p16 proteins from melanoma-prone families are deficient in binding to Cdk4

The tumor suppressor candidate p16INK4 is a cyclin-dependent kinase inhibitor that inhibits cell proliferation. The p16 coding gene is often mutated in glioblastomas, pancreatic adenocarcinomas and melanoma-prone pedigrees, but, until recently, the significance of these allelic variants has remained unclear. Here, we used interaction mating and coprecipitation to measure interaction of seven p16 allelic variants detected in melanoma-prone pedigrees with Cyclin-dependent kinases (Cdks). We found that most variants were deficient in interaction with Cdk4 and Cdk6. One defective variant was found both in cancer prone families and in the control population and therefore previously defined as a common polymorphism. Another variant, which is weakly linked to familial cancer, is only slightly affected in interaction with Cdks. These results are consistent with the idea that p16 allelic variants that decrease Cdk interaction predispose individuals who carry them to an increased risk of cancer. Moreover, they suggest that determination of affinity between p16 mutants and partner proteins may help identify functionally-significant allelic variants not detected by classical human genetic techniques.

Murine chromosomal location of five bHLH-Zip transcription factor genes

The genes for the bHLH-Zip transcription factors Tfap4, Mxi1, Tcfeb, Usf1, and Usf2 have been mapped in mouse by interspecific backcross analysis. Mxi1, Usf1, and Usf2 have been mapped previously by in situ hybridization, but their positions on the meiotic linkage map had not been determined. The other two genes have not previously been mapped in mouse. These transcription factors belong to a growing family of transcriptional regulators, some of which are known to form a complex network of interacting proteins that control cell proliferation and apoptosis. As expected, based on mapping studies of other bHLH-Zip genes, these loci were well distributed among mouse chromosomes. In addition, some of the probes used in this study detected multiple, independently segregating loci, suggesting the possible existence of additional family members or species-specific pseudogenes.

APC binds to the novel protein EB1

Mutations of the APC gene play a critical role in both sporadic and familial forms of colorectal cancer. The vast majority of these mutations result in the loss of the carboxyl terminus of the protein. To further elucidate the function of APC, we searched for cellular proteins that associate with its carboxyl terminus. One million human cDNA clones were screened with the use of the interaction trap two-hybrid system, and 67 clones were found to have a phenotype suggestive of an APC-interacting protein. Nucleotide sequence analysis revealed that 48 of these clones were derived from a single novel named EBI. The association of APC and EB1 proteins was confirmed with in vitro binding assays. mAbs against EB1 were then produced and used to demonstrate the association of APC and EB1 in vivo. The EB1 gene was predicted to encode a 268-amino acid protein without significant homology to proteins with known function. However, searches of nucleotide databases did identify evidence for at least two related human genes and a yeast homologue. This conservation suggests an essential function for EB1 that might provide clues to the mechanism through which APC suppresses colonic neoplasia.

Human RNA polymerase II subunit hsRPB7 functions in yeast and influences stress survival and cell morphology

Using a screen to identify human genes that promote pseudohyphal conversion in Saccharomyces cerevisiae, we obtained a cDNA encoding hsRPB7, a human homologue of the seventh largest subunit of yeast RNA polymerase II (RPB7). Overexpression of yeast RPB7 in a comparable strain background caused more pronounced cell elongation than overexpression of hsRPB7. hsRPB7 sequence and function are strongly conserved with its yeast counterpart because its expression can rescue deletion of the essential RPB7 gene at moderate temperatures. Further, immuno-precipitation of RNA polymerase II from yeast cells containing hsRPB7 revealed that the hsRPB7 assembles the complete set of 11 other yeast subunits. However, at temperature extremes and during maintenance at stationary phase, hsRPB7-containing yeast cells lose viability rapidly, stress-sensitive phenotypes reminiscent of those associated with deletion of the RPB4 subunit with which RPB7 normally complexes. Two-hybrid analysis revealed that although hsRPB7 and RPB4 interact, the association is of lower affinity than the RPB4-RPB7 interaction, providing a probable mechanism for the failure of hsRPB7 to fully function in yeast cells at high and low temperatures. Finally, surprisingly, hsRPB7 RNA in human cells is expressed in a tissue-specific pattern that differs from that of the RNA polymerase II largest subunit, implying a potential regulatory role for hsRPB7. Taken together, these results suggest that some RPB7 functions may be analogous to those possessed by the stress-specific prokaryotic sigma factor rpoS.

