Length-encoded multiplex binding site determination: application to zinc finger proteins

A deterministic code for transcription factor-DNA recognition through computation of binding interfaces

The recognition code between transcription factor (TF) amino acids and DNA bases remains poorly understood. Here, the determinants of TF amino acid-DNA base binding selectivity were identified through the analysis of crystals of TF-DNA complexes. Selective, high-frequency interactions were identified for the vast majority of amino acid side chains (‘structural code’). DNA binding specificities were then independently assessed by meta-analysis of random-mutagenesis studies of Zn finger-target DNA sequences. Selective, high-frequency interactions were identified for the majority of mutagenized residues (‘mutagenesis code’). The structural code and the mutagenesis code were shown to match to a striking level of accuracy (P = 3.1 × 10−33), suggesting the identification of fundamental rules of TF binding to DNA bases. Additional insight was gained by showing a geometry-dictated choice among DNA-binding TF residues with overlapping specificity. These findings indicate the existence of a DNA recognition mode whereby the physical-chemical characteristics of the interacting residues play a deterministic role. The discovery of this DNA recognition code advances our knowledge on fundamental features of regulation of gene expression and is expected to pave the way for integration with higher-order complexity approaches.

From repression domains to designer zinc finger proteins: a novel strategy of intracellular immunization against HIV

Tissue-specific gene regulation of eukaryotic organisms is to a large extent mediated by transcription factors that interact with genomic DNA sequences in a sequence-specific manner. The purpose of this synopsis is to put forward the potential of designer zinc finger proteins in treating infections of human immunodeficiency virus (HIV). Artificial transcription factors containing designer zinc finger structures fused to activator or repressor domains have been designated Transcription Response Modifiers (TRMs). The principle of engineering TRMs has been derived from the analysis of human Krüppel-type zinc finger genes and their products. Our research efforts encompass two fascinating features that are displayed by the human Krüppel-type zinc finger protein KOX1: 1) the Krüppel-type zinc finger domains display rules of sequence-specific DNA recognition, and 2) the evolutionarily conserved Krüppel-associated box (KRAB) presents one of the strongest transcriptional repressors identified so far in mammalian organisms. The KRAB repressor activity is postulated to be mediated through co-repressor molecules, such as Silencing Mediating Protein-1 (SMP-1). Thus, the structural organization and functional analysis of zinc finger proteins revealed principles of zinc finger transcription factors that are applicable for reducing the viral load in individuals infected with HIV. In this article, a novel concept of generating therapeutic proteins is outlined that might be conceptually promising in modulating gene expressions of any kind.

T-cell therapies for HIV: Preclinical successes and current clinical strategies

Although antiretroviral therapy (ART) has been successful in controlling HIV infection, it does not provide a permanent cure, requires lifelong treatment, and HIV-positive individuals are left with social concerns such as stigma. The recent application of T cells to treat cancer and viral reactivations post-transplant offers a potential strategy to control HIV infection. It is known that naturally occurring HIV-specific T cells can inhibit HIV initially, but this response is not sustained in the majority of people living with HIV. Genetically modifying T cells to target HIV, resist infection, and persist in the immunosuppressive environment found in chronically infected HIV-positive individuals might provide a therapeutic solution for HIV. This review focuses on successful preclinical studies and current clinical strategies using T-cell therapy to control HIV infection and mediate a functional cure solution.

Engineering T Cells to Functionally Cure HIV-1 Infection

Despite the ability of antiretroviral therapy to minimize human immunodeficiency virus type 1 (HIV-1) replication and increase the duration and quality of patients' lives, the health consequences and financial burden associated with the lifelong treatment regimen render a permanent cure highly attractive. Although T cells play an important role in controlling virus replication, they are themselves targets of HIV-mediated destruction. Direct genetic manipulation of T cells for adoptive cellular therapies could facilitate a functional cure by generating HIV-1-resistant cells, redirecting HIV-1-specific immune responses, or a combination of the two strategies. In contrast to a vaccine approach, which relies on the production and priming of HIV-1-specific lymphocytes within a patient's own body, adoptive T-cell therapy provides an opportunity to customize the therapeutic T cells prior to administration. However, at present, it is unclear how to best engineer T cells so that sustained control over HIV-1 replication can be achieved in the absence of antiretrovirals. This review focuses on T-cell gene-engineering and gene-editing strategies that have been performed in efforts to inhibit HIV-1 replication and highlights the requirements for a successful gene therapy-mediated functional cure.

Role of promoter DNA sequence variations on the binding of EGR1 transcription factor

In response to a wide variety of stimuli such as growth factors and hormones, EGR1 transcription factor is rapidly induced and immediately exerts downstream effects central to the maintenance of cellular homeostasis. Herein, our biophysical analysis reveals that DNA sequence variations within the target gene promoters tightly modulate the energetics of binding of EGR1 and that nucleotide substitutions at certain positions are much more detrimental to EGR1-DNA interaction than others. Importantly, the reduction in binding affinity poorly correlates with the loss of enthalpy and gain of entropy-a trend indicative of a complex interplay between underlying thermodynamic factors due to the differential role of water solvent upon nucleotide substitution. We also provide a rationale for the physical basis of the effect of nucleotide substitutions on the EGR1-DNA interaction at atomic level. Taken together, our study bears important implications on understanding the molecular determinants of a key protein-DNA interaction at the cross-roads of human health and disease.

Prediction of DNA-binding proteins from relational features

Background

The process of protein-DNA binding has an essential role in the biological processing of genetic information. We use relational machine learning to predict DNA-binding propensity of proteins from their structures. Automatically discovered structural features are able to capture some characteristic spatial configurations of amino acids in proteins.

Results

Prediction based only on structural relational features already achieves competitive results to existing methods based on physicochemical properties on several protein datasets. Predictive performance is further improved when structural features are combined with physicochemical features. Moreover, the structural features provide some insights not revealed by physicochemical features. Our method is able to detect common spatial substructures. We demonstrate this in experiments with zinc finger proteins.

Conclusions

We introduced a novel approach for DNA-binding propensity prediction using relational machine learning which could potentially be used also for protein function prediction in general.

