|
Mark D. Garfinkel
|
|
| Notice: | The abstracts below describe research in progress and should be regarded as privileged communications. They are not to be permanently saved-to-disk, printed, modified, or cited in research publications. Address all inquiries to Mark D. Garfinkel . |
| 1. Determination of the Consensus DNA-Binding Sites Recognized by the OVO Zinc Finger Domain | Sanggyu Lee |
| Abstract: | The ovo genetic function is implicated in regulating a hierarchy
of genes that together mediate viability and sex-appropriate differentiation
of female germline cells. Consistent with the genetic data is the observation
that the shavenbaby-ovo gene region produces overlapping transcripts
coding for at least three proteins that contain an identical set of four
zinc finger motifs at the carboxy termini. We cloned a partial copy of the
ovo cDNA that contains 179 codons, including all four zinc fingers,
into bacterial expression vectors. Overproduced protein was purified and
used to select for ovo-binding sites from a high-degeneracy oligonucleotide
starting population with PCR amplification of protein-bound DNA oligomers.
Gel-mobility shift assays confirmed that 21 oligomers with OVO-binding properties
were recovered. DNA sequencing and computational analysis identified a nine-base-long
consensus sequence shared among the bound oligomers. This finding has intriguing
implications for the nature of the OVO zinc finger-DNA recognition interaction. References: Lee, S. (1998) Ph.D. dissertation, Illinois Institute of Technology; S. Lee and M.D. Garfinkel (1999a) Characterization of Drosophila OVO protein-DNA binding specificity using random DNA oligomers suggests zinc finger degeneration. Nucl. Acids Res., in press. |
| 2. In vitro Dissection of the Germline Sex Determination Hierarchy | Sanggyu Lee |
| Abstract: | We used bacterially expressed OVO zinc finger domain in gel mobility shift
assays and DNase footprinting with promoter fragments from five different
Drosophila genes involved in oogenesis. OVO protein binds multiple
binding sites in close proximity to the transcription start sites of its
own germline promoters and that of the otu gene. In addition, OVO
protein binds to promoter sequences of two other genes that genetically
interact with ovo mutations, and to a fifth gene whose expression
is sex-regulated in the germline. The 13 OVO binding sites thus revealed
reduce to a consensus sequence that is nearly identical to the consensus
derived from oligomer selection. In four of the five genes more than one
binding site was observed, and these sites were generally arrayed as interrupted
direct repeats. Since gene transcription and zinc finger binding
are both asymmetric with respect to DNA, this has clear implications for
the mode of OVO activity on target promoters. References: Lee, S. (1998) Ph.D. dissertation, Illinois Institute of Technology; S. Lee and M.D. Garfinkel (1999b) Manuscript submitted for publication. |
| 3. X-Ray Crystallography of the OVO DNA-Binding Domain and its Complexes with DNA | Gregory J. Sahli, Andrew Howard, Mark D. Garfinkel |
| Abstract: | The unexpected finding that the OVO protein recognizes a nine-base-wide target sequence, at least in vitro, suggests that one of the four zinc fingers might be non-functional for DNA binding. We are purifying the 179-amino-acid carboxy terminal fragment of the OVO protein and plan to determine its atomic-level structure by X-ray crystallography. |
| 4. In vitro Mutagenesis of the OVO Protein | Gerrie Farman, David Campbell |
| Abstract: | We are using site-directed mutagenesis and bacterial overexpression to manufacture and purify mutant proteins lacking individual zinc fingers for in vitro assays of DNA binding. Germline transformation of Drosophila with mutant proteins will allow assessment of their biological activity in vivo. |
| 5. Functional Domains of the OVO Proteins | Kyung-Won Park |
| Abstract: | The ovo genetic function is associated with two protein isoforms, OVO-B and OVO-A, that possess the same carboxy terminal DNA-binding domain. Published data suggests that these isoforms have antagonistic effects on transcription from two different target genes. We are fusing different portions of the OVO isoforms to the lexA DNA-binding domain for expression in yeast cells that carry lexO::lacZ, UASG::lexO::lacZ and lexO::LEU2 reporter genes to identify which protein segments possess "activator domain" and "repressor domain" properties. |
| 6. In vivo Dissection of the Germline Sex Determination Hierarchy | Akram Abou-Zied |
| Abstract: | We plan to assess the physiological significance of OVO binding to various germline promoters by making site-directed mutations in the binding sites, and using the resulting promoters to drive E. coli lacZ expression in genetically transformed Drosophila. |
| 7. Genetic Approaches to trans-Acting Regulatory Loci Affecting Expression of the ovo Germline-Specific Promoter | Demet Sag Ozkol |
| Abstract: | In 1994, Oliver et al. reported that the germline promoter apparently
responsible for the ovo function, when used to drive expression of
E. coli lacZ, was transcribed at high levels in germline cells possessing
the 2X:2A karyotype characteristic of female, and at low levels in germline
cells possessing the XY:2A karyotype characteristic of male. This differential
transcription did not depend upon karyotype-appropriate somatic sexual
differentiation: chromosomal females transformed into phenotypic males through
the action of transformer mutations retained high levels of ovo
promoter activity in germline cells. Likewise, chromosomal males transformed
into phenotypic females through the action of hs-tra cDNA retained
low levels of ovo promoter activity. These results indicate a germline-cell-autonomous
role for X:A ratio in determining ovo germline promoter activity.
