine the epistatic interaction of htk with Notch or Su(H), we examined whether htk could rescue loss- or gain-of-function phenotypes of Notch or Su(H). A severe wing notching phenotype was caused by a reduction in Notch signaling when Notch-RNAi (NIRM) was expressed at the wing margin (Fig. S4A). The wing-nicking phenotype was significantly rescued when HA-tagged htk was overexpressed in this background (Fig. S4B). By contrast, overexpression of Notch-ICD or Su(H)VP16 at the wing margin driven by C96-GAL4 resulted in irregular wing margin bristles (Fig. S4C-E). This phenotype was significantly rescued by decreasing the expression of htk using htk-RNAi in the same background (Fig. S4H-J,V-W). Thus, overexpression of htk can rescue the wing phenotypes caused by Notch loss of function, and a reduction in htk expression leads to rescue of the gain-of-function effects of Notch and Su(H)VP16 in the wing margin (Fig. S4A-J,V-W).

Additionally, when HA-htk was coexpressed with Notch-ICD or Su(H)VP16 at the wing margin using a C96-GAL4 driver, there was an increase in the wing margin irregularities, reflecting the synergistic behavior of htk with Notch and Su(H)VP16 (Fig. S4M-O).

We further confirmed this epistatic interaction by looking at the expression of the Notch downstream target, Cut (Neumann and Cohen, 1996). Loss of Cut expression was observed when Notch-RNAi was expressed at the DV boundary of the wing disc using a C96-GAL4 driver (Fig. S4Q). The reduced activity of Notch signaling, as indicated by the reduction in Cut expression, was partially rescued when HA-htk was expressed in the same background (Fig. S4R,X). By contrast, reducing the expression of htk using htk-RNAi partially rescued the ectopic Cut expression caused by overexpression of Su(H)VP16 at the DV boundary of wing disc using the C96-GAL4 line (Fig. S4T-U,Y).

To further confirm the role of Htk in Notch-Su(H)-mediated downstream target gene expression, we overexpressed Notch-ICD or Su(H)VP16 in htk mutant clonal cells using the MARCM technique (Lee and Luo, 2001). In UAS-Notch-ICD and UAS-Su(H)VP16 lines, Notch-ICD and Su(H)VP16 were under the UAS promoter and tub-GAL4 was used to drive Notch-ICD and Su(H)VP16 globally; however, the presence of GAL80 inhibited GAL4-induced expression of Notch-ICD and UAS-Su(H)VP16 in all cells except those in which GAL80 was eliminated because of FLP-FRT-mediated somatic recombination events. In the same mutant clonal cells, htk gene function was also eliminated. As a result, these Notch-ICD and Su(H)VP16-overexpressing clonal cells, which were also marked with GFP, were for mutated the htk gene. MARCM analysis of wild-type clones without htk mutation was also carried out. The status of the Notch downstream target, Cut, was examined in the clonal cells. Without the htk mutation, Notch-ICD and Su(H)VP16-expressing clones showed ectopic Cut expression (Fig. 5A-D,I-L). Interestingly, in the absence of htk, Notch-ICD and Su(H)VP16 were unable to show the complete activity of Notch, and a reduction in ectopic Cut expression was observed (Fig. 5E-H,M-P).

In addition to MARCM analysis, Su(H)VP16 was overexpressed at the DV boundary using a C96-GAL4 driver; in this background, FLP-FRT-mediated htk null clones were generated. Many of these htk null clones also showed similar results displaying reduction of ectopic Cut expression (Fig. 5Q-T). Thus, together these results confirmed that htk is required for complete Notch signaling activity, and NICD or Su(H) cannot execute all the functions of Notch in the absence of htk.