Probing the DNA-binding affinity and specificity of designed zinc finger proteins

Engineered transcription factors and endonucleases based on designed Cys(2)His(2) zinc finger domains have proven to be effective tools for the directed regulation and modification of genes. The introduction of this technology into both research and clinical settings necessitates the development of rapid and accurate means of evaluating both the binding affinity and binding specificity of designed zinc finger domains. Using a fluorescence anisotropy-based DNA-binding assay, we examined the DNA-binding properties of two engineered zinc finger proteins that differ by a single amino acid. We demonstrate that the protein with the highest affinity for a particular DNA site need not be the protein that binds that site with the highest degree of specificity. Moreover, by comparing the binding characteristics of the two proteins at varying salt concentrations, we show that the ionic strength makes significant and variable contributions to both affinity and specificity. These results have significant implications for zinc finger design as they highlight the importance of considering affinity, specificity, and environmental requirements in designing a DNA-binding domain for a particular application.

Zinc-finger-based artificial transcription factors and their applications

Artificial transcription factors (ATFs) are potentially a powerful molecular tool to modulate endogenous target gene expression in living cells and organisms. To date, many DNA-binding molecules have been developed as the DNA-binding domains for ATFs. Among them, ATFs comprising Cys(2)His(2)-type zinc-finger proteins (ZFPs) as the DNA-binding domain have been extensively explored. The zinc-finger-based ATFs specifically recognize targeting sites in chromosomes and effectively up- and downregulate expression of their target genes not only in vitro, but also in vivo. In this review, after briefly introducing Cys(2)His(2)-type ZFPs, I will review the studies of endogenous human gene regulation by zinc-finger-based ATFs and other applications as well.

Site-specific integration of retroviral DNA in human cells using fusion proteins consisting of human immunodeficiency virus type 1 integrase and the designed polydactyl zinc-finger protein E2C

During the life cycle of retroviruses, establishment of a productive infection requires stable joining of a DNA copy of the viral RNA genome into host cell chromosomes. Retroviruses are thus promising vectors for the efficient and stable delivery of genes in therapeutic protocols. Integration of retroviral DNA is catalyzed by the viral enzyme integrase (IN), and one salient feature of retroviral DNA integration is its lack of specificity, as many chromosomal sites can serve as targets for integration. Despite the promise for success in the clinic, one major drawback of the retrovirus-based vector is that any unintended insertion events from the therapy can potentially lead to deleterious effects in patients, as demonstrated by the development of malignancies in both animal and human studies. One approach to directing integration into predetermined DNA sites is fusing IN to a sequence-specific DNA-binding protein, which results in a bias of integration near the recognition site of the fusion partner. Encouraging results have been generated in vitro and in vivo using fusion protein constructs of human immunodeficiency virus type 1 IN and E2C, a designed polydactyl zinc-finger protein that specifically recognizes an 18-base pair DNA sequence. This review focuses on the method for preparing infectious virions containing the IN fusion proteins and on the quantitative PCR assays for determining integration site specificity. Efforts to engineer IN to recognize specific target DNA sequences within the genome may lead to development of effective retroviral vectors that can safely deliver gene-based therapeutics in a clinical setting.

Energy-based prediction of amino acid-nucleotide base recognition

Despite decades of investigations, it is not yet clear whether there are rules dictating the specificity of the interaction between amino acids and nucleotide bases. This issue was addressed by determining, in a dataset consisting of 100 high-resolution protein-DNA structures, the frequency and energy of interaction between each amino acid and base, and the energetics of water-mediated interactions. The analysis was carried out using HINT, a non-Newtonian force field encoding both enthalpic and entropic contributions, and Rank, a geometry-based tool for evaluating hydrogen bond interactions. A frequency- and energy-based preferential interaction of Arg and Lys with G, Asp and Glu with C, and Asn and Gln with A was found. Not only favorable, but also unfavorable contacts were found to be conserved. Water-mediated interactions strongly increase the probability of Thr-A, Lys-A, and Lys-C contacts. The frequency, interaction energy, and water enhancement factors associated with each amino acid-base pair were used to predict the base triplet recognized by the helix motif in 45 zinc fingers, which represents an ideal case study for the analysis of one-to-one amino acid-base pair contacts. The model correctly predicted 70.4% of 135 amino acid-base pairs, and, by weighting the energetic relevance of each amino acid-base pair to the overall recognition energy, it yielded a prediction rate of 89.7%.

Exploring the recognition of quadruplex DNA by an engineered Cys2-His2 zinc finger protein

We have recently described an engineered zinc finger protein (Gq1) that binds with high specificity to the intramolecular G-quadruplex formed by the human telomeric sequence 5'-(GGTTAG)(5)-3', and that inhibits the activity of the enzyme telomerase in vitro. Here we report site-directed mutagenesis, biophysical, and molecular modeling studies that provide new insights into quadruplex recognition by the zinc finger scaffold. We show that any one finger of Gq1 can be replaced with the corresponding finger of Zif268, without significant loss of quadruplex affinity or quadruplex versus duplex discrimination. Replacement of two fingers, with one being finger 2, of Gq1 by Zif268 results in significant impairment of quadruplex recognition and loss of discrimination. Molecular modeling suggests that the zinc fingers of Gq1 can bind to the human parallel-stranded quadruplex structure in a stable arrangement, whereas Zif268-quadruplex models show significantly weaker binding energy. Modeling also suggests that an important role of the key protein finger residues in the Gq1-quadruplex complex is to maintain Gq1 in an optimum conformation for quadruplex recognition.

Human immunodeficiency virus type 1 incorporated with fusion proteins consisting of integrase and the designed polydactyl zinc finger protein E2C can bias integration of viral DNA into a predetermined chromosomal region in human cells

In vitro studies using fusion proteins consisting of human immunodeficiency virus type 1 integrase (IN) and a synthetic polydactyl zinc finger protein E2C, a sequence-specific DNA-binding protein, showed that integration of retroviral DNA can be biased towards a contiguous 18-bp E2C-recognition site. To determine whether the fusion protein strategy can achieve site-specific integration in vivo, viruses were prepared by cotransfection and various IN-E2C fusion proteins were packaged in trans into virions. The resulting viruses incorporated with the IN-E2C fusion proteins were functional and capable of performing integration at a level ranging from 1 to 24% of that of viruses containing wild-type (WT) IN. Two of the more infectious viruses, which contained E2C fused to either the N (E2C/IN) or to the C (IN/E2C) terminus of IN, were tested for their ability to direct integration into a unique E2C-binding site present within the 5' untranslated region of erbB-2 gene on human chromosome 17. The copy number of proviral DNA was measured using a quantitative real-time nested-PCR assay, and the specificity of directed integration was determined by comparing the number of proviruses within the vicinity of the E2C-binding site to that in the whole genome. Viruses containing IN/E2C fusion proteins had sevenfold higher preference for integrating near the E2C-binding site than those viruses containing WT IN, whereas viruses containing E2C/IN had 10-fold higher preference. The results indicated that the IN-E2C fusion protein strategy is capable of directing integration of retroviral DNA into a predetermined chromosomal region in the human genome.