Two classes of proteins dependent on either the presence or absence of thyroid hormone for interaction with the thyroid hormone receptor

The thyroid hormone (T3) receptors (TRs) are hormone-dependent transcription factors that regulate expression of a variety of specific target genes. To help elucidate the mechanisms that underlie this transcriptional regulation and other potential TR activities, we used the yeast interaction trap to isolate clones encoding proteins that specifically interact with the ligand binding domain of the rat TR beta. Several such proteins, called Trips (TR-interacting proteins), were isolated from independent selections carried out either in the presence or absence of T3. Surprisingly, all of the Trips were dependent on hormone for interaction with the TR, with some interacting only when T3 is present and others only when it is absent. Nearly all of the Trips also show similar ligand-dependent interaction with the retinoid X receptor (RXR), but none interact with the glucocorticoid receptor under any conditions. The sequences of three of the Trips predict specific functional roles: one is an apparent human homolog of a yeast transcriptional coactivator, one is a new member of a class of nonhistone chromosomal proteins, and one contains a conserved domain associated with ubiquitination of specific target proteins. Consistent with the pleiotropic effects of TR and RXR, several other Trips show significant amino acid sequence similarity with proteins involved in various regulatory pathways. The inherent transcriptional activity of the Trips was tested in yeast, and a chimeric protein consisting of a fusion of Trip4 to the bacterial LexA repressor protein is a relatively strong transcriptional activator. Similar LexA fusions to Trip9 and Trip10 had no transcriptional activity on their own but, when coexpressed with both TR and RXR, conferred T3-dependent activation to a reporter gene controlled by LexA binding sites. We suggest that this indirect T3 response provides a novel mechanism for hormonal activation of gene expression, and that studies of the Trips will provide important insights into the specific mechanisms of action of TRs and other receptors.

Interaction mating reveals binary and ternary connections between Drosophila cell cycle regulators

We characterized interactions between Drosophila melanogaster cell cycle regulatory proteins by a yeast interaction-mating technique. The results were displayed as two-dimensional matrices that revealed individual binary interactions between proteins. Each protein (Cdi, cyclin-dependent kinase interactor) interacted with a distinct spectrum of cyclin-dependent kinases (Cdk) from Drosophila and other organisms. Some Cdis interacted with other Cdis, indicating that these proteins may form trimeric complexes that include Cdks. Similar analysis of interaction matrices may be generally useful in detecting other multiprotein complexes and in establishing connectivity between individual complex members. Moreover, such analysis may also help assign function to newly identified proteins, identify domains involved in protein-protein interactions, and aid the dissection of genetic regulatory networks.

Groucho is required for Drosophila neurogenesis, segmentation, and sex determination and interacts directly with hairy-related bHLH proteins

We have used the interaction trap, a yeast two-hybrid system, to identify proteins interacting with hairy, a basic-helix-loop-helix (bHLH) protein that represses transcription during Drosophila embryonic segmentation. We find that the groucho (gro) protein binds specifically to hairy and also to hairy-related bHLH proteins encoded by deadpan and the Enhancer of split complex. The C-terminal WRPW motif present in all these bHLH proteins is essential for this interaction. We demonstrate that these associations reflect in vivo maternal requirements for gro during neurogenesis, segmentation, and sex determination, three processes regulated by the above bHLH proteins, and we propose that gro is a transcriptional corepressor recruited to specific target promoters by hairy-related bHLH proteins.

Applications of interaction traps/two-hybrid systems to biotechnology research

Two-hybrid methods provide a simple and sensitive means to detect the interaction between two proteins in living cells. Their use has resulted in the isolation of new proteins and has facilitated characterization of particular protein-protein interactions. These techniques have already resulted in the identification of important targets for pharmaceutical intervention, and it is likely that their extension in coming years will facilitate the development of new drugs.

