Insensitive is a corepressor for Suppressor of Hairless and regulates Notch signalling during neural development

Abstract

The Notch intracellular domain functions as a co-activator for the DNA-binding protein Suppressor of Hairless (Su(H)) to mediate myriad cell fate decisions. Notch pathway activity is balanced by transcriptional repression, mediated by Su(H) in concert with its Drosophila corepressor Hairless. We demonstrate that the Drosophila neural BEN-solo protein Insensitive (Insv) is a nuclear factor that inhibits Notch signalling during multiple peripheral nervous system cell fate decisions. Endogenous Insv was particularly critical when repressor activity of Su(H) was compromised. Reciprocally, ectopic Insv generated several Notch loss-of-function phenotypes, repressed most Notch targets in the E(spl)-C, and opposed Notch-mediated activation of an E(spl)m3-luc reporter. A direct role for Insv in transcriptional repression was indicated by binding of Insv to Su(H), and by strong chromatin immunoprecipitation of endogenous Insv to most E(spl)-C loci. Strikingly, ectopic Insv fully rescued sensory organ precursors in Hairless null clones, indicating that Insv can antagonize Notch independently of Hairless. These data shed first light on the in vivo function for a BEN-solo protein as an Su(H) corepressor in the Notch pathway regulating neural development.

Figure 1

Insensitive (Insv) is nuclear protein expressed in the Drosophila peripheral nervous system. (A) Model of the external mechanosensory organ lineage. A proneural cluster (PNC, blue) is differentiated from other epidermal cells (grey) by spatially patterned activity of bHLH activator proteins Achaete (Ac) and Scute (Sc). A presumptive sensory organ precursor (SOP) or pI cell activates Notch signalling in non-SOP cells of the PNC, which then adopt an ordinary epidermal fate. The SOP undergoes a series of asymmetric divisions in which the sister cells are net senders (in red) or receivers (in blue) of Notch signalling. A glial cell is apoptotic leaving four cells in the mature sensory organ. Timepoints for dissection are indicated in hours after puparium formation (APF), and cell-specific markers used in this study are indicated. (B) Late third instar wing disc stained for Insv, revealing expression in all SOPs. (C) insv[23B] disc stained for Insv (in green, C) and Senseless (Sens, in red, C′), lack of Insv demonstrates antibody specificity. (D) 14 h APF notum stained for Insv (in green, D) and Sens (in red, D′); merge shows nuclear colocalization of these proteins (D″). (E) Sensory development at the 2-cell stage reveals colocalization of Insv and Cut in both cells. (F) Sensory development at the 4-cell stage reveals that Insv is maintained at a higher level in one cell (asterisks) in each cluster (dotted ovals). (G) Insv is extinguished before full expression of terminal cell markers, but we could observe colocalization of Insv with nascent Elav, but never with Prospero (Pros). This identified the last sensory cell to express Insv as the neuron.

Figure 2

Mutual suppression of insv and N[ts1] phenotypes indicates that endogenous Insv opposes Notch function. (A) Normal pattern of microchaetes in wild-type. Dashed line indicates the midline; there are five rows of microchaetes from the midline to the dorsocentral macrochaete row. (B) insv[23B] female seemingly shows normal notum development, although ‘holes’ in the microchaete field (arrows) hint at a lower sensory organ density. (C) A normal notum pattern is seen in insv[23B] female carrying an insv genomic rescue transgene. (D) Quantification reveals decreased numbers of microchaetes in insv females, and its rescue by either of two insv rescue transgene insertions. Student's two-tailed t-tests were performed; n=59 wt, 65 insv, 18 rescue#1, 13 rescue#2 animals. (E) N[ts1] females cultured at 22°C exhibit an increase in the density of microchaetes, with ∼6–7 disorganized rows of microchaetes from the midline to the dorsocentral macrochaete row. (F) N[ts1]; insv[23B] females cultured at 22°C show a relatively normal pattern of microchaetes. (G) Abdomen of insv[23B] female cultured at 25°C contains a number of double-socketed organs (circles). (H) Scanning electron micrograph illustrating the conversion of the shaft (sh) cell into a socket (so) cell in insv double-socketed organs. (I) N[ts1]; insv[23B] female abdomen cultured at 25°C shows a strong reduction in double sockets. (J) Quantitation of double socketing in insv (n=50), insv; P[insv] (n=40), and N[ts1]; insv (n=48) mutants. (K) insv loss-of-function suppressed the pupal lethality of N[ts1] animals raised at 25°C. Fifteen females and males were used in each condition.