Histone peptide AKRHRK enhances H2O2-induced DNA damage and alters its site specificity

Histone proteins are involved in compaction of DNA and the protection of cells from oxygen toxicity. However, several studies have demonstrated that the metal-binding histone reacts with H(2)O(2), leading to oxidative damage to a nucleobase. We investigated whether histone can accelerate oxidative DNA damage, using a minimal model for the N-terminal tail of histone H4, CH(3)CO-AKRHRK-CONH(2), which has a metal-binding site. This histone peptide enhanced DNA damage induced by H(2)O(2) and Cu(II), especially at cytosine residues, and induced additional DNA cleavage at the 5'-guanine of GGG sequences. The peptide also enhanced the formation of 8-oxo-7,8-dihydro-2'-deoxyguanosine and ESR spin-trapping signal from H(2)O(2) and Cu(II). Cyclic redox reactions involving histone-bound Cu(II) and H(2)O(2), may give rise to multiple production of radicals leading to multiple hits in DNA. It is noteworthy that the histone H4 peptide with specific sequence AKRHRK can cause DNA damage rather than protection under metal-overloaded condition.

Quantitative analysis of EGR proteins binding to DNA: assessing additivity in both the binding site and the protein

Background

Recognition codes for protein-DNA interactions typically assume that the interacting positions contribute additively to the binding energy. While this is known to not be precisely true, an additive model over the DNA positions can be a good approximation, at least for some proteins. Much less information is available about whether the protein positions contribute additively to the interaction.

Results

Using EGR zinc finger proteins, we measure the binding affinity of six different variants of the protein to each of six different variants of the consensus binding site. Both the protein and binding site variants include single and double mutations that allow us to assess how well additive models can account for the data. For each protein and DNA alone we find that additive models are good approximations, but over the combined set of data there are context effects that limit their accuracy. However, a small modification to the purely additive model, with only three additional parameters, improves the fit significantly.

Conclusion

The additive model holds very well for every DNA site and every protein included in this study, but clear context dependence in the interactions was detected. A simple modification to the independent model provides a better fit to the complete data.

Protein-DNA Recognition Patterns and Predictions

Structural data on protein-DNA complexes provide clues for understanding the mechanism of protein-DNA recognition. Although the structures of a large number of protein-DNA complexes are known, the mechanisms underlying their specific binding are still only poorly understood. Analysis of these structures has shown that there is no simple one-to-one correspondence between bases and amino acids within protein-DNA complexes; nevertheless, the observed patterns of interaction carry important information on the mechanisms of protein-DNA recognition. In this review, we show how the patterns of interaction, either observed in known structures or derived from computer simulations, confer recognition specificity, and how they can be used to examine the relationship between structure and specificity and to predict target DNA sequences used by regulatory proteins.

Quantitative evaluation of protein-DNA interactions using an optimized knowledge-based potential

Computational evaluation of protein–DNA interaction is important for the identification of DNA-binding sites and genome annotation. It could validate the predicted binding motifs by sequence-based approaches through the calculation of the binding affinity between a protein and DNA. Such an evaluation should take into account structural information to deal with the complicated effects from DNA structural deformation, distance-dependent multi-body interactions and solvation contributions. In this paper, we present a knowledge-based potential built on interactions between protein residues and DNA tri-nucleotides. The potential, which explicitly considers the distance-dependent two-body, three-body and four-body interactions between protein residues and DNA nucleotides, has been optimized in terms of a Z-score. We have applied this knowledge-based potential to evaluate the binding affinities of zinc-finger protein–DNA complexes. The predicted binding affinities are in good agreement with the experimental data (with a correlation coefficient of 0.950). On a larger test set containing 48 protein–DNA complexes with known experimental binding free energies, our potential has achieved a high correlation coefficient of 0.800, when compared with the experimental data. We have also used this potential to identify binding motifs in DNA sequences of transcription factors (TF). The TFs in 79.4% of the known TF–DNA complexes have accurately found their native binding sequences from a large pool of DNA sequences. When tested in a genome-scale search for TF-binding motifs of the cyclic AMP regulatory protein (CRP) of Escherichia coli, this potential ranks all known binding motifs of CRP in the top 15% of all candidate sequences.

Reduction in DNA-binding affinity of Cys2His2 zinc finger proteins by linker phosphorylation

Cys(2)His(2) zinc finger proteins make up the largest class of transcription factors encoded in the genomes of higher eukaryotes. Recent studies of the Ikaros transcription factor demonstrated that this zinc finger protein undergoes cell cycle-dependent changes in association with DNA that seem to be due to phosphorylation of Thr or Ser residues in the linker regions connecting adjacent zinc finger domains. The high degree of conservation of this linker sequence within the Cys(2)His(2) superfamily suggested a common mechanism for the cell cycle-dependent modulation of DNA-binding affinity throughout this large class of transcription factors. The effects of linker phosphorylation on DNA-binding affinity were investigated through a direct comparison of the DNA-binding properties of four synthetic zinc finger proteins produced by native chemical ligation. The four proteins, comprising three zinc finger domains joined by two consensus Thr-Gly-Glu-Lys-Pro linkers, correspond to all four possible combinations of linker Thr phosphorylation states. Fluorescence-based DNA-binding studies of a specific DNA-binding site revealed that phosphorylation of a single linker reduced binding affinity approximately 40-fold, whereas phosphorylation of both linkers reduced binding affinity 130-fold. These results with purified components demonstrate that linker phosphorylation does, indeed, produce a significant reduction in DNA-binding affinity and support a model wherein a single cell cycle-dependent Ser/Thr kinase could simultaneously inactivate a large number of zinc finger transcription factors.

