Figures
Abstract
AraR is a transcription factor involved in the regulation of carbon catabolism in Bacillus subtilis. This regulator belongs to the vast GntR family of helix-turn-helix (HTH) bacterial metabolite-responsive transcription factors. In this study, AraR-DNA specific interactions were analysed by an in vitro missing-contact probing and validated using an in vivo model. We show that amino acid E30 of AraR, a highly conserved residue in GntR regulators, is indirectly responsible for the specificity of amino acid-base contacts, and that by mutating this residue it will be possible to achieve new specificities towards DNA contacts. The results highlight the importance in DNA recognition and binding of highly conserved residues across certain families of transcription factors that are located in the DNA-binding domain but not predicted to specifically contact bases on the DNA. These new findings not only contribute to a more detailed comprehension of AraR-operator interactions, but may also be useful for the establishment of a framework of rules governing protein-DNA recognition.
Citation: Correia IL, Franco IS, de Sá-Nogueira I (2014) Towards Novel Amino Acid-Base Contacts in Gene Regulatory Proteins: AraR – A Case Study. PLoS ONE 9(11): e111802. https://doi.org/10.1371/journal.pone.0111802
Editor: Albert Jeltsch, Universität Stuttgart, Germany
Received: August 1, 2014; Accepted: October 7, 2014; Published: November 3, 2014
Copyright: © 2014 Correia et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper.
Funding: This work was supported by the Fundação para a Ciência e a Tecnologia [grant numbers PPCDT/BIA-MIC/61140/2004 and PTDC/AGR-AAM/102345/2008 to I.S.N.; PEst-OE/BIA/UI0457/2011 to Centro de Recursos Microbiológicos; fellowship SFRH/BD/29469/2006 to I.L.C.]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Protein–DNA binding is a process fundamental to life as it masters many genetic activities such as transcription, recombination, DNA replication and repair. The specific interaction between transcription factors and their cognate DNA sites is critical for regulation of gene expression in cells. Understanding how these different proteins are able to find and bind selectively to only one, or just a small number, specific sequence(s) out of the millions of nucleotides present in a genome is a major goal of molecular biology. The recognition principles of protein–DNA interfaces are guided by the complex interplay of noncovalent interactions [1], [2], [3], [4]. In general, DNA recognition follows two paradigms, direct and indirect readout. In the case of direct readout, proteins form contacts such as, hydrogen bonds and van der Waals contacts, mainly in the major, and to a lesser extent also the minor, groove of the DNA to the edges of the base pairs to probe the DNA sequence [1], [2], [3], [4]. Indirect readout occurs through protein contacts to the DNA that depend on base pairs that are not directly contacted by the protein in which the sequence-dependent deformability or structural differences between DNA molecules contribute to their discrimination. A DNA–protein “recognition code”, although of great utility in molecular biology, remains elusive and improbable. While it is clear that a single recognition code does not exist there is some evidence for the existence of a degenerated code whereby one group of bases displays tendency to interact with a certain group of amino acids [4], [5], [6]. In recent years, researchers have addressed this issue by strengthening a comprehensive framework of the rules governing protein–DNA interactions. Different strategies have been described for the construction of Zinc-fingers (ZFs) and TAL (transcription activator-like) proteins with new binding specificities [7], [8]. Nevertheless, there is not a simple one-to-one correspondence between protein and DNA sequences, thus direct readout alone is insufficient to justify the specificities of protein-DNA interactions.