Immunocytochemical analysis revealed that endogenous Htk and overexpressed Notch-ICD colocalized in the same nuclear compartment (Fig. 1F-I). These two proteins overlapped in the interchromatin and chromatin space within the nucleus (Fig. 1J-M). In the nucleus, Notch forms an activation complex to express downstream target genes (Fortini and Artavanis-Tsakonas, 1994; Struhl and Adachi, 1998; Borggrefe and Oswald, 2016). Thus, we investigated whether Htk is a component of the Notch-ICD activation complex in the nucleus. If it is, Htk should co-immunoprecipitate Su(H), a component of the Notch-ICD activation complex. Co-immunoprecipitation experiments with HA-tagged Htk revealed that HA-Htk was able to pull down endogenous Su(H) when Htk and Notch-ICD were coexpressed in larval wing discs (Fig. 6A). Additionally, HA-Htk was sufficient to immunoprecipitate endogenous Su(H) when only endogenous Notch was present (Fig. 6A). This confirmed that Htk and Su(H) belong to the same activation complex. Given that Htk physically interacted with Notch-ICD (Fig. 1), we concluded that Htk is a component of the Notch-Su(H) transcription complex.

We also explored the mechanism of regulation of Notch signaling by htk. The nuclear localization, DNA-binding ability and physical interaction of Htk with Notch-ICD prompted us to hypothesize that Htk binds to the promoter sequences of Notch targets and cooperates with Notch-ICD for activation of these downstream targets. We performed chromatin immunoprecipitation (ChIP) of expressed HA-tagged Htk using HA beads followed by PCR using primers (Fig. S5J) for promoter sequences of the well-established Notch targets, E(spl) complex genes. ChIP experiments can identify the in vivo association of transcription factors with regulatory elements. We observed that Htk binds to the promoter sequences of Notch target genes (Fig. 6B). To further confirm that Htk is a component of the activation complex and not the repressor complex, we observed the status of two of these E(spl) complex genes, E(spl)m8 and E(spl)mβ, in htk loss-of-function clones. LacZ reporter stocks were used to verify their expression, which was downregulated in htk loss-of-function clones (Fig. S5A-F). Furthermore, real-time analysis showed that the expression of E(spl) complex genes increased with the increase in Htk expression (Fig. 6C), confirming our hypothesis that Htk is an important component of the Notch co-activation complex. Thus, we showed that Htk interacts with the Notch-ICD and is recruited to promoters of Notch target genes to activate their expression (Fig. 6D).

The Notch signaling pathway regulates a variety of cellular processes. Despite the plethora of information about this conserved signaling pathway, the intricate regulatory mechanism of Notch activation is far from complete. Here, we report for the first time that the chromatin-modeling protein Htk is a novel interactor of Notch-ICD. Htk is a Chromo, ARID and Tudor domain-containing protein, and these domains render this protein with putative DNA-binding activity. Htk was identified as an interacting partner of Notch-ICD in a yeast two-hybrid screen; Htk was found to physically interact with Notch-ICD and to interact genetically with mutants of Notch pathway components, especially those of the transcription factor complex. Htk is a nuclear protein that colocalizes with expressed Notch-ICD inside the nucleus. Our loss-of-function and the complementary gain-of-function analyses indicated that Htk is involved in the regulation of Notch signaling. We also showed that Htk is a component of the Notch activation complex. In the absence of Notch signaling, CSL [CBF-1/RBPJ in vertebrates, Su(H) in Drosophila melanogaster, and lag-1 in Caenorhabditis elegans] remains associated with the repressor complex and actively represses the transcription of target genes. Upon activation of Notch signaling, CSL binds to Notch-ICD, displaces the CSL-associated repressor complex and recruits co-activators, leading to the transcriptional activation of Notch target genes (Fortini and Artavanis-Tsakonas, 1994; Struhl and Adachi, 1998). It has been predicted that CSL-mediated transcription of downstream target genes is switched on or off depending on the molecular signature on the chromatin created by associated activator or repressor complexes, respectively (Borggrefe and Oswald, 2016; Giaimoa et al., 2017). Surprisingly the DNA-binding affinity of CSL is extremely low. There is no evidence to show that either co-repressor complex binding or Notch-ICD-activator complex binding to CSL has any impact on its DNA-binding affinity. However, it was hypothesized that other DNA-binding proteins in the CSL activator complex might facilitate its chromatin association (Giaimoa et al., 2017). Our results showed that the DNA-binding protein Htk is present in the Su(H) activator complex and leads to the activation of transcription of target genes, such as those in the E(Spl) complex. Epistasis analysis confirmed that Notch-ICD and Su(H) require Htk to execute the complete function of Notch. We hypothesize that the association of Htk in the Notch-ICD-Su(H) activator complex stabilizes and sustains the binding of Su(H) with DNA and, as a result, transcriptional activation of Notch target genes takes place. Thus, our results establish a novel mode of regulation of Notch signaling by the chromatin-modeling protein Htk.