Figure 3

Insv exhibits genetic interactions with Su(H) and numb that reflects its antagonism of Notch-directed cell fate decisions. (A) Misexpression of Su(H) using Eq-gal4 causes sensory organ loss due to failure to specify SOPs (see also Supplementary Figure S1). (B) Misexpression of Su(H) in an insv mutant background causes enhanced SOP loss. (C–E) 14 h APF left heminota stained for the SOP marker Sens. (C) Wild-type pattern of SOP cells. (D) Eq>myc-Su(H) shows reduced SOPs. (E) Eq>myc-Su(H) in an insv[23B] mutant background yields enhanced SOP loss. (F) Cellular interpretation of bristle loss phenotype. (G) Activation of hs-numb during division of microchaete SOP cells (14 h APF) results in extensive bristle loss due to pIIA → pIIB transformation. (H) Ectopic Numb is unable to drive cell fate transformations effectively in an insv mutant. (I–K) Notum sensory lineages at 24 h APF stained for the sheath cell marker Prospero and the neural marker Elav. (I) Wild-type sensory organs contain single Pros+ and Elav+ cells. (J) hs-numb pulsed at 37°C for 1 h at 14 h APF yields organs with pairs of Pros+ and Elav+ cells, due to pIIA → pIIB conversion. (K) Similar elevation of numb in an insv mutant fails to yield efficient cell fate conversions. (L) Cellular interpretation of bristle loss phenotype induced by ectopic Numb, and its rescue by absence of Insv.

Figure 4

Insv exhibits potent genetic interactions with Hairless, an Su(H) corepressor. (A) H[E31]/+ heterozygote exhibits dominant double socketing of a few macrochaetes (arrow); (A′) scanning electron micrograph (SEM) illustrates a double socket in this genotype more clearly. (B) insv; H[E31]/+ shows double socketing of nearly all external sensory organs; (B′) SEM illustrates a double socket in this genotype more clearly. See also Supplementary Figure S2 for additional allelic combinations that yielded identical results. (C) The complete double socketing in insv; H[E31]/+ is fully rescued back to the H[E31]/+ phenotype upon introduction of one copy of an insv genomic transgene. (D, E) Triple label of sensory organs at 32 h APF to mark socket cells [Su(H)+ nuclei], shaft cells (large D-Pax2+ nuclei), sheath cells (small D-Pax2+ nuclei), and neurons (Elav+ nuclei). (D) Most sensory organs contain the normal complement of differentiated cells in nota doubly heterozygous for insv and H, including single socket (arrow), shaft (large arrowhead) and sheath (small arrowhead) cells. (E) insv; H[E31]/+ notum exhibits complete double socketing at the expense of shaft specification. (F, G) Cellular interpretation of double-socketing phenotype.

Figure 5

Ectopic Insv inhibits Notch pathway activity. (A) Wild-type adult notum showing normal pattern of mechanosensory bristles; anterior and posterior dorsocentral positions (aDC and pDC) are labelled, as is the microchaete field (μC). (B) sca-Gal4>UAS-insv notum exhibits macrochaete tufting, especially at aDC and pDC. (C) Eq-Gal4>UAS-insv notum exhibits a strong increase in microchaete density. (D) In 14 h APF wild-type notum, Insensitive (Insv) colocalizes with Senseless (Sens) in all pI cells. The regular columns of microchaete pI cells are easily discerned; the midline is indicated with a dashed line. (E) Following misexpression of insv using Eq-gal4, the regular arrangement of pI columns is disturbed, and there are many intervening pI cells as marked by Sens. (F) Wild-type adult wing. (G) vg-Gal4>UAS-insv wing exhibits extensive notching. (H) ptc-Gal4>UAS-insv disc, in wing margin cells that overlap the domain of ectopic Insv (H) there is a gap in Sens (H′) and Cut (H″).