Rch1, a protein that specifically interacts with the RAG-1 recombination-activating protein

RAG1 and RAG2 are lymphoid-specific genes that together induce V(D)J recombinase activity in a variety of nonlymphoid cell types. While no other lymphoid-specific factors are required to induce recombination, other factors with more widespread expression patterns have been implicated in the reaction. However, none of these factors have been cloned, and their relationship to the RAG proteins is unclear. Using the yeast two-hybrid assay, we have identified RCH1, a gene encoding a protein of molecular weight 58,000 that interacts specifically with RAG-1. The predicted Rch1 protein sequence is 47% identical to yeast SRP1, a protein associated with the nuclear envelope. A truncated form of Rch1, which retains the ability to interact with RAG-1, reduces V(D)J recombination activity in HeLa cells.

Specific DNA recognition and intersite spacing are critical for action of the bicoid morphogen

We examined DNA site recognition by Bicoid and its importance for pattern formation in developing Drosophila embryos. Using altered DNA specificity Bicoid mutants and appropriate reporter genes, we show that Bicoid distinguishes among related DNA-binding sites in vivo by a specific contact between amino acid 9 of its recognition alpha-helix (lysine 50 of the homeodomain) and bp 7 of the site. This result is consistent with our earlier results using Saccharomyces cerevisiae but differs from that predicted by crystallographic analysis of another homeodomain-DNA interaction. Our results also demonstrate that Bicoid binds directly to those genes whose transcription it regulates and that the amino acid 9 contact is necessary for Bicoid to direct anterior pattern formation. In both Drosophila embryos and yeast cells, Bicoid requires multiple binding sites to activate transcription of target genes. We find that the distance between binding sites is critical for Bicoid activation but that, unexpectedly, this critical distance differs between Drosophila and S. cerevisiae. This result suggests that Bicoid activation in Drosophila might require an ancillary protein(s) not present in S. cerevisiae.

Mapping of two genes encoding members of a distinct subfamily of MAX interacting proteins: MAD to human chromosome 2 and mouse chromosome 6, and MXI1 to human chromosome 10 and mouse chromosome 19

Both the MAD and the MXI1 genes encode basic-helix-loop-helix-leucine zipper (bHLH-Zip) transcription factors which bind Max in vitro, forming a sequence-specific DNA-binding complex similar to the Myc-Max heterodimer. Mad and Myc compete for binding to Max. In addition, Mad has been shown to act as a transcriptional repressor while Myc appears to function as an activator. Mxi1 also appears to lack a transcriptional activation domain. Therefore, Mxi1 and Mad might antagonize Myc function and are candidate tumor suppressor genes. We report here the mapping of the MAD and MXI1 genes in human and mouse by fluorescence in situ hybridization (FISH) and by recombination mapping. The MAD gene was mapped to human chromosome 2 at band p13 by FISH and to mouse chromosome 6 by meiotic mapping. The MXI1 gene was mapped to human chromosome 10 at band q25 and on mouse chromosome 19 at region D by FISH. There was a second site of hybridization on mouse chromosome 2 at region C, which may represent a pseudogene or a related sequence. The mapping results confirm regions of conservation between human chromosome 2p13 and mouse chromosome 6 and between chromosome 10q25 and mouse chromosome 19D. Human chromosomes 2p13 and 10q25 have been involved in specific tumors where the role of Mad and Mxi1 can now be investigated.

Cdi1, a human G1 and S phase protein phosphatase that associates with Cdk2

We used the interaction trap, a yeast genetic selection for interacting proteins, to isolate human cyclin-dependent kinase interactor 1 (Cdi1). In yeast, Cdi1 interacts with cyclin-dependent kinases, including human Cdc2, Cdk2, and Cdk3, but not with Ckd4. In HeLa cells, Cdi1 is expressed at the G1 to S transition, and the protein forms stable complexes with Cdk2. Cdi1 bears weak sequence similarity to known tyrosine and dual specificity phosphatases. In vitro, Cdi1 removes phosphate from tyrosine residues in model substrates, but a mutant protein that bears a lesion in the putative active site cysteine does not. Overexpression of wild-type Cdi1 delays progression through the cell cycle in yeast and HeLa cells; delay is dependent on Cdi1 phosphatase activity. These experiments identify Cdi1 as a novel type of protein phosphatase that forms complexes with cyclin-dependent kinases.