Recent reports revealed that Sin3A physically binds with Htk (Spain et al., 2010) and also with its human homologs ARID4A and ARID4B (Lai et al., 2001; Fleischer et al., 2003; Suryadinata et al., 2011; Wan et al., 2015). The histone deacetylase Sin3A acts as a negative regulator of transcription (Silverstein and Ekwall, 2005; Kadosh and Struhl, 1998). Sin3A is a binding partner of Su(H), and is a part of a transcription repressor complex involved in suppression of Notch target gene expression in the absence of Notch protein (Zhou et al., 2000). However, here we established that Htk is involved in the transcription activation complex to turn on Notch downstream target gene expression in the presence of Notch protein. The exact role of Htk in regulating the switch from repressor to activation complex is yet to be determined. However, we hypothesize that, similar to Su(H), Htk might also be present in both the repressor and activator complexes and might be required to stabilize and sustain the binding of Su(H) to the promoter sequences of Notch downstream target genes in both complexes.

The data presented here revealed Htk to be a positive regulator of the Notch signaling pathway in Drosophila. Htk has been previously reported to interact with a repressor protein, such as Sin3A (Spain et al., 2010); in addition, the mammalian orthologs of Htk, ARID4A and ARID4B, are members of the chromatin-remodeling complex and function as transcriptional repressors in different contexts (Patsialou et al., 2005). By contrast, we showed that loss of function of Htk decreased Notch signaling, whereas its gain of function upregulated Notch activity, suggesting a transcriptional activator function of Htk in regulating the Notch pathway. This is not the first report showing the positive regulation of Notch signaling by a repressor protein. According to a recent report, a well-established transcriptional repressor protein, the histone deacetylase HDAC1, acts as a positive regulator of Notch signaling (Wang et al., 2018). It has also been shown that HDAC1, along with the HDAC1-associated transcriptional co-repressor Atrophin (Atro), regulates Notch protein levels by promoting Notch transcription (Wang et al., 2018). Similarly, our results also revealed a previously unidentified transcriptional activator function of the predicted repressor complex protein Htk in regulating Notch downstream target gene expression during development.

Similar to CSL in Notch signaling activation, the switch from the repression to activation function of transcription factors was also observed in other signaling pathways, such as T cell factor/lymphoid enhancer factor (TCF/LEF) in the Wnt signaling pathway. Htk might have a similar function in these signaling pathways, although the full spectrum of its function remains to be explored. Given that it is a chromatin-binding protein, Htk could regulate the expression of a range of genes during development. Thus, future studies should explore its function in regulating the activity of different signaling pathways, such as the Wnt, Hippo, Janus kinase (JNK) and Hedgehog (Hh) pathways.