Figure 6

Insv inhibits the expression of Notch target genes. (A) qrt-PCR was used to demonstrate downregulation of N/Su(H) target genes in the E(spl)-C, in embryos overexpressing Insv under hs-Gal4 control. Both bHLH-R and Bearded family genes were repressed by Insv; in contrast, the Kazal protease encoded by E(spl)m1 is not regulated by Notch signalling and was not affected by ectopic insv. Values are the mean from two independent RNA extractions, each quantified in quadruplicate. (B) Cell culture assays demonstrate that Insv antagonizes transcriptional activation of E(spl)m3-luc by NICD; total amounts of transfected plasmids were equalized with empty expression vectors. Value is the mean from three independent experiments, each of which included four independent transfections. Student's two-tailed t-tests were performed.

Figure 7

Insv is a direct partner of Su(H). (A) GST-pulldown assays showed that GST–Insv could specifically pulldown Su(H), relative to a large excess of GST alone. (B) Domain structure of Insv, with predicted coiled-coil domain in yellow and BEN domain in blue. GST-pulldown assays demonstrate that the Insv-BEN domain alone can mediate protein–protein interaction with Su(H). (C) ChIP assays. Chromatin from 2.5 to 6.5 h embryos was immunoprecipitated with antibodies to endogenous Insv, and the indicated target amplicons were checked by qPCR. Insv was bound to the Su(H)-regulated promoters of 10 genes in the E(spl)-C, including both Bearded family and bHLH-R genes. Insv did not associate with multiple control promoters, and the strong 20–30-fold ChIP enrichments to E(spl)m4 and E(spl)m8 were abolished in chromatin prepared from insv homozygous mutant embryos.

Figure 8

Insensitive can rescue sensory organ development in Hairless null clones. (A) SEM of adult notum bearing H null MARCM clones (hs-flp, tub-Gal4, UAS-GFP; FRT82B H[E31]/FRT82B tub-Gal80); × 205. (B) Close-up of H clonal territory (boxed region in A) showing lack of external sensory organ structures; × 1150. (C) SEM of adult notum bearing H null MARCM clones expressing Insv (hs-flp, tub-Gal4, UAS-GFP; UAS-insv; FRT82B H[E31]/FRT82B tub-Gal80); × 205. (D) Closeup of H clonal territory expressing Insv (boxed region in C) showing lack of external sensory organ structures; × 1150. Circles highlight multisocket organs bearing 2–4 sockets; inset shows a multisocket organ with a shaft. (E–G) Fourteen hours APF nota bearing MARCM clones marked by GFP (green) and stained for Hindsight to mark SOPs (red) and DAPI as a counterstain; all of the (′) panels show the clone border as a dashed line. (E, E′) Control clone shows normal patterning of SOPs; some have just divided and are visible as stained pairs of cells. (F, F′) H MARCM clone exhibits complete lack of SOP specification in the mutant territory. Reciprocally, there is a non-autonomous effect in that wild-type cells bordering the clones are more likely to adopt the SOP fate (asterisks). (G) H MARCM clone expressing Insv exhibits rescue of SOP specification and depletion of wild-type SOPs neighbouring the clone; in fact the density of SOPs is greater than in wt. (H–J) Thirty hours APF nota bearing MARCM clones marked by GFP (green) and stained for Cut to mark all sensory organ cells (red), Elav to mark neurons (white), and DAPI as a counterstain (blue). (H) Control clones show that each cluster of four sensory cells includes one neuron. (I) H null clones fail to differentiate any sensory lineage cells, consistent with the failure to specify SOPs; the non-autonomous effect persists (asterisks). (J) H MARCM clone expressing Insv exhibits sensory lineages with multiple large cells determined to be Su(H)-expressing sockets (see D and Supplementary Figure S6); however, about half of the organs could differentiate neurons (arrows).