Intermolecular and Intramolecular Readout Mechanisms in Protein–DNA Recognition

Protein-DNA recognition plays an essential role in the regulation of gene expression. Regulatory proteins are known to recognize specific DNA sequences directly through atomic contacts (intermolecular readout) and/or indirectly through the conformational properties of the DNA (intramolecular readout). However, little is known about the respective contributions made by these so-called direct and indirect readout mechanisms. We addressed this question by making use of information extracted from a structural database containing many protein-DNA complexes. We quantified the specificity of intermolecular (direct) readout by statistical analysis of base-amino acid interactions within protein-DNA complexes. The specificity of the intramolecular (indirect) readout due to DNA was quantified by statistical analysis of the sequence-dependent DNA conformation. Systematic comparison of these specificities in a large number of protein-DNA complexes revealed that both intermolecular and intramolecular readouts contribute to the specificity of protein-DNA recognition, and that their relative contributions vary depending upon the protein-DNA complexes. We demonstrated that combination of the intermolecular and intramolecular energies derived from the statistical analyses lead to enhanced specificity, and that the combined energy could explain experimental data on binding affinity changes caused by base mutations. These results provided new insight into the relationship between specificity and structure in the process of protein-DNA recognition, which would lead to prediction of specific protein-DNA binding sites.

Fusion proteins consisting of human immunodeficiency virus type 1 integrase and the designed polydactyl zinc finger protein E2C direct integration of viral DNA into specific sites

In order to establish a productive infection, a retrovirus must integrate the cDNA of its RNA genome into the host cell chromosome. While this critical process makes retroviruses an attractive vector for gene delivery, the nonspecific nature of integration presents inherent hazards and variations in gene expression. One approach to alleviating the problem involves fusing retroviral integrase to a sequence-specific DNA-binding protein that targets a defined chromosomal site. We prepared proteins consisting of wild-type or truncated human immunodeficiency virus type 1 (HIV-1) integrase fused to the synthetic polydactyl zinc finger protein E2C. The purified fusion proteins bound specifically to the 18-bp E2C recognition sequence as analyzed by DNase I footprinting. The fusion proteins were catalytically active and biased integration of retroviral DNA near the E2C-binding site in vitro. The distribution was asymmetric, and the major integration hot spots were localized within a 20-bp region upstream of the C-rich strand of the E2C recognition sequence. Integration bias was not observed with target plasmids bearing a mutated E2C-binding site or when HIV-1 integrase and E2C were added to the reaction as separate proteins. The results demonstrate that the integrase-E2C fusion proteins offer an efficient approach and a versatile framework for directing the integration of retroviral DNA into a predetermined DNA site.

Expanding the DNA-recognition repertoire for zinc finger proteins beyond 20 amino acids

The design of DNA binding domains based on the Cys2His2 zinc finger motif has proven to be a successful strategy for the specific recognition of novel DNA sequences. Although considerable effort has been devoted to the generation of zinc finger proteins with widely varying DNA-binding preferences, only a limited number of potential DNA binding sites have been targeted with a high degree of specificity. These restrictions on zinc finger design appear to be a consequence of the limited repertoire of side-chain lengths and functionalities available with the 20 proteinogenic amino acids. To demonstrate that these limitations can be overcome through the use of "unnatural" amino acids, expressed protein ligation was employed to incorporate the amino acid citrulline into a single position within a three-zinc finger protein. As anticipated, the resulting semisynthetic protein specifically recognizes adenine in the appropriate position of its binding site.

Human zinc fingers as building blocks in the construction of artificial transcription factors

We describe methods for generating artificial transcription factors capable of up- or downregulating the expression of genes whose promoter regions contain the target DNA sequences. To accomplish this, we screened zinc fingers derived from sequences in the human genome and isolated 56 zinc fingers with diverse DNA-binding specificities. We used these zinc fingers as modular building blocks in the construction of novel, sequence-specific DNA-binding proteins. Fusion of these zinc-finger proteins with either a transcriptional activation or repression domain yielded potent transcriptional activators or repressors, respectively. These results show that the human genome encodes zinc fingers with diverse DNA-binding specificities and that these domains can be used to design sequence-specific DNA-binding proteins and artificial transcription factors.

The 5'-untranslated RNA of the human dhfr minor transcript alters transcription pre-initiation complex assembly at the major (core) promoter

The human dhfr minor transcript is distinguished from the predominant dhfr mRNA by an approximately 400 nucleotide extension of the 5'-untranslated region, which corresponds to the major (core) promoter DNA (its template). Based on its unusual sequence composition, we hypothesized that the minor transcript 5'-UTR might be capable of altering transcription pre-initiation complex assembly at the core promoter, through direct interactions of the RNA with specific regulatory polypeptides or the promoter DNA itself. We found that the minor transcript 5'-UTR selectively sequesters transcription factor Sp3, and to a lesser extent Sp1, preventing their binding to the dhfr core promoter. This allows a third putative transcriptional regulatory protein, which is relatively resistant to sequestration by the minor transcript RNA, the opportunity to bind the dhfr core promoter. The selective sequestration of Sp3 > Sp1 by the minor transcript 5'-UTR involves an altered conformation of the RNA, and a structural domain of the protein distinct from that required for binding to DNA. As a consequence, the minor transcript 5'-UTR inhibits transcription from the core promoter in vitro (in trans) in a concentration-dependent manner. These results suggest that the dhfr minor transcript may function in vivo (in cis) to regulate the transcriptional activity of the major (core) promoter.

Rationalization of gene regulation by a eukaryotic transcription factor: calculation of regulatory region occupancy from predicted binding affinities

DNA-binding proteins regulate gene expression by binding preferentially to a set of related sequences. In order to quantify the correlation between gene regulation and the presence of sequence motifs, the affinity of a transcription factor for each variant of the binding site must be known or predicted. In addition, the contribution of multiple binding sites to the regulation of a single gene must be modeled. To predict the affinity of the yeast Leu3 transcription factor for genomic-binding sites, we measured the in vitro equilibrium dissociation constants of 43 binding-site variants and established that the free energy of binding can be approximated as a sum of free energy contributions from each base-pair. This allows the prediction of an equilibrium dissociation constant for all potential binding sites in the genome and, therefore, their fractional occupancy at some assumed concentration of free Leu3. From the occupancy of individual sites, the probability that at least one site is occupied within a defined segment upstream of a gene was calculated for all genes in yeast. We find that this probability is substantially better correlated with regulation by Leu3 than is the number of binding sites. This is true whether the number of binding sites is based on a consensus site definition of the binding site or by enumeration of all variants that have a predicted K(d) value below some threshold. The occupancy calculation was best able to rationalize the Leu3-regulated gene set over a Leu3 concentration range that spans the K(d) values for the best sites.