The Notch–Htk interaction might be involved in regulating a spectrum of cellular processes. Recently, it was revealed that loss of htk suppresses the TDP-43-mediated age-dependent neurodegeneration seen in amyotrophic lateral sclerosis (ALS) (Sreedharan et al., 2015). Accumulation of the nuclear RNA-binding protein TDP-43 (encoded by TARDBP) in the cytoplasm is the histopathological signature of degenerating neurons in ALS. Investigations of gene expression patterns that accompany TDP-43-induced neurotoxicity in the Drosophila system showed the strong upregulation of cell cycle regulators and Notch target genes (Zhan et al., 2013). Mutations in Notch pathway components also extended the lifespan of TDP-43 transgenic lines (Zhan et al., 2013). Thus, Notch activation has deleterious effect in TDP-43 flies. Given that the Sreedharan group reported the suppression of TDP-43 toxicity by htk mutations, it is tempting to speculate that Htk might also have an important role in the activation of Notch signaling that leads to the TDP-43-mediated neurodegeneration in ALS. If this is the case, then Htk could be considered as a future therapeutic target for ALS.

A 393-bp Drosophila Notch cDNA (accession number M11664) fragment encoding amino acids 1765-1895 containing NLS was cloned in frame with the sequence encoding the LexA DNA-binding domain of the bait vector. This construct was used as a bait to screen oligo(dT)-primed D. melanogaster 0-24 h embryo cDNA libraries cloned in pGAD prey vectors containing GAL4 activation domains. A yeast two-hybrid screen was carried out using a mating approach (Fromont-Racine et al., 1997) with L40ΔGAL4 and Y187 yeast strains. His+ colonies were selected on media lacking tryptophan, leucine and histidine, and all positive pGAD plasmids from His+ colonies were isolated and sequenced to identify interactors (as described by Mukherjee et al., 2005).

For co-immunoprecipitation experiments, Notch-ICD was overexpressed in larval salivary glands under the control of a sgs-GAL4 driver. Salivary glands were dissected and homogenized in lysis buffer [25 Mm Tris (pH 8.0), 27.5 mM NaCl, 20 mM KCl, 25 mM sucrose, 10 mM EDTA, 10 mM EGTA, 1 mM DTT, 10% glycerol, 0.5% Tergitol solution, 1 mM PMSF and 1× protease inhibitor (Roche)]. Crude lysate containing 3 mg of total protein was mixed with 5 µl anti-Htk antibody and 20 µl protein A/G beads and kept on an end-over-end rotator overnight at 4°C. No antibody was added to the control samples. Beads were collected after washing three times with lysis buffer, separated on a 12% denaturing SDS polyacrylamide gel and transferred onto Immuno-Blot PVDF membranes (Bio-Rad). After washing for 10 min in TBST [Tris base (50 mM), NaCl (150 mM) and Tween-20 (0.1%)] followed by blocking (4% skimmed milk in TBST) for 30 min, blots were probed with a mouse anti-Notch antibody [C17.9C6, 1:3000 dilution, Developmental Studies Hybridoma Bank (DSHB)]. After washing again for three times in TBST and blocking (4% skimmed milk in TBST) for 30 min, a goat anti-mouse IgG-AP conjugate (Molecular Probes) diluted at 1:2000 in blocking solution was added for 90 min followed by three washings in TBST. Color was detected by Sigma FAST BCIP/NBT (Sigma).

Similarly, lysates were prepared from larval wing discs expressing Notch-ICD and HA-Htk, and only HA-Htk using a vg-GAL4 driver, followed by immunoprecipitation with anti-HA affinity beads (Sigma), and anti-Notch antibody along with A/G beads. Monoclonal mouse anti-Notch (C17.9C6) antibodies at a 1:3000 dilution were used to detect Notch; mouse anti-HA antibodies at a 1:1000 dilution (Sigma) were used to detect HA; and rabbit anti-Su(H) antibodies at a 1:1000 dilution (Santa Cruz Biotechnology) was used to detect Su(H). Goat anti-mouse IgG-AP conjugates or anti-rabbit IgG-AP conjugates at a 1:2000 dilution (Molecular Probes) were used as secondary antibodies.

