Reciprocal Self-Assembly of Peptide-DNA Conjugates into a Programmable Sub-10-nm Supramolecular Deoxyribonucleoprotein.

To overcome the limitations of molecular assemblies, the development of novel supramolecular building blocks and self-assembly modes is essential to create more sophisticated, complex, and controllable aggregates. The self-assembly of peptide-DNA conjugates (PDCs), in which two orthogonal self-assembly modes, that is, ß-sheet formation and Watson-Crick base pairing, are covalently combined in one supramolecular system, is reported. Despite extensive research, most self-assembly studies have focused on using only one type of building block, which restricts structural and functional diversity compared to multicomponent systems. Multicomponent systems, however, suffer from poor control of self-assembly behaviors. Covalently conjugated PDC building blocks are shown to assemble into well-defined and controllable nanostructures. This controllability likely results from the decrease in entropy associated with the restriction of the molecular degrees of freedom by the covalent constraints. Using this strategy, the possibility to thermodynamically program nano-assemblies to exert gene regulation activity with low cytotoxicity is demonstrated.

An archaeal RadA paralog influences presynaptic filament formation.

Recombinases of the RecA family play vital roles in homologous recombination, a high-fidelity mechanism to repair DNA double-stranded breaks. These proteins catalyze strand invasion and exchange after forming dynamic nucleoprotein filaments on ssDNA. Increasing evidence suggests that stabilization of these dynamic filaments is a highly conserved function across diverse species. Here, we analyze the presynaptic filament formation and DNA binding characteristics of the Sulfolobus solfataricus recombinase SsoRadA in conjunction with the SsoRadA paralog SsoRal1. In addition to constraining SsoRadA ssDNA-dependent ATPase activity, the paralog also enhances SsoRadA ssDNA binding, effectively influencing activities necessary for presynaptic filament formation. These activities result in enhanced SsoRadA-mediated strand invasion in the presence of SsoRal1 and suggest a filament stabilization function for the SsoRal1 protein.

Gel-based fluorescence resonance energy transfer (gelFRET) analysis of nucleoprotein complex architecture.

A gel-based fluorescence resonance energy transfer (gelFRET) assay was developed for analysis of the architecture of nucleoprotein complexes. gelFRET is based on fluorescence analysis of nucleoprotein complexes separated by polyacrylamide gel electrophoresis. These complexes are separated from free components and nonspecific complexes, enabling fluorescence analysis of complexes containing all components in stoichiometric proportions. gelFRET can be used to investigate the structural organization of nucleoprotein complexes through comparison of the relative efficiencies of energy transfer from donor fluorophores linked to different positions on DNA to an acceptor fluorophore linked to a unique position on the binding protein. We have applied gelFRET to analysis of the orientation of binding by heterodimeric transcription factors. By using Fos-Jun heterodimers as a model system we have identified the structural determinants that control the orientation of heterodimer binding. gelFRET can be applied to studies of a variety of biological processes that influence the proximity of two sites within a complex, such as the assembly of transcription regulatory complexes.

Nucleoprotein gene tracking: localization of specific HIV-1 genes to subchromatin nucleoprotein complexes containing endonuclease activity in HIV-1-infected human cells.

We developed a technique with which to isolate specific subchromatin deoxyribonucleoprotein/ribonucleoprotein precursor complexes containing discrete genes from intact mammalian nuclei by direct restriction enzyme treatment with MspI. These nucleoprotein complexes can be further fractionated and purified by two-dimensional isoelectric focusing/sodium dodecyl sulfate-polyacrylamide gel electrophoresis. After electroelution and removal of detergent, virtually thousands of nucleoprotein complexes can be examined for the presence of tightly bound genes and enzymatic activities. Nucleoprotein gene tracking procedures were applied to study the acidic nucleoprotein complexes from steady-state human H9 cells uninfected or infected with human immunodeficiency virus type 1 (HIV-1) virus. The purified nucleoprotein complexes were screened for the presence of loosely and tightly associated HIV-1 gene sequences (pol, env, tat, and rev) using a DNA hybridization protocol. In HIV-1-infected cells, four specific nucleoprotein complexes out of several hundred were found to contain tightly bound HIV-1 pol gene sequences after purification. By contrast, the other HIV-1 gene sequences (env, tat, and rev) were not tightly bound to any of the nucleoprotein complexes in HIV-infected cells. The observations suggest that certain HIV-1 genes associate with specific chromatin nucleoprotein complexes, regardless of their pattern of DNA integration into the human genome. At least two of the HIV-1 pol-containing nucleoprotein complexes of apparent M(r) approximately 94,000, pI approximately 6.5, and M(r) approximately 47,000, pI approximately 5.1 contain DNA endonuclease activity. This was confirmed in the present study, using linearized pUC19 plasmid substrate. The technique can be used to study a variety of problems concerning the association of specific genes and enzymes with specific nucleoprotein complexes J. Cell. Biochem. Suppls. 32/33:158-165, 1999.

