Separable transcriptional regulatory domains within Otd control photoreceptor terminal differentiation events

Abstract

Orthodenticle (Otd)-related transcription factors are essential for anterior patterning and brain morphogenesis from Cnidaria to Mammals, and genetically underlie several human retinal pathologies. Despite their key developmental functions, relatively little is known regarding the molecular basis of how these factors regulate downstream effectors in a cell- or tissue-specific manner. Many invertebrate and vertebrate species encode two to three Otd proteins, whereas Drosophila encodes a single Otd protein. In the fly retina, Otd controls rhabdomere morphogenesis of all photoreceptors and regulates distinct Rhodopsin-encoding genes in a photoreceptor subtype-specific manner. Here, we performed a structure-function analysis of Otd during Drosophila eye development using in vivo rescue experiments and in vitro transcriptional regulatory assays. Our findings indicate that Otd requires at least three distinct transcriptional regulatory domains to control photoreceptor-specific rhodopsin gene expression and photoreceptor morphogenesis. Our results also uncover a previously unknown role for Otd in preventing co-expression of sensory receptors in blue vs. green-sensitive R8 photoreceptors. Sequence analysis indicates that many of the transcriptional regulatory domains identified here are conserved in multiple Diptera Otd-related proteins. Thus, these studies provide a basis for identifying shared molecular pathways involved in a wide range of developmental processes.

Figure 1

Otd protein sequence and alignment with other insect orthologs. A) D. melanogaster Otd sequence used for these studies. The DNA-binding homeodomain (HD, bolded), repeated SP residues (italicized), and the GNS, ANS, and GV sequences denoted in B (underlined) are indicated. The DLCYP sequence before the HD (underlined) is not conserved in other Otd orthologs, but is present in a subset of Otd ESTs deposited into Flybase (data not shown). The LDY motif at the C terminus that matches residues within the vertebrate Otx-tail (Freund et al., 1997b) is underlined. B) Aligned areas of conservation from various insect Otd orthologs. Patterned boxes correspond to % identity to D. melanogaster Otd. D, Drosophila; mel, melanogaster; sim, simulans; sec, sechelia; yak, yakuba; ere, erecta; wil, willistoni; vir, virilis; ana, ananassae; pse, pseudoobscura; gri, grimshawi; moj, mojanvensis.

Figure 2

Otd deletions differentially regulate rhodopsin promoter activity in vitro. A) Diagram of Otd deletion constructs used for rhodopsin promoter reporter assays. OtdFL = aa 1-550, OtdΔN = aa 69-500, OtdΔC” = aa 1-529 (removes the LDY tail), OtdΔC = aa 1-430, OtdΔNC = aa 69-430, OtdΔA = 1-137 + 216-550, Otd ΔB = 1-215 + 430-550, OtdΔAB, 1-137 + 430-550. Luciferase reporter assays using minimal promoters for Rh3 (A), Rh5 (B), and Rh6 (C) with each Otd deletion constructs. Average values from multiple replicas for each promoter are plotted and tested for statistical difference from OtdFL values. * = p<0.05; ** = p<0.01 E) Rh6 promoter activity using Otd deletions alone (left axis) or Otd deletions with full-length Sens (right axis). Note that Sens increases promoter activity with OtdFL and Otd^AB by 10-fold, while with Otd^N, over 75-fold increased activation is observed with Sens compared to Otd^N alone. No activation is observed with or without Sens with Otd ^C. F) Diagram of Otd fragments fused to the GAL4-DNA Binding Domain (DBD). N= aa 1-70, A = aa 129-215, B = aa 216-430, AB = 129-430, ABC = 129-550, BC = aa 215-550, C = aa 430-550, C’ = aa 490-550, C” = aa 529-550. G) Graph representing average activity from multiple replicas of different Otd-GAL4 DBD fusions from on a UAS-Luciferase reporter, normalized to the Gal4-DBD alone. * = p<0.05; ** = p<0.01. Inset represents the A, B, AB, and C” values from the larger graph. Several fragments (N, ABC, BC, C. and C’) significantly activate the promoter, whereas only the AB fragment significantly represses promoter activity.

