odd-skipped is expressed in multiple tissues during Drosophila embryogenesis.

The odd-skipped (odd) gene encodes a zinc finger protein that represses other segmentation genes in the early Drosophila embryo. Though odd is initially expressed in a striped pattern that reflects its function within the segmentation hierarchy, it is also expressed in a variety of patterns during later stages of embryogenesis. To identify the cells and tissues that correspond to these latter patterns, we examined the distribution of the Odd protein at all embryonic stages. Our results indicate that Odd is a specific and persistent marker for subsets of cells in developing mesoderm, ectoderm, and neural tissue. We conclude that Odd is a useful tool for studying cell specification, cell migrations and morphogenetic movements during organogenesis of the heart, gut and central nervous system.

Comparison of the structure and expression of odd-skipped and two related genes that encode a new family of zinc finger proteins in Drosophila.

The odd-skipped (odd) gene, which was identified on the basis of a pair-rule segmentation phenotype in mutant embryos, is initially expressed in the Drosophila embryo in seven pair-rule stripes, but later exhibits a segment polarity-like pattern for which no phenotypic correlate is apparent. We have molecularly characterized two embryonically expressed odd-cognate genes, sob and bowel (bowl), that encode proteins with highly conserved C2H2 zinc fingers. While the Sob and Bowl proteins each contain five tandem fingers, the Odd protein lacks a fifth (C-terminal) finger and is also less conserved among the four common fingers. Reminiscent of many segmentation gene paralogues, the closely linked odd and sob genes are expressed during embryogenesis in similar striped patterns; in contrast, the less-tightly linked bowl gene is expressed in a distinctly different pattern at the termini of the early embryo. Although our results indicate that odd and sob are more likely than bowl to share overlapping developmental roles, some functional divergence between the Odd and Sob proteins is suggested by the absence of homology outside the zinc fingers, and also by amino acid substitutions in the Odd zinc fingers at positions that appear to be constrained in Sob and Bowl

bowel, an odd-skipped homolog, functions in the terminal pathway during Drosophila embryogenesis.

The terminal genes of Drosophila specify non-segmented regions of the larval body that are derived from the anterior and posterior regions of the early embryo. Terminal class genes include both maternal-effect loci (typified by the receptor tyrosine kinase torso) that encode components of a signal transduction cascade and zygotic genes (e.g. tailless and huckebein) that are transcribed at the poles of the embryo in response to the local activation of the pathway. We have characterized a zygotic gene, bowel, that was identified as a zinc finger homolog of the pair-rule segmentation gene odd-skipped. bowel transcripts are initially expressed at both poles of the blastoderm embryo and in a single cephalic stripe. This pattern depends upon torso and tailless activity, but is not affected in huckebein mutants. We isolated and sequenced five mutations that affect the bowel protein, including a nonsense mutation upstream of the zinc fingers and a missense mutation in a putative zinc-chelating residue. bowel mutants die as late embryos with defects in terminal derivatives including the hindgut and proventriculus. Our results indicate that the developmental roles of odd-skipped and bowel have diverged substantially, and that bowel represents a new member of the terminal hierarchy that acts downstream of tailless and mediates a subset of tailless functions in the posterior of the embryo.

Molecular analysis of odd-skipped, a zinc finger encoding segmentation gene with a novel pair-rule expression pattern.

odd-skipped (odd) is one of eight known pair-rule genes that establish portions of alternating segments during Drosophila embryogenesis; odd mutant embryos exhibit pattern defects in anterior regions of odd-numbered segments. P element transposon tagging was used to clone 25 kb of DNA from the odd genomic region. Molecular analysis of phenotypic revertants confirmed that the P element used to tag the locus was responsible for the corresponding odd mutation, and significant structural changes were identified in two additional odd mutants. Several cDNA clones derived from a 2.2 kb embryonic transcript were isolated and the longest was sequenced. The predicted odd protein of 392 amino acids is highly basic and contains four tandem Cys-Cys/His-His zinc finger repeats, consistent with a presumed function for odd as a DNA binding protein and transcriptional regulator. In situ hybridization analysis indicated that odd transcripts accumulate in a dynamic pattern during early embryogenesis, with two temporally distinct modes of expression. The first mode results in a 'pair-rule' pattern of seven stripes at the blastoderm stage, representing the expected double segment periodicity. During gastrulation, the seven primary stripes are supplemented by secondary stripes which appear in alternate segments, resulting in the equivalent labeling of every segment in the extended germ band. Similar double to single segment transitions have now been reported for four of the six pair-rule genes analyzed.

Gene activities and segmental patterning in Drosophila: analysis of odd-skipped and pair-rule double mutants.

The odd-skipped (odd) gene is required to generate anterior regions of the odd-numbered segments in Drosophila; homozygous embryos show pattern deletions that are always less than a segment in width and are associated with mirror-image duplications of adjacent regions. To define further the role of odd and determine how it interacts with other segmentation genes, we have described the effects of combining odd with mutations at other pair-rule loci. We have observed phenotypic suppression in double-mutant combinations with even-skipped (eve), paired, sloppy paired, and engrailed (en). In the most thoroughly characterized combination (odd eve), both naked cuticle and specific denticle rows are restored that would normally have been deleted by one of the two mutants alone. In the odd en double mutant, we observe nearly complete suppression of the odd phenotype, such that the mirror image duplications are eliminated and the odd-numbered denticle bands are restored. We conclude that the requirements of pattern elements for specific gene activities are not absolute, and propose mechanisms by which these genes interact to specify cell fates.