DNA-binding domains of Fos and Jun do not induce DNA curvature: an investigation with solution and gel methods.

We demonstrate the use of a DNA minicircle competition binding assay, together with DNA cyclization kinetics and gel-phasing methods, to show that the DNA-binding domains (dbd) of the heterodimeric leucine zipper protein Fos-Jun do not bend the AP-1 target site. Our DNA constructs contain an AP-1 site phased by 1-4 helical turns against an A-tract-directed bend. Competition binding experiments reveal that (dbd)Fos-Jun has a slight preference for binding to linear over circular AP-1 DNAs, independent of whether the site faces in or out on the circle. This result suggests that (dbd)Fos-Jun slightly stiffens rather than bends its DNA target site. A single A-tract bend replacing the AP-1 site is readily detected by its effect on cyclization kinetics, in contrast to the observations for Fos-Jun bound at the AP-1 locus. In contrast, comparative electrophoresis reveals that Fos-Jun-DNA complexes, in which the A-tract bend is positioned close (1-2 helical turns) to the AP-1 site, show phase-dependent variations in gel mobilities that are comparable with those observed when a single A-tract bend replaces the AP-1 site. Whereas gel mobility variations of Fos-Jun-DNA complexes decrease linearly with increasing Mg2+ contained in the gel, the solution binding preference of (dbd)Fos-Jun for linear over circular DNAs is independent of Mg2+ concentration. Hence, gel mobility variations of Fos-Jun-DNA complexes are not indicative of (dbd)Fos-Jun-induced DNA bending (upper limit 5 degrees) in the low salt conditions of gel electrophoresis. Instead, we propose that the gel anomalies depend on the steric relationship of the leucine zipper region with respect to a DNA bend.

Characterization of the ATF/CREB site and its complex with GCN4.

We have studied DNA minicircles containing the ATF/CREB binding site for GCN4 by using a combination of cyclization kinetics experiments and Monte Carlo simulations. Cyclization rates were determined with and without GCN4 for DNA constructs containing the ATF/CREB site separated from a phased A-tract multimer bend by a variable length phasing adaptor. The cyclization results show that GCN4 binding does not significantly change the conformation of the ATF/CREB site, which is intrinsically slightly bent toward the major groove. Monte Carlo simulations quantitate the ATF/CREB site structure as an 8 degrees bend toward the major groove in a coordinate frame near the center of the site. The ATF/CREB site is underwound by 53 degrees relative to the related AP-1 site DNA. The effect of GCN4 binding can be modeled either as a decrease in the local flexibility, corresponding to an estimated 60% increase in the persistence length for the 10-bp binding site, or possibly as a small decrease (1 degrees) in intrinsic bend angle. Our results agree with recent electrophoretic and crystallographic studies and demonstrate that cyclization and simulation can characterize subtle changes in DNA structure and flexibility.

A previously unidentified host protein protects retroviral DNA from autointegration.

Integration of a DNA copy of the viral genome into a host chromosome is an essential step in the retrovirus life cycle. The machinery that carries out the integration reaction is a nucleoprotein complex derived from the core of the infecting virion. To successfully integrate into host DNA, the viral DNA within this complex must avoid self-destructive integration into itself, a reaction termed autointegration. We have previously shown [Lee, M. S. and Craigie, R. (1994) Proc. Natl. Acad. Sci. USA 91, 9823-9827] that viral nucleoprotein complexes isolated from Moloney murine leukemia virus-infected cells exhibit a barrier to autointegration. This autointegration barrier could be destroyed by stripping factors from the complexes and subsequently restored by incubation with a host cell extract, but not by incubation with an extract of disrupted virions. We have now used this autointegration barrier reconstitution assay to purify the host factor from uninfected NIH 3T3 fibroblasts. It is a single polypeptide of 89 aa that does not match any previously identified protein. The identity of the protein was confirmed by expressing it in Escherichia coli and demonstrating the activity of the heterologously expressed protein in the reconstitution assay.