Figure 3

Developing an Otd rescue paradigm. A,B) otd^1.6 Gal4 drives reporter expression that mimics endogenous Otd expression in wild type or otd^uvi photoreceptors. Late 3^rd instar imaginal discs whole mounts (A-A”) and cryosections from adult yw; otd^1.6-GAL4; UAS-nuclear LacZ heads (B-B”), and adult otd^uvi; otd^1.6-GAL4; UAS-nuclear LacZ heads (C-C”) were co-immunostained for a pan-neuronal nuclear marker, Elav (white, A,B,C), Otd (green, A’, B’, C’), and β-galactosidase (purple, A”, B”, C”). The otd^1.6 Gal4 transgene drives a lacZ reporter expression in photoreceptors after neuronal specification at the same time of onset as endogenous Otd expression (A), and is maintained in adult wild type (B) and otd^uvi mutant (C) photoreceptors. To analyze Otd rescues, 8 μm cryosections of adult retinas from control (D-G), otd^uvi ; pWIZ (H-K), otd^uvi; otd^1.6-GAL4/UAS-OtdFL; pWIZ rescues (L-O), yw⁶⁷/otd^uvi; otd^1.6-GAL4, pWIZ; TM2/TM6B (P), yw⁶⁷/otd^uvi; otd^1.6-GAL4, pWIZ/UAS-OtdFL; UAS-OtdFL/TM2 (Q), yw⁶⁷; otd^1.6-GAL4, pWIZ; UAS-Melt/TM2 (R), otd^uvi; otd^1.6-GAL4, pWIZ/CyO; UAS-melted/TM2 (S), and otd^uvi; otd^1.6-GAL4, pWIZ/UAS-OtdFL; UAS-melted/TM2 (T) were imaged with DIC (D,H,L), stained with FITC-phalloidin to visualize actin-rich rhabdomeres (E, I,M), or immunostained with Rh3 (purple, F,J,N; blue, P,Q), Rh4 (green, F,J,N), Rh5 (purple, G,K,O,R-T; red, P,Q), or Rh6 (green, G,K,O-T). The distal and proximal layers of the retina where the R7 and R8 layer, respectively, normally form is indicated with a dotted line (F,G,J,K,N,O,R-T). Wild-type and Otd rescues exhibit elongated rhabdomeres (bracket, D,L) and a normal rhabdomere topology with six large OPR rhabdomeres arranged in a trapezoid, and a smaller central R7 rhabdomere (E,M), while otd^uvi rhabdomeres fail to elongate beyond ~1/3 of the retinal depth (bracket, H), and are often misshapen or duplicated (I). Rh3 and Rh5 are absent in otd^uvi mutants (J,K), but are restored in the rescue (N,O). (P, Q) A low Rh5:Rh6 ratio is observed in otd^uvi heterozygous eyes, with Rh3-expressing R7s miscoupled with Rh6-expressing R8s (arrows, P). A normal Rh5:Rh6 ratio is restored by misexpressing Otd and appropriate Rh3/Rh5 coupling is observed (yellow arrowheads, Q). (R-T) UAS-melted overexpression in control (R) and OtdFL rescues (T), but not otd^uvi mutants (S) increases Rh5 expression and decreases Rh6 expression in R8 cells.