Designed transcription factors as tools for therapeutics and functional genomics

The paucity of tools that control expression of specific genes in vivo represents a major limitation of functional genomics in mammals; most available small-molecule regulators of transcription-e.g. histone deacetylase inhibitors-exert pan-genomic effects. Recent developments in understanding the role of chromatin in regulating the genome, and of protein-DNA interactions have allowed the development of designed transcription factors that regulate specific genes in vivo (Reik et al., Curr Opin Genet Dev 2002;12:233). These proteins contain two modules: (i) a zinc finger protein (ZFP)-based DNA-binding domain (DBD) designed to recognize a specific sequence (for example, a motif in the promoter of a certain gene); (ii) a functional module (for example, a transcriptional activation or repression domain). Recent data describe the use of such designed transcription factors to regulate a variety of clinically relevant gene targets in human cells: these include MDR1, erythropoietin, erbB-2 and erbB-3, VEGF, and PPARgamma. In the case of VEGF (Liu et al., J Biol Chem 2001;276:11323), proportional upregulation by the designed transcription factor of all three distinct splice isoforms generated by this locus was observed, illuminating the utility of endogenous gene control in therapeutic settings (proper isoform ratio is essential for the proangiogenic function of VEGF). In the case of PPARgamma, use of a transcriptional repressor designed to downregulate the expression of two PPARgamma isoforms allowed "mutation-free reverse genetics" analysis that illuminated a unique role for the PPARgamma2 isoform in adipogenesis (Ren et al., Genes Dev 2002;16:27). The ability to selectively activate or repress specific mammalian genes in vivo using designed transcription factors thus has considerable promise in clinical and in basic science settings.

Is there a code for protein-DNA recognition? Probab(ilistical)ly

Transcriptional regulation of all genes is initiated by the specific binding of regulatory proteins called transcription factors to specific sites on DNA called promoter regions. Transcription factors employ a variety of mechanisms to recognise their DNA target sites. In the last few decades, attempts have been made to describe these mechanisms by general sets of rules and associated models. We give an overview of these models, starting with a historical review of the somewhat controversial issue of a "recognition code" governing protein-DNA interaction. We then present a probabilistic framework in which advantages and disadvantages of various models can be discussed. Finally, we conclude that simplifying assumptions about additivity of interactions are sufficiently justified in many situations (and can be suitably extended in other situations) to allow a unifying concept of a "probabilistic code" for protein-DNA recognition to be defined.

Integration site selection by lentiviruses: biology and possible control

Retroviruses integrate into naked DNA in a generally sequence nonspecific fashion, but closer study reveals a variety of forces that influence target site selection. Primary sequence of the target plays a small but detectable role. Proteins bound to target DNA can inhibit integration by blocking access of integration complexes or stimulate integration by distorting DNA. An important example of the latter is DNA distortion in nucleosomal DNA. In vivo integration has not yet been convincingly shown to be biased in favor of any identifiable sequence features, though this could still change in future studies. Many applications of retroviral vectors could be facilitated by targeting integration in vivo to predetermined sites. Towards this end, several groups have studied the properties of fusions of integrase proteins to sequence-specific DNA-binding domains. To date such studies establish that targeting can work well in reactions in vitro, but a variety of obstacles complicate applications in vivo. However, naturally occurring retrotransposons do carry out highly targeted integration using retrovirus-like integrase proteins, fueling long-term hopes for targeting with retroviral integrases as well.

A novel zinc finger transcription factor with two isoforms that are differentially repressed by estrogen receptor-alpha

Estrogen receptor-alpha (ERalpha) can induce the expression of genes in response to estrogen by binding to estrogen response elements in the promoters of target genes. There is growing evidence that ERalpha can alter patterns of gene expression in response to ligand by regulating the activity of other factors through a direct protein-protein interaction. To identify other factors that are regulated by ERalpha, a yeast two-hybrid screen was performed that identified a novel Cys(2)His(2) zinc finger protein named ZER6. The ZER6 protein contains a Kruppel-associated box domain and six Cys(2)His(2) zinc fingers. Transcripts from the ZER6 gene can have alternate 5' exons and encode either a p71 or p52 isoform. The p52-ZER6 protein interacts strongly with ERalpha in the presence of 17beta-estradiol, whereas the p71-ZER6 isoform has a HUB-1 amino-terminal domain that inhibits the interaction with ERalpha. A consensus ZER6 binding element was defined using PCR-assisted binding site selection. In COS-1 cells, both the p52 and p71 isoforms can activate transcription through the ZER6 binding element; however, in the presence of ERalpha, transactivation by the p52 isoform is specifically repressed. Overexpression of the p52 isoform was able to abrogate activation by p71-ZER6. Expression of ZER6 was largely restricted to the mammary gland with a lower level of expression in the kidney. We conclude that ZER6 is a novel zinc finger transcription factor in which regulation of transcription in hormone-responsive cells can be controlled by the relative level of expression of two distinct isoforms.

The autoimmune regulator (AIRE) is a DNA-binding protein

The autoimmune regulator (AIRE) protein is a putative transcription regulator with two plant homeodomain-type zinc fingers, a putative DNA-binding domain (SAND), and four nuclear receptor binding LXXLL motifs. We have shown here that in vitro, recombinant AIRE can form homodimers and homotetramers that were also detected in thymic protein extracts. Recombinant AIRE also oligomerizes spontaneously upon phosphorylation by cAMP dependent protein kinase A or protein kinase C. Similarly, thymic AIRE protein is phosphorylated at the tyrosine and serine/threonine residues. AIRE dimers and tetramers, but not the monomers, can bind to G-doublets with the ATTGGTTA motif and the TTATTA-box. Competition assays revealed that sequences with one TTATTA motif and two tandem repeats of ATTGGTTA had the highest binding affinity. These findings demonstrate that AIRE is an important DNA binding molecule involved in immune regulation.