AraR is a homodimeric transcription factor involved in the regulation of carbon catabolism in Bacillus subtilis. The protein displays a chimeric organization, consisting of two functional domains with different phylogenetic origins [9], [10]: a small N-terminal DNA-binding domain (DBD) comprising a winged helix–turn–helix (HTH) motif belonging to the GntR family of transcriptional regulators [11] and a larger C-terminal domain homologous to that of the GalR/LacI family of bacterial regulators and sugar-binding proteins [12]. Recently, the three-dimensional crystal structure of the AraR C-terminal domain [13] and the DNA-binding domain [14] were independently solved. AraR typifies one of the GntR-subfamilies of proteins (reviewed in [15]). The GntR superfamily is one of the largest groups of HTH bacterial metabolite-responsive transcription factors (Pfam family: PF00392; Prosite Family PS50949) and GntR-like regulators are widespread in bacteria and are known to control many fundamental cellular processes, such as primary metabolism, motility, development, antibiotic production, antibiotic resistance, plasmid transfer and virulence (reviewed in [15]).
The control in gene expression exerted by AraR is modulated by the presence of the inducer L-arabinose. Binding of AraR to L-arabinose leads to induction of expression of the ara regulon (Figure 1), which is composed of at least thirteen genes. The products of these genes include the regulator itself, extracellular and intracellular catabolic enzymes involved in the degradation of arabinose-, galactose- and xylose-containing polysaccharides, uptake of these sugars into the cell and further catabolism of L-arabinose and arabinose oligomers [9], [16], [17], [18]. In the absence of inducer, AraR recognizes and binds at least eight palindromic operator sequences (ara boxes), located in the five known arabinose-inducible promoters (Figure 1). Three of these promoters contain two ara boxes: the promoter of the araABDLMNPQ-abfA operon (boxes ORA1 and ORA2), of araE (ORE1 and ORE2) and of abf2 (ORX1 and ORX2). In the cases of the genes araR and abnA, a single box is present (ORR3 and ORB1) (Figure 1). AraR binding to the promoters displaying two boxes is cooperative, requiring in phase and properly spaced operators, and involves the formation of a small loop in the DNA. These two mechanistically diverse modes of action of AraR result in distinct levels of transcriptional regulation, as cooperative binding to two ara boxes results in a high level of repression while interaction with a single operator allows a more flexible control [10], [18], [19].
The genes are represented as black arrows pointing at the direction of transcription. The AraR repressor, in the absence of the effector molecule - arabinose - binds to palindromic sequences (At(T/A)tGTaCGTAcaa(A/T)T consensus depicted, bottom left) found in the promoter region of the ara genes. The AraR protein is shown as a dimer. The eight AraR boxes are represented as white rectangles. Binding to the different operators may either be cooperative or uncooperative.
Previous studies have mapped the functional domains of AraR and characterized the C-terminal region involved in effector binding and dimerization [20]. Moreover, guided by molecular modelling we identified amino acids potentially involved in DNA binding and the effect of their substitution revealed key residues necessary for the DNA binding and regulatory activity in vivo and in vitro [21]. In addition, important bases for AraR-DNA interactions in both arms of the palindromic operator sequences were also identified [21]. In this work we studied AraR-DNA specific interactions using methodologies designed to detect direct or indirect interactions between the atoms/residues of the interacting partners, both in vitro and in vivo. AraR mutant proteins displaying a moderate effect in AraR-DNA interaction and single point mutations in the operator DNA leading to partial derepression of gene expression were probed. The results obtained provide valuable information concerning the specific interaction of AraR-DNA and insights into the binding of GntR regulators in general.
Materials and Methods
Strains and growth conditions
Escherichia coli DH5α (Gibco BRL) was used as host for routine molecular cloning work. E. coli strains were grown in LB [22] medium and the antibiotics ampicillin (100 µg ml−1) and tetracycline (12 µg ml−1) were added when appropriated. B. subtilis strains used in this study (Table 1) were grown in liquid in LB or C-minimal medium [23] and chloramphenicol (5 mg ml−1), kanamycin (10 mg ml−1) or erythromycin (1 mg ml−1) were added when appropriate. The B. subtilis and E. coli cells were transformed as described previously [7]. The Amy phenotype was tested by detection of starch hydrolysis on tryptose blood agar base medium (Difco) plates, containing 1% (w/v) of potato starch, with an I2–KI solution as described previously [9]. The Thr phenotype was determined by growth on Spizizen minimal medium [24] supplemented with 2% (w/v) of glucose, 0.2% (w/v) potassium glutamate, 3 mM MgSO4, and 2% (w/v) agar.
