A chemical phylogeny of group I introns based upon interference mapping of a bacterial ribozyme.

Despite its small size, the 205 nt group I intron from Azoarcus tRNA(Ile) is an exceptionally stable self-splicing RNA. This IC3 class intron retains the conserved secondary structural elements common to group I ribozymes, but lacks several peripheral helices. These features make it an ideal system to establish the conserved chemical basis of group I intron activity. We collected nucleotide analog interference mapping (NAIM) data of the Azoarcus intron using 14 analogs that modified the phosphate backbone, the ribose sugar, or the purine base functional groups. In conjunction with a complete interference set collected on the Tetrahymena group I intron (IC1 class), these data define a "chemical phylogeny" of functional groups that are important for the activity of both introns and that may be common chemical features of group I intron catalysts. The data identify the functional moieties most likely to play a conserved role as ligands for catalytic metal ions, the substrate helix, and the guanosine cofactor. These include backbone functional groups whose nucleotide identity is not conserved, and hence are difficult to identify by standard phylogenetic sequence comparisons. The data suggest that both introns utilize an equivalent set of long range tertiary interactions for 5'-splice site selection between the P1 substrate helix and its receptor in the J4/5 asymmetric bulge, as well as an equivalent set of 2'-OH groups for P1 helix docking into most of the single stranded segment J8/7. However, the Azoarcus intron appears to make an alternative set of interactions at the base of the P1 helix and at the 5'-end of the J8/7. Extensive differences were observed within the intron peripheral domains, particularly in P2 and P8 where the Azoarcus data strongly support the proposed formation of a tetraloop-tetraloop receptor interaction. This chemical phylogeny for group I intron catalysis helps to refine structural models of the RNA active site and identifies functional groups that should be carefully investigated for their role in transition state stabilization.

A specific monovalent metal ion integral to the AA platform of the RNA tetraloop receptor.

Metal ions are essential for the folding and activity of large catalytic RNAs. While divalent metal ions have been directly implicated in RNA tertiary structure formation, the role of monovalent ions has been largely unexplored. Here we report the first specific monovalent metal ion binding site within a catalytic RNA. As seen crystallographically, a potassium ion is coordinated immediately below AA platforms of the Tetrahymena ribozyme P4-P6 domain, including that within the tetraloop receptor. Interference and kinetic experiments demonstrate that potassium ion binding within the tetraloop receptor stabilizes the folding of the P4-P6 domain and enhances the activity of the Azoarcus group I intron. Since a monovalent ion binding site is integral to the tetraloop receptor, a tertiary structural motif that occurs frequently in RNA, monovalent metal ions are likely to participate in the folding and activity of a wide diversity of RNAs.

Effect of base composition on DNA bending by phosphate neutralization.

Of the many forces involved in DNA bending by proteins, we have focused on the possible role of asymmetric phosphate neutralization due to interactions between the negatively charged phosphate backbone of duplex DNA and cationic amino acids of an approaching protein. The resulting unbalanced charge distribution along the duplex DNA is thought to induce the double helix to collapse toward the neutralized surface. Previous work has confirmed that DNA bending (approximately 20.7 +/- 4 degrees) is induced by asymmetric incorporation of six uncharged racemic methylphosphonate analogs partially neutralizing one face of GC-rich duplex DNA. We have now analyzed DNA duplexes with similar patches of methylphosphonate linkages in an AT-rich sequence context and again observe bending toward the neutralized face, to an extent (20 +/- 0.6 degrees) comparable to that observed for neutral patches in GC-rich DNA. The similar induced bend angles in AT-rich and GC-rich contexts does not reveal increased flexibility in AT-rich sequences, or a particular propensity of A-T base pairs to roll toward the minor groove in the tested sequences.

Effects of phosphate neutralization on the shape of the AP-1 transcription factor binding site in duplex DNA.

Previous electrophoretic experiments suggest that the AP-1 site in duplex DNA bends in response to the pattern of amino acid charges distal to the basic region in bound bZIP proteins. The extent and direction of apparent DNA bending are consistent with the prediction that DNA will collapse locally upon asymmetric phosphate charge neutralization. To prove that asymmetric phosphate neutralization could produce the observed degree of DNA bending, the present experiments partially substitute anionic phosphate diesters in the AP-1 site with various numbers of neutral methylphosphonate linkages. DNA bending is induced toward the neutralized face of DNA. The degree of DNA bending induced by methylphosphonate substitution (approximately 3.5 degrees per neutralized phosphate) is comparable to that induced by GCN4 variants carrying increasing numbers of additional basic amino acids. It is plausible, therefore, that asymmetric phosphate neutralization is the cause of DNA bending in such complexes.