A novel DNA recognition mode by the NF-kappa B p65 homodimer.

The crystal structure of the NF-kappa B p65 (RelA) homodimer in complex with a DNA target has been determined to 2.4 A resolution. The two p65 subunits are not symmetrically disposed on the DNA target. The homodimer should optimally bind to a pseudo-palindromic nine base pair target with each subunit recognizing a 5'GGAA-3' half site separated by a central A-T base pair. However, one of the subunits (subunit B) encounters a half site of 5'-GAAA-3'. The single base-pair change from G-C to A-T results in highly unfavorable interactions between this half site and the base contacting protein residues in subunit B, which leads to an 18 degrees rotation of the N-terminal terminal domain from its normal conformation. Remarkably, subunit B retains all the interactions with the sugar phosphate backbone of the DNA target. This mode of interaction allows the NF-kappa B p65 homodimer to recognize DNA targets containing only one cognate half site. Differences in the sequence of the other half site provide variations in conformation and affinity of the complex.

Natural resistance of acute myeloid leukemia cell lines to mitoxantrone is associated with lack of apoptosis.

The purpose of this study was to characterize mitoxantrone-induced cytotoxicity in KG1a and TF-1, two P-glycoprotein expressing AML cell lines which display early differentiation phenotypes, compared to more mature HL-60 and U937 cells. KG1a and TF-1 cells were found to be 30-40-fold more resistant to mitoxantrone than HL-60 and U937 cells. Uptake and efflux of mitoxantrone were similar for all cell lines. Moreover, a potent P-glycoprotein blocker (PSC833) had no impact on either accumulation or efflux. No differences were found in the appearance and removal of mitoxantrone-induced DNA-protein complexes. These results suggest that resistance of KG1a and TF-1 cells is not related to a decreased interaction between mitoxantrone and topoisomerase II. Further studies showed that the mechanisms of cell death were different for sensitive and resistant cell lines. Thus, mitoxantrone induced rapid apoptotic cell death in sensitive cells as indicated by characteristic morphological changes and both high molecular weight and internucleosomal DNA fragmentation. In contrast, mitoxantrone induced a G2-M block in resistant cells followed by a progressive loss of viability with necrotic features. Neither oligonucleosomal nor large DNA fragments were detected in these cells during a post-treatment period of up to 96 h. Finally, drug-induced activation of the AP-1 transcription factor was higher in resistant cell lines than in sensitive ones whereas activation of NF-kappaB was comparable. Therefore, our study provides evidence that certain AML cells display natural resistance to mitoxantrone which is independent of drug transport and drug-target interactions but appears to be associated with the inability of the drug to induce apoptosis in these cells.

The Rep78 gene product of adeno-associated virus (AAV) self-associates to form a hexameric complex in the presence of AAV ori sequences.

The Rep78 and Rep68 proteins of adeno-associated virus (AAV) are replication initiator proteins that bind the viral replicative-form origin of replication, nick the origin in a site- and strand-specific fashion, and mediate vectorial unwinding of the DNA duplex via an ATP-dependent helicase activity, thus initiating a strand displacement mechanism of viral DNA replication. Genetic and biochemical studies have identified Rep mutants that demonstrate a trans-dominant negative phenotype in vitro and in vivo, suggesting the possibility that multimerization of Rep is essential for certain replicative functions. In this study, we have investigated the ability of the largest of the Rep proteins, Rep78, to self-associate in vitro and in vivo. Self-association of Rep78 in vivo was demonstrated through the use of a mammalian two-hybrid system. Rep-Rep protein interaction was confirmed in vitro through coimmunoprecipitation experiments with a bacterially expressed maltose-binding protein-Rep78 fusion protein in combination with [35S]methionine-labeled Rep78 synthesized in a coupled in vitro transcription-translation system. Mapping studies with N- and C-terminal truncation mutant forms of Rep indicate that amino acid sequences required for maximal self-association occur between residues 164 and 484. Site-directed mutagenesis identified two essential motifs within this 321-amino-acid region: (i) a putative alpha-helix bearing a 3,4-hydrophobic heptad repeat reminiscent of those found in coiled-coil domains and (ii) a previously recognized nucleoside triphosphate-binding motif. Deletion of either of these regions from the full-length polypeptide resulted in severe impairment of Rep-Rep interaction. In addition, gel filtration chromatography and protein cross-linking experiments indicated that Rep78 forms a hexameric complex in the presence of AAV ori sequences.

