Fos and Jun bend the AP-1 site: effects of probe geometry on the detection of protein-induced DNA bending.

The effect of Fos and Jun binding on the structure of the AP-1 recognition site is controversial. Results from phasing analysis and phase-sensitive detection studies of DNA bending by Fos and Jun have led to opposite conclusions. The differences between these assays, the length of the spacer between two bends and the length of the sequences flanking the bends, are investigated here using intrinsic DNA bend standards. Both an increase in the spacer length as well as a decrease in the length of flanking sequences resulted in a reduction in the phase-dependent variation in electrophoretic mobilities. Probes with a wide separation between the bends and short flanking sequences, such as those used in the phase-sensitive detection studies, displayed no phase-dependent mobility variation. This shape-dependent variation in electrophoretic mobilities was reproduced by complexes formed by truncated Fos and Jun. Results from ligase-catalyzed cyclization experiments have been interpreted to indicate the absence of DNA bending in the Fos-Jun-AP-1 complex. However, truncated Fos and Jun can alter the relative rates of inter- and intramolecular ligation through mechanisms unrelated to DNA bending, confounding the interpretation of cyclization data. The analogous phase- and shape-dependence of the electrophoretic mobilities of the Fos-Jun-AP-1 complex and an intrinsic DNA bend confirm that Fos and Jun bend DNA, which may contribute to their functions in transcription regulation.

Sensitive detection of bacterial transcription initiation sites and differentiation from RNA processing sites in the pheromone-induced plasmid transfer system of Enterococcus faecalis.

A method was developed to detect 5' ends of bacterial RNAs expressed at low levels and to differentiate newly initiated transcripts from processed transcripts produced in vivo. The procedure involves use of RNA ligase to link a specific oligoribonucleotide to the 5' ends of cellular RNAs, followed by production of cDNA and amplification of the gene of interest by PCR. The method was used to identify the precise sites of transcription initiation within a 10-kb region of the pheromone-inducible conjugative plasmid pCF10 of Enterococcus faecalis. Results confirmed the 5' end of a very abundant, constitutively produced transcript (from prgQ) that had been mapped previously by primer extension and defined the initiation point of a less abundant, divergently transcribed message (from prgX). The method also showed that the 5' end of a pheromone-inducible transcript (prgB) that had been mapped by primer extension was generated by processing rather than new initiation. In addition, the results provided evidence for two promoters, 3 and 5 kb upstream of prgB, and indicated that only the transcripts originating 5 kb upstream may be capable of extending to prgB.

The effects of sequence context on DNA curvature.

Recent experiments have exposed significant discrepancies between experimental data and predictive models for DNA structure. These results strongly suggest that DNA structural parameters incorporated in the models are not always sufficient to account for the influence of sequence context and of specific ion effects. In an attempt to evaluate these two effects, we have investigated repetitive DNA sequences with the sequence motif GAGAG.CTCTC located in different helical phasing arrangements with respect to poly(A) tracts and GGGCCC.GGGCCC sequence motifs. Methods used are ligase-mediated cyclization and gel mobility experiments along with DNase I cutting and chemical probe studies. The results provide new evidence for curvature in poly(A) tracts. They also show that the sequence context in which bending and flexible sequence elements are found is an important aspect of sequence-dependent DNA conformation. Although dinucleotide models generally have good predictive power, this work demonstrates that in some instances sequence elements larger than the dinucleotide must be taken into account, and hence it provides a starting point for the appropriate modification and refinement of existing structural models for DNA.

DUB-1, a deubiquitinating enzyme with growth-suppressing activity.

Cytokines regulate cell growth by inducing the expression of specific target genes. Using the differential display method, we have cloned a cytokine-inducible immediate early gene, DUB-1 (for deubiquitinating enzyme). DUB-1 is related to members of the UBP superfamily of deubiquitinating enzymes, which includes the oncoprotein Tre-2. A glutathione S-transferase-DUB-1 fusion protein cleaved ubiquitin from a ubiquitin-beta-galactosidase protein. When a conserved cysteine residue of DUB-1, required for ubiquitin-specific thiol protease activity, was mutated to serine (C60S), deubiquitinating activity was abolished. Continuous expression of DUB-1 from a steroid-inducible promoter induced growth arrest in the G1 phase of the cell cycle. Cells arrested by DUB-1 expression remained viable and resumed proliferation upon steroid withdrawal. Our results suggest that DUB-1 regulates cellular growth by modulating either the ubiquitin-dependent proteolysis or the ubiquitination state of an unknown growth regulatory factor(s).

Copy-up mutants of the plasmid RK2 replication initiation protein are defective in coupling RK2 replication origins.

