Salivary gland determination in Drosophila: a salivary-specific, fork head enhancer integrates spatial pattern and allows fork head autoregulation.

In the early Drosophila embryo, a system of coordinates is laid down by segmentation genes and dorsoventral patterning genes. Subsequently, these coordinates must be interpreted to define particular tissues and organs. To begin understanding this process for a single organ, we have studied how one of the first salivary gland genes, fork head (fkh), is turned on in the primordium of this organ, the salivary placode. A placode-specific fkh enhancer was identified 10 kb from the coding sequence. Dissection of this enhancer showed that the apparently homogeneous placode is actually composed of at least four overlapping domains. These domains appear to be developmentally important because they predict the order of salivary invagination, are evolutionarily conserved, and are regulated by patterning genes that are important for salivary development. Three dorsoventral domains are defined by EGF receptor (EGFR) signaling, while stripes located at the anterior and posterior edges of the placode depend on wingless signaling. Further analysis identified sites in the enhancer that respond either positively to the primary activator of salivary gland genes, SEX COMBS REDUCED (SCR), or negatively to EGFR signaling. These results show that fkh integrates spatial pattern directly, without reference to other early salivary gland genes. In addition, we identified a binding site for FKH protein that appears to act in fkh autoregulation, keeping the gene active after SCR has disappeared from the placode. This autoregulation may explain how the salivary gland maintains its identity after the organ is established. Although the fkh enhancer integrates information needed to define the salivary placode, and although fkh mutants have the most extreme effects on salivary gland development thus far described, we argue that fkh is not a selector gene for salivary gland development and that there is no master, salivary gland selector gene. Instead, several genes independently sense spatial information and cooperate to define the salivary placode.

The Drosophila Pax gene eye gone is required for embryonic salivary duct development.

What are the developmental mechanisms required for conversion of an undifferentiated, two-dimensional field of cells into a patterned, tubular organ? In this report, we describe the contribution of the Drosophila Pax gene eye gone to the development of the embryonic salivary glands and ducts. eye gone expression in salivary tissues is controlled by several known regulators of salivary fate. After the initial establishment of the salivary primordium by Sex combs reduced, fork head excludes eye gone expression from the pregland cells so that its salivary expression is restricted to the posterior preduct cells. trachealess, in contrast, activates eye gone expression in the posterior preduct cells. We have previously described the process by which fork head and the EGF receptor pathway define the border between the gland and duct primordia. Here we show that eye gone is required for the subdivision of the duct primordium itself into the posterior individual duct and the anterior common duct domains. In the absence of eye gone, individual ducts as well as the precursor of the adult salivary glands, the imaginal ring, are absent. We took advantage of this ductless phenotype to show that Drosophila larvae do not have an obligate requirement for salivary glands and ducts. In addition to its role in the salivary duct, eye gone is required in the embryo for the development of the eye-antennal imaginal disc and the chemosensory antennal organ.

The Tec29 tyrosine kinase is required during Drosophila embryogenesis and interacts with Src64 in ring canal development.

Tec29 encodes the only known Drosophila member of the Tec tyrosine kinases. By identifying the first mutations in Tec29 (formerly Src29A), we show that it is essential for head involution during embryogenesis and for ring canal development during oogenesis. Tec29 mutant egg chambers are defective in transfer of cytoplasm from the accessory nurse cells through the ring canals into the oocyte. Growth of the mutant ring canals is arrested, and they lack the strong phosphotyrosine localization seen in wild-type ring canals. Mutants lacking the Drosophila Src homolog Src64 show the same phenotype, and we show that Src64 is required for the localization of Tec29 to the ring canals. This interaction is similar to that between vertebrate Src and Tec kinases and suggests that Tec29 is an effector of Src64 that modifies ring canal components required for growth.

Salivary duct determination in Drosophila: roles of the EGF receptor signalling pathway and the transcription factors fork head and trachealess.