A mutation in the largest subunit of RNA polymerase II alters RNA chain elongation in vitro.

An in vitro transcription system which utilized a semisynthetic DNA template (Kadesch, T. R., and Chamberlin, M. J. (1982) J. Biol. Chem. 257, 5286-5295) was developed and used to compare RNA chain elongation by wild type and mutant RNA polymerases II of Drosophila. With this template, all of the active polymerases rapidly initiated RNA chains at synthetic single-stranded sites at the ends of the DNA, and then entered a long (15 to 20 min) period of elongation through duplex regions of template before any measurable termination occurred. A comparison of wild type and mutant polymerase activities during this elongation phase indicated that a mutation to amanitin resistance reduces the rate at which the enzyme elongates transcripts. The reduced elongation rate of the mutant was associated with an altered substrate Km. Because the polymerase II mutation is in the largest enzyme subunit (Greenleaf, A. L. (1983) J. Biol. Chem. 258, 13403-13406), these results demonstrate a functional role for this subunit during RNA chain elongation.

Immunological studies of RNA polymerase II using antibodies to subunits of Drosophila and wheat germ enzyme.

We induced goat antibodies to Drosophila RNA polymerase II and rabbit antibodies to the isolated 215,000-dalton and 140,000-dalton polymerase II subunits (P215 and P140, respectively). Similarly, we induced rabbit antibodies to wheat germ RNA polymerase II and to the 220,000-dalton subunit and 140,000-dalton subunit (P220 and P140, respectively). Anti-polymerase antibodies precipitated the homologous native enzyme and inhibited its activity in vitro, while several of the anti-subunit sera did neither. The anti-Drosophila P215 serum specifically labeled RNA polymerase II fixed in situ on polytene chromosomes. We reacted the antibodies with polymerase subunits separated by sodium dodecyl sulfate gel electrophoresis and electrophoretically transferred to nitrocellulose ("protein blotting"). Each antibody to whole polymerase reacted with multiple subunits, while the anti-subunit sera each reacted specifically with the subunit employed as immunogen. The anti-subunit sera also cross-reacted with the analogous subunit from several heterologous polymerases II (from yeast, wheat germ, Drosophila, and calf thymus), demonstrating shared subunit-specific determinants in polymerase II from widely divergent organisms. The anti-polymerase sera also showed cross-reactivity with subunits of heterologous enzymes, but only in one case did the cross-reactivity involve subunits other than the two largest ones. Specifically, the goat anti-Drosophila polymerase serum displayed easily detectable cross-reactivity with four low molecular weight subunits of calf thymus polymerase II, providing a unique demonstration of antigenic relatedness of small RNA polymerase II subunits from different higher eukaryotes.

Properties of mutationally altered RNA polymerases II of Drosophila.

We tested and compared several in vitro properties of wild type and mutant RNA polymerases II from Drosophila melanogaster, using several different mutants of a single X-linked genetic locus, RpIIC4 (Greenleaf, A. L., Weeks, J. R., Voelker, R. A., Ohnishi, S., and Dickson, B. (1980) Cell 21, 785-792); the mutants tested included the original amanitin-resistant mutant, C4, which is nonconditional, plus the temperature-sensitive mutants A9, C20, E28, and 1Fb40. Using a tritium-labeled amanitin derivative, we demonstrated that C4 polymerase has a reduced binding affinity for amanitin. The C4 polymerase was as stable to thermal denaturation as the wild type enzyme, and the two enzymes had similar specific activities, ionic strength and Mn2+ requirements, and apparent Km values for UTP and GTP when assayed in the presence of Mn2+. However, with Mg2+ as the divalent cation, C4 polymerase was less active than wild type and had 2-fold higher apparent Km values for UTP and GTP. Three of the temperature-sensitive mutants, A9, C20, and E28, were derived from the amanitin-resistant mutant C4; the polymerase II activities from these mutants displayed resistance to alpha-amanitin in vitro identical with that of the C4 enzyme. C20, E28, and 1Fb40 polymerases were markedly less stable to thermal denaturation in vitro than wild type polymerase. The results presented indicate that the mutations at the RNA polymerase locus (RpIIC4-) directly alter the structure of the enzyme, providing conclusive evidence that the locus is a structural gene for a polymerase II subunit.

Alpha-amanitin-resistant D. melanogaster with an altered RNA polymerase II.

Following EMS mutagenesis we recovered a mutant of D. melanogaster that grows at concentrations of alpha-amanitin lethal to wild-type. To our knowledge this mutant represents the first example of an amanitin-resistant eucaryotic organism. The amanitin resistance of the mutant (AmaC4) is due to an alteration in its DNA-dependent RNA polymerase II, which is approximately 250 times less sensitive to inhibition by amanitin than the wild-type polymerase II whether tested in nuclei, in partially-fractionated extracts or as a highly purified enzyme. While the wild-type enzyme activity is inhibited 50% by 2.1 x 10(-8) M alpha-amanitin, inhibition of 50% of the AmaC4 RNA polymerase II activity requires a toxin concentration of 5.6 x 10(-6) M. The mutation responsible for the amanitin resistance of AmaC4 is on the X chromosome near the vermillion locus.