A Genetic Screen for Novel Components of the Ras/Mitogen-Activated Protein Kinase Signaling Pathway That Interact With the yan Gene of Drosophila Identifies split ends, a New RNA Recognition Motif-Containing Protein

Corresponding author: Ilaria Rebay, Whitehead Institute, 9 Cambridge Center, Cambridge, MA 02142., rebay{at}wi.mit.edu (E-mail)

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The receptor tyrosine kinase (RTK) signaling pathway is used reiteratively during the development of all multicellular organisms. While the core RTK/Ras/MAPK signaling cassette has been studied extensively, little is known about the nature of the downstream targets of the pathway or how these effectors regulate the specificity of cellular responses. Drosophila yan is one of a few downstream components identified to date, functioning as an antagonist of the RTK/Ras/MAPK pathway. Previously, we have shown that ectopic expression of a constitutively active protein (yan^ACT) inhibits the differentiation of multiple cell types. In an effort to identify new genes functioning downstream in the Ras/MAPK/yan pathway, we have performed a genetic screen to isolate dominant modifiers of the rough eye phenotype associated with eye-specific expression of yan^ACT. Approximately 190,000 mutagenized flies were screened, and 260 enhancers and 90 suppressors were obtained. Among the previously known genes we recovered are four RTK pathway components, rolled (MAPK), son-of-sevenless, Star, and pointed, and two genes, eyes absent and string, that have not been implicated previously in RTK signaling events. We also isolated mutations in five previously uncharacterized genes, one of which, split ends, we have characterized molecularly and have shown to encode a member of the RRM family of RNA-binding proteins.

DURING development, multicellular organisms must coordinate the growth, differentiation, and maintenance of many different cell types. To achieve this, each cell must continually integrate a complex array of external signals, including both inductive and inhibitory cues, and then translate these instructions into spatially and temporally appropriate developmental responses. Many of the signaling mechanisms regulating these decisions are used repeatedly, generating different cellular responses in different developmental contexts.

The Drosophila compound eye is an ideal tissue in which to study the molecular mechanisms underlying cell-cell communication (RUBIN 1988 ; ZIPURSKY and RUBIN 1994 ). The adult eye is composed of a regular array of ~800 ommatidial units, each one containing ~20 cells recruited in a stereotyped sequence of inductive interactions that involves extensive networks of intercellular signaling events (TOMLINSON and READY 1986 ). The same evolutionarily conserved signaling mechanisms used reiteratively during the development of all multicellular organisms, including receptor tyrosine kinase (RTK), Notch, hedgehog, wingless, and transforming growth factor-ß (TGF-ß) pathways, play essential roles in establishing the architecture of the fly eye (CAGAN and READY 1989 ; HEBERLEIN et al. 1993 ; DICKSON and HAFEN 1994 ; TREISMAN and RUBIN 1995 ; PIGNONI and ZIPURSKY 1997 ). Because the Drosophila eye is not required for viability or fertility of the adult, these essential signaling pathways can be perturbed and manipulated in an eye-specific manner without affecting the development or viability of the animal as a whole.

We have been focusing our investigations on the RTK-mediated signaling pathway that regulates a broad range of developmental events including mitogenesis, cell fate specification, and differentiation (MARSHALL 1994 ; MARUTA and BURGESS 1994 ; VAN DER GEER et al. 1994 ). Extensive investigations from numerous laboratories have shown that RTKs signal through an evolutionarily conserved pathway that involves the GTPase Ras and the mitogen-activated protein kinase (MAPK) cascade of serine/threonine kinases (MARSHALL 1994 ; ZIPURSKY and RUBIN 1994 ). In Drosophila, the RTK/Ras/MAPK cascade acts inductively to promote cell fate specification and differentiation of numerous embryonic and adult tissues including the central nervous system (CNS), muscles, tracheal system, wing veins, neuronal and nonneuronal cells in the eye, and oogenesis (SHILO 1992 ; DICKSON and HAFEN 1994 ; DUFFY and PERRIMON 1994 ; SCHNORR and BERG 1996 ; ENGSTROM et al. 1997 ; LEE and MONTELL 1997 ; SCHWEITZER and SHILO 1997 ; BUFF et al. 1998 ). While genetic and biochemical screens have identified many of the components required for Ras activation and for transmission of this information through the MAPK signaling cascade, the downstream effector molecules that coordinate different physiological responses, and thereby impart specificity to the pathway, have remained largely elusive.

A small number of genes have been identified as downstream targets of activated MAPK in Drosophila. These include two Ets-domain transcription factors, pointed and yan, the AP-1 transcription factor D-jun, and the cell death gene hid (O'NEILL et al. 1994 ; KOCKEL et al. 1997 ; BERGMANN et al. 1998 ). However, little is known about how these proteins regulate gene expression in response to RTK-initiated signals or whether the Ras/MAPK signal alone regulates their activity. We have been investigating the function of one of these genes, yan, in an effort to understand more about how RTK-mediated signals are transduced to the nucleus.

yan was originally identified as a negative regulator of R7 photoreceptor neuron differentiation in the developing eye, acting as an antagonist to the proneural signal mediated by the sevenless RTK signaling pathway (LAI and RUBIN 1992 ). Subsequent studies have shown that yan functions as a fairly general inhibitor of differentiation, preventing both neuronal and nonneuronal cell types from responding inappropriately to external signals in multiple developmental contexts (REBAY and RUBIN 1995 ). Experiments in cultured cells suggest yan functions as a transcriptional repressor, and that phosphorylation of yan by MAPK negatively regulates this repressor activity (O'NEILL et al. 1994 ).

Mutation of the phosphoacceptor residues in all eight of the MAPK phosphorylation consensus sites in yan produces a stable protein that cannot be properly "turned off" (REBAY and RUBIN 1995 ). This allele is referred to as yan^ACT for "activated" yan. Overexpression of yan^ACT inhibits differentiation of multiple neuronal and nonneuronal cell types throughout development, indicating that yan function is not restricted to the sevenless (sev) RTK pathway, which operates only in a discrete subset of cells in the developing eye (REBAY and RUBIN 1995 ). Work from several labs has confirmed that in addition to its role in antagonizing the sev-mediated signal, yan also functions downstream of and negatively regulates at least two other known RTK pathways, the epidermal growth factor receptor and the fibroblast growth factor receptor pathways (SHILO 1992 ; REBAY and RUBIN 1995 ; SAMAKOVLIS et al. 1996 ). Thus, because yan functions as a transcriptional repressor downstream of multiple RTK pathways, it offers an ideal tool with which to address the issue of how specificity of response to RTK-mediated signaling events is generated in different developmental contexts.

We have undertaken an extensive genetic screen to identify dominant modifiers of the rough eye phenotype associated with eye-specific overexpression of the yan^ACT transgene. Our screen was modeled on previous successful screening efforts that have been pivotal in identifying components of the RTK/Ras/MAPK cascade. Each screen has taken a slightly different starting point, beginning with the RTK and progressing down the pathway (for example, SIMON et al. 1991 ; DICKSON et al. 1996 ; KARIM et al. 1996 ). While there has been some overlap in the genes identified, each new screen has also isolated unique genes, and thereby has made significant contributions to our understanding of how the signal is transduced from the RTK, to Ras, Raf, and the MAPK cascade. Thus, by approaching the pathway from various angles, new information has been discovered at each step. Centering the screen around one of the most downstream components in this conserved pathway should facilitate identification of some of the elusive downstream effectors and regulators of the RTK pathway.

In this article we describe the results of a genetic modifier screen in which we isolated 15 complementation groups that dominantly enhance the yan^ACT rough eye phenotype and 6 complementation groups that dominantly suppress it. These genes include known components of the pathway such as rolled/MAPK, the guanine nucleotide exchange factor son of sevenless, and the Ets family transcription factor pointed, as well as several novel genes, which, based on genetic interaction data, are likely to encode relevant new components of the Ras/MAPK pathway. We report preliminary molecular analysis of one of these novel genes, called split ends, that suggests it encodes a member of the RNA recognition motif (RRM) family of proteins.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

yan^ACT modifier screen:
To facilitate chromosomal linkage analysis and establishment of balanced stocks, the GMR-yan^ACT and sev-yan^ACT transgenes (REBAY and RUBIN 1995 ) were mobilized onto CyO and TM3Sb balancer chromosomes. Independent insertion events in which the yan^ACT phenotype cosegregated with the balancer were screened for strength, dose sensitivity, and variability of phenotype. Based on these criteria, two lines, Sev-yan^ACT on CyO (SCY) and GMR-yan^ACT on CyO (GCY), were selected as suitable for use in the modifier screen. Two other lines, sev-yan^ACT on TM3Sb (STY) and GMR-yan^ACT on TM3Sb (GTY), were selected for use in determining chromosomal linkage.

Male w¹¹¹⁸ flies isogenic for the second and third chromosomes were fed 25 mM EMS or treated with 4000 R of X rays (1000 sec at 115 kEV, 5 mA), and then mated to w¹¹¹⁸;Sco/SCY or w¹¹¹⁸;Sco/GCY females. The eyes of F₁ progeny were examined under a dissecting microscope for enhancement or suppression of the SCY or GCY phenotype. Potential modifiers were backcrossed to w¹¹¹⁸;Sco/SCY or w¹¹¹⁸;Sco/GCY as appropriate, and the F₂ progeny were rescored for suppression or enhancement. In the F₂ generation, linkage of the mutation to the second chromosome could be determined if all non-Sco flies bearing the SCY or GCY chromosome showed the appropriate modification; in this case a balanced stock, w¹¹¹⁸;modifier/SCY or GCY, was established. If the modifier segregated randomly with respect to the SCY or GCY chromosome, this indicated linkage of the mutation to the third chromosome. To establish a balanced stock, flies of the genotype w¹¹¹⁸;+/SCY or GCY;modifier/+ were crossed to w¹¹¹⁸;CxD/STY or GTY. In the next generation, a balanced stock of w¹¹¹⁸;modifier/STY or GTY was established by selecting STY or GTY flies showing the appropriate enhancement or suppression. x-chromosome-linked modifiers were too difficult to maintain in the yan^ACT backgrounds, which on their own exhibit reduced fertility, and were lost. Therefore, only the second and third chromosomes have been screened for modification of the yan^ACT phenotype.