Previous studies (Schupbach, 1982, 1985) had also shown a germline-cell-autonomous
role for the X:A ratio in determining gamete sexual differentiation; however
the X-chromosome counting mechanism in germline cells apparently uses different
"numerator" loci than those operating in the soma. We used fly
strains carrying ovo::lacZ fusion genes as in vivo reporters
for identifying X-chromosome regions required for proper expression of the
ovo germline promoter. Males carrying ovo::lacZ fusion genes
are mated to deletion/balancer females from each of the Bloomington "X-chromosome
deficiency kit" strains, and ovaries from female progeny heterozygous
for each deletion (experimentals) and female progeny carrying the balancer
chromosome (controls) are dissected. These ovaries are used for histochemical
staining reactions with X-gal, and for preparation of soluble extracts for
a highly sensitive and quantitative spectrophotometric lacZ activity
assay. Six X-chromosomal regions appear to contain trans-acting positive
regulators of ovo promoter activity, and may identify germline X-numerator
genes; another five X-chromosomal regions appear, unexpectedly, to contain
trans-acting negative regulators of ovo promoter activity.
References: Oliver, B., J. Singer, V. Laget, G. Pennetta and D. Pauli (1994) Development 120: 3185-3195; Schupbach, T. (1982) Developmental Biology 89: 117-127; Schupbach, T. (1985) Genetics 109: 529-548. |
| 8. Yeast Expression Cloning Strategies for Identification of Drosophila Target Genes Regulated by the OVO-B Protein | Ira Sigar |
| Abstract: | Genetic evidence suggests that the ovo function may be required for
transcription regulation of a gene hierarchy involved in germline sex determination.
Several genes (i.e., ovarian tumor, sans fille, Sex-lethal,
bag-of-marbles and orb) are candidates for being downstream
targets of that regulation. Other experiments underway (see Abstract 2,
6) are designed, ultimately, to test whether the known genetic downstream
loci are direct biochemical targets for the ovo function. This may
not be the case; instead, one or more presently unknown intermediary genes
may be regulated directly by the ovo function and these may regulate
expression of the known targets. Mastick et al. (1995) have developed a general method of cloning genes and gene fragments that contain sites recognized by sequence-specific transcription factors. In their technique, yeast cells are used as hosts for two types of expression vectors. From one plasmid, expression of an arbitrary Drosophila transcription factor is induced by choice of growth medium. The other plasmid is used to clone a library of small fragments of Drosophila genomic DNA adjacent to a selectable prototrophic gene, which is expressed only if the adjoining Drosophila DNA fragment contains a binding site recognized by the inducibly expressed heterologous transcription factor. Library clones containing a large number of different Drosophila genomic DNA fragments inserted next to the selectable marker are transformed with the plasmid expressing the desired transcription factor. On suitable medium, a minority of clones should grow and these should contain Drosophila genomic DNA fragments recognized by the Drosophila transcription factor. We are constructing yeast expression vectors harboring the entire cDNA open reading frame of each protein type coded by the svb-ovo gene region. Using this method, we should be able to identify fragments of the Drosophila genome that contain DNA-binding sites recognized by the Drosophila OVO protein. The resulting molecular probes will be used to isolate entire Drosophila genes whose expression is modulated by the ovo protein and which are thus likely to be required for sex-dimorphic aspects of germline function. Reference: Mastick, G.S., R. McKay, T. Oligino, K. Donovan and A.J. López (1995) Genetics 139: 349-353. |
| 9. Ectopic Expression in Larval Salivary Glands as a Tool for Identifying Transcription Factor Target Genes | Akram Abou-Zied, Katie J. Woolley, Mark D. Garfinkel |
| Abstract: | The polytene chromosomes of the late larval salivary gland afford a unique
opportunity for identifying genomic targets for DNA-binding proteins. Specific
immunolocalization of proteins bound to their target sites was first detected
over twenty years ago, in the pioneering studies of S.C.R. Elgin and colleagues.
