Activation of hypodermal differentiation in the Caenorhabditis elegans embryo by GATA transcription factors ELT-1 and ELT-3.

The Caenorhabditis elegans GATA transcription factor genes elt-1 and elt-3 are expressed in the embryonic hypodermis (also called the epidermis). elt-1 is expressed in precursor cells and is essential for the production of most hypodermal cells (22). elt-3 is expressed in all of the major hypodermal cells except the lateral seam cells, and expression is initiated immediately after the terminal division of precursor lineages (13). Although this expression pattern suggests a role for ELT-3 in hypodermal development, no functional studies have yet been performed. In the present paper, we show that either elt-3 or elt-1 is sufficient, when force expressed in early embryonic blastomeres, to activate a program of hypodermal differentiation even in blastomeres that are not hypodermal precursors in wild-type embryos. We have deleted the elt-3 gene and shown that ELT-3 is not essential for either hypodermal cell differentiation or the viability of the organism. We showed that ELT-3 can activate hypodermal gene expression in the absence of ELT-1 and that, conversely, ELT-1 can activate hypodermal gene expression in the absence of ELT-3. Overall, the combined results of the mutant phenotypes, initial expression times, and our forced-expression experiments suggest that ELT-3 acts downstream of ELT-1 in a redundant pathway controlling hypodermal cell differentiation.

Coordination of ges-1 expression between the Caenorhabditis pharynx and intestine.

We have previously shown that the Caenorhabditis elegans gut-specific esterase gene (Ce-ges-1) has the unusual ability to be expressed in different modules of the embryonic digestive tract (anterior pharynx, posterior pharynx, and rectum) depending on sequence elements within the Ce-ges-1 promoter. In the present paper, we analyze the expression of the ges-1 homolog (Cb-ges-1) from the related nematode Caenorhabditis briggsae and show that Cb-ges-1 also has the ability to switch expression between gut and pharynx + rectum. The control of this expression switch centres on a tandem pair of WGATAR sites in the Cb-ges-1 5'-flanking region, just as it does in Ce-ges-1. We use sequence alignments and subsequent deletions to identify a region at the 3'-end of both Ce-ges-1 and Ce-ges-1 that acts as the ges-1 cryptic pharynx enhancer whose activity is revealed by removal of the 5' WGATAR sites. This region contains a conserved binding site for PHA-4 (the C. elegans ortholog of forkhead/HNF3 alpha, beta,gamma factors), which is expressed in all cells of the developing pharynx and a subset of cells of the developing rectum. We propose a model in which the normal expression of ges-1 is controlled by the gut-specific GATA factor ELT-2. We propose that, in the pharynx (and rectum), PHA-4 is normally bound to the ges-1 3'-enhancer sequence but that the activation function of PHA-4 is kept repressed by a (presently unknown) factor binding in the vicinity of the 5' WGATAR sites. We suggest that this control circuitry is maintained in Caenorhabditis because pharyngeal expression of ges-1 is advantageous only under certain developmental or environmental conditions.

Homologous tails? Or tales of homology?

Classical mutations at the mouse Brachyury (T) locus were discovered because they lead to shortened tails in heterozygous newborns. no tail (ntl) mutants in the zebrafish, as their name suggests, show a similar phenotype. In Drosophila, mutants in the brachyenteron (byn) gene disrupt hindgut formation. These genes all encode T-box proteins, a class of sequence-specific DNA binding proteins and transcription factors. Mutations in the C. elegans mab-9 gene cause massive defects in the male tail because of failed fate decisions in two tail progenitor cells. In a recent paper, Woollard and Hodgkin have cloned the mab-9 gene and found that it too encodes a T-box protein, similar to Brachyury in vertebrates and brachyenteron in Drosophila. The authors suggest that their results support models for an evolutionarily ancient role for these genes in hindgut formation. We will discuss this proposal and try to decide whether the gene sequences, gene interactions and gene expression patterns allow any conclusions to be made about the rear end of the ancestral metazoan.

Direct visualization of the elt-2 gut-specific GATA factor binding to a target promoter inside the living Caenorhabditis elegans embryo.