Complementation tests based on lethality were performed among all mutations mapping to the same chromosome. Single hits are those mutations that complement all other mutations on that chromosome. A number of mutations are viable; these have not been mapped, and therefore it is not possible to know how many different genes they represent. Two alleles for each defined complementation group were mapped meiotically relative to b pr c px sp on the second chromosome and ru h th st cu sr e ca on the third. Both the lethality and enhancement or suppression of yan^ACT were mapped. Lethal complementation tests with deficiencies, P-element insertions, and other mutations in the region were used to further refine the map position. Noncomplementation was as follows (refer to Table 1): EY2-1 and clift(eya1); EY2-2 and rolled^S135 (KARIM et al. 1996 ); EY2-3 and son of sevenless^e2H (SIMON et al. 1991 ); EY2-4 and Star^ES2-4e220 (NEUFELD et al. 1998 ); EY2-5 and Df(2L)ast1; EY2-7 and P elements l(2)03350, l(2)k06805, l(2)k07612, l(2)k07721, l(2)k08102, l(2)k13233, l(2)k13624,l(2)k15612, l(2)k06703, l(2)k10325, l(2)k13601; EY2-8 and Df(2L)CX1; EY2-9 and sin3^l(2)7401 (NEUFELD et al. 1998 ); EY2-10 and dead ringer¹ (GREGORY et al. 1996 ); EY3-1 and pointed⁷⁸²⁵⁷⁸ (O'NEILL et al. 1994 ); EY3-2 and Df(3R)e-R1; EY3-3 and Df(3R)e-R1; SY2-1 and SS2-1^s35 (NEUFELD et al. 1998 ); SY3-1 and SR3-4a^S-192 (KARIM et al. 1996 ); SY3-2 and TAF110^S117 (KARIM et al. 1996 ); SY3-3 and glass¹; SY3-4 and string^eAS7a (VERHEYEN et al. 1996 ).

Histology:
Flies were prepared for scanning electron microscopy by fixation for 2 hr in 1% glutaraldehyde, 1% paraformaldehyde in 1 M cacodylate buffer followed by dehydration through an ethanol series and critical point drying. Fixation and sectioning of adult eyes was as described by TOMLINSON et al. 1987 .

Molecular analysis of EY2-7:
Plasmid rescue (PIRROTTA 1986 ) was used to isolate genomic DNA flanking l(2)7721 and l(2)7612. An ~20-kb BglII fragment on one side of the P-element insertion site and an ~25-kb SacII site on the other were recovered. Cytological examination of X-ray-induced EY2-7 alleles revealed one allele, designated xLS1000, to be a cytologically visible inversion. To determine whether either of the two plasmid rescue DNA fragments crossed the xLS1000 breakpoint, chromosomal in situ hybridization on xLS1000 chromosomes was performed. The SacII fragment crossed the xLS1000 breakpoint, while the BglII fragment was entirely distal (data not shown). The entire SacII genomic fragment and various smaller subclones of this fragment were used as probes to screen cDNA libraries prepared from eye-antennal imaginal discs (made by A. Cowman), adult heads (obtained from the lab of Y. N. Jan), and embryos (made by L. Hong and G. M. Rubin). To obtain cDNAs spanning the entire coding sequence, four successive rounds of screening were performed, each using a probe taken from the most 3' sequence of the longest cDNA at hand. The cDNAs were sonicated. Random ~1-kb fragments were subcloned into Bluescript and then sequenced with Universal and Reverse primers on an ABI 373 sequencer. An ~2-kb gap in the coding sequence (nucleotide coordinates 12018–14125), apparent from comparison of our sequence to the Berkeley Drosophila Genome Project (BDGP) genomic sequence (GenBank accession no. AC005334), was filled in by RT-PCR using the Marathon cDNA amplification kit (Clontech, Palo Alto, CA). Primer pairs used to amplify two overlapping fragments covering the region were: primer pair 1: 5' ATGGTAAGCCAACCATCACC 3' and 5' TCTGGTTACAGGAGCTGTTACGGG-3'; primer pair 2: 5' ATCATGATACAGAGGACGAAACGG 3' and 5' AGTATCCGTGAGCAACATGTGACCG 3'. Sequencing primers were designed based on the BDGP genomic sequence, and both strands of the RT-PCR products were sequenced directly without subcloning. To confirm the existence of the noncoding cDNAs j1 and 2A1, RT-PCR was performed from adult heads and embryos. Primers were designed to span the introns in both cDNAs: for cDNA j1: primers 5' CGAAACGGCAAGTGCTTAAGCGAAAG 3' and 5' AACAGTTCGTTGTGGATAGCAGCACC 3'. For cDNA 2A1: primers 5' TGGTAAGCTATCAAGAACATATCTGC 3' and 5' TACCTTCCCACACAGATCGTACTCC3'. For negative control, primers were designed in the predicted introns: for cDNA j1: intron primers 5' AGTGAAAGGCGTGTGGTAGATCACG 3' and 5' TACCTTCCCACACAGATCGTACTCC 3'; for cDNA 2A1: intron primers 5' AATACGACTTCAAGATGAGTCGTCG 3' and 5' TTTCGGAAGTCTGTGAATCGATAGC 3'. If our cDNAs were artifactual, no PCR product would be recovered in the case of j1, which has an ~12-kb predicted intron, and a larger than expected product would be amplified for 2A1, which has a much smaller predicted intron. For both j1 and 2A1, bands of the expected sizes were obtained, indicating that these two cDNAs represent bona fide spliced mRNA species (data not shown). No bands were amplified with the intronic primers. The approximate position of the introns was determined by comparing our cDNA sequence to BDGP genomic sequence.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A genetic screen for modifiers of yan^ACT:
Overexpression of a yan cDNA in which all eight putative MAPK phosphorylation consensus sites were mutated to a nonphosphorylatable form, referred to as yan^ACT, inhibits the differentiation of multiple neuronal and nonneuronal cell types (REBAY and RUBIN 1995 ). Previously, we have described the effects of expressing the yan^ACT transgene in the developing Drosophila eye using the sevenless (sev) and GMR promoters (Figure 1, A–C; FORTINI et al. 1992 ; HAY et al. 1994 ; REBAY and RUBIN 1995 ). sev-yan^ACT flies have rough eyes characterized by the loss of the R3, R4, and R7 photoreceptor neurons and the nonneuronal cone cells. GMR-yan^ACT flies have more severely disrupted eyes due to inhibition of differentiation of photoreceptors R1–R8, the cone cells, and other accessory cells of the eye. Both sev-yan^ACT and GMR-yan^ACT rough eye phenotypes are sensitive to the copy number of the transgene (data not shown). That is, flies carrying two copies of the transgene have more severely disrupted eyes than flies with only one copy. The dose sensitivity of the phenotype suggested that both sev-yan^ACT and GMR-yan^ACT genetic backgrounds might represent good starting points for a genetic screen designed to identify potential regulators or targets of yan activity.

View larger version (117K): In this window In a new window Download PPT slide   Figure 1. The yanACT-based genetic screen. Scanning electron micrographs (SEM) of the starting phenotypes used in the genetic modifier screen. (A) Wildtype, (B) sev-yanACT/+, and (C) GMR-yanACT/+. (D) Schematic representation of the strategy for the screen. w1118 males were mutagenized with either EMS or X rays and were mated to virgin females carrying the yanACT transgene either under sevenless or GMR promoter control (SCY and GCY, respectively). In the subsequent (F1) generation, progeny were scored for enhancement or suppression of the rough eye phenotype. An asterisk indicates a mutagenized chromosome.

An important feature of this type of genetic screen is that it is designed to isolate dominant modifiers of a specific phenotype. The expectation is that a twofold reduction in activity of a pathway relevant gene, achieved by mutating one of the two copies present in the diploid genome, will dominantly enhance or suppress the starting background phenotype. In contrast, in a wild-type background, a twofold reduction in gene activity will produce dominant phenotypes only rarely. Thus in a wild-type background, screens designed to study the phenotypic consequences of reduction in gene activity will require at least two generations to produce a homozygous mutant and a visible phenotype. However, many of the genes involved in signal transduction pathways are essential for viability, necessitating either analysis of lethal embryonic phenotypes or clonal analysis in particular tissues. While such F₂ screens have been extremely successful at isolating genes functioning in specific developmental processes (for example, NUSSLEIN-VOLHARD et al. 1984 ; DUFFY et al. 1998 ) the genetic modifier screen has proven to be a highly efficient method of dissecting signaling pathways in genetically manipulable organisms (for example, in Drosophila, SIMON et al. 1991 ; DICKSON et al. 1996 ; KARIM et al. 1996 ; VERHEYEN et al. 1996 ). In addition, the ease of a dominant F₁ screen allows large numbers of flies to be screened, thereby approaching saturation.

In our screen, because yan is a negatively acting factor in the RTK pathway, expression of the stable yan^ACT product reduces overall signaling output by the pathway. Thus in the yan^ACT background, a twofold reduction in activity of positively acting genes elsewhere in the pathway will further reduce signaling output, thereby exacerbating the eye phenotype. Such mutations will be recovered as enhancers. Conversely, halving the activity of negatively acting components of the pathway will increase the overall output of the pathway. Such mutations will be recovered as suppressors of yan^ACT or as flies with less severely disrupted eyes.

Approximately 80,000 progeny from EMS-mutagenized flies and 110,000 progeny from X-ray-mutagenized flies were screened for ability to dominantly modify the yan^ACT phenotype. The scheme of the screen is diagrammed in Figure 1D. Approximately half of the screen was performed in the sev-yan^ACT background and half in the GMR-yan^ACT background (Figure 2A and Figure D). Both enhancers and suppressors of the rough eye phenotype were recovered from each portion of the screen (Figure 2B, Figure C, Figure E, and Figure F). A total of 260 enhancer mutations was recovered, 177 on the second chromosome and 83 on the third chromosome. Also, 90 suppressor mutations were recovered, 49 on the second chromosome and 21 on the third chromosome. Complementation analysis placed 215 of the mutations into 21 complementation groups (Table 1). The remaining 135 mutations include both "single hit" alleles that do not fall into any of the complementation groups and viable mutations. These 135 mutations have not been mapped or characterized further.

View larger version (212K): In this window In a new window Download PPT slide   Figure 2. Isolation of enhancers and suppressors of yanACT. SEMs of both the starting yanACT phenotypes and examples of recovered modifiers determined by the genetic tests as likely to encode RTK pathway relevant products. (A) SCY/+; (B) EY2-3/SCY; (C) SY2-2/SCY; (D) GCY/+; (E) EY2-3/GCY; (F) SY2-2/GCY. For enhancer mutations, the eye is smaller and more disorganized than the parental strain. For suppressor mutations, the eye is larger and the ommatidial lattice is more regular in appearance. EY, enhancer of yanACT; SY, suppressor of yanACT.

Genetic tests to determine which are the "pathway relevant" groups:
One of the most important aspects of a genetic interaction screen is to design suitable secondary screens that will distinguish pathway relevant mutations from inevitable background. These secondary tests serve as a sort of "genetic triage" that allows rapid classification of the isolated mutations and prioritization of further efforts to understand the role of the genes in the particular context being studied. Our secondary screens were designed to identify those mutations that specifically affect the RTK/Ras/MAPK/yan pathway. All alleles from the 21 complementation groups were tested, and the results for two alleles representative of each group are shown in Table 1. The results of these genetic tests, together with the identification of several known components of the RTK signaling pathway among the candidate genes isolated, indicate that the yan^ACT interaction screen appears to have worked as designed.