More recently, Zink and Paro (1989) and Chinwalla et al. (1995) have
shown specific binding by Polycomb and trithorax to previously
known and putative-new target genes. In all these cases, the polytene-bound
protein was expressed from endogenous genes that are normally transcribed
in the third-instar larval salivary glands. To broaden the scope of this
method, we have designed and built constructs that should allow expression
of OVO-derived fusion proteins under the control of the salivary gland secretion
(glue) gene, Sgs-3. This gene is abundantly transcribed during the
third instar, and its protein accumulates to nearly 10% of gland protein.
sgs-3 transcriptional control sequences are modular, with cleanly
separable enhancers and a single promoter that can be combined to systematically
vary reporter expression over a 100-fold range (Roark et al. 1990).
OVO-binding to putative target genes will be observed by indirect immunofluorescence
with anti-OVO antisera, and GFP-tagged fusion proteins. References: Zink, B. and R. Paro (1989) Nature 337: 468-471; Roark et al. (1990) Developmental Biology 139: 121-133; Chinwalla et al. (1995) EMBO Journal 14: 2056-2065. |
| 10. Identification and Characterization of the Mid-Embryonic shavenbaby Promoter Region | James McNulty, Ahmed Mirza, Lorna Mosse |
| Abstract: | The svb function, required for proper morphogenesis of denticle belts
and dorsal hairs in the larval epidermis, apparently depends upon a ~7-kb
mRNA expressed during a brief, 3-hour, period near the middle of embryogenesis.
This period is particularly significant since the known pattern formation
genetic hierarchy operating in the early embryo (i.e., gap, pair-rule, and
segment-polarity genes) have essentially completed their function at this
time. The svb mRNA is thus a candidate for being one of the earliest
differentiation-specific functions, as distinct from determination functions,
transcribed in the developing larval epidermis. Previous results (Garfinkel et al., 1994) showed that the svb mRNA differed in structure from the 5-kb germline mRNAs that serve the ovo function in at least three ways: (i) an alternative polyadenylation site is used in the mid-embryonic mRNA causing a 1.1-kb-longer 3' untranslated region; (ii) an alternative RNA splicing donor at the 3' end of Exon 2, causing a 534-nt-shorter open reading frame; (iii) an uncharacterized 5' leader exon. Size estimates of the mid-embryonic mRNA suggests that the "Exon 0" could be as long as 1.5-kb, certainly sufficient to introduce a novel in-frame translation start codon, so that the SVB protein would have both a net deletion from the amino-terminal sequences and a net addition relative to the OVO proteins. In order to make this determination, we used two methods in attempts to clone the missing Exon 0. A high-complexity cDNA library cloned in lambda-gt11, made from 9-12-hr-old Drosophila embryos, the period at which the putative svb mRNA is most abundant, was screened by plaque-filter hybridization to recover clones homologous with appropriate svb-ovo region probes. In the second approach, high-complexity phage DNA purified from the same library was used as a template for polymerase chain reaction using a series of primers chosen from within Exon 2, and from the phage sequences surrounding the cloning site. Fragments of cDNA with expected properties (i.e., originating from genomic DNA distal to the proximal breakpoint of Df(1)biD2, and with embryonic tissue in situ hybridization patterns) are being recovered with both techniques, and will be subjected to further tests of their authenticity. In addition, several of us are determining the DNA sequence of the corresponding svb region genomic DNA, and in doing so are making deletion derivatives useful for in vivo promoter mapping experiments. One of our ultimate aims is to generate full-length svb cDNA minigenes for use in several Drosophila transformation-complementation experiments, and for use in biochemical and genetic tests of protein function using methods described in Abstracts 1 and 2. Another of our ultimate aims is a comprehensive analysis of how the svb promoter is regulated by embryonic pattern formation genes, and perhaps other regulatory loci as well. |
| 11. Sequence Analysis of Apparent Point Mutations Disrupting ovo and svb Functions | Sangita Parande, Mark D. Garfinkel |
| Abstract: | Mutations at the shavenbaby-ovo gene region fall into three classes: (I) mutations specific for the female-germline function called ovo, (II) mutations specific for the sex-independent ectodermal function called shavenbaby, and (III) mutations that simultaneously disrupt both genetic functions. By definition, mutations are alterations in gene structure that identify physiologically important components of gene product structure. Earlier studies (e.g., Garfinkel et al., 1992) used whole-genome Southern (WGS) blotting of genomic DNA to show that transposable element insertions accounted for many Class I and many Class III mutations, while a solitary Class II mutation was due to a WGS-detectable breakpoint of a cytologically visible chromosomal deletion. The different classes mapped to discrete locations of genomic DNA, in an interdigitated arrangement, which in combination with cDNA clone structures and RNA gel blot data supported the view of the shavenbaby-ovo gene region as comprised of