Organogenesis in Drosophila embryos begins at 4-5 hours of development as the expression of organ-specific genes is initiated. The salivary primordium, which occupies the ventral epidermis of parasegment 2, is among the earliest to be defined. It is soon divided into two distinct regions: the more dorsal pregland cells and the more ventral preduct cells. We show that it is the opposing activities of the Drosophila EGF receptor (DER) signaling pathway and the Fork head transcription factor that distinguish these cell types and set up the boundary between them. DER signaling acts ventrally to block fork head expression in the preduct cells, thereby restricting gland identity to the more dorsal cells. Fork head in turn blocks expression of duct-specific genes in the pregland cells, thereby restricting duct identity to the more ventral cells. A third regulatory activity, the Trachealess transcription factor, is also required to establish the identity of the preduct cells, but we show that it acts independently or downstream from the DER:fork head confrontation. In trachealess mutants, subdivision of the salivary primordium occurs normally and the dorsal cells form glands, but the ventral cells are undetermined. We present a model proposing that trachealess is the crucial duct-specific gene that Fork head represses to distinguish pregland from preduct cells.

The Broad-Complex directly controls a tissue-specific response to the steroid hormone ecdysone at the onset of Drosophila metamorphosis.

In Drosophila, all of the major metamorphic transitions are regulated by changes in the titer of the steroid hormone ecdysone. Here we examine how a key regulator of metamorphosis and primary ecdysone response gene, the Broad-Complex, transmits the hormonal signal to one of its targets, the Sgs-4 glue gene. We show that Broad-Complex RNAs accumulate in mid third instar larval salivary glands prior to Sgs-4 induction, as expected for the products of a gene that regulates the timing of Sgs-4 activation. The Broad-Complex codes for a family of zinc finger transcriptional regulators. We have identified a number of binding sites for these proteins in sequences known to regulate the timing of Sgs-4 induction, and have used these sites to derive a binding consensus for each protein. Some of these binding sites are required in vivo for Sgs-4 activity. In addition, rbp+, a genetically defined Broad-Complex function that is required for Sgs-4 induction, acts through these Broad-Complex binding sites. Thus, the Broad-Complex directly mediates a temporal and tissue-specific response to ecdysone as larvae become committed to metamorphosis.

A transcriptional switch between the Pig-1 and Sgs-4 genes of Drosophila melanogaster.

Pig-1 and Sgs-4 are a pair of closely linked and divergently transcribed Drosophila melanogaster genes, which are both expressed in larval salivary glands but at different times during development. While Sgs-4 is expressed at high levels only at the end of the third instar, Pig-1 exhibits a major peak of expression during late second and early third instar. Thus, Pig-1 expression declines as Sgs-4 expression is induced. In this paper, we show that three adjacent elements located within the short region between these genes can account for the switch from Pig-1 to Sgs-4 expression. A 170-bp segment acts as an enhancer to direct Sgs-4 expression in late-third-instar salivary glands. A 64-bp sequence located just upstream from the enhancer can modify its temporal specificity so that it works throughout the third instar. Expression induced at mid-third instar by a combination of these two elements can be repressed by a negative regulatory sequence located still further upstream. We present evidence suggesting that the changing interactions between these regulatory elements and the Sgs-4 and Pig-1 promoters lead to the correct pattern of expression of the two genes.

Organogenesis in Drosophila melanogaster: embryonic salivary gland determination is controlled by homeotic and dorsoventral patterning genes.

We have investigated Drosophila salivary gland determination by examining the effects of mutations in pattern forming genes on the salivary gland primordium. We find that the anterior-posterior extent of the primordium, a placode of columnar epithelial cells derived from parasegment 2, is established by the positive action of the homeotic gene Sex combs reduced (Scr). Embryos mutant for Scr lack a detectable placode, while ectopic Scr expression leads to the formation of ectopic salivary glands. In contrast, the dorsal-ventral extent of the placode is regulated negatively. Functions dependent on the decapentaplegic product place a dorsal limit on the placode, while dorsal-dependent genes act to limit the placode ventrally. We propose a model in which these pattern forming genes act early to determine the salivary gland anlage by regulating the expression of salivary gland determining genes, which in turn control genes that are involved in salivary gland morphogenesis.

Characterization of the memory gene dunce of Drosophila melanogaster.