In analyzing the transcriptional networks that regulate development, one ideally would like to determine whether a particular transcription factor binds directly to a candidate target promoter inside the living embryo. Properties of the Caenorhabditis elegans elt-2 gene, which encodes a gut-specific GATA factor, have allowed us to develop such a method. We previously have shown, by means of ectopic expression studies, that elt-2 regulates its own promoter. To test whether this autoregulation is direct, we fused green fluorescent protein (GFP) close to the C terminus of elt-2 in a construct that contains the full elt-2 promoter and the full elt-2 zinc finger DNA binding domain; the construct is expressed correctly (i.e., only in the gut lineage) and is able to rescue the lethality of an elt-2 null mutant. Multicopy transgenic arrays of this rescuing elt-2::GFP construct were integrated into the genome and transgenic embryos were examined when the developing gut has 4-8 cells; the majority of these embryonic gut nuclei show two discrete intense foci of fluorescence. We interpret these fluorescent foci as the result of ELT-2::GFP binding directly to its own promoter within nuclei of the developing gut lineage. Numerous control experiments, both genetic and biochemical, all support this conclusion and support the specificity of the binding. The approach should be applicable to studying other transcription factors binding target promoters, all within the living C. elegans embryo.

ELT-3: A Caenorhabditis elegans GATA factor expressed in the embryonic epidermis during morphogenesis.

We have identified a gene encoding a new member of the Caenorhabditis elegans GATA transcription factor family, elt-3. The predicted ELT-3 polypeptide contains a single GATA-type zinc finger (C-X2-C-X17-C-X2-C) along with a conserved adjacent basic region. elt-3 mRNA is present in all stages of C. elegans development but is most abundant in embryos. Reporter gene analysis and antibody staining show that elt-3 is first expressed in the dorsal and ventral hypodermal cells, and in hypodermal cells of the head and tail, immediately after the final embryonic cell division that gives rise to these cells. No expression is seen in the lateral hypodermal (seam) cells. elt-3 expression is maintained at a constant level in the epidermis until the 2(1/2)-fold stage of development, after which reporter gene expression declines to a low level and endogenous protein can no longer be detected by specific antibody. A second phase of elt-3 expression in cells immediately anterior and posterior to the gut begins in pretzel-stage embryos. elt-1 and lin-26 are two genes known to be important in specification and maintenance of hypodermal cell fates. We have found that elt-1 is required for the formation of most, but not all, elt-3-expressing cells. In contrast, lin-26 function does not appear necessary for elt-3 expression. Finally, we have characterised the candidate homologue of elt-3 in the nematode Caenorhabditis briggsae. Many features of the elt-3 genomic and transcript structure are conserved between the two species, suggesting that elt-3 is likely to perform an evolutionarily significant function during development.

Reprogramming of early embryonic blastomeres into endodermal progenitors by a Caenorhabditis elegans GATA factor.

The END-1 GATA factor has been implicated in specifying endoderm in Caenorhabditis elegans and is the earliest known zygotic protein expressed in the lineage of E, the clonal endoderm progenitor. We report that ubiquitous end-1 expression during a critical period in embryogenesis causes all non-endodermal lineages to produce endoderm instead of ectoderm and/or mesoderm. END-1 expression bypasses the requirement for maternal SKN-1 and the maternal Wnt signaling pathway in endoderm formation. This suggests that a primary function of these maternal factors is to regulate zygotic end-1 expression, which is then sufficient to initiate the entire program for endoderm development.

Anterior-posterior patterning within the Caenorhabditis elegans endoderm.

The endoderm of higher organisms is extensively patterned along the anterior/posterior axis. Although the endoderm (gut or E lineage) of the nematode Caenorhabditis elegans appears to be a simple uniform tube, cells in the anterior gut show several molecular and anatomical differences from cells in the posterior gut. In particular, the gut esterase ges-1 gene, which is normally expressed in all cells of the endoderm, is expressed only in the anterior-most gut cells when certain sequences in the ges-1 promoter are deleted. Using such a deleted ges-1 transgene as a biochemical marker of differentiation, we have investigated the basis of anterior-posterior gut patterning in C. elegans. Although homeotic genes are involved in endoderm patterning in other organisms, we show that anterior gut markers are expressed normally in C. elegans embryos lacking genes of the homeotic cluster. Although signalling from the mesoderm is involved in endoderm patterning in other organisms, we show that ablation of all non-gut blastomeres from the C. elegans embryo does not affect anterior gut marker expression; furthermore, ectopic guts produced by genetic transformation express anterior gut markers generally in the expected location and in the expected number of cells. We conclude that anterior gut fate requires no specific cell-cell contact but rather is produced autonomously within the E lineage. Cytochalasin D blocking experiments fully support this conclusion. Finally, the HMG protein POP-1, a downstream component of the Wnt signalling pathway, has recently been shown to be important in many anterior/posterior fate decisions during C. elegans embryogenesis (Lin, R., Hill, R. J. and Priess, J. R. (1998) Cell 92, 229-239). When RNA-mediated interference is used to eliminate pop-1 function from the embryo, gut is still produced but anterior gut marker expression is abolished. We suggest that the C. elegans endoderm is patterned by elements of the Wnt/pop-1 signalling pathway acting autonomously within the E lineage.