Electrostatic effects in DNA bending by GCN4 mutants.

DNA architecture has been shown to be important for cellular processes such as activation of transcription, recombination, and replication. Many proteins reconfigure the shape of duplex DNA upon binding. Previous experiments have shown that some members of the eukaryotic bZIP family of DNA binding proteins appear to bend DNA, while others do not. We are exploring the role of electrostatic effects in DNA bending by bZIP proteins. The yeast bZIP transcription factor GCN4 does not induce DNA bending in vitro. Previously we substituted basic residues for three neutral amino acids in GCN4 to produce a GCN4 derivative that bends DNA by approximately 15 degrees. This result is consistent with a model of induced DNA bending wherein excess positive charge in proximity to one face of the double helix neutralizes local phosphate diester anions resulting in a laterally-asymmetric charge distribution along the DNA. Such an unbalanced charge distribution can result in collapse of the DNA toward the neutralized surface. We now present a more comprehensive analysis of electrostatic effects in DNA bending by GCN4 derivatives. It is shown that the direction and extent of DNA bending by these derivatives are a linear function of the charges of the amino acids adjacent to the basic domain of the protein. This relation holds over the charge range +6 (16 degrees bend toward the minor groove) to -6 (25 degrees bend toward the major groove).

DNA bending by GCN4 mutants bearing cationic residues.

Transcription activation is thought to require DNA bending to promote the interaction of upstream activators and the basal transcription machinery. Previous experiments have shown that some members of the bZIP family of DNA binding proteins bend DNA, while others do not. We are exploring the possibility that electrostatic effects play a role in these differences. The yeast bZIP transcription factor GCN4 does not induce DNA bending in vitro. Substitution of basic residues for three neutral amino acids of GCN4 confers the ability to bend DNA. This result is consistent with a model of induced DNA bending wherein excess positive charge in proximity to one face of the double helix neutralizes local phosphate diester anions resulting in a laterally asymmetric charge distribution along the DNA. Previous data suggest that such an unbalanced charge distribution results in collapse of the DNA toward the neutralized surface. Interpretations of the present data are discussed. Our result supports the hypothesis that electrostatic interactions can play a key role in DNA bending by bZIP proteins.

Effects of neutralization pattern and stereochemistry on DNA bending by methylphosphonate substitutions.

Asymmetric phosphate neutralization has been hypothesized to play a role in DNA bending by proteins. Neutralization is thought to involve salt bridges between the negatively charged phosphate backbone of duplex DNA and the cationic amino acids of an approaching protein. According to this model, the resulting unbalanced charge distribution along the duplex DNA induces the double helix to collapse toward the neutralized surface. Previous work has confirmed that DNA bending is induced by the asymmetric incorporation of racemic methylphosphonate linkages creating a neutral region on one face of duplex DNA. Neutralization was accomplished by substitution of three consecutive phosphodiesters on each strand, arranged across one minor groove of the DNA (a total of six neutralized phosphates). We now measure DNA bending induced by a more diffuse patch of neutralization (alternating neutralized and anionic phosphates) and explore the effect of methylphosphonate stereochemistry. DNA duplexes with patches of alternating methylphosphonate and phosphodiester linkages are less bent than DNAs wherein consecutive phosphates are neutralized. Furthermore, duplexes neutralized by incorporation of pure (RP)-methylphosphonate isomers are bent approximately 30% less than duplexes neutralized by racemic methylphosphonates.

Role of asymmetric phosphate neutralization in DNA bending by PU.1.

The PU.1 transcription factor is a member of the Ets family of DNA binding proteins. PU.1 binds to DNA via a loop-helix-loop domain and functions in the differentiation of hematopoietic cells. The structure of a PU.1-DNA complex was recently reported (Kodandapani, R., Pio, F., Ni, C.-Z., Piccialli, G., Klemsz, M., McKercher, S., Maki, R., and Ely, K. (1996) Nature 380, 456-460). The DNA in this complex is deformed by 8 degrees as it curves around the protein. The pattern of electrostatic contacts between PU.1, and its DNA binding site suggests that laterally asymmetric phosphate neutralization accompanies PU.1 binding. Because of our previous studies showing that such neutralization can induce bending in naked DNA, we have explored the effect of phosphate neutralization by substituting neutral methylphosphonate internucleoside linkages at relevant positions within DNA containing the PU.1 binding sequence. Consistent with the prediction that DNA will collapse toward its partially neutralized surface, DNA neutralized at seven positions to simulate PU.1 binding is observed to bend by 28 degrees . The directions of DNA curvature are slightly different in the co-crystal versus the partially neutralized duplexes. The electrostatic component of the binding energy appears more than enough to account for the DNA bending observed in the PU.1-DNA complex.