Non-independence of Mnt repressor-operator interaction determined by a new quantitative multiple fluorescence relative affinity (QuMFRA) assay

Salmonella bacteriophage repressor Mnt belongs to the ribbon-helix-helix class of transcription factors. Previous SELEX results suggested that interactions of Mnt with positions 16 and 17 of the operator DNA are not independent. Using a newly developed high-throughput quantitative multiple fluorescence relative affinity (QuMFRA) assay, we directly quantified the relative equilibrium binding constants (K(ref)) of Mnt to operators carrying all the possible dinucleotide combinations at these two positions. Results show that Mnt prefers binding to C, instead of wild-type A, at position 16 when wild-type C at position 17 is changed to other bases. The measured K(ref) values of double mutants were also higher than the values predicted from single mutants, demonstrating the non-independence of these two positions. The ability to produce a large number of quantitative binding data simultaneously and the potential to scale up makes QuMFRA a valuable tool for the large-scale study of macromolecular interaction.

Structural insights on Smad function in TGFbeta signaling

TGFbeta signaling plays a central role in regulating a broad range of cellular responses in a variety of organisms. TGFbeta signaling from the cell membrane to the nucleus is mediated by the Smad family of proteins. During the past five years of intense investigation, key events in TGFbeta signaling have been documented at the molecular and cellular level. Recent structural studies have improved our understanding of how specificity is generated in the TGFbeta signaling pathways. Despite this progress, significant questions remain regarding the precise mechanisms of signaling and point to the urgent need for well-controlled biochemical studies. Rather than giving a comprehensive review on Smad-mediated TGFbeta signaling, this review focuses on functional insights provided by recent structural studies and discusses several existing controversies.

DNA recognition by Cys2His2 zinc finger proteins

Cys2His2 zinc fingers are one of the most common DNA-binding motifs found in eukaryotic transcription factors. These proteins typically contain several fingers that make tandem contacts along the DNA. Each finger has a conserved beta beta alpha structure, and amino acids on the surface of the alpha-helix contact bases in the major groove. This simple, modular structure of zinc finger proteins, and the wide variety of DNA sequences they can recognize, make them an attractive framework for attempts to design novel DNA-binding proteins. Several studies have selected fingers with new specificities, and there clearly are recurring patterns in the observed side chain-base interactions. However, the structural details of recognition are intricate enough that there are no general rules (a "recognition code") that would allow the design of an optimal protein for any desired target site. Construction of multifinger proteins is also complicated by interactions between neighboring fingers and the effect of the intervening linker. This review analyzes DNA recognition by Cys2His2 zinc fingers and summarizes progress in generating proteins with novel specificities from fingers selected by phage display.

A novel four zinc-finger protein targeted against p190(BcrAbl) fusion oncogene cDNA: utilisation of zinc-finger recognition codes

A three zinc-finger protein that binds specifically to the cDNA representing the unique fusion gene BCR:Abl, associated with acute lymphoblastic leukaemia, has previously been characterised. At this breakpoint, a sequence homology of 8/9 bp exists between the BCR:Abl (fusion) and c-ABL: (parental) target sequences. We show that the three zinc-finger protein discriminates poorly between the fusion (BCR:Abl) and parental (ABL:) sequence (K:(d)s of 42.8 and 65.1 nM, respectively). In order to improve the discriminatory properties of this protein, and to demonstrate the utility of current zinc-finger databases, we have added a fourth zinc-finger to the original three zinc-finger protein. This fourth finger recognises a 3 bp subsite derived from the BCR: portion of the breakpoint and is not present in c-ABL: This novel four finger protein, which now recognises a 12 bp sequence, demonstrates improved specific binding to BcrAbl (K:(d )= 17 nM). More significantly we have shown that there is now enhanced discrimination between BcrAbl and ABL: sequences by the four finger protein than the original three finger protein.

Insights into the molecular recognition of the 5'-GNN-3' family of DNA sequences by zinc finger domains

In order to construct zinc finger domains that recognize all of the possible 64 DNA triplets, it is necessary to understand the mechanisms of protein/DNA interactions on the molecular level. Previously we reported 16 zinc finger domains which had been characterized in detail to bind specifically to the 5'-GNN-3' family of DNA sequences. Artificial transcription factors constructed from these domains can regulate the expression of endogenous genes. These domains were created by phage-display selection followed by site-directed mutagenesis. A total of 84 mutants of a three-domain zinc finger protein have been analyzed for their DNA-binding specificity. Here, we report the results of this systematic and extensive mutagenesis study. New insights into zinc finger/DNA interactions were obtained by combining specificity data with computer modeling and comparison with known structural data from NMR and crystallographic studies. This analysis suggests that unusual cross-strand and inter-helical contacts are made by some of these proteins, and the general orientation of the recognition helix to the DNA is flexible, even when constrained by flanking zinc finger domains. These findings disfavor the utility of existing simple recognition codes and suggest that highly specific domains cannot be obtained from phage display alone in most cases, but only in combination with rational design. The molecular basis of zinc finger/DNA interaction is complex and its understanding is dependent on the analysis of a large number of proteins. This understanding should enable us to refine rapidly the specificity of other zinc finger domains, as well as polydactyl proteins constructed with these domains to recognize extended DNA sequences.

DNA binding specificity studies of four ETS proteins support an indirect read-out mechanism of protein-DNA recognition

Members of the ETS family of transcription factors are involved in several developmental and physiological processes, and, when overexpressed or misexpressed, can contribute to a variety of cancers. Each family member has a conserved DNA-binding domain that recognizes DNA sequences containing a G-G-A trinucleotide. Discrimination between potential ETS-binding sites appears to be governed by both the nucleotides flanking the G-G-A sequence and protein-protein interactions. We have used an adaptation of the "length-encoded multiplex" approach (Desjarlais, J. R., and Berg, J. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 11099-11103) to define DNA binding specificities for four ETS proteins: Fli-1, SAP-1, PU.1, and TEL. Our results support a model in which cooperative effects among neighboring bases flanking the central G-G-A site contribute to the formation of stable ETS/DNA complexes. These results are consistent with a mechanism for specific DNA binding that is partially governed by an indirect read-out of the DNA sequence, in which a sequence-specific DNA conformation is sensed or induced.

Combining structure-based design with phage display to create new Cys(2)His(2) zinc finger dimers

BACKGROUND

Several strategies have been reported for the design and selection of novel DNA-binding proteins. Most of these studies have used Cys(2)His(2) zinc finger proteins as a framework, and have focused on constructs that bind DNA in a manner similar to Zif268, with neighboring fingers connected by a canonical (Krüppel-type) linker. This linker does not seem ideal for larger constructs because only modest improvements in affinity are observed when more than three fingers are connected in this manner. Two strategies have been described that allow the productive assembly of more than three canonically linked fingers on a DNA site: connecting sets of fingers using linkers (covalent), or assembling sets of fingers using dimerization domains (non-covalent).