DNA manipulation and construction of plasmids
DNA manipulations were carried out as described by Sambrook et al. [25]. Restriction enzymes were purchased from MBI Fermentas and used according to the manufacturer's instructions. DNA was eluted from agarose gels with GFX gel band purification kit (Amersham Pharmacia Biotech). DNA sequencing was performed with ABI PRIS BigDye Terminator Ready Reaction Cycle Sequencing kit (Applied Biosystems). PCR amplifications were done using high-fidelity Phusion DNA polymerase (Finnzymes) and the resulting products purified by QIAquick PCR purification kit (Qiagen).
For the construction of plasmids pMI35 and pMI36, bearing substitutions E30A and Y5F, respectively, the mutated araR alleles were amplified by PCR with primers ARA1 and ARA73 (Table 2), using as template chromosomal DNA from strains IQB568 and IQB571 [21], respectively. The PCR products were digested with EcoRI-BamHI (or EcoRI-BglII) and independently subcloned in the respective pLS30 sites [20]. The obtained plasmids were then digested with ScaI, which allows the occurrence of a double crossover recombination event at amyE locus of the B. subtilis chromosome (Table 1).
Plasmids pMI37, pMI45, pMI46 and pMI48, contain respectively, the wild-type of the araABDLMNPQ-abfA operon promoter and the same promoter bearing mutation ORA1 (T16→G), ORA1 (A1→C) or ORA1 (T6→G), respectively, fused to a lacZ gene. These plasmids were constructed by insertion of the 204-bp BamHI-EcoRI DNA fragment from pLM32 [10], pLM67, pLM68, pLM65 [21], respectively, into the same sites of pDG1663 [26], to generate an ORA1A2 -lacZ fusion, suitable for a double crossover recombination event at thrC locus of the B. subtilis chromosome (Table 1). To create abf2 promoter - lacZ fusions the wild-type and the mutated abf2 promoter, ORX1 (T6→G) were inserted into the vector pDG1663 to yield plasmids pMI64 and pMI63, respectively. For construction of pMI64, a 291-bp EcoRI-BamHI DNA fragment from pRIT1 [18] bearing the abf2 wild-type promoter was subcloned into those sites of pDG1663. Mutagenesis of the abf2 promoter, ORX1 (T6→G), was achieved by PCR overlap extension, regions immediately upstream and downstream of mutagenesis target region were amplified in two independent PCR experiments, using primers ARA87 and ARA542 (PCR1) using as template chromosomal DNA of B. subtilis 168T+ and primers ARA541 and ARA73 (PCR2), using pRIT1 as template. The products were joined by overlapping PCR, with primers ARA87 and ARA73 (Table 2), and the resulting fragment was digested with BamHI and EcoRI and cloned into pDG1663 BamHI-EcoRI, yielding pMI63.
β-Galactosidase assays
B. subtilis strains were grown in C-minimal medium supplemented with 1% (w/v) casein hydrolysate in the presence and absence of 0.4% (w/v) L-arabinose, as previously reported [9]. Samples of cell culture were collected and analysed 2 h after the addition of L-arabinose, β-Galactosidase activity was measured using the substrate p-nitrophenyl-β-D-galactoside (ONPG) and expressed in Miller units, the ratio of β-galactosidase activity in the presence and absence of inducer was taken as a measure of AraR repression in the analysed strains (Repression Index) as described previously [9].