The GATA-factor elt-2 is essential for formation of the Caenorhabditis elegans intestine.

The Caenorhabditis elegans elt-2 gene encodes a single-finger GATA factor, previously cloned by virtue of its binding to a tandem pair of GATA sites that control the gut-specific ges-1 esterase gene. In the present paper, we show that elt-2 expression is completely gut specific, beginning when the embryonic gut has only two cells (one cell cycle prior to ges-1 expression) and continuing in every cell of the gut throughout the life of the worm. When elt-2 is expressed ectopically using a transgenic heat-shock construct, the endogenous ges-1 gene is now expressed in most if not all cells of the embryo; several other gut markers (including a transgenic elt-2-promoter::lacZ reporter construct designed to test for elt-2 autoregulation) are also expressed ectopically in the same experiment. These effects are specific in that two other C. elegans GATA factors (elt-1 and elt-3) do not cause ectopic gut gene expression. An imprecise transposon excision was identified that removes the entire elt-2 coding region. Homozygous elt-2 null mutants die at the L1 larval stage with an apparent malformation or degeneration of gut cells. Although the loss of elt-2 function has major consequences for later gut morphogenesis and function, mutant embryos still express ges-1. We suggest that elt-2 is part of a redundant network of genes that controls embryonic gut development; other factors may be able to compensate for elt-2 loss in the earlier stages of gut development but not in later stages. We discuss whether elements of this regulatory network may be conserved in all metazoa.

pha-4 is Ce-fkh-1, a fork head/HNF-3alpha,beta,gamma homolog that functions in organogenesis of the C. elegans pharynx.

The C. elegans Ce-fkh-1 gene has been cloned on the basis of its sequence similarity to the winged-helix DNA binding domain of the Drosophila fork head and mammalian HNF-3alpha,beta,gamma genes, and mutations in the zygotically active pha-4 gene have been shown to block formation of the pharynx (and rectum) at an early stage in embryogenesis. In the present paper, we show that Ce-fkh-1 and pha-4 are the same gene. We show that PHA-4 protein is present in nuclei of essentially all pharyngeal cells, of all five cell types. PHA-4 protein first appears close to the point at which a cell lineage will produce only pharyngeal cells, independently of cell type. We show that PHA-4 binds directly to a 'pan-pharyngeal enhancer element' previously identified in the promoter of the pharyngeal myosin myo-2 gene; in transgenic embryos, ectopic PHA-4 activates ectopic myo-2 expression. We also show that ectopic PHA-4 can activate ectopic expression of the ceh-22 gene, a pharyngeal-specific NK-2-type homeodomain protein previously shown to bind a muscle-specific enhancer near the PHA-4 binding site in the myo-2 promoter. We propose that it is the combination of pha-4 and regulatory molecules such as ceh-22 that produces the specific gene expression patterns during pharynx development. Overall, pha-4 can be described as an 'organ identity factor', completely necessary for organ formation, present in all cells of the organ from the earliest stages, capable of integrating upstream developmental pathways (in this case, the two distinct pathways that produce the anterior and posterior pharynx) and participating directly in the transcriptional regulation of organ specific genes. Finally, we note that the distribution of PHA-4 protein in C. elegans embryos is remarkably similar to the distribution of the fork head protein in Drosophila embryos: high levels in the foregut/pharynx and hindgut/rectum; low levels in the gut proper. Moreover, we show that pha-4 expression in the C. elegans gut is regulated by elt-2, a C. elegans gut-specific GATA-factor and possible homolog of the Drosophila gene serpent, which influences fork head expression in the fly gut. Overall, our results provide evidence for a highly conserved pathway regulating formation of the digestive tract in all (triploblastic) metazoa.