All fly stocks were maintained on standard cornmeal/yeast/molasses/agar medium at 25°C. Oregon-R flies were used as wild-type controls. The htk-null mutant alleles, htk⁷¹ FRT19A/FM7, htk³⁹ FRT19A/FM7 and htk⁴⁷ FRT19A/FM7 were provided by Dr Jemeen Sreedharan, Babraham Institute, Cambridge, UK (Sreedharan et al., 2015). Notch pathway components were provided by Prof. S. Artavanis-Tsakonas, Department of Cell Biology, Harvard Medical School, USA. We used the following stocks for our studies:

For genetic interactions: UAS-dominant-negative Notch (UAS-DNN7)/CyO; C96-GAL4/TM6B, Su(H)T4/CyO (gain-of-function allele), Su(H)1/CyO (loss-of-function allele) (BL417) and C96-GAL4, UAS-MamH/TM6B were used.

For epistatic studies: UAS-Notch-RNAi/CyO (UAS-NIRM), UAS-Su(H)VP16 (II) (provided by Prof. Sarah Bray, University of Cambridge, UK) and UAS-Su(H)VP16/TM3 ser (provided by Prof. Jessica Treisman, School of Medicine, NYU, USA) were used.

For loss- and gain-of-function studies: UAS-htk-RNAi (BL31574), UAS-Notch-RNAi/CyO, NRE-eGFP (BL30728) and UAS-HA-htk/TM3, Ser (Fly Line ID: F000657) were used. UAS-HA-htk stock (which expresses a functional but truncated form of the Htk protein) was ordered from FlyORF.

GAL4 driver lines: en-GAL4, C96-GAL4, Sgs3-GAL4, ap-GAL4/CyO, MS1096-GAL4, vg-GAL4, ptc-GAL4, ey-GAL4 and GMR-GAL4 were ordered from the Bloomington Drosophila Stock Centre.

For mosaic generation: hs-FLP neoFRT19A Ubi-RFP (BL31418) and neoFRT19A (BL1709) were used.

To generate somatic clones, we used the FLP/FRT system. Males of the hs-FLP neoFRT19A Ubi-RFP strain were crossed with htk⁷¹ FRT19A/FM7, htk⁴⁷ FRT19A/FM7 and htk³⁹ FRT19A/FM7 females. Heat shock was given at 37°C for 60 min at 24 h after egg laying (AEL), and third-instar female larvae were analyzed for mutant clones. Given that we observed similar phenotypes in all the htk alleles, we only show results for the htk⁷¹ mutation.

To generate MARCM-derived clones and htk⁷¹ mutant clones in an Su(H)VP16 overexpression background, the following flies were generated by appropriate genetic crosses:

hs-FLP neoFRT19A Ubi-RFP/Y; +/+; C96-GAL4

hs-FLP neoFRT19A tub-GAL80/Y; +/+; tub-GAL4 UAS-GFP/TM6c sb

htk⁷¹ FRT19A/FM7; UAS-Su(H)VP16

htk⁷¹ FRT19A/FM7; UAS-NICD

neoFRT19A/FM7; UAS-Su(H)VP16

neoFRT19A/FM7; UAS-NICD

The MARCM system was used to generate GFP-marked htk⁷¹ mutant clones overexpressing either Notch-ICD or Su(H)VP16. htk⁷¹ FRT19A/FM7; UAS-Su(H)VP16 and htk⁷¹ FRT19A/FM7; UAS-NICD females were crossed to males hs-FLP neoFRT19A tub-GAL80/Y; +/+; tub-GAL4 UAS-GFP/TM6c sb. In parallel, a control experiment was carried out in which neoFRT19A/FM7; UAS-Su(H)VP16 and neoFRT19A/FM7; UAS-NICD females were used for crosses. Heat shock was given at 37°C for 60 min at 24 h AEL and third-instar female larvae were analyzed for GFP-marked clones.