The dunce (dnc) gene of Drosophila melanogaster encodes cAMP phosphodiesterase (PDEase) and is required for learning/memory and female fertility. The gene is structurally complex, demonstrated in part by Northern blotting experiments which detected multiple RNAs ranging in size from 4.2 to 9.6 kb (1 kb = 10(3) bases or base-pairs). To characterize these RNAs and to understand their sequence heterogeneity, we isolated and analyzed 29 new and independent cDNA clones representing the dnc RNAs. Restriction mapping, hybridization analysis and sequence determination of these cDNA clones and the corresponding genomic exons resolved these into six different classes. Exons defined by the cDNA clones are distributed over more than 148 kb of genomic DNA, with some exons being used alternatively among the RNAs. The RNAs are transcribed from at least three initiation sites: two of these were mapped by parallel S1-nuclease and primer extension experiments. In addition, some of the heterogeneity is generated by using varying lengths of a 3'-untranslated trailer sequence. Altogether, the results indicate that the size and sequence heterogeneity of dnc transcripts results from transcription initiation at multiple sites, alternative splicing, and processes which generate different 3' ends. The existence of multiple protein products is suggested by the alternative use of exons which code for portions of the open reading frame. The protein variation potentially includes N-terminal differences coded for by transcript-specific 5' exons and internal differences arising from the optional inclusion of a 39 base-pair exon and from the alternative use of two 3' splice sites separated by six base-pairs. Expression of a cDNA clone in yeast containing a large portion of the open reading frame produced cAMP PDEase activity identical in properties to the Drosophila enzyme affected by the dnc mutation. The results suggest that the remarkable structural complexity of dnc may reflect an intricate control of the spatial and/or temporal expression of various isoforms of cAMP PDEase.

Noncoordinate expression of Drosophila glue genes: Sgs-4 is expressed at many stages and in two different tissues.

The glue genes of Drosophila melanogaster comprise a family of genes expressed at high levels in the salivary glands of late third instar larvae in response to the insect hormone ecdysone. We present evidence that, in contrast to the other glue genes, Sgs-4 is turned on throughout Drosophila development and is not expressed exclusively in the larval salivary glands. Larvae transformed with an Sgs-4/Adh (alcohol dehydrogenase) hybrid gene exhibit Sgs-4-directed Adh expression in the larval proventriculus as well as in the salivary glands as early as the first instar. Sgs-4-specific RNA can be detected at very low levels during all stages of development. During late third instar, levels of Sgs-4 RNA in the salivary glands increase several-thousand-fold, thereby accounting for the large amounts of Sgs-4 protein present in the glue produced by the salivary glands. This pattern of expression is unique to the Sgs-4 gene. While expression of several of the other glue genes can be detected in embryos and early larvae, they appear to be expressed neither throughout development nor in the larval proventriculus. Appearance of the glue gene RNAs in mid third instar salivary glands is noncoordinate, even for the chromosomally clustered genes Sgs-3, Sgs-7, and Sgs-8.

Functional redundancy in the tissue-specific enhancer of the Drosophila Sgs-4 gene.

During the last day of larval development, the Sgs-4 glue gene of Drosophila melanogaster is expressed at high levels in a single tissue, the larval salivary glands. As shown by transformation experiments and by DNA sequence analysis of Sgs-4 underproducing strains, an essential regulatory region for Sgs-4 expression lies between 149 and 568 bp upstream from the transcribed part of the gene. This region shows the positional independence of a transcriptional enhancer and directs at least three regulatory activities: tissue specificity, developmental timing and high-level expression. Here we use a transient transformation assay to identify three elements within this enhancer that are involved in tissue specificity. For at least this regulatory activity the enhancer is internally redundant. Any pairwise combination of the three elements is sufficient to direct salivary gland expression, although none of the three can act alone.

At least two genes reside within a large intron of the dunce gene of Drosophila.