Mxi1, a protein that specifically interacts with Max to bind Myc-Max recognition sites

We used the interaction trap to isolate a novel human protein that specifically interacts with Max. This protein, Mxi1 (for Max interactor 1), contains a bHLH-Zip motif that is similar to that found in Myc family proteins. Mxi1 interacts specifically with Max to form heterodimers that efficiently bind to the Myc-Max consensus recognition site. When bound to DNA by a LexA moiety in yeast, Mxi1 does not stimulate transcription. mxi1 mRNA is expressed in many tissues, and its expression is elevated in U-937 myeloid leukemia cells that have been stimulated to differentiate. These facts are consistent with a model in which Mxi1-Max heterodimers indirectly inhibit Myc function in two ways: first, by sequestering Max, thus preventing the formation of Myc-Max heterodimers, and second, by competing with Myc-Max heterodimers for binding to target sites.

Fused protein domains inhibit DNA binding by LexA

Many studies of transcription activation employ fusions of activation domains to DNA binding domains derived from the bacterial repressor LexA and the yeast activator GAL4. Such studies often implicitly assume that DNA binding by the chimeric proteins is equivalent to that of the protein donating the DNA binding moiety. To directly investigate this issue, we compared operator binding by a series of LexA-derivative proteins to operator binding by native LexA, by using both in vivo and in vitro assays. We show that operator binding by many proteins such as LexA-Myc, LexA-Fos, and LexA-Bicoid is severely impaired, while binding of other LexA-derivative proteins, such as those that carry bacterially encoded acidic sequences ("acid blobs"), is not. Our results also show that DNA binding by LexA derivatives that contain the LexA carboxy-terminal dimerization domain (amino acids 88 to 202) is considerably stronger than binding by fusions that lack it and that heterologous dimerization motifs cannot substitute for the LexA88-202 function. These results suggest the need to reevaluate some previous studies of activation that employed LexA derivatives and modifications to recent experimental approaches that use LexA and GAL4 derivatives to detect and study protein-protein interactions.

A yeast transcription assay defines distinct rel and dorsal DNA recognition sequences

Recent data have demonstrated that vRel, cRel, Dorsal, and NF-kappa B are members of a larger family of DNA-binding regulatory proteins. Rel proteins interact to form homo- and heterodimers that recognize specific sites on DNA, and it is likely that such protein-protein and protein-DNA interactions contribute to proper regulation of target gene expression by these proteins. Here we describe the use of a yeast transcription activation assay to study binding of three Rel family proteins to their native binding sites. These results show that the vRel and cRel proteins recognize two known NF-kappa B binding sites; the Dorsal protein does not recognize NF-kappa B sites, but does recognize related sites upstream of the Drosophila zerknüllt gene. Our experiments demonstrate that the members of this protein family recognize similar, but not identical, sites in the promoters of target genes, and we are able to identify a particular nucleotide that is apparently involved in the DNA-protein interaction. We exploit the properties of LexA fusion proteins to study the dimerization and DNA-contacting domains of cRel. Our results suggest that the cRel protein forms homodimers and that dimer formation may be necessary for cRel to bind DNA. Finally, our results show that transcription activation by these proteins is cooperative; such cooperativity may be important for correct temporal and spatial regulation of target gene expression.

A genetic model for interaction of the homeodomain recognition helix with DNA

The Bicoid homeodomain protein controls anterior development in the Drosophila embryo by binding to DNA and regulating gene expression. With the use of genetic assays in yeast, the interaction between the Bicoid homeodomain and a series of mutated DNA sites was studied. These experiments defined important features of homeodomain binding sites, identified specific amino acid-base pair contacts, and suggested a model for interaction of the recognition alpha-helices of Bicoid and Antennapedia-class homeodomain proteins with DNA. The model is in general agreement with results of crystallographic and magnetic resonance studies, but differs in important details. It is likely that genetic studies of protein-DNA interaction will continue to complement conventional structural approaches.