Mutations affecting lysine-35 of gpNu1, the small subunit of bacteriophage lambda terminase, alter the strength and specificity of holoterminase interactions with DNA.

The small subunit of lambda terminase, gpNu1, contains a low-affinity ATPase activity that is stimulated by nonspecific dsDNA. The location of the gpNu1 ATPase center is suggested by a sequence match between gpNu1 (29-VLRGGGKG-36) and the phosphate-binding loop, or P-loop (GXXXXGKT/S), of known ATPase. The proposed P-loop of gpNu1 is just downstream of a putative helix-turn-helix DNA-binding motif, located between residues 5 and 24. Published work has shown that changing lysine-35 of the proposed P-loop of gpNu1 alters the response of the ATPase activity to DNA, as follows. The changes gpNu1 k35A and gpNu1 K35D increase the level of DNA required for maximal stimulation of the gpNu1 ATPase by factors of 2- and 10-fold, respectively. The maximally stimulated ATPase activities of the mutant enzymes are indistinguishable from that of the wild-type enzyme. In the present work, the effects of changing lysine-35 on the cos-cleavage and DNA-packaging activities of terminase were examined. In vitro, the gpNu1 K35A enzyme cleaved cos as efficiently as the wild-type enzyme, but required a 2-fold increased level of substrate DNA for saturation, suggesting a slight reduction in DNA affinity. In a crude DNA-packaging system using cleaved lambda DNA as substrate, the gpNu1 K35A enzyme had a 10-fold defect. In vivo, lambda Nu1 K35A showed a 2-fold reduction in cos cleavage, but no packaged DNA was detected. The primary defect of the gpNu1 K35A enzyme was concluded to be in a post-cos-cleavage step of DNA packaging. In in vitro cos-cleavage experiments, the gpNu1 K35D enzyme had a 10-fold increased requirement for saturation by substrate DNA. Furthermore, the cos-cleavage activity of gpNu1 K35D enzyme was strongly inhibited by the presence of nonspecific DNA, indicating that the gpNu1 K35D enzyme is unable to discriminate effectively between cos and nonspecific DNA. No cos cleavage was observed in vivo for lambda Nu1 K35D, a result consistent with the discrimination defect found in vitro for the gpNu1 K35D enzyme. In a crude packaging system the gpNu1 K35D enzyme had a 200-fold defect; in a purified packaging system, the gpNu1 K35D enzyme was found to be unable to discriminate between lambda DNA and nonspecific phage T7 DNA, a result indicating that the gpNu1 K35D enzyme is also defective in discriminating between lambda DNA and nonspecific DNA during DNA packaging.

Detection of DNA bending in a DNA-PAP1 protein complex by fluorescence resonance energy transfer.

The structure of DNA in a DNA-protein complex was studied by means of fluorescence resonance energy transfer (FRET) method. Oligonucleotide phosphorothioates were labeled with fluorescein and eosin to obtain a donor- and acceptor-labeled DNA. The formation of a complex of the DNA with PAP1(70), which is a DNA binding site fragment derived from transcription regulatory protein, PAP1, of fission yeast, was confirmed by gel retardation analysis and fluorescence measurements. FRET of the donor- and acceptor-labeled DNA with and without PAP1(70) indicated that the DNA in the complex was bent about 26 degrees toward the protein-binding surface.

In organello footprinting analysis of rat mitochondrial DNA: protein interaction upstream of the Ori-L.

An in organello footprinting approach has been used to probe a protein-DNA interaction of a nuclear coded 25 kDa protein, previously isolated in our laboratory, that binds "in vitro" a region within the ND2 gene, located upstream of the Ori-L. Footprinting studies with the purine-modifying reagent dimethyl sulfate and the pirimidine-modifying reagent potassium permanganate were carried out in isolated mitochondria from rat liver. Dimethyl sulfate footprinting has allowed the detection of a protein-DNA interaction within the curved ND2 region with contact sites located in both the strands. Potassium permanganate footprinting allowed detection of an adjacent permanganate-reactive region. We hypothesize that the permanganate-reactive region is a single stranded DNA due to a profound helix distortion induced by a 25 kDa protein binding to the nearest region.