Test 1: To distinguish potential pathway relevant mutations from those that affect the strength of the promoters driving yan^ACT expression: Because our screen involves expressing the yan^ACT transgene under the control of eye-specific promoter elements, one class of modifiers we expected to isolate and wanted to eliminate was that affecting the strength of the promoters. For example, because the yan^ACT phenotypes are dose sensitive, if we recover a mutation that increases the strength of the promoter element, this will result in increased yan^ACT expression and an enhancement of the rough eye phenotype. Such a mutation does not affect yan or Ras pathway activity directly, and is unlikely to be relevant to the signaling pathway. Thus, the first genetic test was designed to separate the pathway relevant mutants from the promoter-specific mutants. We reasoned that pathway relevant mutants will modify both the GMR-yan^ACT and sev-yan^ACT rough eye phenotypes, whereas promoter-specific mutations will modify one background only. Therefore, all modifiers obtained from the GMR-yan^ACT half of the screen were crossed to the sev-yan^ACT background, and conversely all modifiers obtained from the sev-yan^ACT half of the screen were crossed to the GMR-yan^ACT background.

On the basis of this test, 11 of the 21 complementation groups appear relevant to Ras/MAPK/yan signaling (these are underlined in Table 1). In fact, for the complementation groups that passed test 1, alleles were isolated both from the sev-yan^ACT and the GMR-yan^ACT halves of the screen, confirming these results in an unbiased manner. Further confirming that this test was working as planned, among the 11 pathway relevant groups are known components of the RTK pathway including rolled/MAPK, son of sevenless, Star, and pointed (ROGGE et al. 1991 ; SIMON et al. 1991 ; KLAMBT 1993 ; BIGGS et al. 1994 ; O'NEILL et al. 1994 ). Isolation of these genes as enhancers of the yan^ACT phenotype is consistent with their known roles as positive factors in the RTK pathway.

Two of the 11 pathway relevant groups were found to be allelic to previously characterized genes whose role, if any, in the RTK signaling pathway has not yet been determined. EY2-1 is allelic to eyes absent (eya), a gene known to function within a hierarchy of genes essential for determining eye fate during development (BONINI et al. 1993 , BONINI et al. 1997 ). SY3-4 is allelic to string, the Drosophila homologue of the cdc25 phosphatase cell cycle regulator (EDGAR and O'FARRELL 1989 ). The remaining RTK pathway candidate groups, EY2-5, EY2-6, EY2-7, and EY3-5, appear to be uncharacterized genes.

Among the known genes classified as promoter specific are TAF110, a coactivating factor in PolII transcription (HOEY et al. 1993 ), glass, a transcription factor essential for expression of the GMR element (ELLIS et al. 1993 ; HAY et al. 1994 ), dead ringer, a homeodomain-related protein (GREGORY et al. 1996 ), and sin3A, a transcriptional repressor (WANG et al. 1990 ). Other genes in this category that have been isolated in previous screens are SY2-1, which is allelic to SS2-1, and SY3-1, which is allelic to SR3-4a and SS3-4 (KARIM et al. 1996 ; NEUFELD et al. 1998 ). The genes corresponding to the EY2-8, EY3-2, EY3-3, and EY3-4 complementation groups have to our knowledge not been identified previously.

Test 2: To determine which groups interact with an endogenous gain-of-function yan allele: In the second genetic test to determine which were the relevant yan-interacting genes, a gain-of-function yan allele (yan^S2382; KARIM et al. 1996 ) was crossed to all alleles of the 21 complementation groups (Table 1). The phenotypes of yan^S2382 animals are similar to, although milder than, those seen with the yan^ACT transgene. Homozygous yan^S2382 flies have slightly roughened eyes, reflecting an increase in stability of the yan^S2382 protein product that inhibits differentiation of the photoreceptor neurons. Sequence determination of the molecular lesion associated with the mutation revealed the predicted protein to be a truncated form of yan that lacks the C-terminal 162 amino acids (REBAY and RUBIN 1995 ). The sequence deleted in yan^S2382 covers one of the PEST-rich regions of the protein that is postulated to be critical for normal downregulation of yan in response to Ras/MAPK activation. Thus yan^ACT, a gain-of-function allele generated by mutating the phosphoacceptor residue in all eight putative MAPK sites to alanine, and yan^S2382, a truncation allele, although they each cause very different alterations in the protein, both result in similar phenotypes in the developing eye. Modifiers that interact strongly with both gain-of-function yan backgrounds are likely to encode proteins relevant to yan function.

The results of this second test overlapped almost perfectly with the first test, further supporting our logic that those groups "passing" the tests represent likely RTK/yan pathway relevant genes. We found that 9/11 of the genes determined to be pathway relevant by test 1 interacted strongly with yan^S2382, whereas none of the promoter-specific groups affected yan^S2382. The two groups (indicated in Table 1 with an asterisk) that did not modify the yan^S2382 background but are not promoter specific are EY2-5 and SY3-4 (string). One possibility is that these two groups are relevant to the Ras/MAPK/yan pathway, but that functional differences between yan^ACT and the yan^S2382 gain-of-function allele create differences in sensitivity to second site modifiers. An alternative possibility is that these two genes act at the level of promoter strength, but act similarly on both the GMR and sev elements. Interestingly, EY2-5, a group of two alleles, was placed in the category of genes passing test 1 because one allele was isolated from the sev-yan^ACT part of the screen and the other from the GMR-yan^ACT part. However, neither EY2-5 allele modifies the eye phenotype in the other promoter background.

Test 3: To determine which groups interact with other RTK pathway components: The third test was based on the assumption that if a modifier of yan^ACT is involved in the RTK signaling pathway, it is likely to interact with other components of the pathway in addition to yan. We therefore tested whether the modifiers of yan^ACT interacted with two eye-specific transgenes, sev promoter-driven dominant negative Ras (Ras^N17) and sev promoter-driven activated Ras (Ras^V12), as well as with a hypomorphic Raf alleles which at 18° produces occasional hemizygous males with rough eyes (Raf^HM7; KARIM et al. 1996 ). Furthermore, the direction of interaction, whether enhancement or suppression, should be consistent with the modifier acting as a positive or negative factor in the pathway. For example, the expectation for an allele that enhances yan^ACT is that it should enhance dominant negative Ras^N17 and hypomorphic Raf^HM7, but suppress activated Ras^V12. All 11 groups identified as RTK pathway relevant showed consistent interactions with additional components of the pathway, whereas the promoter-specific groups did not (Table 1).

The alleles we isolated of known components of the RTK pathway behaved generally as expected or as shown previously by others. For example, EY2-2 (rolled/MAPK) alleles, which were isolated as enhancers of yan^ACT, strongly suppress Ras^V12, enhance Ras^N17, and are synthetic lethal with Raf^HM7. The direction of modification is in each case consistent with rolled functioning as a positively acting component of the RTK pathway. Several complementation groups exhibited similar patterns of interactions in these crosses, strongly suggesting an involvement as positively acting factors in the pathway. For example, most alleles of EY2-1 (eyes absent) suppress Ras^V12 (although several, including G130, show no interaction), enhance Ras^N17, and are either synthetic lethal with Raf^HM7 or else produce hemizygous "escaper" males with an enhanced rough eye phenotype (data not shown). EY2-7 exhibited a similar pattern of interactions except all alleles were synthetic lethal with Raf^HM7.

Other groups exhibited less consistent patterns of interactions in the various Ras pathway backgrounds. For complementation groups consisting of small numbers of alleles, there were allele-specific differences in behavior that made classification difficult. With large complementation groups of 10 or more alleles, it was easier to establish the general pattern of interaction, despite an occasional allele-specific inconsistency. SY3-4 (string) alleles showed no interaction with Ras^N17 or Raf^HM7, but 1 of the 2 alleles strongly enhanced Ras^V12 while the other showed no interaction. EY2-5 and EY2-6 also showed some contradictory patterns of interaction. One allele of EY2-5 and 1 allele of EY2-6 were found to enhance both Ras^N17 and Ras^V12, while the other allele showed no interaction. Other groups, such as EY2-3 (sos), EY3-1 (pointed), and EY3-5, enhanced Ras^N17 but showed little or no interaction with Ras^V12. These inconsistencies could simply be indicative of the different sensitivities of a particular genetic background to modification by reduction of dosage of a specific gene, or could suggest that the interaction may be more complex, possibly involving both positive and negative regulatory feedback loops. It should be stressed that these initial tests are used to prioritize future efforts as to which genes to investigate in detail. Definitive proof of involvement in RTK/yan-mediated signaling events must await further genetic and biochemical characterization of the function of the genes.

Molecular analysis of EY2-7:
EY2-7 comprises a group of 37 alleles, of which 18 are EMS induced and 19 are X-ray induced. Meiotic mapping localized the gene to the distal end of the left arm of the second chromosome. To define the genomic location of the gene more precisely, we screened deficiencies and lethal P-element insertions in the region for failure to complement EY2-7. Df(2L)PMF47c and 11 lethal P-element insertions fail to complement EY2-7 and also enhance yan^ACT (data not shown).

To determine whether the P-element insertions were responsible for the lethal noncomplementation, the P elements were excised from all 11 lines. For all P alleles, both viable and lethal excision events were recovered. The viable lines, presumably precise excision events in which the P elements were removed without causing chromosomal deletions, complemented EY2-7 alleles, and no longer enhanced the yan^ACT rough eye phenotype, indicating that they no longer carried an EY2-7 mutation (data not shown). The lethal lines, presumably imprecise excisions in which the chromosomal region surrounding the original P-element insertion site was disrupted, failed to complement EY2-7 alleles and enhanced yan^ACT(data not shown). These results suggested that the P alleles were inserted in or near the EY2-7 gene and could be used as molecular probes with which to clone the gene.

We isolated ~45 kb of genomic DNA flanking the insertion sites of the two P elements showing the strongest enhancement of yan^ACT (see MATERIALS AND METHODS). We have subsequently determined that all 11 P-element lines are inserted within the same ~4-kb region (Figure 3A; data not shown). A 25-kb fragment of genomic DNA flanking the P-element insertion sites was found to cross the inversion breakpoint of a cytologically visible X-ray-induced EY2-7 allele (Figure 3A; data not shown). Southern blot restriction fragment polymorphism (RFLP) analyses using this genomic fragment as a probe revealed polymorphisms in multiple EY2-7 alleles (data not shown). Together, these data suggested the EY2-7 gene would reside within the 25 kb of genomic DNA. This genomic fragment was therefore used as a probe to screen an embryonic cDNA library. A single class of cDNAs was isolated (referred to as class I; Figure 3A), the largest of which was ~6.5 kb in length. Restriction mapping placed the cDNA ~20 kb away from the site of P-element insertion. In situ hybridization showed that the cDNA crossed the EY2-7 inversion allele breakpoint (data not shown). Sequence analysis revealed that the class I cDNA contained a 6-kb open reading frame (ORF) but lacked a stop codon. To obtain the full-length cDNA, several additional rounds of screening were needed to compile an ~18-kb contig.