The broad host range plasmid RK2 replicates and regulates its copy number in a wide range of Gram-negative bacteria. The plasmid-encoded trans-acting replication protein TrfA and the origin of replication oriV are sufficient for controlled replication of the plasmid in all Gram-negative bacteria tested. The TrfA protein binds specifically to direct repeat sequences (iterons) at the origin of replication. A replication control model, designated handcuffing or coupling, has been proposed whereby the formation of coupled TrfA-oriV complexes between plasmid molecules results in hindrance of origin activity and, consequently, a shut-down of plasmid replication under conditions of higher than normal copy number. Therefore, according to this model, the coupling activity of an initiation protein is essential for copy number control and a copy-up initiation protein mutant should have reduced ability to form coupled complexes. To test this model for plasmid RK2, two previously characterized copy-up TrfA mutations, trfA-254D and trfA-267L, were combined and the resulting copy-up double mutant TFrfA protein TrfA-254D/267L was characterized. Despite initiating runaway (uncontrolled) replication in vivo, the copy-up double-mutant TrfA protein exhibited replication kinetics similar to the wild-type protein in vitro. Purified TrfA-254D, TrfA-267L, and TrfA-254D/267L proteins were then examined for binding to the iterons and for coupling activity using an in vitro ligase-catalyzed multimerization assay. It was found that both single and double TrfA mutant proteins exhibited substantially reduced (single mutants) or barely detectable (double mutant) levels of coupling activity while not being diminished in their capacity to bind to the origin of replication. These observations provide direct evidence in support of the coupling model of replication control.

Swiveling-domain mechanism for enzymatic phosphotransfer between remote reaction sites.

The crystal structure of pyruvate phosphate dikinase, a histidyl multiphosphotransfer enzyme that synthesizes adenosine triphosphate, reveals a three-domain molecule in which the phosphohistidine domain is flanked by the nucleotide and the phosphoenolpyruvate/pyruvate domains, with the two substrate binding sites approximately 45 angstroms apart. The modes of substrate binding have been deduced by analogy to D-Ala-D-Ala ligase and to pyruvate kinase. Coupling between the two remote active sites is facilitated by two conformational states of the phosphohistidine domain. While the crystal structure represents the state of interaction with the nucleotide, the second state is achieved by swiveling around two flexible peptide linkers. This dramatic conformational transition brings the phosphocarrier residue in close proximity to phosphoenolpyruvate/pyruvate. The swiveling-domain paradigm provides an effective mechanism for communication in complex multidomain/multiactive site proteins.

Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis lesions.

Mutations in the human Cu,Zn superoxide dismutase gene (SOD1) are found in 20% of kindreds with familial amyotrophic lateral sclerosis. Transgenic mice (line G1H) expressing a human SOD1 containing a mutation of Gly-93 --> Ala (G93A) develop a motor neuron disease similar to familial amyotrophic lateral sclerosis, but transgenic mice (line N1029) expressing a wild-type human SOD1 transgene do not. Because neurofilament (NF)-rich inclusions in spinal motor neurons are characteristic of amyotrophic lateral sclerosis, we asked whether mutant G1H and/or N1029 mice develop similar NF lesions. NF inclusions (i.e., spheroids, Lewy body-like inclusions) were first detected in spinal cord motor neurons of the G1H mice at 82 days of age about the time these mice first showed clinical evidence of disease. Other neuronal intermediate filament proteins (alpha-internexin, peripherin) also accumulated in these spheroids. The onset of accumulations of ubiquitin immunoreactivity in the G1H mice paralleled the emergence of vacuoles and NF-rich spheroids in neurons, but they did not colocalize exclusively with spheroids. In contrast, NF inclusions were not seen in the N1029 mice until they were 132 days old, and ubiquitin immunoreactivity was not increased in the N1029 mice even at 199 days of age. Astrocytosis in spinal cord was associated with a marked increase in glial fibrillary acidic protein immunoreactivity in the G1H mice, but not in the N1029 mice. Finally, comparative studies revealed a striking similarity between the cytoskeletal pathology in the G1H transgenic mice and in patients with amyotrophic lateral sclerosis. These findings link a specific SOD1 mutation with alterations in the neuronal cytoskeleton of patients with amyotrophic lateral sclerosis. Thus, neuronal cytoskeletal abnormalities may be implicated in the pathogenesis of human familial amyotrophic lateral sclerosis.

Structure of a gene encoding a cytosolic acetyl-CoA carboxylase of hexaploid wheat.