Caenorhabditis elegans LET-502 is related to Rho-binding kinases and human myotonic dystrophy kinase and interacts genetically with a homolog of the regulatory subunit of smooth muscle myosin phosphatase to affect cell shape.

We have identified two genes associated with the hypodermal cell shape changes that occur during elongation of the Caenorhabditis elegans embryo. The first gene, called let-502, encodes a protein with high similarity to Rho-binding Ser/Thr kinases and to human myotonic dystrophy kinase (DM-kinase). Strong mutations in let-502 block embryonic elongation, and let-502 reporter constructs are expressed in hypodermal cells at the elongation stage of development. The second gene, mel-11, was identified by mutations that act as extragenic suppressors of let-502. mel-11 encodes a protein similar to the 110- to 133-kD regulatory subunits of vertebrate smooth muscle myosin-associated phosphatase (PP-1M). We suggest that the LET-502 kinase and the MEL-11 phosphatase subunit act in a pathway linking a signal generated by the small GTP-binding protein Rho to a myosin-based hypodermal contractile system that drives embryonic elongation. LET-502 may directly regulate the activity of the MEL-11 containing phosphatase complex and the similarity between LET-502 and DM-kinase suggests a similar function for DM-kinase.

Modulation of gene expression in the embryonic digestive tract of C. elegans.

The Caenorhabditis elegans digestive tract is composed of four distinct modules derived from separate cell lineages: anterior pharynx from the ABa lineage, posterior pharynx from the MS lineage, gut from the E lineage, and rectum from the ABp lineage. The C. elegans gut esterase gene (ges-1) is normally expressed in the embryonic gut or E lineage. However, expression ges-1 can be switched into cells of the embryonic pharynx and tail by virtue of deleting a tandem pair of WGATAR sites in the ges-1 promoter. Here, we use both laser ablation experiments and genetic analysis to show that cells expressing the WGATAR-deleted ges-1 transgene belong to all three nongut lineages of the digestive tract: ABa, MS, and ABp. We also show that the molecular size and spatial distribution of ges-1 mRNA transcripts produced by either the WGATAR-deleted ges-1 transgene or the undeleted ges-1 control transgene appear correctly regulated, suggesting that the spatial switch in ges-1 expression occurs at the level of transcription initiation. We further show that both the WGATAR-deleted and the undeleted ges-1 transgenes respond appropriately to mutations in a series of maternal effect genes (skn-1, mex-1, pie-1, and pop-1) that alter early blastomere fate. Moreover, the pharynx/tail expression of the WGATAR-deleted ges-1 transgene is abolished by mutations in the zygotic gene pha-4. Finally, we use imprecise transposon excision to produce two independent C. elegans strains with 1- to 2-kb deletions that remove the tandem WGATAR sites from the promoter of the endogenous chromosomal ges-1 gene: in both of these strains, ges-1 is not expressed in the embryonic gut but is expressed in cells of the embryonic pharynx; pharynx expression is weak but incontrovertible. Overall, our results validate previous transgenic analysis of ges-1 control and show further that ges-1 appears to be regulated in a system-specific, rather than a lineage-specific, manner. The multiple facets of ges-1 expression provide an opportunity to investigate how a multicomponent organ system such as the digestive tract is established from diverse cell lineages.

A fork head/HNF-3 homolog expressed in the pharynx and intestine of the Caenorhabditis elegans embryo.

We have cloned a member of the fork head/HNF-3 family of transcription factors from the nematode Caenorhabditis elegans. Within the predicted DNA binding domain, this gene, called Ce-fkh-1, is 75-78% identical to the Drosophila fork head and rat liver HNF-3 alpha, beta, and gamma genes. Ce-fkh-1 mRNA is highly enriched in embryos. The Ce-fkh-1 gene produces three major transcripts: the longest mRNA retains its original 5'-end but two shorter mRNAs are trans-spliced at the beginning of exons 2 and 3, respectively. In situ hybridization and transgenic Ce-fkh-1::lacZ reporter constructs indicate that the Ce-fkh-1 gene is expressed in both pharynx and intestine of the embryo, beginning at the midproliferation stage. A second phase of Ce-fkh-1 expression occurs in cells of the larval somatic gonad. The pharynx-gut expression of Ce-fkh-1 in the C. elegans embryo is compared with expression of fork head throughout the gut of Drosophila embryos and with expression of HNF-3 (alpha beta gamma) in the endoderm of mammalian embryos. Such conserved patterns of gene expression point to universal features of gastrulation and of digestive tract formation.