View larger version (349K): In this window In a new window Download PPT slide   Figure 3. Molecular characterization of EY2-7 (spen). (A) The 50-kb genomic region encompassing the spen gene is depicted as a black line. The numbers underneath indicate the size of the region in kilobases. The scale of the drawing is indicated with a short line representing 2 kb. The P-element lines that fail to complement EY2-7 mutations are inserted over an ~4-kb region. The position of the inversion allele, xLS1000, is indicated with a pair of lines at ~15 kb along the genomic sequence, and is marked as In for inversion. Three cDNAs, j1, 2A1, and class I, are indicated as thick black lines above the genomic line. j1 and 2A1 are noncoding cDNAs; class I corresponds to the spen gene. Introns are depicted as open triangles above the cDNAs. The predicted spen protein product derived from the class I cDNA is depicted underneath. Three structural domains, RRM, black box, coiled-coil (CC, striped box), and C-terminal conserved region (c-term, gray box) are indicated. The scale bar beneath indicates 500 amino acids. (B) The spen RRM motif is shown aligned with RRMs from several predicted proteins. The positions of conserved residues that define the RRM consensus sequence are indicated with an asterisk above the amino acid. The amino acid consensus sequence defining the RRM motif is indicated above the asterisk for the first RRM of spen. The RNP1 hexapeptide and RNP2 octapeptide regions for each RRM are indicated with a line above the amino acid sequence. The three spen RRMs are indicated with circled numbers 1, 2, and 3. In the alignment, identical amino acids are boxed and highlighted in dark gray. Similar amino acids are boxed and highlighted in light gray. Alignment was done using MacVector's ClustalW program. Accession numbers for all GenBank sequences are as follows: Dm 44a, AC005448; Hs spen, AL096858; Mm RRM, CAB01562 (the aligned sequence for Mm RRM is a composite of the GenBank sequence and sequence from a partial cDNA isolated from a mouse teratocarcinoma library; I. REBAY, unpublished results; sequence available upon request); Ce RRM, AF067616. (C) The conserved spen C terminus is shown aligned with C-terminal domains from several predicted proteins. Identical amino acids are boxed and highlighted in dark gray. Similar amino acids are boxed and highlighted in light gray. Alignment was done using MacVector's ClustalW program. Accession numbers for all GenBank sequences are as follows: Hs spen, AL096858; Dm 44a, AC005448; Ce CAA91320, Z66511. (D) Sequence of the composite full-length spen cDNA. Nucleotide sequence is indicated above; conceptual translation of the ORF is indicated below. The putative RRM region is underlined with a solid line. The predicted coiled-coil region is underlined with a dashed line. The conserved C-terminal region is highlighted in gray. Two putative potential polyadenylation signal sequences are highlighted in bold. GenBank accession no. AF184612.

Because the 5' end of our cDNA was so far away from the P-element insertion site, we wanted to rule out the possibility that the true EY2-7 gene lay within the intervening ~20 kb of genomic DNA. To test whether there were additional transcripts in this region, we subdivided the intervening ~20 kb into five different subclones, and screened several cDNA libraries with each fragment. Two cDNA clones were isolated, one from an eye imaginal disc library (2A1) and the other from an adult head library (j1; Figure 3A). When sequenced, both were found to lack ORFs greater than ~250 amino acids in any reading frame. Because we isolated so many EMS alleles of EY2-7, we reasoned the gene was unlikely to encode a short polypeptide, and did not consider these noncoding cDNAs likely candidates to encode EY2-7. We have used RT-PCR to confirm that the noncoding cDNAs we isolated reflect actual transcripts rather than possible genomic contaminants in the libraries (see MATERIALS AND METHODS).

Another possibility was that the EY2-7 transcript was within this ~20-kb genomic region, but was of such low abundance that even extensive screening of the libraries would fail to detect it. However, confirming that our lack of success in finding a candidate cDNA within the ~20-kb genomic region was not simply due to its being an extremely low abundance message, but rather due to the lack of coding transcripts within this region, the BDGP has submitted complete genomic sequence covering this region (GenBank accession no. AC005334). Conceptual translation of this ~20-kb genomic DNA confirms the absence of any sizeable ORFs (data not shown), suggesting that there are no additional coding sequences residing in between the P-element insertion sites and our class I cDNA.

On the basis of this information, our best cDNA candidate for EY2-7 resides ~20 kb away from the P-element insertion sites. We think it is likely that the noncoding cDNAs we isolated may represent parts of a large and complex 5' untranslated region (5'UTR) or possibly alternatively spliced 5' ends of EY2-7. Supporting this idea, the insertion sites for several of the noncomplementing P-element alleles fall within one of these cDNAs (j1; Figure 3A).

Complementation tests show that EY2-7 is allelic to the gene split ends (spen) (KOLODZIEJ et al. 1995 ). Mutations in EY2-7 fail to complement two alleles of spen that have been molecularly characterized and shown to have lesions within the coding sequence of the EY2-7/spen cDNA that result in predicted truncated protein products (B. KUANG and P. KOLODZIEJ, unpublished results). This indicates that our class I cDNA is indeed encoded by the EY2-7 gene. From now on we will refer to EY2-7 as spen.

spen encodes a predicted protein of 5476 amino acids. Database searches indicate spen belongs to a family of RRM proteins. The RRM is a loosely conserved RNA binding domain of ~22 conserved amino acids spread over an 80–100-amino-acid-long region (for review see BURD and DREYFUSS 1994 ). The most highly conserved sequences within the RRM motif are two ribonucleoprotein (RNP) domains designated RNP2 (a hexapeptide) and RNP1 (an octapeptide). Other conserved residues are scattered throughout the domain and include primarily hydrophobic amino acids.

Spen contains three RRM motifs in tandem toward the N terminus of the protein (amino acids ~500–750). The spen RRMs are most similar to RRMs in several novel proteins of unknown function. These include proteins predicted from conceptual translation of the BDGP Drosophila genomic sequence and the Caenorhabditis elegans genomic sequence, and proteins predicted from conceptual translation of human and mouse EST sequences (Figure 3B). The first RRM of spen is most divergent from the RRM consensus, and is also more loosely conserved among other spen-like proteins. Homology between spen and other known RRM proteins in Drosophila or other species is less striking and strictly limited to residues defining the RRM consensus sequence (not shown). Thus, spen may define a new subclass of RRM proteins.

Other motifs in spen include a predicted coiled coil region over amino acids 1857–1922 (probability 1.0) and amino acids 1979–2014 (probability 0.589) that could be suggestive of protein-protein interactions (BERGER et al. 1995 ; WOLF et al. 1997 ) and a highly conserved C-terminal domain of unknown function that is found in proteins from worms to humans (Figure 3A and Figure C). Otherwise, the spen protein sequence appears novel. Because some of the sequences identified by the database searches correspond to short EST sequences (for example, Mm RRM), it will be necessary to isolate full-length cDNA clones in order to determine whether these proteins contain both the RRM and the C-terminal domain. However, both motifs are found in a second Drosophila protein and in a human protein (referred to as Dm 44A and Hs spen, respectively, in Figure 3B and Figure C), suggesting that Drosophila spen is a member of a novel family of proteins defined by both the RRM and C-terminal motifs.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have performed a dominant modifier screen for enhancers and suppressors of the rough eye phenotype associated with expression of the yan^ACT transgene in the developing Drosophila eye with the goal of isolating novel components of the RTK/Ras/MAPK pathway. Genetic tests suggest 11 of the 21 complementation groups recovered are Ras pathway relevant genes. We have found that 4 of these 11 groups are allelic to known RTK/Ras pathway genes. Isolation of known Ras pathway signaling components confirms that both the screen and the genetic tests are working as designed. That is, known Ras pathway genes are isolated as modifiers of yan^ACT, and these genes can be functionally distinguished from the promoter-specific factors that were also recovered in the screen.

In addition to known pathway components, we have also isolated a set of seven genes that have not been implicated previously in RTK signaling events, but appear likely to play important roles in this pathway, based on the genetic tests. Thus, by initiating a genetic modifier screen based on one of the most downstream components of the RTK pathway identified to date, we have successfully isolated new candidate signaling molecules that were not recovered in previous RTK pathway-based screens. These seven genes include two genes with known functions in other developmental contexts, eyes absent and string, as well as five previously uncharacterized genes. We have initiated the molecular characterization of one of these novel genes, called split ends, and show it encodes a member of the RRM family of RNA-binding proteins.

Isolation of known RTK pathway genes:
Among the mutations in known RTK pathway genes isolated in our screen, 44 alleles of the Ets domain transcription factor pointed were recovered as strong enhancers of yan^ACT. Genetically, pointed is a positive regulator of RTK/Ras-mediated signals, functioning downstream of MAPK in multiple developmental contexts including the developing oocyte, the embryonic CNS, and the photoreceptors of the eye (KLAMBT 1993 ; BRUNNER et al. 1994 ; O'NEILL et al. 1994 ; SCHOLZ et al. 1997 ). Experiments in cultured cells have demonstrated that pointed is a transcriptional activator whose activity is stimulated by the Ras/MAPK pathway (O'NEILL et al. 1994 ). Because pointed and yan are members of the same family of transcription factors and will bind to the same DNA sequences in vitro, it has been postulated that cellular responses to RTK signaling may involve a competition between yan and pointed for access to certain promoters. An alternate model is that yan and pointed represent two different branches of gene regulation downstream of MAPK. Efforts to distinguish between these hypotheses have been thwarted by our lack of knowledge of the identities of these downstream targets. In addition, other than homology and parallel placement in the Ras pathway, there was no evidence supporting a functional interaction between yan and pointed. Our isolation of pointed alleles as dominant enhancers of yan^ACT is a strong indication of an intimate interaction between yan and pointed. Future genetic and biochemical investigations will be required to understand more fully the mechanisms underlying the genetic relationship between these two transcription factors in RTK signaling events.

Our screen also identified a group of 35 enhancers allelic to Star. Star encodes a membrane protein of unknown biochemical function that is thought to assist in processing of the epidermal growth factor receptor (Egfr) ligand Spitz, thereby promoting signaling (KOLODKIN et al. 1994 ; PICKUP and BANERJEE 1999 ). Star mutations exhibit a dominant rough eye phenotype, suggesting the enhancement of yan^ACT could simply represent an additive phenotype rather than a synergistic interaction. On the other hand, the enhancement of yan^ACT is extremely strong, reducing the eye to a tiny slit (data not shown). This phenotype is stronger than would be expected if the phenotypes were simply additive, and this suggests that there may be a synergistic aspect to the interaction. Mutations in Star have been isolated in numerous, but not all, genetic screens based on rough eye phenotypes, again suggesting there may be meaningful specificity to the interaction (DICKSON et al. 1996 ; KARIM et al. 1996 ; VERHEYEN et al. 1996 ; NEUFELD et al. 1998 ). Assuming Star functions at the level of regulating the production of an activated ligand for Egfr (PICKUP and BANERJEE 1999 ), the interaction with yan is likely to be indirect unless regulatory feedback loops or direct transcriptional control is involved.

The third known RTK pathway gene isolated in our screen is rolled. rolled encodes the MAPK that functions downstream of multiple RTKs in Drosophila, and has been shown to phosphorylate yan directly in vitro (BIGGS et al. 1994 ; BRUNNER et al. 1994 ). Because the yan^ACT product lacks the eight MAPK consensus phosphorylation sites and can thus presumably no longer be phosphorylated by MAPK, it is likely that enhancement of the yan^ACT eye phenotype by mutations in rolled reflects either an interaction with the endogenous wild-type yan protein or with other pathway components such as pointed. For example, reducing the activity of MAPK would increase the stability of the endogenous pool of yan protein, thereby exacerbating the rough eye phenotype associated with yan^ACT. Reduction in MAPK activity would also result in a reduction in levels of pointed activation, which, based on our isolation of pointed mutations in the screen, causes a strong enhancement of the yan^ACT rough eye phenotype.