The dunce locus of Drosophila melanogaster is considered to house a gene involved in memory, because flies carrying lesions at the locus have shortened memory of several different conditioned behaviours. Our recent partial characterization of the gene at the molecular level, along with prior genetic and biochemical evidence, recently provides compelling evidence that the gene codes for the enzyme cAMP phosphodiesterase. The observation that the gene encodes at least six overlapping poly(A)+ RNA molecules ranging in size from 4.2 to 9.5 kilobases (kb) (ref. 8), suggests that the gene is extraordinarily complex. Here we provide the sequence of a dunce complementary DNA clone and the corresponding genomic coding regions which show that the organization of the gene is elaborate. The cDNA clone defines dunce exons which are separated by a large intron of 79 kb. More importantly, at least two other genes are shown to reside within the large intron, including the well-defined glue protein gene, Sgs-4. The location of dunce exons relative to the molecular breakpoints of chromosomal aberrations with defined cytological positions indicates that the dunce gene extends over more than five polytene chromosome bands.

Developmental regulation by an enhancer from the Sgs-4 gene of Drosophila.

A cis acting regulatory region has previously been identified 300-500 bp upstream of the Drosophila glue protein gene, Sgs-4. The functional capabilities of this region have now been examined by fusing it to the Drosophila Adh gene and determining the pattern of expression from the fused construct after transformation. The results show that the Sgs-4 sequences between -150 and -568 are able to direct Adh expression in late third-instar salivary glands, the appropriate tissue and timing for Sgs-4 expression. In addition, the Sgs-4 sequence elevates Adh expression in the anterior midgut and fat body, despite the fact that Sgs-4 is not normally expressed there. All three regulatory activities, tissue specificity, timing and enhancement, show the positional flexibility of enhancer elements. In addition, the Sgs-4 and Adh regulatory elements combine to direct expression in novel spatial/temporal combinations in which neither would normally be expressed.

The structure of hobo transposable elements and their insertion sites.

The hobo transposable elements of Drosophila form a family of 3.0-kb elements and their deletion derivatives. Their distribution is consistent with the model that 3.0-kb elements are functionally complete but that smaller hobos are defective and require complete elements in trans for transposition. The sequence of one 3.0-kb element is presented; it has several interesting features, including a 1.9-kb open reading frame downstream from potential TATA and CAT sequences. Comparison of 11 independent insertion sites shows that in every case the hobo element has integrated at and duplicated either the sequence NNNNNNAC or CTTTNNNN. There is evidence that an eight nucleotide sequence internal to hobo that matches both of these sequences has been used as an insertion site for a second hobo element, as the first step in the creation of an internal deletion derivative. Structural similarities between hobo and the eukaryotic transposable elements P, Ac, 1723, and Tam3, found in widely divergent host organisms, suggest that they all transpose by a common mechanism.

Cis-acting sequences which regulate expression of the Sgs-4 glue protein gene of Drosophila.

The Sgs-4 glue protein gene of Drosophila is expressed only in third-instar larval salivary glands. Previous work suggests that a regulatory region lies 5' and remote to the gene, as indicated by a region of tissue-specific DNase I hypersensitivity and by underproducing mutants with DNA lesions in the hypersensitive region. Here we demonstrate by germ line transformation of cloned fragments containing Sgs-4 that the sequences between 840 bp 5' and 130 bp 3' to the gene are sufficient for Sgs-4 activity. When 5' sequence was removed to -392, activity was eliminated, thereby verifying the existence of essential sequences far upstream. Fragments that are active include, in addition to the capacity for normal levels of expression, three other cis-acting regulatory activities: developmental timing, tissue specificity, and dosage compensation. In contrast, the fragments tested did not specify formation of the puff with which Sgs-4 is normally associated. As shown by chromosomal rearrangements, the region required for puffing is limited to 16-19 kb surrounding the gene.

A transposable element inserted just 5' to a Drosophila glue protein gene alters gene expression and chromatin structure.

The Drosophila Sgs-4 gene directs the developmentally regulated production of a glue protein in the salivary glands of mature larvae. Previous work suggests that Sgs-4 expression requires a remote upstream region that becomes hypersensitive to DNAase I digestion when the gene is active. Here we describe a variant Sgs-4 locus that has a 1.3 kb DNA insert separating the gene from the remote hypersensitive region. This insert defines a new family of transposable elements that we call hobo. Expression from the variant locus is reduced 50 to 100 fold, and rather than the one, normal Sgs-4 transcript, there are now four transcripts, two starting within the hobo element. These multiple transcripts are still expressed only in late larval salivary glands, implying that developmental regulation is unaffected by changes in the site of transcript initiation. When Sgs-4 is active, the remote DNAase I-hypersensitive region, now even more remote due to the insert, still forms over its normal sequences. In contrast, new hypersensitive sites form within hobo near the starts of the new transcripts.