A gut-to-pharynx/tail switch in embryonic expression of the Caenorhabditis elegans ges-1 gene centers on two GATA sequences.

The Caenorhabditis elegans ges-1 gene (gut esterase No. 1) is expressed only in the intestinal lineage, beginning when the developing gut has only four to eight cells. We analyze the sequence requirements for this tissue-specific gene regulation by injecting deleted/mutated constructs of the ges-1 gene into a viable ges-1 (null) strain of worms and assaying heritably transformed embryos by esterase histochemistry. Many deletion constructs accurately reconstitute the wildtype gut-specific ges-1 expression. However, deletions in the neighborhood of 1100 bp upstream of the ges-1 ATG abolish ges-1 expression in the developing gut, while at the same time activating ges-1 expression in cells of the pharynx/tail that appear to belong to the sister lineage of the gut. Deletions of a 36-bp DNA region containing two tandem WGATAR sequences are sufficient to cause this gut-to-pharynx/tail switch in expression pattern. Deletion of either one of the WGATAR sites or deletion of an adjoining downstream region directs ges-1 expression only in a restricted set of cells of the anterior gut. The ges-1 GATA region acts like a gut-specific enhancer in that: (i) it restores ges-1 gut expression when reinserted elsewhere into the GATA-deleted ges-1 gene; and (ii) multiple copies direct gut expression of an hsp16-lacZ reporter gene. The ges-1 GATA-region also acts as the site of the pharynx/tail repression in that reinsertion elsewhere into the GATA-deleted ges-1 construct causes repression of ges-1 in the pharynx/tail. However, multiple copies of the GATA region are not able to repress the heat-induced expression of an hsp16-lacZ reporter gene, suggesting that the pharynx/tail repression mechanism is specific to the ges-1 environment. Finally, mutation rather than deletion of the individual GATA sequences suggests that gut activation and pharynx/tail repression may be due to separate factors. We present a molecular model that summarizes these results. The ges-1 control circuitry appears surprisingly complex for what might have been expected to be the simplest possible example of a nonessential gene expressed early in a clonal embryonic lineage.

elt-2, a second GATA factor from the nematode Caenorhabditis elegans.

We have previously shown that a tandem pair of (A/T)GATA(A/G) sequences in the promoter region of the Caenorhabditis elegans gut esterase gene (ges-1) controls the tissue specificity of ges-1 expression in vivo. The ges-1 GATA region was used as a probe to screen a C. elegans cDNA expression library, and a gene for a new C. elegans GATA-factor (named elt-2) was isolated. The longest open reading frame in the elt-2 cDNA codes for a protein of M(r) 47,000 with a single zinc finger domain, similar (approximately 75% amino acid identity) to the C-terminal fingers of all other two-fingered GATA factors isolated to date. A similar degree of relatedness is found with the single-finger DNA binding domains of GATA factors identified in invertebrates. An upstream region in the ELT-2 protein with the sequence C-X2-C-X16-C-X2-C has some of the characteristics of a zinc finger domain but is highly diverged from the zinc finger domains of other GATA factors. The elt-2 gene is expressed as an SL1 trans-spliced message, which can be detected at all stages of development except oocytes; however, elt-2 message levels are 5-10-fold higher in embryos than in other stages. The genomic clone for elt-2 has been characterized and mapped near the center of the C. elegans X chromosome, ELT-2 protein, produced by in vitro transcription-translation, binds to ges-1 GATA-containing oligonucleotides similar to a factor previously identified in C. elegans embryo extracts, both as assayed by electrophoretic migration and by competition with wild type and mutant oligonucleotides. However, there is as yet no direct evidence that elt-2 does or does not control ges-1.

The C. elegans neuronally expressed homeobox gene ceh-10 is closely related to genes expressed in the vertebrate eye.