The final known RTK component we isolated is son of sevenless (sos), the guanine nucleotide exchange factor (GEF) for Ras (ROGGE et al. 1991 ; SIMON et al. 1991 ; CHARDIN et al. 1993 ). We were somewhat surprised to isolate alleles of sos because prior to initiating the genetic screen, we had tested all available RTK pathway mutations, including sos null mutations, for interaction with our yan^ACT background (I. REBAY and G. M. RUBIN, unpublished results). We found that mutations in sos exhibited a barely detectable enhancement of the rough eye phenotype, and thus we did not expect to isolate sos in our screen. However, mapping and complementation tests showed that EY2-3, a complementation group of 10 alleles that includes several extremely strong enhancers of yan^ACT (Figure 2B and Figure E), is allelic to sos. In addition, sos null alleles are strong suppressors of the RasV12 background, whereas our EY2-3 alleles show little or no suppression (F. CHEN and I. REBAY, unpublished observations). On the basis of the difference in interaction patterns between null sos alleles and our EY2-3 alleles, we think it is likely that our screen has selected for a particular class of sos alleles. In addition to the Ras GEF domain, sos also contains a so-called Dbl homology domain that is thought to serve as a GEF for the Rho family of small GTPases that included Ras, Rho, and Cdc42 (NIMNUAL et al. 1998 ; SOISSON et al. 1998 ). Thus, one possibility is that our EY2-3 alleles have normal Ras GEF function, but have altered Rho GEF function. If correct, this hypothesis implies a role for yan downstream of the Rho family GTPases in addition to its role downstream of the Ras pathway.

Consistent with the idea that our yan-based screen may be identifying genes that function in pathways other than the Ras/MAPK cascade and that yan itself may be a downstream target of multiple signaling pathways, yan has been shown previously to be a downstream target of Jun kinase (Jnk; RIESGO-ESCOVAR and HAFEN 1997 ). Jnk is a member of the MAPK family, and is thought to function downstream of the Rac/Rho/Cdc42 GTPases (NOSELLI 1998 ). In vitro, Jnk can directly phosphorylate yan on the same sites phosphorylated by rolled/MAPK, and based on genetic interactions between yan and basket (Jnk), this phosphorylation presumably destabilizes and inactivates yan (RIESGO-ESCOVAR and HAFEN 1997 ). It is possible that sos, being capable of serving as a GEF for both Ras and Rho family GTPases, may be a key mediator of cross-talk between these signaling cascades, and that yan, a direct target of at least two MAPK family members, may represent an additional downstream integration point for the different signals. Further investigation of points of intersection between the Ras, and the Rac, Rho, and Cdc42 signaling pathways will be required to elucidate the mechanisms underlying these interactions.

Isolation of string and eyes absent, two known genes that have not been implicated previously in RTK signaling events:
Among the remaining seven pathway relevant candidates are two genes of known function that have not been implicated previously in RTK/Ras-mediated signaling events. One of these groups, SY3-4, is allelic to the phosphatase string, a Drosophila homologue of the yeast cell cycle gene cdc25 (EDGAR and O'FARRELL 1989 ). string regulates the G2-M transition in dividing cells by dephosphorylating and activating the cdc2 kinase, thereby allowing formation of cyclin/cdc2 complexes that promote S phase (EDGAR and O'FARRELL 1990 ). On the basis of the direction of interaction with yan^ACT, string would be postulated to be an antagonist of Ras signaling. Previous suggestions of a possible antagonistic relationship between string and Ras signaling came from a screen for modifiers of roughex, a negative regulator of G1 progression in the developing eye, which identified string as a suppressor and ras1 as an enhancer (THOMAS et al. 1994 ).

yan itself has been implicated in cell cycle control (ROGGE et al. 1995 ). Whereas hypomorphic yan mutations are semi-viable and have an extra photoreceptor phenotype (LAI and RUBIN 1992 ), null mutations in yan are embryonic lethal, with the embryos dying as a result of overproliferation of cells in the dorsal neuroectoderm (ROGGE et al. 1995 ). Thus, depending on the developmental context, yan regulates not only the transition between undifferentiated and differentiated cell types, but also the choice between differentiation and cell division. Recovery of string alleles in our screen could reflect cross-talk between cell cycle and differentiation pathways that occurs in part at the level of transcriptional regulation. Thus, it is possible that the downstream targets of yan will include cell cycle regulators such as string or that yan expression and stability may be linked to cell cycle controls. Alternatively, string could have postmitotic functions essential to differentiation.

To our knowledge, string has not been isolated in any other Ras pathway screens; however, it was isolated, again as a suppressor, in a screen for modifiers of activated Notch (VERHEYEN et al. 1996 ). The direction of interaction suggests string acts as a positive regulator of Notch signaling. Expression of N^ACT and yan^ACT have similar developmental consequences in that both inhibit or delay differentiation of the cell types in which they are expressed (FORTINI et al. 1993 ; REBAY and RUBIN 1995 ). Isolation of string in both screens could indicate a point of cross-talk between the Notch and the RTK/Ras pathway. Alternatively, isolation of string as a suppressor of both Notch^ACT and yan^ACT could have more to do with the similar terminal phenotype of these two backgrounds rather than reflecting direct interactions with the two pathways. Supporting the first hypothesis, cross-talk between the Notch and RTK pathways has been reported by numerous labs (CAGAN and READY 1989 ; ROGGE et al. 1995 ; KARIM et al. 1996 ; VERHEYEN et al. 1996 ; PRICE et al. 1997 ). Despite all the genetic interaction data, the mechanisms whereby the Notch and RTK pathways intersect remain to be determined. Experiments designed to study signaling by both pathways in vivo have suggested an antagonistic relationship (DE CELIS and BRAY 1997 ; PRICE et al. 1997 ; GREENWALD 1998 ; MILLER and CAGAN 1998 ), which would be consistent with string acting as a negative regulator of Ras signal transduction and a positive regulator of Notch signal transduction.

The second gene with a previously defined function that may be relevant to the RTK pathway on the basis of our genetic tests is eyes absent (eya; BONINI et al. 1993 ). eya encodes a novel nuclear protein of unknown function that has been shown recently by several labs to function in a hierarchy of "master eye regulatory genes" that are required to specify and promote differentiation of eye tissue (CHEN et al. 1997 ; PIGNONI et al. 1997 ; HALDER et al. 1998 ). However, on the basis of expression pattern and phenotypes, it is possible that eya plays additional roles in development independent of its role in determining competence to become eye tissue (BONINI et al. 1998 ; LEISERSON et al. 1998 ). One possibility is that eya could be directly complexed with yan, and could direct its transcriptional repressor activity in certain tissues. However, preliminary yeast two-hybrid experiments have failed to indicate yan-eya protein-protein interactions (A. WILLIAMS and I. REBAY, unpublished results). An alternate possibility to be investigated is transcriptional regulation of eya by yan. Given the genetic interactions we have observed between eya and yan^ACT, it will be interesting to investigate the possible role of eya in RTK/yan-mediated signaling events in the embryo and developing eye. It could be that in order to differentiate as eye tissue, a developing cell must receive both a "general" differentiation signal from the RTK pathway and a more specific eye fate specification signal.

Isolation of potential new components of the RTK/Ras/Yan signaling pathway:
Four of the remaining five pathway relevant complementation groups, EY2-5, EY2-6, EY3-5, and SY2-2, are in the initial stages of characterization. Potential roles for these genes include downstream transcriptional targets of yan, upstream kinases or phosphatases, proteins directly complexed with yan that regulate its activity, or elements of intersecting signaling pathways that cross-talk with the Ras pathway at the level of yan (Figure 4). We are particularly interested in SY2-2 as it is one of the few suppressor groups recovered in the screen that passed both genetic tests, strongly suppressing GMR-yan^ACT, Sev-yan^ACT, and yan^S2382 (Figure 2C and Figure D; Table 1). Based on the direction of our screen, suppressor mutations are expected to be negative regulators of Ras signal transduction. In addition, both alleles of SY2-2 interact very strongly, in the expected direction, with all Ras pathway backgrounds tested except GMR-sina and GMR-ttk (Table 1). Therefore, because of the unique nature of SY2-2 in the screen and because of the strength and consistency of its genetic behavior, a primary focus in the future will be to determine what it encodes. To date, complementation analyses with available Ras pathway and other candidate genes have been uninformative.

View larger version (14K): In this window In a new window Download PPT slide   Figure 4. Potential functions for yan-interacting genes isolated in the screen. A simplified RTK pathway is depicted. Arrows indicate a positive direction of interaction, and blunt arrows indicate a negative direction of interaction. Enhancers of yanACT [E(yan)], predicted to be positively acting components of the pathway, could potentially function as downstream targets of yan's transcriptional activity, as cofactors binding to yan that regulate its transcriptional activity, as components of other intersecting signaling pathways that feed into the RTK pathway at the level of yan, as other downstream targets of MAPK that function in parallel or in conjunction with yan, or as regulators of the RTK/Ras/MAPK cascade upstream of yan.

spen encodes a novel member of the RRM family of RNA-binding proteins:
We have cloned the fifth pathway relevant complementation group, spen (EY2-7), and found that it encodes a member of the RRM family of RNA-binding proteins (BURD and DREYFUSS 1994 ). The RRM sequence is an 80–100-amino-acid stretch of sequence that forms an RNA-binding domain (RBD). This motif is characterized by a bipartite RNP consensus sequence that forms the RNA-binding surface of the domain, as well as several predominantly hydrophobic amino acids dispersed throughout the RRM that are essential for the overall structure of the domain. RRM-containing proteins mediate a variety of post-transcriptional RNA-processing reactions including mRNA splicing, stability, and transport. In Drosophila, RNA-processing reactions and the RRM-containing proteins that mediate them are critical to multiple developmental processes. For example, in sex determination, sex-specific alternate splicing reactions are regulated by the RRM-containing proteins tra2 and Sxl (BELL et al. 1988 ; AMREIN et al. 1994 ). Establishment of dorsoventral polarity in the oocyte requires input from the RRM protein squid to localize the mRNA encoding the Egfr ligand gurken (MATUNIS et al. 1994 ). Another RRM protein, elav, is expressed in all Drosophila neurons and is required for determination and maintenance of the neuronal fate, although the precise mechanism of function is not clear (KOUSHIKA et al. 1996 ; YANNONI and WHITE 1997 ).

spen has three RRM motifs in tandem within the first 1000 amino acids of the protein. Database searches identify a number of predicted proteins from either genomic sequencing efforts or EST projects that have very similar RRMs. For these proteins, the homology within the RRM region extends beyond the conserved residues that define the motif. Homology between spen and other RRM-containing proteins is less striking and is limited to the residues that comprise the motif (data not shown). Because the RRM domain is thought to mediate specificity of RNA-binding interaction, it is possible that the spen class of RRM proteins interacts with similar substrates (BURD and DREYFUSS 1994 ). Further experiments will be required to determine whether spen actually binds RNA, what its in vivo targets are, and what the function of this activity is in the context of RTK signaling events during development.