Inhibition of X-ray-induced exencephaly by protease inhibitors

We report that protease inhibitors can reduce the incidence of radiation-induced exencephaly in mice. Previous studies from this and other laboratories have demonstrated that protease inhibitors, in particular antipain and the Bowman-Birk inhibitor, are effective anticarcinogenic agents in a variety of in vivo systems and inhibit cell transformation in vitro. Given our results, further investigation is warranted into preventive effects of protease inhibitors on the inhibition of neural tube defects.

Oncogenic transformation by vrel requires an amino-terminal activation domain

The mechanism by which the products of the v-rel oncogene, the corresponding c-rel proto-oncogene, and the related dorsal gene of Drosophila melanogaster exert their effects is not clear. Here we show that the v-rel, chicken c-rel, and dorsal proteins activated gene expression when fused to LexA sequences and bound to DNA upstream of target genes in Saccharomyces cerevisiae. We have defined two distinct activation regions in the c-rel protein. Region I, located in the amino-terminal half of rel and dorsal proteins, contains no stretches of glutamines, prolines, or acidic amino acids and therefore may be a novel activation domain. Lesions in the v-rel protein that diminished or abolished oncogenic transformation of avian spleen cells correspondingly affected transcription activation by region I. Region II, located in the carboxy terminus of the c-rel protein, is highly acidic. Region II is not present in the v-rel protein or in a transforming mutant derivative of the c-rel protein. Our results show that the oncogenicity of Rel proteins requires activation region I and suggest that the biological function of rel and dorsal proteins depends on transcription activation by this region.

DNA specificity of the bicoid activator protein is determined by homeodomain recognition helix residue 9

Formation of anterior structures in the Drosophila embryo requires the product of the gene bicoid. The bicoid protein contains a homeodomain and may exert its effects in early development by regulating transcription of the gap gene, hunchback (hb). Consistent with this view, we have demonstrated that DNA-bound Bicoid fusion proteins stimulate gene expression. We used the gene activation phenotype in yeast to study DNA recognition by the Bicoid homeodomain. We found that a single amino acid replacement at position 9 of the recognition helix was sufficient to switch the DNA specificity of the Bicoid protein. The altered specificity Bicoid mutants recognized DNA sites bound by Ultrabithorax, fushi tarazu, and other related homeo-domain proteins. Our results suggest that DNA specificity in Bicoid and Antennapedia class proteins is determined by recognition helix residue 9.

Gene activation and DNA binding by Drosophila Ubx and abd-A proteins

The Ubx and abd-A gene products are required for proper development of thoracic and abdominal structures in Drosophila. We expressed LexA-Ubx and LexA-abdA fusion proteins in yeast. These proteins activated expression of target genes that carried either upstream LexA operators or upstream Ubx binding sites. Both proteins contain homeodomains. Experiments with mutant fusion proteins show that the homeodomain is not required for the proteins to form dimers or enter the nucleus, and that, when DNA binding is provided by the LexA moiety, the homeodomain is not required for gene activation. Our results suggest that the homeodomain is necessary for these proteins to bind Ubx sites, but that the homeodomain does not contact DNA exactly like bacterial helix-turn-helix proteins. Finally, our data suggest that gene activation by these proteins is a simple consequence of their binding to DNA, while negative gene regulation requires that these proteins act together with other Drosophila gene products.

DNA-bound Fos proteins activate transcription in yeast

We constructed genes encoding the DNA binding region of the bacterial LexA repressor fused to the v-fos and c-fos oncogene products. The resulting LexA-Fos fusion proteins activated transcription in yeast. Transcription activation by these proteins was as strong as transcription activation by proteins native to yeast. LexA-Fos fusion proteins only activated transcription of genes when they were bound to LexA binding sites inserted upstream of those genes. Transcription was activated less strongly by similar proteins in which the DNA binding region of LexA was fused to vMyc and cMyc. Transcription was not activated by native LexA or by proteins containing the DNA binding domain of LexA fused to bacteriophage 434 repressor or yeast MAT alpha 2 protein. These results demonstrate that Fos proteins activate eukaryotic gene expression when they are bound to promoter DNA, and thus suggest that Fos proteins exert some of their effects because they stimulate transcription of cellular genes. Regulation of transcription by Fos and Myc proteins in yeast provides a phenotype that may facilitate genetic analysis of the function of these proteins in higher organisms.