Figure 4

Mapping domains within Otd involved in Rh3 and Rh5 activation. Whole-mounted retinas or agarose-embedded cryosections from freshly eclosed flies were co-immunostained with A) R7 opsins Rh3 (purple, and black-and-white channel below) and Rh4 (green) or B) R8 opsins Rh5 (purple, and black-and-white channel below) and Rh6 (green). Flies analyzed had the following genotype: otd^uvi; otd^1.6-GAL4, pWIZ; UAS-Otd transgene. The UAS-deletions used are listed above. otd^uvi controls carried the otd^1.6-GAL4 driver that lacked a UAS-Otd transgene. Empty inner photoreceptors, identified by co-staining of Rh3-Rh6 (shown) with phalloidin (not shown), are circled for constructs that partially restore expression, whereas constructs in which no Rh3 or Rh5 activation was observed were left uncircled. Arrows indicate ommatidia in which two opsins are co-expressed in the same cell. All constructs except OtdDelta;NC, OtdDelta;AB, and OtdHD are able to rescue at least some Rh3 expression, whereas no Rh5 expression is detected in flies rescued with OtdDelta;N, OtdDelta;C, OtdDelta;NC, OtdDelta;ABC, and OtdHD. Co-expression of Rh5 and Rh6 are observed with OtdDelta;A, OtdDelta;B, OtdDelta;AB.

Figure 5

A) Rh6 repression in outer photoreceptors requires Otd’s C-terminus. Cryosections of agarose-embedded heads from otd^uvi flies or those rescued with various Otd deletions (UAS-transgenes listed above; genotypes as in Fig 4). Sections were stained with FITC-phalloidin (purple, top; white, center) to recognize the rhabdomeres, and Rh6 (green, top; white, bottom). Rh6 is restricted to the rhabdomeres of a single central photoreceptor (the R8) when rescued with OtdFL and OtdDelta;A (shown), as well as OtdDelta;N, OtdDelta;B, and OtdDelta;AB (not shown). Constructs lacking the C-terminus (OtdDelta;C, OtdDelta;NC, OtdDelta;ABC, and OtdHD), however, show Rh6 expression outside of the actin-rich rhabdomeres in multiple outer photoreceptors. B) Optical sections from the top third (R7) or bottom third (R8) of retinas from otd^uvi flies or those rescued with various Otd deletions (UAS-transgenes listed above; genotypes as in Fig 4). Sections were analyzed for duplicated rhabdomeres (arrows), missing rhabdomeres (arrowheads), and the ability to elongate fully formed rhabdomeres to the R8 layer. Note that few rhabdomeres extend to the R8 layer in otd^uvi mutants, and individual rhabdomeres are difficult to discern in the R7 layer. OtdDelta;N rescues behave similarly to OtdFL, with defined trapezoids at the R7 and R8 layer and little to no duplicated rhabdomeres. Defined trapezoids are frequently found in the R7 layer with all other constructs, including the HD, but duplicated rhabdomeres are frequent (arrows), and some ommatidia without a full complement of 7 photoreceptors are observed (yellow arrowheads indicate areas which lack a fully-developed rhabdomere). Constructs lacking the C-terminus also often fail to extend rhabdomeres into the R8 layer (circles highlight ommatidia lacking a full complement of photoreceptors).

Figure 6

Summary of in vivo mapping results. A) Otd deletion constructs used for otd^uvi rescues. “-“ represents a failure to rescue the otd^uvi mutant phenotype, “++” indicates that the rescued phenotype is similar to OtdFL rescues, “+” represents phenotypes that provide intermediate rescue, and “+/−“ indicates poor, but partial rescue. Rh6 repression refers to the ability (++) or inability (−) to repress Rh6 in OPRs. * indicates that Rh5 expression is only observed in the presence of Melt (Supp Fig 3), and ** indicates that Rh5 and Rh6 are co-expressed in a subpopulation of R8 cells. B) Model of Otd’s functional domains. Both the N- and the C-terminus both participate in Rh3 and Rh5 expression, but the N-terminus is sufficient to activate Rh3, and the C-terminus is essential for Rh5 expression (Fig 4). Removing the N-terminus also leads to expanded Rh6 expression in pR8s (Fig 4B) and suppresses Otd/Sens-mediated synergism on the Rh6 promoter (Fig 2E). The homeodomain not only binds target sequences to regulate opsin expression (Tahayato et al, 2003), but also participates in maintaining some aspects of rhabdomere morphology. The AB region, when deleted, causes high numbers of R8 cells to co-express Rh5 and Rh6 (Fig 4B), whereas the C region contributes to Otd’s ability to repress Rh6 in OPRs (Fig 5A).