To generate htk⁷¹ mutant clones in a Su(H)VP16 overexpression background, hs-FLP neoFRT19A Ubi-RFP/Y; +/+; C96-GAL4 males were crossed with htk⁷¹ FRT19A/FM7; UAS-Su(H)VP16 females (experimental) and neoFRT19A/FM7; UAS-Su(H)VP16 females (controls). The remainder of the procedure was as previously detailed.

Eye imprints using nail polish were made for analyzing ommatidial defects, and were examined under differential interference contrast (DIC) optics on a Nikon Eclipse Ni microscope.

Total RNA was isolated from imaginal discs of third instar larvae using TRIZOL reagent following the manufacturer's recommended protocol (Sigma-Aldrich). RNA (1 μg) was treated with DNAse using 1.0 µl 10× reaction buffer, 0.5 µl DNAse (New England Biolabs) and 0.5 μl RNAse inhibitor (1,000 U/ml), and volume was adjusted to 10 µl using DEPC MilliQ water. The reaction mixture was incubated at 37°C for 1 h, then 1 µl of 25 mM EDTA was added followed by a 10 min incubation at 65°C. For single-stranded cDNA preparation, 10 µl of DNAse-treated RNA was mixed with 1 µl M-MuLV reverse transcriptase enzyme (New England Biolabs), 2 µl of 60 μM random primers, 2 µl of 10× M-MuLV buffer, 1 µl of 10mM dNTP and 1 µl of RNase inhibitor, and nuclease-free water was used to make the total volume to 20 µl. Mix was incubated at 25°C for 5 min followed by a 42°C incubation for 1 h. This cDNA was used as a template DNA for semiquantitative and RT quantitative (RT-q)PCR. RT-qPCR was carried out as per the manufacturer's protocol (Applied Biosystems). A total of 10 μl of the reaction volume included 5 μl 2× SYBR green, 0.25 μl of each forward and reverse primer, and 1 μl cDNA; PCR was performed using an ABI 7500 instrument. Data were normalized to rps17 before calculating the relative fold change. Primers for RT-qPCR can be found in Table S1.

For immunostaining, imaginal discs from third-instar larvae were dissected in cold PBS, followed by 20-min incubation in 4% paraformaldehyde. Tissues were then washed four times in washing solution [a mixture of 1×PBS, 0.2% Triton-X-100 and 0.1% bovine serum albumin (BSA)] for 10 min each, followed by incubation in blocking solution (Tri-PBS with 0.1% BSA and 8% normal goat serum) for 60 min. Primary antibodies were diluted in blocking solution, added to the ovaries and incubated overnight at 4°C. After four washes in washing solution and 60-min blocking, secondary antibodies were added at a 1:200 dilution and incubated for 90 min at room temperature. This was followed by four washes in washing solution for 15 min each. After a PBS wash, 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI) (1 µg/ml) was added for 20 min to the tissues. After a final dissection in cold PBS, samples were mounted in 1,4-diazabicyclo[2.2.2]octane (DABCO). The following primary antibodies were used: rabbit polyclonal anti-Htk (1:100) (Singh et al., 2018), anti-Notch raised in mouse (1:300) (C17.9C6), mouse anti-Wg (1:100) (4D4), mouse anti-Cut (1:100) (2B10; DSHB), mouse anti-HA (1:100) (Sigma) and mouse anti-βGAL (1.5:100) (Promega). The secondary antibodies used were: Alexa Fluor 555 conjugated goat anti-mouse IgG (1:200), Alexa Fluor 488 conjugated goat anti-mouse IgG (1:200), Alexa Fluor 555 conjugated goat anti-rabbit IgG (1:200), Alexa Fluor 555 conjugated goat anti-rat IgG (1:200) and Alexa Fluor 488 conjugated goat anti-rabbit IgG (1:200) (all from Molecular Probes). Fluorescent images were obtained with a Carl Zeiss LSM780 confocal microscope.