RESULTS

Using a combination of structure-based design and phage display, we have developed a new dimerization system for Cys(2)His(2) zinc fingers that allows the assembly of more than three fingers on a desired target site. Zinc finger constructs employing this new dimerization system have high affinity and good specificity for their target sites both in vitro and in vivo. Constructs that recognize an asymmetric binding site as heterodimers can be obtained through substitutions in the zinc finger and dimerization regions.

CONCLUSIONS

Our modular zinc finger dimerization system allows more than three Cys(2)His(2) zinc fingers to be productively assembled on a DNA-binding site. Dimerization may offer certain advantages over covalent linkage for the recognition of large DNA sequences. Our results also illustrate the power of combining structure-based design with phage display in a strategy that assimilates the best features of each method.

The artificial zinc finger coding gene 'Jazz' binds the utrophin promoter and activates transcription

Up-regulation of utrophin gene expression is recognized as a plausible therapeutic approach in the treatment of Duchenne muscular dystrophy (DMD). We have designed and engineered new zinc finger-based transcription factors capable of binding and activating transcription from the promoter of the dystrophin-related gene, utrophin. Using the recognition 'code' that proposes specific rules between zinc finger primary structure and potential DNA binding sites, we engineered a new gene named 'Jazz' that encodes for a three-zinc finger peptide. Jazz belongs to the Cys2-His2 zinc finger type and was engineered to target the nine base pair DNA sequence: 5'-GCT-GCT-GCG-3', present in the promoter region of both the human and mouse utrophin gene. The entire zinc finger alpha-helix region, containing the amino acid positions that are crucial for DNA binding, was specifically chosen on the basis of the contacts more frequently represented in the available list of the 'code'. Here we demonstrate that Jazz protein binds specifically to the double-stranded DNA target, with a dissociation constant of about 32 nM. Band shift and super-shift experiments confirmed the high affinity and specificity of Jazz protein for its DNA target. Moreover, we show that chimeric proteins, named Gal4-Jazz and Sp1-Jazz, are able to drive the transcription of a test gene from the human utrophin promoter.

Characterization of Drosophila OVO protein DNA binding specificity using random DNA oligomer selection suggests zinc finger degeneration

The Drosophila melanogaster ovo locus codes for several tissue- and stage-specific proteins that all possess a common C-terminal array of four C(2)H(2)zinc fingers. Three fingers conform to the motif framework and are evolutionarily conserved; the fourth diverges considerably. The ovo genetic function affects germ cell viability, sex identity and oogenesis, while the overlapping svb function is a key selector for epidermal structures under the control of wnt and EGF receptor signaling. We isolated synthetic DNA oligomers bound by the OVO zinc finger array from a high complexity starting population and derived a statistically significant 9 bp long DNA consensus sequence, which is nearly identical to a consensus derived from several Drosophila genes known or suspected of being regulated by the ovo function in vivo. The DNA consensus recognized by Drosophila OVO protein is atypical for zinc finger proteins in that it does not conform to many of the 'rules' for the interaction of amino acid contact residues and DNA bases. Additionally, our results suggest that only three of the OVO zinc fingers contribute to DNA-binding specificity.

Altered sequence specificity identified from a library of DNA-binding small molecules

BACKGROUND

The ability to target specific DNA sequences using small molecules has major implications for basic research and medicine. Previous studies revealed that a bis-intercalating molecule containing two 1,4,5,8-napthalenetetracarboxylic diimides separated by a lysine-tris-glycine linker binds to DNA cooperatively, in pairs, with a preference for G + C-rich sequences. Here we investigate the binding properties of a library of bis-intercalating molecules that have partially randomized peptide linkers.

RESULTS

A library of bis-intercalating derivatives with varied peptide linkers was screened for sequence specificity using DNase I footprinting on a 231 base pair (bp) restriction fragment. The library mixtures produced footprints that were generally similar to the parent bis-intercalator, which bound within a 15 bp G + C-rich repeat above 125 nM. Nevertheless, subtle differences in cleavage enhancement bands followed by library deconvolution revealed a derivative with novel specificity. A lysine-tris-beta-alanine derivative was found to bind preferentially within a 19 bp palindrome, without substantial loss of affinity.

CONCLUSIONS

Synthetically simple changes in the bis-intercalating compounds can produce derivatives with novel sequence specificity. The large size and symmetrical nature of the preferred binding sites suggest that cooperativity may be retained despite modified sequence specificity. Such findings, combined with structural data, could be used to develop versatile DNA ligands of modest molecular weight that target relatively long DNA sequences in a selective manner.

Nucleic acid-binding properties of low-density lipoproteins: LDL as a natural gene vector

The role of apo B-100 as a transcription factor is indicated by the presence of regions in its primary structure that are similar to the DNA-binding domains of the transcription factors ISGF3gamma, STATs, IRFs, and SREBPs as well as by the presence of 11 RNA-binding KH domains. The Apo B-100 sequence also contains numerous bipartite nuclear localization sequences (NLS). A modified gel shift assay was used to show binding of highly purified preparations of human LDLs to fragmented genomic DNA, plasmid DNA, synthetic oligonucleotides (ISRE, 5'-GGGAAACCGAAACTG and E/C, E-box motif and CCAAT, adipocyte-specific genes promoter site), and total RNA from human liver. LDL was observed to bind preferentially to plasmid DNA containing the hCMV IE2 promoter region. In experiments using human liver total RNA, RNA for five different genes was recovered from LDL and VLDL bands. Gene transfection experiments using human skin fibroblast cells were used to study the gene transfer capacity of LDL. Cells transfected with a pEGFP-N1 plasmid DNA and LDL expressed functional GFP, as indicated by fluorescence, at approximately 3 hrs after transfection. Our results strongly support an alternative role for apo B-100, in toto or perhaps as functional fragments, in the control of gene expression and as gene transfer vector.