Two additional structural motifs of note in the spen protein are the C-terminal ~170 amino acids and a region of predicted coiled-coil near the middle of the protein. The coiled-coil domain is likely to mediate protein-protein interactions, either with spen itself or with other proteins (BERGER et al. 1995 ; WOLF et al. 1997 ). One possibility is that spen interacts directly with yan via this domain. Alternatively, the interaction between spen and yan could be less direct. Homology searches using the C terminus of the spen detect several proteins, several of which, including a Drosophila protein predicted from the BDGP genomic sequence and a human protein compiled from multiple overlapping ESTs, also have spen-class RRMs. Although the structure of the C terminus of spen is not homologous to any protein domain of known function, the high degree of conservation of this region between fly, worm, and mammalian proteins suggests it is likely to have a conserved function. Apart from the RRM, coiled-coil, and C-terminal domains, the rest of the spen protein is novel, and to date it is uninformative in terms of providing clues as to function. Because the protein is so large (5476 amino acids), structure-function analyses and determination of the molecular lesions associated with the various spen alleles will be required to determine whether there are other important functional domains not detected by sequence homology. Investigation of the function of spen homologues in other species may also be informative.

Our isolation of spen as an enhancer of yan^ACT suggests it may play a role as a positive regulator of the RTK/Ras pathway. Preliminary results indicate spen is a nuclear protein broadly expressed in most tissues and enriched in neuronal lineages (F. CHEN and I. REBAY, unpublished results). We currently do not know whether spen functions upstream or downstream of yan. One possibility is that spen might regulate the stability of the yan transcript. It has been postulated that the mechanism for downregulating yan activity involves post-translational modifications of the protein, namely phosphorylation by activated MAPK, that subsequently targets yan for degradation. Such post-translational regulation of yan would presumably need to be reinforced at the transcriptional and/or translational level. Thus, spen might play a role in destabilizing yan mRNA in response to Ras signaling. This would be consistent with our isolating mutations in spen as enhancers of yan^ACT. Alternatively, spen could be transcriptionally regulated by yan, and could play a role in splicing, stability, or transport of other downstream effector genes. Future phenotypic, genetic, and biochemical characterization of spen will be necessary to understand its role in Ras/yan signaling events.

In conclusion, our genetic screen has identified several genes that are capable of modulating Ras/MAPK/yan-mediated signaling events. Among these genes are known RTK pathway elements, thereby validating the screening approach and suggesting that those genes that had not been implicated previously in RTK signal transduction may also be relevant to the pathway. Our hope is that further analysis of these new genes will improve our understanding of RTK pathway function in different contexts throughout development, and will shed light on the mechanisms whereby the RTK-mediated signals are integrated with other developmental signals to effect coordinated and context-appropriate cellular responses.

	ACKNOWLEDGMENTS

We thank Amy Beaton, Nancy Bonini, Felix Karim, Tom Neufeld, Elizabeth O'Neill, Robert Saint, Amy Tang, Esther Verheyen, David Wassarman, and the Bloomington Stock Center for generously providing Drosophila stocks. We acknowledge the Berkeley Drosophila Genome Project for providing genomic sequence in the EY2-7 region. We thank everyone in the Rubin lab for help and advice throughout the beginning of this project. This manuscript was improved with the comments of Terry Orr-Weaver and Rick Fehon. I.R. is a recipient of a Burroughs Wellcome Fund Career Award in the Biomedical Sciences and is a Rita Allen Foundation Scholar. This work was supported in part by National Institutes of Health grant GM-33135 to G.M.R.

Manuscript received July 19, 1999; Accepted for publication October 11, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMREIN, H., M. L. HEDLEY, and T. MANIATIS, 1994  The role of specific protein-RNA and protein-protein interactions in positive and negative control of pre-mRNA splicing by Transformer 2. Cell 76:735-746[Medline].

BELL, L. R., E. M. MAINE, P. SCHEDL, and T. W. CLINE, 1988  Sex-lethal, a Drosophila sex determination switch gene, exhibits sex-specific RNA splicing and sequence similarity to RNA binding proteins. Cell 55:1037-1046[Medline].

BERGER, B., D. B. WILSON, E. WOLF, T. TONCHEV, and M. MILLA et al., 1995  Predicting coiled coils by use of pairwise residue correlations. Proc. Natl. Acad. Sci. USA 92:8259-8263[Abstract/Free Full Text].

BERGMANN, A., J. AGAPITE, K. MCCALL, and H. STELLER, 1998  The Drosophila gene hid is a direct molecular target of Ras-dependent survival signaling. Cell 95:331-341[Medline].

BIGGS, W. H. R., K. H. ZAVITZ, B. DICKSON, A. VAN DER STRATEN, and D. BRUNNER et al., 1994  The Drosophila rolled locus encodes a MAP kinase required in the sevenless signal transduction pathway. EMBO J. 13:1628-1635[Medline].

BONINI, N. M., W. M. LEISERSON, and S. BENZER, 1993  The eyes absent gene: genetic control of cell survival and differentiation in the developing Drosophila eye. Cell 72:379-395[Medline].

BONINI, N. M., Q. T. BUI, G. L. GRAY-BOARD, and J. M. WARRICK, 1997  The Drosophila eyes absent gene directs ectopic eye formation in a pathway conserved between flies and vertebrates. Development 124:4819-4826[Abstract].

BONINI, N. M., W. M. LEISERSON, and S. BENZER, 1998  Multiple roles of the eyes absent gene in Drosophila. Dev. Biol. 196:42-57[Medline].

BRUNNER, D., K. DUCKER, N. OELLERS, E. HAFEN, and H. SCHOLZ et al., 1994  The ETS domain protein pointed-P2 is a target of MAP kinase in the sevenless signal transduction pathway. Nature 370:386-389[Medline].

BUFF, E., A. CARMENA, S. GISSELBRECHT, F. JIMENEZ, and A. M. MICHELSON, 1998  Signalling by the Drosophila epidermal growth factor receptor is required for the specification and diversification of embryonic muscle progenitors. Development 125:2075-2086[Abstract].

BURD, C. G. and G. DREYFUSS, 1994  Conserved structures and diversity of functions of RNA-binding proteins. Science 265:615-621[Abstract/Free Full Text].

CAGAN, R. L. and D. F. READY, 1989  Notch is required for successive cell decisions in the developing Drosophila retina. Genes Dev. 3:1099-1112[Abstract/Free Full Text].

CHARDIN, P., J. H. CAMONIS, N. W. GALE, L. VAN AELST, and J. SCHLESSINGER et al., 1993  Human Sos1: a guanine nucleotide exchange factor for Ras that binds to GRB2. Science 260:1338-1343[Abstract/Free Full Text].

CHEN, R., M. AMOUI, Z. ZHANG, and G. MARDON, 1997  Dachshund and eyes absent proteins form a complex and function synergistically to induce ectopic eye development in Drosophila. Cell 91:893-903[Medline].

DE CELIS, J. F. and S. BRAY, 1997  Feed-back mechanisms affecting Notch activation at the dorsoventral boundary in the Drosophila wing. Development 124:3241-3251[Abstract].

DICKSON, B. and E. HAFEN, 1994  Genetics of signal transduction in invertebrates. Curr. Opin. Genet. Dev. 4:64-70[Medline].

DICKSON, B. J., A. VAN DER STRATEN, M. DOMINGUEZ, and E. HAFEN, 1996  Mutations modulating Raf signaling in Drosophila eye development. Genetics 142:163-171[Abstract].

DUFFY, J. B. and N. PERRIMON, 1994  The torso pathway in Drosophila: lessons on receptor tyrosine kinase signaling and pattern formation. Dev. Biol. 166:380-395[Medline].

DUFFY, J. B., D. A. HARRISON, and N. PERRIMON, 1998  Identifying loci required for follicular patterning using directed mosaics. Development 125:2263-2271[Abstract].

EDGAR, B. A. and P. H. O'FARRELL, 1989  Genetic control of cell division patterns in the Drosophila embryo. Cell 57:177-187[Medline].

EDGAR, B. A. and P. H. O'FARRELL, 1990  The three postblastoderm cell cycles of Drosophila embryogenesis are regulated in G2 by string. Cell 62:469-480[Medline].

ELLIS, M. C., E. M. O'NEILL, and G. M. RUBIN, 1993  Expression of Drosophila glass protein and evidence for negative regulation of its activity in non-neuronal cells by another DNA-binding protein. Development 119:855-865[Abstract].

ENGSTROM, L., E. NOLL, and N. PERRIMON, 1997  Paradigms to study signal transduction pathways in Drosophila. Curr. Top. Dev. Biol. 35:229-261[Medline].

FORTINI, M. E., M. A. SIMON, and G. M. RUBIN, 1992  Signalling by the sevenless protein tyrosine kinase is mimicked by Ras1 activation. Nature 355:559-561[Medline].

FORTINI, M. E., I. REBAY, L. A. CARON, and S. ARTAVANIS-TSAKONAS, 1993  An activated Notch receptor blocks cell-fate commitment in the developing Drosophila eye. Nature 365:555-557[Medline].

GREENWALD, I., 1998  LIN-12/Notch signaling: lessons from worms and flies. Genes Dev. 12:1751-1762[Free Full Text].

GREGORY, S. L., R. D. KORTSCHAK, B. KALIONIS, and R. SAINT, 1996  Characterization of the dead ringer gene identifies a novel, highly conserved family of sequence-specific DNA-binding proteins. Mol. Cell. Biol. 16:792-799[Abstract].

HALDER, G., P. CALLAERTS, S. FLISTER, U. WALLDORF, and U. KLOTER et al., 1998  Eyeless initiates the expression of both sine oculis and eyes absent during Drosophila compound eye development. Development 125:2181-2191[Abstract].

HAY, B. A., T. WOLFF, and G. M. RUBIN, 1994  Expression of baculovirus P35 prevents cell death in Drosophila.. Development 120:2121-2129[Abstract].

HEBERLEIN, U., T. WOLFF, and G. M. RUBIN, 1993  The TGF beta homolog dpp and the segment polarity gene hedgehog are required for propagation of a morphogenetic wave in the Drosophila retina. Cell 75:913-926[Medline].

HOEY, T., R. O. WEINZIERL, G. GILL, J. L. CHEN, and B. D. DYNLACHT et al., 1993  Molecular cloning and functional analysis of Drosophila TAF110 reveal properties expected of coactivators. Cell 72:247-260[Medline].

KARIM, F. D., H. C. CHANG, M. THERRIEN, D. A. WASSARMAN, and T. LAVERTY et al., 1996  A screen for genes that function downstream of Ras1 during Drosophila eye development. Genetics 143:315-329[Abstract].

KLAMBT, C., 1993  The Drosophila gene pointed encodes two ETS-like proteins which are involved in the development of the midline glial cells. Development 117:163-176[Abstract/Free Full Text].

KOCKEL, L., J. ZEITLINGER, L. M. STASZEWSKI, M. MLODZIK, and D. BOHMANN, 1997  Jun in Drosophila development: redundant and nonredundant functions and regulation by two MAPK signal transduction pathways. Genes Dev. 11:1748-1758[Abstract/Free Full Text].