Mapping the contacts of yeast TFIIIB and RNA polymerase III at various distances from the major groove of DNA by DNA photoaffinity labeling.

The structure of the Saccharomyces cerevisiae RNA polymerase III transcription complex on the SUP4 tRNATyr gene was probed at distances of approximately 10 to approximately 23 A from the C-5 methyl of thymidine in the major groove of DNA using photoreactive aryl azides attached to deoxyuridine by variable chain lengths. The nucleotide analogs contained an azidobenzoyl group attached with chain lengths that were incrementally increased by approximately 4. 3 A by inserting 1-3 glycine residues into the chain. Another photoreactive deoxyuridine analog was made that contained a butyl chain (ABU-dUMP) to assess the effect of the chain's hydrophobicity on its ability to photoaffinity label the transcription complex. These nucleotide analogs were incorporated at base pairs (bp) -26/-21, -17, or -3/-2 on the nontranscribed strand of the SUP4 tRNATyr gene along with an [alpha-32P]dNMP by primer extension using an immobilized single-stranded DNA template annealed to specific oligonucleotides. The 27-kDa subunit of TFIIIB or the TATA box binding protein was photoaffinity labeled at bp -26/-21 with nucleotide analogs containing a approximately 19- or approximately 23-A chain and not with shorter chains of approximately 10 to approximately 15 A in length. The B" subunit of TFIIIB (Mr = 90 kDa) was photoaffinity labeled at bps -26/-21 with DNA containing a approximately 14-A chain and not with shorter or longer chains. Cross-linking of the B" subunit was inhibited by binding of RNA polymerase III (Pol III) to the TFIIIB-DNA complex and suggested that Pol III binding causes a conformational change in the TFIIIB-DNA complex resulting in the displacement of the 90-kDa subunit at bps -26/-21. Next, the chain length dependence of photoaffinity labeling the 34-kDa subunit of Pol III at bps -17 and -3/-2 indicated that the 34-kDa subunit of Pol III is slightly removed from the major groove at bp -17 in the initiation complex and makes closer contact at bps -3/-2 in a stalled elongation complex.

NMR studies of the Escherichia coli Trp repressor.trpRs operator complex.

To understand the specificity of the Escherichia coli Trp repressor for its operators, we have begun to study complexes of the protein with alternative DNA sequences, using 1H-NMR spectroscopy. We report here the 1H-NMR chemical shifts of a 20-bp oligodeoxynucleotide containing the sequence of a symmetrised form of the trpR operator in the presence and absence of the holorepressor. Deuterated protein was used to assign the spectrum of the oligodeoxynucleotide in a 37-kDa complex with the Trp holorepressor. Many of the resonances of the DNA shift on binding to the protein, which suggests changes in conformation throughout the sequence. The largest changes in shifts for the aromatic protons in the major groove are for A15 and G16, which are thought to hydrogen bond to the protein, possibly via water molecules. We have also examined the effect of DNA binding on the corepressor, tryptophan, in this complex. The indole proton resonance of the tryptophan undergoes a downfield shift of 1.2 ppm upon binding of DNA. This large shift is consistent with hydrogen bonding of the tryptophan to the phosphate backbone of the trpR operator DNA, as in the crystal structure of the holoprotein with the trp operator.

Photoaffinity labelling of a DNA-binding site on the globular domain of histone H5.

We have labelled a DNA-binding site on the globular domain of histone H5 (GH5) by ultraviolet-activated cross-linking of a self-complementary 5-bromodeoxyuridine (5BrU)-substituted oligonucleotide with the sequence 5'-AGCGA5BrUATCGCT-3'. Cross-linking was to His62, mainly to the protein backbone. This observation provides further support for the mode of binding of GH5 to DNA proposed on the basis of the similarity between the X-ray crystal structure of GH5 and the DNA-bound structures of catabolite activator protein and hepatic nuclear factor 3 gamma [Ramakrishnan, V. (1994) Curr. Opin. Struct. Biol. 4. 44-50].

Activation of replication origins in phi29-related phages requires the recognition of initiation proteins to specific nucleoprotein complexes.