An entire gene encoding wheat (var. Hard Red Winter Tam 107) acetyl-CoA carboxylase [ACCase; acetyl-CoA:carbon-dioxide ligase (ADP-forming), EC 6.4.1.2] has been cloned and sequenced. Comparison of the 12-kb genomic sequence with the 7.4-kb cDNA sequence reported previously revealed 29 introns. Within the coding region, the exon sequence is 98% identical to the known wheat cDNA sequence. A second ACCase gene was identified by sequencing fragments of genomic clones that include the first two exons and the first intron. Additional transcripts were detected by 5' and 3' RACE analysis (rapid amplification of cDNA ends). One set of transcripts had a 5' end sequence identical to the cDNA found previously and another set was identical to the gene reported here. The 3' RACE clones fall into four distinguishable sequence sets, bringing the number of ACCase sequences to six. None of these cDNA or genomic clones encodes a chloroplast targeting signal. Identification of six different sequences suggests that either the cytosolic ACCase genes are duplicated in the three chromosome sets in hexaploid wheat or that each of the six alleles of the cytosolic ACCase gene has a readily distinguishable DNA sequence.

Reconstitution of repair-gap UV mutagenesis with purified proteins from Escherichia coli: a role for DNA polymerases III and II.

Using a cell-free system for UV mutagenesis, we have previously demonstrated the existence of a mutagenic pathway associated with nucleotide-excision repair gaps. Here, we report that this pathway can be reconstituted by using six purified proteins: UvrA, UvrB, UvrC, DNA helicase II, DNA polymerase III core, and DNA ligase. This establishes the minimal requirements for repair-gap UV mutagenesis. DNA polymerase II could replace DNA polymerase III, although less effectively, whereas DNA polymerase I, the major repair polymerase, could not. DNA sequence analysis of mutations generated in the in vitro reaction revealed a spectrum typical of mutations targeted to UV lesions. These observations suggest that repair-gap UV mutagenesis is performed by DNA polymerase III, and to a lesser extent by DNA polymerase II, by filling-in of a rare class of excision gaps that contain UV lesions.

Surface hydrophobic residues of multiubiquitin chains essential for proteolytic targeting.

Ubiquitin conjugation is a signal for degradation of eukaryotic proteins by the 26S protease. Conjugation of a homopolymeric multiubiquitin chain to a substrate lysine residue results in 10-fold faster degradation than does conjugation of monoubiquitin, but the molecular basis of enhanced targeting by chains is unknown. We show that ubiquitin residues L8, I44, and V70 are critical for targeting. Mutation of pairs of these residues to alanine had little effect on attachment of ubiquitin to substrates but severely inhibited degradation of the resulting conjugates. The same mutations blocked the binding of chains to a specific subunit (S5a) of the regulatory complex of the 26S protease. The side chains implicated in this binding--L8, I44, and V70--form repeating patches on the chain surface. Thus, hydrophobic interactions between these patches and S5a apparently contribute to enhanced proteolytic targeting by multiubiquitin chains.

Signal-induced degradation of I kappa B alpha requires site-specific ubiquitination.

The inhibitor protein I kappa B alpha controls the nuclear import of the transcription factor NF-kappa B. The inhibitory activity of I kappa B alpha is regulated from the cytoplasmic compartment by signal-induced proteolysis. Previous studies have shown that signal-dependent phosphorylation of serine residues 32 and 36 targets I kappa B alpha to the ubiquitin-proteasome pathway. Here we provide evidence that lysine residues 21 and 22 serve as the primary sites for signal-induced ubiquitination of I kappa B alpha. Conservative Lys-->Arg substitutions at both Lys-21 and Lys-22 produce dominant-negative mutants of I kappa B alpha in vivo. These constitutive inhibitors are appropriately phosphorylated but fail to release NF-kappa B in response to multiple inducers, including viral proteins, cytokines, and agents that mimic antigenic stimulation through the T-cell receptor. Moreover, these Lys-->Arg mutations prevent signal-dependent degradation of I kappa B alpha in vivo and ubiquitin conjugation in vitro. We conclude that site-specific ubiquitination of phosphorylated I kappa B alpha at Lys-21 and/or Lys-22 is an obligatory step in the activation of NF-kappa B.

The human papilloma virus 16E6 gene sensitizes human mammary epithelial cells to apoptosis induced by DNA damage.

Programmed cell death (apoptosis) is a normal physiological process, which could in principle be manipulated to play an important role in cancer therapy. The key importance of p53 expression in the apoptotic response to DNA-damaging agents has been stressed because mutant or deleted p53 is so common in most kinds of cancer. An important strategy, therefore, is to find ways to induce apoptosis in the absence of wild-type p53. In this paper, we compare apoptosis in normal human mammary epithelial cells, in cells immortalized with human papilloma virus (HPV), and in mammary carcinoma cell lines expressing wild-type p53, mutant p53, or no p53 protein. Apoptosis was induced with mitomycin C (MMC), a DNA cross-linking and damaging agent, or with staurosporine (SSP), a protein kinase inhibitor. The normal and HPV-transfected cells responded more strongly to SSP than did the tumor cells. After exposure to MMC, cells expressing wild-type p53 underwent extensive apoptosis, whereas cells carrying mutated p53 responded weakly. Primary breast cancer cell lines null for p53 protein were resistant to MMC. In contrast, two HPV immortalized cell lines in which p53 protein was destroyed by E6-modulated ubiquitinylation were highly sensitive to apoptosis induced by MMC. Neither p53 mRNA nor protein was induced in the HPV immortalized cells after MMC treatment, although p53 protein was elevated by MMC in cells with wild-type p53. Importantly, MMC induced p21 mRNA but not p21 protein expression in the HPV immortalized cells. Thus, HPV 16E6 can sensitize mammary epithelial cells to MMC-induced apoptosis via a p53- and p21-independent pathway. We propose that the HPV 16E6 protein modulates ubiquitin-mediated degradation not only of p53 but also of p21 and perhaps other proteins involved in apoptosis.