DNA sequence changes in an upstream DNase I-hypersensitive region are correlated with reduced gene expression.

Previous experiments have identified a region that is required for the expression of the Drosophila glue protein gene Sgs-4 and is located 300-500 base pairs upstream from the structural gene. The chromatin in this region changes conformation and becomes hypersensitive to DNase I digestion when the gene becomes active, a change that apparently induces additional conformational changes near the site of transcription initiation. To learn more about the DNA sequence requirements for the function of this region, we analyzed three naturally occurring Sgs-4 under-producers. In two of these strains, a single base pair change within the hypersensitive region is correlated with a 50% reduction in the amount of Sgs-4 RNA produced. Another strain, which has multiple 5' lesions, is severely reduced in Sgs-4 expression and in the DNase hypersensitivity of the upstream region. Several of the sequence changes in this extreme underproducer lie near hypersensitive sites, suggesting that they inhibit the appearance of the normal DNase hypersensitive conformation.

Association of a Drosophila transposable element of the roo family with chromosomal deletion breakpoints.

A 9.3 kb transposable element of the roo family has been found inserted 3' to the Sgs-4 glue protein gene of Drosophila. The X chromosome which carries this insert also carries wDZL, a dominant, unstable allele of the white locus caused by the insertion of the 13 kb wDZL element. Three deletions isolated from the wDZL strain have molecular breakpoints 3' to Sgs-4 that are associated with the roo element. Though the deletions eliminate much of the DNA between white and Sgs-4, none of the distal breakpoints fall at or near the wDZL element. The results suggest that this copia-like element, which is structurally similar to an integrated retrovirus, is capable of promoting chromosomal deletions.

A complex of interacting DNAase I-hypersensitive sites near the Drosophila glue protein gene, Sgs4.

The chromatin structure adjacent to the Drosophila glue protein gene Sgs4 changes drastically when the gene is active. In nuclei from embryos or tissue culture cells in which Sgs4 is inactive, there are three DNAase I-hypersensitive sites 3' to the gene, but none near its 5' end. In the nuclei of late third instar salivary glands, Sgs4 is actively transcribed, and a complex of five DNAase I-hypersensitive sites appears 5' to the gene. Two of the sites are near the point of transcription initiation, at -70 and +30. The others are much farther from the gene at -330, -405 and -480 and are affected by small deletions: one deletion reduces expression about 50-fold and removes sequences corresponding to the -330 hypersensitive site; another makes no Sgs4 RNA and removes sequences corresponding to two hypersensitive sites, -405 and -480. Thus the hypersensitive sites, or DNA sequences within 50 bp of them, seem to be required for normal gene expression Distinct changes are seen in the chromatin from salivary glands of these mutant strains. The first strain is simply missing the -330 hypersensitive site, while the second is missing all of the tissue-specific 5' sites, even though sequences corresponding to three of them remain. This suggests that hierarchical interactions among the regions 5' to Sgs4 are required for its full expression. A sequence of 14 bp at the most prominent hypersensitive site (-405) is closely related to sequences 5' to several other eucaryotic genes.

Molecular limits on the size of a genetic locus in Drosophila melanogaster.

This report places outer limits on the size of the DNA region required for expression of a Drosophila gene. This region, termed the unit of expression, includes not only the structural gene but also any cis-acting sequences that modulate its activity. The locus we have chosen, Sgs-4, codes for one of the glue proteins secreted by larval Drosophila salivary glands. Cytological deficiencies have been identified that eliminate sequences on one side or the other of Sgs-4 without affecting its expression. The ends of these deficiencies have been localized accurately with respect to restriction endonuclease sites in and near the locus. These endpoints limit the Sgs-4 structural gene and essential flanking sequences to a 16- to 19-kilobase region of the X chromosome. The results also show that there is no DNA sequence rearrangement in the Sgs-4 region during development of either the polytene larval salivary glands or adult flies.