KOLODKIN, A. L., A. T. PICKUP, D. M. LIN, C. S. GOODMAN, and U. BANERJEE, 1994  Characterization of Star and its interactions with sevenless and EGF receptor during photoreceptor cell development in Drosophila. Development 120:1731-1745[Abstract].

KOLODZIEJ, P. A., L. Y. JAN, and Y. N. JAN, 1995  Mutations that affect the length, fasciculation, or ventral orientation of specific sensory axons in the Drosophila embryo. Neuron 15:273-286[Medline].

KOUSHIKA, S. P., M. J. LISBIN, and K. WHITE, 1996  ELAV, a Drosophila neuron-specific protein, mediates the generation of an alternatively spliced neural protein isoform. Curr. Biol. 6:1634-1641[Medline].

LAI, Z. C. and G. M. RUBIN, 1992  Negative control of photoreceptor development in Drosophila by the product of the yan gene, an ETS domain protein. Cell 70:609-620[Medline].

LEE, T. and D. J. MONTELL, 1997  Multiple Ras signals pattern the Drosophila ovarian follicle cells. Dev. Biol. 185:25-33[Medline].

LEISERSON, W. M., B. SEYMOUR, and N. M. BONINI, 1998  Dual functions of the Drosophila eyes absent gene in the eye and embryo. Mech. Dev. 73:193-202[Medline].

MARSHALL, C. J., 1994  MAP kinase kinase kinase, MAP kinase kinase and MAP kinase. Curr. Opin. Genet. Dev. 4:82-89[Medline].

MARUTA, H. and A. W. BURGESS, 1994  Regulation of the Ras signalling network. Bioessays 16:489-496[Medline].

MATUNIS, E. L., R. KELLEY, and G. DREYFUSS, 1994  Essential role for a heterogeneous nuclear ribonucleoprotein (hnRNP) in oogenesis: hrp40 is absent from the germ line in the dorsoventral mutant squid. Proc. Natl. Acad. Sci. USA 91:2781-2784[Abstract/Free Full Text].

MILLER, D. T. and R. L. CAGAN, 1998  Local induction of patterning and programmed cell death in the developing Drosophila retina. Development 125:2327-2335[Abstract].

NEUFELD, T. P., A. H. TANG, and G. M. RUBIN, 1998  A genetic screen to identify components of the sina signaling pathway in Drosophila eye development. Genetics 148:277-286[Abstract/Free Full Text].

NIMNUAL, A. S., B. A. YATSULA, and D. BAR-SAGI, 1998  Coupling of Ras and Rac guanosine triphosphatases through the Ras exchanger Sos. Science 279:560-563[Abstract/Free Full Text].

NOSELLI, S., 1998  JNK signaling and morphogenesis in Drosophila. Trends Genet. 14:33-38[Medline].

NUSSLEIN-VOLHARD, C., E. WIESCHAUS, and H. KULDING, 1984  Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. Roux's Arch Dev. Biol. 193:267-282.

O'NEILL, E. M., I. REBAY, R. TJIAN, and G. M. RUBIN, 1994  The activities of two Ets-related transcription factors required for Drosophila eye development are modulated by the Ras/MAPK pathway. Cell 78:137-147[Medline].

PICKUP, A. T. and U. BANERJEE, 1999  The role of Star in the production of an activated ligand for the EGF receptor signaling pathway. Dev. Biol. 205:254-259[Medline].

PIGNONI, F. and S. L. ZIPURSKY, 1997  Induction of Drosophila eye development by decapentaplegic. Development 124:271-278[Abstract].

PIGNONI, F., B. HU, K. H. ZAVITZ, J. XIAO, and P. A. GARRITY et al., 1997  The eye-specification proteins So and Eya form a complex and regulate multiple steps in Drosophila eye development. Cell 91:881-891[Medline].

PIRROTTA, V., 1986 Cloning Drosophila genes, pp. 83–110 in Drosophila: A Practical Approach, edited by D. R. ROBERTS. Oxford University Press, Washington, DC.

PRICE, J. V., E. D. SAVENYE, D. LUM, and A. BREITKREUTZ, 1997  Dominant enhancers of Egfr in Drosophila melanogaster: genetic links between the Notch and Egfr signaling pathways. Genetics 147:1139-1153[Abstract].

REBAY, I. and G. M. RUBIN, 1995  Yan functions as a general inhibitor of differentiation and is negatively regulated by activation of the Ras1/MAPK pathway. Cell 81:857-866[Medline].

RIESGO-ESCOVAR, J. R. and E. HAFEN, 1997  Drosophila Jun kinase regulates expression of decapentaplegic via the ETS-domain protein Aop and the AP-1 transcription factor DJun during dorsal closure. Genes Dev. 11:1717-1727[Abstract/Free Full Text].

ROGGE, R., P. J. GREEN, J. URANO, S. HORN-SABAN, and M. MLODZIK et al., 1995  The role of yan in mediating the choice between cell division and differentiation. Development 121:3947-3958[Abstract].

ROGGE, R. D., C. A. KARLOVICH, and U. BANERJEE, 1991  Genetic dissection of a neurodevelopmental pathway: son of sevenless functions downstream of the sevenless and EGF receptor tyrosine kinases. Cell 64:39-48[Medline].

RUBIN, G. M., 1988  Drosophila melanogaster as an experimental organism. Science 240:1453-1459[Abstract/Free Full Text].

SAMAKOVLIS, C., N. HACOHEN, G. MANNING, D. C. SUTHERLAND, and K. GUILLEMIN et al., 1996  Development of the Drosophila tracheal system occurs by a series of morphologically distinct but genetically coupled branching events. Development 122:1395-1407[Abstract].

SCHNORR, J. D. and C. A. BERG, 1996  Differential activity of Ras1 during patterning of the Drosophila dorsoventral axis. Genetics 144:1545-1557[Abstract].

SCHOLZ, H., E. SADLOWSKI, A. KLAES, and C. KLAMBT, 1997  Control of midline glia development in the embryonic Drosophila CNS. Mech. Dev. 62:79-91[Medline].

SCHWEITZER, R. and B. Z. SHILO, 1997  A thousand and one roles for the Drosophila EGF receptor. Trends Genet. 13:191-196[Medline].

SHILO, B. Z., 1992  Roles of receptor tyrosine kinases in Drosophila development. FASEB J. 6:2915-2922[Abstract].

SIMON, M. A., D. D. BOWTELL, G. S. DODSON, T. R. LAVERTY, and G. M. RUBIN, 1991  Ras1 and a putative guanine nucleotide exchange factor perform crucial steps in signaling by the sevenless protein tyrosine kinase. Cell 67:701-716[Medline].

SOISSON, S. M., A. S. NIMNUAL, M. UY, D. BAR-SAGI, and J. KURIYAN, 1998  Crystal structure of the Dbl and pleckstrin homology domains from the human Son of sevenless protein. Cell 95:259-268[Medline].

THOMAS, B. J., D. A. GUNNING, J. CHO, and L. ZIPURSKY, 1994  Cell cycle progression in the developing Drosophila eye: roughex encodes a novel protein required for the establishment of G1. Cell 77:1003-1014[Medline].

TOMLINSON, A. and D. READY, 1986  Neuronal differentiation in the Drosophila ommatidium. Dev. Biol. 120:366-376.

TOMLINSON, A., D. D. BOWTELL, E. HAFEN, and G. M. RUBIN, 1987  Localization of the sevenless protein, a putative receptor for positional information, in the eye imaginal disc of Drosophila. Cell 51:143-150[Medline].

TREISMAN, J. E. and G. M. RUBIN, 1995  wingless inhibits morphogenetic furrow movement in the Drosophila eye disc. Development 121:3519-3527[Abstract].

VAN DER GEER, P., T. HUNTER, and R. A. LINDBERG, 1994  Receptor protein-tyrosine kinases and their signal transduction pathways. Annu. Rev. Cell Biol. 10:251-337.

VERHEYEN, E. M., K. J. PURCELL, M. E. FORTINI, and S. ARTAVANIS-TSAKONAS, 1996  Analysis of dominant enhancers and suppressors of activated Notch in Drosophila. Genetics 144:1127-1141[Abstract].

WANG, H., I. CLARK, P. R. NICHOLSON, I. HERSKOWITZ, and D. J. STILLMAN, 1990  The Saccharomyces cerevisiae SIN3 gene, a negative regulator of HO, contains four paired amphipathic helix motifs. Mol. Cell. Biol. 10:5927-5936[Abstract/Free Full Text].

WOLF, E., P. S. KIM, and B. BERGER, 1997  MultiCoil: a program for predicting two- and three-stranded coiled coils. Protein Sci. 6:1179-1189[Medline].

YANNONI, Y. M. and K. WHITE, 1997  Association of the neuron-specific RNA binding domain-containing protein ELAV with the coiled body in Drosophila neurons. Chromosoma 105:332-341[Medline].

ZIPURSKY, S. L. and G. M. RUBIN, 1994  Determination of neuronal cell fate: lessons from the R7 neuron of Drosophila.. Annu. Rev. Neurosci. 17:373-397[Medline].

	NOTE ADDED IN PROOF

: The isolation and sequence of spen was recently reported by E. L. WIELLETTE, K. W. HARDING, K. A. MACE, M. R. RONSHAUGEN, F. Y. WANG and W. MCGINNIS (1999, spen encodes an RNP motif protein that interacts with Hox pathways to repress the development of head-like sclerites in the Drosophila trunk. Development 126: 5373–5385).