Electrophoretic mobility shift assay (EMSA)
DNA fragments carrying the operator sequences ORA1A2 wild-type and mutants ORA1 A1→C, G5→T, T6→G, and T16→G were amplified by PCR, with primers ARA262 and ARA263, using plasmids pLM51, pLM61, pLM62 and pLM58 [21], respectively, as template. Overexpression and protein purification of the AraR wild-type and mutant variants (Y5F and E30A) were performed as described previously [20].
The assays were performed as described in Franco et al. [21], DNA fragments were radiolabelled with [γ-32P] dATP using T4 Polynucleotide Kinase. The protein-DNA binding reaction was carried out in a volume of 10 µl containing 12 mM HEPES-KOH pH 7.6, 10 mM MgCl2, 0.5% (w/v) BSA, 1 mM DTT, 10% Glycerol (v/v), 200 mM NaCl, 4 mM Na2HPO4, 4 mM NaH2PO4, 0.4 mM EDTA, a 200-fold molar excess of competitor DNA (polydIdC), 1 nM of labelled DNA and increasing concentrations of wild-type or mutant AraR proteins, and incubated at room temperature for 30 min. The reaction mixtures were then submitted to electrophoresis on a native 8% polyacrylamide gel containing Tris-glycine buffer (25 mM Tris, 200 mM glycine, pH 8.9) and run at 100 V for ∼1 h. Gels were vacuum dried and exposed on a Phosphorimager screen before analysis with a Molecular Dynamics Storm 860 Imager and ImageQuant version 5.0.
The determination of the dissociation constants, Kd values, was obtained using the GraphPad Prism software and the “one site total binding” model, following the equation Y = Bmax.X/(Kd+X)+NS.X, with X = AraR concentration, Y = bound protein, Bmax is the maximum specific binding and NS is the slope of nonspecific binding. Concentrations of AraR were determined assuming a pure dimeric protein. Differences between Kd were analyzed by Mann Whitney U test using SPSS software, P<0.05 was considered as the level of statistical significance. The value 0.057 (Table 3) was considered moderate evidence against the null hypothesis [H0: On average there is no difference in binding affinity of the two DNA fragments (mutant DNA fragment vs wild-type DNA fragment)]. The association constant (Kass) is calculated from Kd = 1/Kass, and the Gibbs free energy (ΔG°) by ΔG° = −RT ln Kass.
Results
Probing amino acid-base contacts in vitro
In a previous study aimed at understanding the specific properties of the interaction AraR-operator sequences, we substituted amino acids, in or near the winged-HTH motif, which according to the model were predicted to contact DNA [20], [21], and the effects of these substitutions on the ability of AraR to function in vivo and on the DNA-binding affinities in vitro were determined [20], [21]. Conversely, mutational analysis of the AraR-binding sites was used to determine the base-specific requirements for transcriptional regulation in vivo and DNA binding in vitro. These experiments showed that specific AraR residues and operator bases are crucial to achieve a high level of regulatory activity, while others display variable contributions to DNA binding. In order to characterize in detail the AraR-DNA specific interaction we used the loss-of-contact approach [27]. In this study we initially used an in vitro missing-contact probing [28], [29] using electrophoretic mobility shift assay (EMSA) to determine the binding affinities of AraR and mutant proteins to a DNA fragment bearing the promoter of the metabolic operon with two operators (ORA1-ORA2) and the same fragment comprising single base pair substitutions in the ORA1 box (AATTGTTCGTACAAAT). The rationale of these experiments was as following: a certain amino acid alteration leads to an increase in Kd for the wild-type operator (Figure 2A); if this increment is the consequence of a lost direct or indirect interaction between that particular amino acid and a specific base, when we use a DNA fragment with a substitution in that particular base we expect no major effect in the Kd, when compared to the wild-type DNA, because a particular contact had already been lost and quantified (Figure 2B); in contrast, if the amino acid exchanged is not involved in contacts with the specific mutated base we will expect an additional increase in Kd (Figure 2C).