Zinc finger peptides for the regulation of gene expression

Zinc fingers are small DNA-binding peptide motifs that were discovered in this laboratory. These motifs can be used as modular building blocks for the construction of larger protein domains that recognise and bind to specific DNA sequences. Phage display has been used to create a large library of different zinc fingers from which selections were made for binding to a given DNA sequence. From this database there have been elucidated elements of recognition rules that relate the amino acid sequence of a finger to its preferred DNA binding site. Control of gene expression using designed zinc finger peptides has been demonstrated by the specific inhibition of an oncogene mouse cell line and also by switching on genes in expression plasmids. These experiments demonstrate that zinc finger DNA-binding domains can be engineered de novo to recognise given DNA sequences. Five to six individual zinc fingers linked together would recognise a DNA sequence 15-18 bp in length, sufficiently long to constitute a rare address in the human genome. By adding functional groups to the engineered DNA-binding domains, e.g. silencing domains, novel transcription factors can be generated to up- or downregulate expression of a target gene. Among potential applications are the repression of oncogene expression and the disruption of the reproductive cycle of virus infection.

Binding studies with mutants of Zif268. Contribution of individual side chains to binding affinity and specificity in the Zif268 zinc finger-DNA complex

The Zif268 zinc finger-DNA complex has served as a model system for understanding how Cys2His2 type zinc fingers recognize DNA. Structural studies of the Zif268-DNA complex revealed that residues at four positions in the alpha helix of each zinc finger play key roles in recognition, but there has been no information about the precise contributions of individual residues. Here we report the results of binding studies involving five mutants of Zif268 that have changes in the base-contacting residues of finger one. These studies let us evaluate the contributions that Arg18 (position -1 of the alpha helix), Asp20 (position 2), Glu21 (position 3), and Arg24 (position 6) make to the overall energy of DNA binding. Our results confirm the important role played by these arginines. By comparing the affinities of the wild type and mutant peptides for various sites, we also prove that Asp20 and Glu21 play important roles in determining binding site specificity.

Quantifying DNA-protein interactions by double-stranded DNA arrays

We have created double-stranded oligonucleotide arrays to perform highly parallel investigations of DNA-protein interactions. Arrays of single-stranded DNA oligonucleotides, synthesized by a combination of photolithography and solid-state chemistry, have been used for a variety of applications, including large-scale mRNA expression monitoring, genotyping, and sequence-variation analysis. We converted a single-stranded to a double-stranded array by synthesizing a constant sequence at every position on an array and then annealing and enzymatically extending a complementary primer. The efficiency of second-strand synthesis was demonstrated by incorporation of fluorescently labeled dNTPs (2'-deoxyribonucleoside 5'-triphosphates) and by terminal transferase addition of a fluorescently labeled ddNTP. The accuracy of second-strand synthesis was demonstrated by digestion of the arrayed double-stranded DNA (dsDNA) on the array with sequence-specific restriction enzymes. We showed dam methylation of dsDNA arrays by digestion with DpnI, which cleaves when its recognition site is methylated. This digestion demonstrated that the dsDNA arrays can be further biochemically modified and that the DNA is accessible for interaction with DNA-binding proteins. This dsDNA array approach could be extended to explore the spectrum of sequence-specific protein binding sites in genomes.

Analysis of zinc fingers optimized via phage display: evaluating the utility of a recognition code

Cys2His2 zinc finger proteins are composed of modular DNA-binding domains and provide an excellent framework for the design and selection of proteins with novel site specificity. Crystal structures of zinc finger-DNA complexes have shown that many Cys2His2 zinc fingers use a conserved docking arrangement that juxtaposes residues at key positions in the "recognition helix" with corresponding base positions in the three to four base-pair subsite. Several groups have proposed that specificity can be explained with a zinc finger-DNA recognition code that correlates specific amino acids at these key positions in the alpha-helix with specific bases in each position of the corresponding subsite. Here, we explore the utility of such a code through detailed studies of zinc finger variants selected via phage display. These proteins provide interesting systems for detailed analysis since they have affinities and specificities for their sites similar to those of naturally occurring DNA-binding proteins. Comparisons are facilitated by the fact that only key DNA-binding residues are varied in each finger while leaving all other regions of the structure unchanged. We study these proteins in detail by (1) selecting their optimal binding sites and comparing these binding sites with sites that might have been predicted from a code; (2) by examining the "evolutionary history" of these proteins during the phage display protocol to look for evidence of context-dependent effects; and (3) by reselecting finger 1 in the presence of the optimized finger 2/finger 3 domains to obtain further data on finger modularity. Our data for optimized fingers and binding sites demonstrate a clear correlation with contacts that would be predicted from a code. However, there are enough examples of context-dependent effects (not explained by any existing code) that selection is the most reliable method for maximizing the affinity and specificity of new zinc finger proteins.

A detailed study of the substrate specificity of a chimeric restriction enzyme

Recently, the crystal structure of the designed zinc finger protein, DeltaQNK, bound to a preferred DNA sequence was reported. We have converted DeltaQNK into a novel site-specific endonuclease by linking it to the Fok I cleavage domain (FN). The substrate specificity and DNA cleavage properties of the resulting chimeric restriction enzyme (DeltaQNK-FN) were investigated, and the binding affinities of DeltaQNK and DeltaQNK-FN for various DNA substrates were determined. Substrates that are bound by DeltaQNK with high affinity are the same as those that are cleaved efficiently by DeltaQNK-FN. Substrates bound by DeltaQNK with lower affinity are cleaved with very low efficiency or not at all by DeltaQNK-FN. The binding of DeltaQNK-FN to each substrate was approximately 2-fold weaker than that for DeltaQNK. Thus, the fusion of the Fok I cleavage domain to the zinc finger motif does not change the DNA sequence specificity of the zinc finger protein and does not change its binding affinity significantly.

Binding properties of the artificial zinc fingers coding gene Sint1

On the basis of the recognition "code" that suggests specific rules between zinc finger's primary structure and the finger's potential DNA binding sites, we have constructed a new three-zinc finger coding gene to target the nine base pair DNA sequence: 5'-TGG-ATG-GAC-3'. This artificial gene named "Sint1" belongs to the Cys2-His2 zinc finger type. The amino acid positions, crucial for DNA binding, have been specifically chosen on the basis of the amino acid/base contacts more frequently represented in the available list of the proposed recognition "code". Here we demonstrate that Sint1 protein binds specifically the double strand "code" DNA target, with a dissociation constant (Kd) comparable to the Kd of the well known Zif268 protein. Sint1 "code" deduced and the "experimental" selected DNA binding sites share five nucleotide positions. Interestingly, Sint1 shows both high affinity and specificity toward the single strand "code" DNA binding site, with a Kd comparable to the corresponding double strand DNA target. Moreover, we prove that Sint1 is able to bind RNA similarly to several natural zinc finger proteins.