This article has been cited by other articles:

				L. H. Jin, J. K. Choi, B. Kim, H. S. Cho, J. Kim, J. Kim-Ha, and Y.-J. Kim Requirement of Split ends for Epigenetic Regulation of Notch Signal-Dependent Genes during Infection-Induced Hemocyte Differentiation Mol. Cell. Biol., March 15, 2009; 29(6): 1515 - 1525. [Abstract] [Full Text] [PDF]

				G. D. Raffel, G. C. Chu, J. L. Jesneck, D. E. Cullen, R. T. Bronson, O. A. Bernard, and D. G. Gilliland Ott1 (Rbm15) Is Essential for Placental Vascular Branching Morphogenesis and Embryonic Development of the Heart and Spleen Mol. Cell. Biol., January 15, 2009; 29(2): 333 - 341. [Abstract] [Full Text] [PDF]

				Y. Song, L. Fee, T. H. Lee, and R. P. Wharton The Molecular Chaperone Hsp90 Is Required for mRNA Localization in Drosophila melanogaster Embryos Genetics, August 1, 2007; 176(4): 2213 - 2222. [Abstract] [Full Text] [PDF]

				A. Simonsen, R. C. Cumming, K. Lindmo, V. Galaviz, S. Cheng, T. E. Rusten, and K. D. Finley Genetic Modifiers of the Drosophila Blue Cheese Gene Link Defects in Lysosomal Transport With Decreased Life Span and Altered Ubiquitinated-Protein Profiles Genetics, June 1, 2007; 176(2): 1283 - 1297. [Abstract] [Full Text] [PDF]

				G. D. Raffel, T. Mercher, H. Shigematsu, I. R. Williams, D. E. Cullen, K. Akashi, O. A. Bernard, and D. G. Gilliland Ott1(Rbm15) has pleiotropic roles in hematopoietic development PNAS, April 3, 2007; 104(14): 6001 - 6006. [Abstract] [Full Text] [PDF]

				S. H. Eun, K. Lea, E. Overstreet, S. Stevens, J.-H. Lee, and J. A. Fischer Identification of Genes That Interact With Drosophila liquid facets Genetics, March 1, 2007; 175(3): 1163 - 1174. [Abstract] [Full Text] [PDF]

				K. L. Pepple, A. E. Anderson, B. J. Frankfort, and G. Mardon A Genetic Screen in Drosophila for Genes Interacting With senseless During Neuronal Development Identifies the Importin moleskin Genetics, January 1, 2007; 175(1): 125 - 141. [Abstract] [Full Text] [PDF]

				J. Jemc and I. Rebay Characterization of the split ends-Like Gene spenito Reveals Functional Antagonism Between SPOC Family Members During Drosophila Eye Development Genetics, May 1, 2006; 173(1): 279 - 286. [Abstract] [Full Text] [PDF]

				E. Hiriart, H. Gruffat, M. Buisson, I. Mikaelian, S. Keppler, P. Meresse, T. Mercher, O. A. Bernard, A. Sergeant, and E. Manet Interaction of the Epstein-Barr Virus mRNA Export Factor EB2 with Human Spen Proteins SHARP, OTT1, and a Novel Member of the Family, OTT3, Links Spen Proteins with Splicing Regulation and mRNA Export J. Biol. Chem., November 4, 2005; 280(44): 36935 - 36945. [Abstract] [Full Text] [PDF]

				R. D. Read, P. J. Goodfellow, E. R. Mardis, N. Novak, J. R. Armstrong, and R. L. Cagan A Drosophila Model of Multiple Endocrine Neoplasia Type 2 Genetics, November 1, 2005; 171(3): 1057 - 1081. [Abstract] [Full Text] [PDF]

				A. L. Adamson, N. Wright, and D. R. LaJeunesse Modeling Early Epstein-Barr Virus Infection in Drosophila melanogaster: The BZLF1 Protein Genetics, November 1, 2005; 171(3): 1125 - 1135. [Abstract] [Full Text] [PDF]

				L. W. Cheng, J. P. M. Viala, N. Stuurman, U. Wiedemann, R. D. Vale, and D. A. Portnoy Use of RNA interference in Drosophila S2 cells to identify host pathways controlling compartmentalization of an intracellular pathogen PNAS, September 20, 2005; 102(38): 13646 - 13651. [Abstract] [Full Text] [PDF]

				M. A. Gerber, A. Shilatifard, and J. C. Eissenberg Mutational Analysis of an RNA Polymerase II Elongation Factor in Drosophila melanogaster Mol. Cell. Biol., September 1, 2005; 25(17): 7803 - 7811. [Abstract] [Full Text] [PDF]

				M. Mutsuddi, B. Chaffee, J. Cassidy, S. J. Silver, T. L. Tootle, and I. Rebay Using Drosophila to Decipher How Mutations Associated With Human Branchio-Oto-Renal Syndrome and Optical Defects Compromise the Protein Tyrosine Phosphatase and Transcriptional Functions of Eyes Absent Genetics, June 1, 2005; 170(2): 687 - 695. [Abstract] [Full Text] [PDF]

				M. Y. Zhu, R. Wilson, and M. Leptin A Screen for Genes That Influence Fibroblast Growth Factor Signal Transduction in Drosophila Genetics, June 1, 2005; 170(2): 767 - 777. [Abstract] [Full Text] [PDF]

				S. J. Silver and I. Rebay Signaling circuitries in development: insights from the retinal determination gene network Development, January 1, 2005; 132(1): 3 - 13. [Abstract] [Full Text] [PDF]

				A. H. Ludewig, C. Kober-Eisermann, C. Weitzel, A. Bethke, K. Neubert, B. Gerisch, H. Hutter, and A. Antebi A novel nuclear receptor/coregulator complex controls C. elegans lipid metabolism, larval development, and aging Genes & Dev., September 1, 2004; 18(17): 2120 - 2133. [Abstract] [Full Text] [PDF]

				O. L. Sierra, S.-L. Cheng, A. P. Loewy, N. Charlton-Kachigian, and D. A. Towler MINT, the Msx2 Interacting Nuclear Matrix Target, Enhances Runx2-dependent Activation of the Osteocalcin Fibroblast Growth Factor Response Element J. Biol. Chem., July 30, 2004; 279(31): 32913 - 32923. [Abstract] [Full Text] [PDF]

				M. G. Voas and I. Rebay The Novel Plant Homeodomain Protein Rhinoceros Antagonizes Ras Signaling in the Drosophila Eye Genetics, December 1, 2003; 165(4): 1993 - 2006. [Abstract] [Full Text] [PDF]

				B. Cen, A. Selvaraj, R. C. Burgess, J. K. Hitzler, Z. Ma, S. W. Morris, and R. Prywes Megakaryoblastic Leukemia 1, a Potent Transcriptional Coactivator for Serum Response Factor (SRF), Is Required for Serum Induction of SRF Target Genes Mol. Cell. Biol., September 15, 2003; 23(18): 6597 - 6608. [Abstract] [Full Text] [PDF]

				S. Szuplewski, B. Kottler, and R. Terracol The Drosophila bZIP transcription factor Vrille is involved in hair and cell growth Development, August 15, 2003; 130(16): 3651 - 3662. [Abstract] [Full Text] [PDF]

				M. Ariyoshi and J. W.R. Schwabe A conserved structural motif reveals the essential transcriptional repression function of Spen proteins and their role in developmental signaling Genes & Dev., August 1, 2003; 17(15): 1909 - 1920. [Abstract] [Full Text] [PDF]

				H. V. Lin, D. B. Doroquez, S. Cho, F. Chen, I. Rebay, and K. M. Cadigan Splits ends is a tissue/promoter specific regulator of Wingless signaling Development, July 15, 2003; 130(14): 3125 - 3135. [Abstract] [Full Text] [PDF]

				T. L. Tootle, P. S. Lee, and I. Rebay CRM1-mediated nuclear export and regulated activity of the Receptor Tyrosine Kinase antagonist YAN require specific interactions with MAE Development, March 1, 2003; 130(5): 845 - 857. [Abstract] [Full Text] [PDF]

				H. K. Shamloula, M. P. Mbogho, A. C. Pimentel, Z. M. A. Chrzanowska-Lightowlers, V. Hyatt, H. Okano, and T. R. Venkatesh rugose (rg), a Drosophila A kinase Anchor Protein, Is Required for Retinal Pattern Formation and Interacts Genetically With Multiple Signaling Pathways Genetics, June 1, 2002; 161(2): 693 - 710. [Abstract] [Full Text] [PDF]

				R. K. Smith, P. M. Carroll, J. D. Allard, and M. A. Simon MASK, a large ankyrin repeat and KH domain-containing protein involved in Drosophila receptor tyrosine kinase signaling Development, January 1, 2002; 129(1): 71 - 82. [Abstract] [Full Text] [PDF]

				N. Aoyagi and D. A. Wassarman Developmental and Transcriptional Consequences of Mutations in Drosophila TAFII60 Mol. Cell. Biol., October 15, 2001; 21(20): 6808 - 6819. [Abstract] [Full Text] [PDF]

				D. R. LaJeunesse, B. M. McCartney, and R. G. Fehon A Systematic Screen for Dominant Second-Site Modifiers of Merlin/NF2 Phenotypes Reveals an Interaction With blistered/DSRF and scribbler Genetics, June 1, 2001; 158(2): 667 - 679. [Abstract] [Full Text] [PDF]

				T. Mercher, M. B.-L. Coniat, R. Monni, M. Mauchauffe, F. N. Khac, L. Gressin, F. Mugneret, T. Leblanc, N. Dastugue, R. Berger, et al. Involvement of a human gene related to the Drosophila spen gene in the recurrent t(1;22) translocation of acute megakaryocytic leukemia PNAS, May 8, 2001; 98(10): 5776 - 5779. [Abstract] [Full Text] [PDF]

				J. L. Hartman IV, B. Garvik, and L. Hartwell Principles for the Buffering of Genetic Variation Science, February 9, 2001; 291(5506): 1001 - 1004. [Abstract] [Full Text] [PDF]

				M. A. Su, R. G. Wisotzkey, and S. J. Newfeld A Screen for Modifiers of decapentaplegic Mutant Phenotypes Identifies lilliputian, the Only Member of the Fragile-X/Burkitt's Lymphoma Family of Transcription Factors in Drosophila melanogaster Genetics, February 1, 2001; 157(2): 717 - 725. [Abstract] [Full Text]

				F Wittwer, A van der Straten, K Keleman, B. Dickson, and E Hafen Lilliputian: an AF4/FMR2-related protein that controls cell identity and cell growth Development, January 3, 2001; 128(5): 791 - 800. [Abstract] [PDF]

				A. Tang, T. Neufeld, G. Rubin, and H. Muller Transcriptional regulation of cytoskeletal functions and segmentation by a novel maternal pair-rule gene, lilliputian Development, January 3, 2001; 128(5): 801 - 813. [Abstract] [PDF]

				L. Wan, T. C. Dockendorff, T. A. Jongens, and G. Dreyfuss Characterization of dFMR1, a Drosophila melanogaster Homolog of the Fragile X Mental Retardation Protein Mol. Cell. Biol., November 15, 2000; 20(22): 8536 - 8547. [Abstract] [Full Text]

				F. Chanut, A. Luk, and U. Heberlein A Screen for Dominant Modifiers of roDom, a Mutation That Disrupts Morphogenetic Furrow Progression in Drosophila, Identifies Groucho and Hairless as Regulators of atonal Expression Genetics, November 1, 2000; 156(3): 1203 - 1217. [Abstract] [Full Text]

				A. M. Huang and G. M. Rubin A Misexpression Screen Identifies Genes That Can Modulate RAS1 Pathway Signaling in Drosophila melanogaster Genetics, November 1, 2000; 156(3): 1219 - 1230. [Abstract] [Full Text]

				M. Therrien, D. K. Morrison, A. M. Wong, and G. M. Rubin A Genetic Screen for Modifiers of a Kinase Suppressor of Ras-Dependent Rough Eye Phenotype in Drosophila Genetics, November 1, 2000; 156(3): 1231 - 1242. [Abstract] [Full Text]

				B. Yun, K. Lee, R. Farka, C. Hitte, and L. Rabinow The LAMMER Protein Kinase Encoded by the Doa Locus of Drosophila Is Required in Both Somatic and Germline Cells and Is Expressed as Both Nuclear and Cytoplasmic Isoforms Throughout Development Genetics, October 1, 2000; 156(2): 749 - 761. [Abstract] [Full Text]

				B Kuang, S. Wu, Y Shin, L Luo, and P Kolodziej split ends encodes large nuclear proteins that regulate neuronal cell fate and axon extension in the Drosophila embryo Development, January 4, 2000; 127(7): 1517 - 1529. [Abstract] [PDF]