complex, partly overlapping transcription units coding for at least two proteins that possess identical carboxy-terminal zinc finger motifs. The same studies also demonstrated that eight gamma-ray-induced Class III mutations were undetectable by WGS, implying that subtle defects in gene structure were able nonetheless to disrupt both genetic functions. In addition to the apparent point mutations induced by gamma irradiation, a series of ethyl methane sulfonate (EMS)-induced and diepoxybutane (DEB)-induced mutations are available, with members in each of the three phenotypic classes. If independent point mutations in each class cause amino acid substitutions that are clustered into "hot spots," they may define genetically separable, functionally distinct, domains of the protein sequences. To identify point mutations, we are employing denaturant-gradient gel electrophoresis and single-strand conformation polymorphism, techniques highly sensitive to even single-base alterations in DNA sequences. |
| 12. X-Ray Scattering Approaches to Indirect Flight Muscle Physiology | Thomas C. Irving, Tenley Archer, Gerrie Farman, David Campbell, Mark D. Garfinkel |
| Abstract: | We are using the BioCAT beamline at Argonne National Laboratory's Advanced Photon Source to generate X-ray scattering images of living, intact, flies at rest and in tethered flight. Insect indirect flight muscle (IFM), like vertebrate striated and cardiac muscles, is a highly ordered supramolecular assembly of protein filaments. IFM is composed of two opposing muscle types, the dorsal-longitudinal muscles (DLM) and the dorsoventral muscles (DVM). When a fly is carefully aligned to the path of tightly focussed monochromatic X-rays, each muscle type generates a diagnostic scattering patterns of spots, arcs, and lines. These images convey detailed structural information about the spacings between different protein filaments within the DVM or the DLM, and about the spacings between different protein subunits along each protein filament that constitute the muscle. Of particular interest are lattice spacing changes related to myosin cross-bridge motion. |
| 13. Radiation-Hardened Flies | Tamara Fedczyna, Micael Gustavsson, Thomas C. Irving, Mark D. Garfinkel |
| Abstract: | The BioCat beamline is capable of depositing over a trillion X-ray photons per second onto a square spot no larger than 100 microns on each side. When that intense X-ray beam is aimed at the IFM of an adult Drosophila fly, radiation damage is rapidly induced, limiting sharply the amount of physiological data that can be recovered during tethered flight. This radiation damage is likely due to induction of oxygen free radicals, which are highly reactive towards protein, nucleic acids and lipid bilayers. Using the Awesome Power of Drosophila Genetics[tm], we are developing fly strains that overproduce in the IFM two enzymes involved in oxygen free radical scavenging, superoxide dismutase (SOD) and catalase (CAT). PCR was used to site-direct-mutagenize the cDNAs encoding each of these enzymes, constructing cassettes that can easily be inserted downstream of the strong IFM-specific promoter-enhancer fragment of the Actin 88F gene. Genetically engineered flies overexpressing SOD and CAT, either singly or together, will be subjected to a variety of tests to ascertain whether free-radical damage is reduced compared to isogenic controls. |
| 14. Computational Molecular Biology of OVO Protein Structure and Genome Targets | Mark D. Garfinkel |
| Abstract: | The OVO proteins of Drosophila share a DNA-binding domain that is
evolutionarily conserved in an interesting way. One cognate gene in the
worm, two cognates in mouse, and three cognates in human all share strong
homology in the first three zinc finger motifs, but are relatively divergent
in the fourth zinc finger motif. Indeed, the putative DNA-base-contacting
residues of the first three fingers are essentially invariant, while the
fourth finger has, arguably, degenerated to a non-functional state (at least
in the fly sequence). I have been using a variety of molecular modelling
software applications to predict the three-dimensional conformations of
each of the OVO DNA-binding domains. My aim is to rationalize the DNA-binding
site preferences we observed experimentally (see Abstracts 1 and 2) in terms
of putative amino acid sidechain-DNA base pair interactions. This computational
work is an adjunct to our ongoing crystallography experimental efforts (see
Abstract 3). Another project I have undertaken is to use profile hidden
Markov model methodology to scan the complete worm genome sequence (see
Clarke and Berg, 1998, for example) for predicted OVO binding sites. Since
the worm cognate of OVO is nearly identical to the fly protein in zinc fingers
1-3, and the fly protein recognizes a 9-bp target sequence, we reasoned
that the worm protein probably has the same specificity. Depending upon
the stringency of the screen, a few dozen to perhaps a few hundred worm
genes are predicted to be potential targets for "wovo" regulation,
and a satisfying fraction of these make biochemical sense in terms of epidermal
differentiation. Reference: Clarke, N.D. and J.M. Berg (1998) Science 282: 2018-2022. |
| To Departmental Home Page | BACK to Garfinkel Index Page | BACK to Research Overview | Garfinkel Personal Page | Web Sites of Scientific Interest |