We describe the homeobox gene ceh-10 from the nematode Caenorhabditis elegans. The homeodomain of ceh-10 is closely related to the homeodomains of two genes recently cloned from the vertebrate retina, Chx10 from mice and Vsx-1 from goldfish. We show that the sequence conservation extends well beyond the homeodomain and includes a region (named the CVC domain) of roughly 60 amino acids immediately C-terminal to the homeodomain. As assayed in transgenic worms, the promoter region of ceh-10 directs expression of a lacZ reporter gene to a small number of neurons. We draw a parallel between the bipolar cells of the inner nuclear layer of the vertebrate retina, which express Chx10 and Vsx-1, and an interneuron in C. elegans called AIY, which expresses ceh-10. AIY receives synaptic input from a sensory cell, just as do bipolar cells of the vertebrate retina. In C. elegans, the sensory cell AFD is not known to be photosensitive but is known to be thermosensitive; moreover, a cell with similar position in the amphids of other nematodes has been suggested indeed to be photosensitive. Our results emphasize the highly conserved nature of sensory regulatory mechanisms and suggest one way in which photosensitive organelles might have originated in evolution.

A carboxylesterase from the parasitic nematode Ascaris suum homologous to the intestinal-specific ges-1 esterase of Caenorhabditis elegans.

We have identified a carboxylesterase in A. suum that appears to be the homolog of the gut-specific C. elegans ges-1 enzyme. The A. suum esterase was purified and its N-terminal sequence found to be 50% identical to the C. elegans ges-1 protein. We have used isoelectric focusing analysis to demonstrate that, unlike the C. elegans ges-1 esterase, the A. suum enzyme is not restricted to the gut but is expressed in a wide range of tissues.

DNA-protein interactions in the Caenorhabditis elegans embryo: oocyte and embryonic factors that bind to the promoter of the gut-specific ges-1 gene.

We describe an experimental system in which to investigate DNA-protein interactions in the early Caenorhabditis elegans embryo. A homogeneous population of developmentally blocked mid-proliferation stage embryos can be produced by exposure to the deoxynucleotide analog fluorodeoxyuridine. These blocked embryos remain viable for days and express a number of biochemical markers of early differentiation, for example, gut granules, the gut esterase ges-1, and two regulatory genes, mab-5 and hlh-1. Using the techniques of gel mobility shift and DNase I footprinting, we show that nuclear extracts prepared from these embryos contain factors that bind to the 5'-promoter sequences of the C. elegans gut-specific ges-1 gene. In particular, we examine a putative gut "activator" region, which was previously identified by deletion-transformation analysis and which contains two copies of a consensus GATA-factor binding sequence. Factors that bind to double-stranded oligonucleotides containing the ges-1 GATA sequences are present predominantly in nuclear extracts of embryos but are found neither in cytoplasmic nor in nuclear extracts of unfertilized oocytes. Two proteins, of 43 and 60 kDa, can be uv-crosslinked to double-stranded oligonucleotides containing the ges-1 GATA sequences. The sizes of these proteins correspond to the sizes expected for the elt-1 protein and for the skn-1 protein, two regulatory factors present in early C. elegans embryos and possible candidates for ges-1 control. However, we show that homozygous deficiency embryos (mDf7/mDf7 embryos and eDf19/eDf19 embryos, both of which lack the elt-1 gene, and nDf41/nDf41 embryos, which have no skn-1 gene), still express the ges-1 esterase. We conclude that neither the elt-1 gene nor the skn-1 gene is necessary zygotically for ges-1 expression. We suggest that neither the elt-1 protein nor the skn-1 protein interacts directly with the ges-1 gene and that the observed binding proteins must correspond to products of other genes. More generally, the present experimental system should allow the biochemical study of any gene expressed during early C. elegans embryogenesis.

Uracil-DNA glycosylase as a probe for protein--DNA interactions.

The DNA repair enzyme Uracil-DNA Glycosylase (UDG) can be used to investigate three different features of protein-DNA interactions. Complexes can be probed by simple protection experiments ('footprinting') or by two kinds of interference assays: a missing thymine site (MT-site) experiment and a missing thymine methyl site (MTM-site) experiment. The three probing methods are assessed using the well-characterized in vitro systems of lambda repressor and lac repressor binding to their respective operator sites. The results obtained with UDG probing agree well with previous probing experiments on the same systems and, in certain cases, extend previous interpretations: for example, comparison of the results obtained with the two interference assays shows that formation of the lac repressor-operator complex requires interactions with the methyl group of one particular thymine residue (T-13) in the operator but also requires interactions with other parts of the thymine base at operator positions 7, 8, 9, 21, 23 and 24. Overall, the properties of UDG recommend it as a versatile and convenient method to investigate DNA-protein interactions both in vitro and possibly in vivo.