In total, 100 heads were dissected from GMR-GAL4, UAS-HA-htk flies in 1× PBS and fixed in 1% formaldehyde solution at room temperature for 20 min. Incubation for 1 min in 0.125 M glycine solution was performed to stop crosslinking followed by three washes in 1× cold PBS. Then, 400 µl nuclear lysis buffer[1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8) and 1× protease inhibitor] was added and heads were homogenized followed by sonication (QSonica, Q700) (50 cycles, amplitude of 90, pulse on 30 s, pulse off 1 min). After centrifugation at 13,000 rpm (18,000 g) for 10 min at 4°C, supernatants (chromatin samples) were collected and stored at −20°C. DNA was purified from 50 µl of this chromatin sample using the phenol/chloroform method, for which one volume of phenol:chloroform:isoamyl alcohol (25:24:1) was added to the sample and mixed thoroughly, followed by centrifugation at 15,000 g at 4°C for 10 min. The upper aqueous phase was removed carefully and transferred to a fresh tube. Sample was mixed with one volume of chloroform:isoamyl alcohol (24:1) by centrifugation at 15,000 g at 4°C for 10 min and the upper aqueous phase collected again. This was followed by DNA precipitation using 3 mM sodium acetate and absolute ethanol, and the fragment size was checked using agarose gel electrophoresis (image not shown). For immunoprecipitation, 200 µl of the chromatin sample was mixed with 25 µl pre-washed HA beads (Sigma) and ChIP dilution buffer [0.01% SDS, 1.1 % Triton X-100, 1.2 mM EDTA, 16.7 mM TrisHCl (pH 8), 167 mM NaCl]. was added to make up the final volume to 300 µl. Mock-IP was used as negative control in which, A/G beads (Santa Cruz Biotechnology) were added instead of HA beads in the chromatin sample. The mixed sample was kept for end-over-end rotation at 4°C overnight. Beads were then washed with a low-salt wash buffer [0.1% SDS, 1% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8), 150 mM NaCl], a high-salt wash buffer [0.1% SDS, 1% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8), 0.5 M NaCl], LiCl wash buffer [25 mM LiCl, 1% NP-40, 1% NaDOC, 1 mM EDTA, 10 mM Tris-Cl (pH 8)] and TE buffer [10 mM Tris-HCl (pH 8), 1 mM EDTA]. Beads were resuspended in 150 µl ChIP elution buffer (50 mM NaHCO₃, 1% SDS) and vortexed gently for 15 min at room temperature, followed by centrifugation at 2000 rpm (425 g) for 4°C; the supernatants were then collected. This step was repeated twice and the samples were pooled; 1 µl of RNase with 18 µl 5 M NaCl was then added and incubated at 67°C for 4-5 h. 25 µl 5× PK buffer [50 mM Tris-HCl (pH 8), 25 mM EDTA.NA₂, 1.25% SDS] and 1.5 µl proteinase K (10 mg/ml) was added and incubated at 45°C for 2 h. Thereafter, the DNA was purified using phenol/chloroform extraction and ethanol precipitation followed by resuspension in 30 μl TE buffer. PCR was performed using 2 µl eluted DNA as a template and the primers are described in Table S2.

The authors thank Prof. Spyros Artavanis-Tsakonas (Department of Cell Biology, Harvard Medical School, USA) for Notch pathway mutant Drosophila stocks. Prof. Jessica Treisman (School of Medicine, NYU, USA) and Prof. Sarah Bray (University of Cambridge, UK) are acknowledged for Su(H)VP16 stocks and Dr Jemeen Sreedharan (Babraham Institute, Cambridge, UK) is acknowledged for providing htk mutant alleles. The Bloomington Drosophila Stock Center is acknowledged for providing fly stocks. We also acknowledge the confocal facility of the Interdisciplinary School of Life Sciences, Banaras Hindu University.