Protein p6 of Bacillus subtilis phage phi29 activates the initiation of viral DNA replication by forming a multimeric nucleoprotein complex at the origins of replication, located at both ends of the linear genome. This activation requires a precise positioning of the protein p6 array with respect to the initiation site. To investigate this activation mechanism, we have purified the phi29 protein p6 counterparts from the related phages Nf and GA-1 and analyzed the formation of complexes with DNA. In the homologous protein p6-DNA complexes the phi29 and Nf protein arrays showed an identical positioning, different than that of the GA-1 protein array. In contrast, in the heterologous complexes the protein showed a different arrangement except in the case of the Nf protein-phi29 DNA complex. We have also purified the proteins involved in the initiation of replication (terminal protein and DNA polymerase) from phages Nf and GA-1 and measured the ability of the different p6 proteins to activate homologous and heterologous replication origins. The results obtained indicate that the activation requires not only the formation of a specific nucleoprotein complex but also its specific recognition by the proteins involved in the initiation of DNA replication.

Transcription factor TFIIH and DNA endonuclease Rad2 constitute yeast nucleotide excision repair factor 3: implications for nucleotide excision repair and Cockayne syndrome.

Nucleotide excision repair (NER) of ultraviolet light-damaged DNA in eukaryotes requires a large number of highly conserved protein factors. Recent studies in yeast have suggested that NER involves the action of distinct protein subassemblies at the damage site rather than the placement there of a "preformed repairosome" containing all the essential NER factors. Neither of the two endonucleases, Rad1-Rad10 and Rad2, required for dual incision, shows any affinity for ultraviolet-damaged DNA. Rad1-Rad10 forms a ternary complex with the DNA damage recognition protein Rad14, providing a means for targeting this nuclease to the damage site. It has remained unclear how the Rad2 nuclease is targeted to the DNA damage site and why mutations in the human RAD2 counterpart, XPG, result in Cockayne syndrome. Here we examine whether Rad2 is part of a higher order subassembly. Interestingly, we find copurification of Rad2 protein with TFIIH, such that TFIIH purified from a strain that overexpresses Rad2 contains a stoichiometric amount of Rad2. By several independent criteria, we establish that Rad2 is tightly associated with TFIIH, exhibiting an apparent dissociation constant < 3.3 x 10(-9) M. These results identify a novel subassembly consisting of TFIIH and Rad2, which we have designated as nucleotide excision repair factor 3. Association with TFIIH provides a means of targeting Rad2 to the damage site, where its endonuclease activity would mediate the 3' incision. Our findings are important for understanding the manner of assembly of the NER machinery and they have implications for Cockayne syndrome.

Protein and DNA requirements for the transcription factor IIIA-induced distortion of the 5 S rRNA gene promoter.

Transcription factor-induced DNA distortion has become a common theme in eukaryotic gene regulation. A number of techniques have been applied to the study of transcription factor-induced DNA bending and flexibility including electron microscopy, circular permutation gel analysis, helical phasing gel analysis and cyclisation kinetics in solution. We have applied these techniques in order to assess the role that specific DNA sequences and protein domains of transcription factor IIIA (TFIIIA) play in the TFIIIA-induced distortion of the Xenopus 5 S ribosomal RNA gene promoter. Electron spectroscopic imaging analysis of TFIIIA:DNA complexes indicate that TFIIIA binding involves compaction of the 5 S promoter into a precise three-dimensional hairpin-shaped structure. This compaction can be detected utilising circular permutation gel analysis and the distortion results in an apparent bend angle of 55 to 60 degrees near the centre of the TFIIIA binding site. Helical phasing analysis demonstrates that the 60 degrees bend angle as measured by circular permutation can be detected as a static bend directed towards the minor groove between bases +63 and +64 of the 5 S rRNA gene. The amplitude of the TFIIIA:5 S gene phasing signal is similar to the phasing signal obtained utilising bacterial CAP:DNA complexes with bend angles of approximately 90 degrees. These results are supported by phased ligase-mediated cyclisation kinetics in solution. Analysis of DNA deletion constructs indicate that the 5' A block of the internal 5 S gene promoter, which is required for transcriptional activity, is also required for TFIIIA-induced distortion of the 5 S gene promoter. Analysis of the N-terminal papain fragment of TFIIIA indicates that the 34 kDa zinc finger DNA binding domain is sufficient for compaction of the 5 S gene promoter. These results are discussed in relation to the modular model of TFIIIA:DNA interaction in which individual zinc fingers contribute to the protein-induced distortion of the DNA helix and overall DNA binding affinity in a complex, non-additive fashion.