Modes of expression and common structural features of the complete phenylalanine ammonia-lyase gene family in parsley.

We describe a complete gene family encoding phenylalanine ammonia-lyase (PAL; EC 4.3.1.5) in one particular plant species. In parsley (Petroselinum crispum), the PAL gene family comprises two closely related members, PAL1 and PAL2, whose TATA-proximal promoter and coding regions are almost identical, and two additional members, PAL3 and PAL4, with less similarity to one another and to the PAL1 and PAL2 genes. Using gene-specific probes derived from the 5' untranslated regions of PAL1/2, PAL3, and PAL4, we determined the respective mRNA levels in parsley leaves and cell cultures treated with UV light or fungal elicitor and in wounded leaves and roots. For comparison, the functionally closely related cinnamate 4-hydroxylase (C4H) and 4-coumarate:CoA ligase (4CL) mRNAs were measured in parallel. The results indicate various degrees of differential responsiveness of PAL4 relative to the other PAL gene family members, in contrast to a high degree of coordination in the overall expression of the PAL, C4H, and 4CL genes. The only significant sequence similarities shared by all four PAL gene promoters are a TATA-proximal set of three putative cis-acting elements (boxes P, A, and L). None of these elements alone, or the promoter region containing all of them together, conferred elicitor or light responsiveness on a reporter gene in transient expression assays. The elements appear to be necessary but not sufficient for elicitor- or light-mediated PAL gene activation, similar to the situation previously reported for 4CL.

Basis for selection of improved carbohydrate-binding single-chain antibodies from synthetic gene libraries.

A technique is described for the simultaneous and controlled random mutation of all three heavy or light chain complementarity-determining regions (CDRs) in a single-chain Fv specific for the O polysaccharide of Salmonella serogroup B. Sense oligonucleotides were synthesized such that the central bases encoding a CDR were randomized by equimolar spiking with A, G, C, and T at a level of 10% while the antisense strands contained inosine in the spiked regions. Phage display of libraries assembled from the spiked oligonucleotides by a synthetic ligase chain reaction demonstrated a bias for selection of mutants that formed dimers and higher oligomers. Kinetic analyses showed that oligomerization increased association rates in addition to slowing dissociation rates. In combination with some contribution from reduced steric clashes with residues in heavy-chain CDR2, oligomerization resulted in functional affinities that were much higher than that of the monomeric form of the wild-type single-chain Fv.

Isolation of a cDNA encoding human holocarboxylase synthetase by functional complementation of a biotin auxotroph of Escherichia coli.

Holocarboxylase synthetase (HCS) catalyzes the biotinylation of the four biotin-dependent carboxylases in human cells. Patients with HCS deficiency lack activity of all four carboxylases, indicating that a single HCS is targeted to the mitochondria and cytoplasm. We isolated 21 human HCS cDNA clones, in four size classes of 2.0-4.0 kb, by complementation of an Escherichia coli birA mutant defective in biotin ligase. Expression of the cDNA clones promoted biotinylation of the bacterial biotinyl carboxyl carrier protein as well as a carboxyl-terminal fragment of the alpha subunit of human propionyl-CoA carboxylase expressed from a plasmid. The open reading frame encodes a predicted protein of 726 aa and M(r) 80,759. Northern blot analysis revealed the presence of a 5.8-kb major species and 4.0-, 4.5-, and 8.5-kb minor species of poly(A)+ RNA in human tissues. Human HCS shows specific regions of homology with the BirA protein of E. coli and the presumptive biotin ligase of Paracoccus denitrificans. Several forms of HCS mRNA are generated by alternative splicing, and as a result, two mRNA molecules bear different putative translation initiation sites. A sequence upstream of the first translation initiation site encodes a peptide structurally similar to mitochondrial presequences, but it lacks an in-frame ATG codon to direct its translation. We anticipate that alternative splicing most likely mediates the mitochondrial versus cytoplasmic expression, although the elements required for directing the enzyme to the mitochondria remain to be confirmed.