Schematic representation of the AraR protein in dark grey, and DNA fragment in light grey comprising one operator sequence in dark grey. Each base is represented by a square. Amino acids in contact with the DNA are depicted as triangles. Open triangles indicate mutated amino acids. Open squares represent mutated base. Arrows denote increase in Kd. A) A certain mutation in an amino acid leads to an increase in Kd for the wild-type operator as consequence of a specific interaction that was lost; B) any DNA position normally contacted by the altered amino acid may be mutated with little or no effect; C) any DNA position not involved in contacts by the altered amino acid when mutated leads to a cumulative increase in Kd.
This methodology, in addition to indicating residues directly involved in contacts with bases may also reveal amino acids whose presence is important to maintain the overall structural arrangement of the protein even though they do not directly contact bases in the DNA. For the experiments we chose AraR mutant proteins, AraR Y5F and AraR E30A, which displayed a moderate effect in AraR-DNA interaction both in vivo and in vitro, and base pair substitutions leading to partial derepression in vivo, A1→C, G5→T, T16→G and T6→G [21]. The results of the EMSA are summarized in Figure 3 and the calculated Kd values are shown in Table 3. The AraR wild-type protein showed a statistical significant decrease in the affinities for a DNA fragment bearing the promoter of the metabolic operon with two operators (ORA1-ORA2), when we compared the wild-type DNA fragment to the same fragment bearing mutations in the ORA1 box. Previously, we have shown that binding of AraR to ORA1-ORA2 is cooperative and a single point mutation in either ORA1 and ORA2 causes an almost complete loss of AraR regulation in vivo [10], [19]. Similarly, in vitro a single-point mutation in ORA1 reduces dramatically the apparent affinity of AraR for the second operator ORA2 [10].
AraR wild type left column; AraR E30A middle column; and AraR Y5F mutant right column. The indicated amounts of AraR protein were used in the binding reactions, AraR was incubated with the 5′-end labelled probe (1 nM) bearing the wild-type or mutant operators ORA1-ORA2 and the protein-DNA complexes were resolved on native 8% polyacrylamide gels. The mutation in each DNA fragment is depicted.
The AraR E30A protein displayed a decrease in the affinity for all mutated operators except for the T6→G operator (Table 3). In fact, AraR E30A showed no additional significant decrease in the affinity, relative to the wild-type operator, when the T6→G operator mutant was used (Figure 3 and Table 3). As T6 in ORA1 is important for protein binding [21], and the T6→G mutation did not reduce the binding affinity of AraR E30A, this suggests that this operator substitution did not further affect the loss of contact of AraR E30A. The Kd of the mutant AraR Y5F for the operator mutations tested revealed a significant a reduction in the affinity compared to the wild-type for G5→T and T6→G, but not for A1→C or T16→G (Figure 3 and Table 3). This could indicate that Y5 might be relevant for the contact of AraR with T16 and A1 of ORA1. Because these nucleotides are located in opposite positions in the palindromic sequence of the operator, this observation suggests that Y5 of one monomer is important for the interaction with A1, while the other contacts T16. However, the crystal structure of the AraR-DNA binding domain bound to ORA1 [14] showed Y5 interacting with the DNA backbone near nucleotide T6 (see below).
In summary, the results obtained in vitro suggest that AraR residue E30 may play an important role in the interaction of the protein with the T6 nucleotide.
In vivo validation of protein-DNA interactions
Since the experimental conditions used to derive Kd values bear little resemblance to intracellular situations, the in vitro results were confirmed by in vivo assays. For this, we constructed B. subtilis strains in order to confront the different araR alleles and mutant DNA operator sequences in the same cell. The different araR alleles were ectopically introduced at the amyE locus of an araR null mutant background. Additionally a transcriptional fusion between the araA promoter, carrying the ORA1-ORA2 operators, and the E. coli lacZ gene, was generated and ectopically introduced at the B. subtilis thrC locus (Figure 4). This genetic system allows us to measure the regulatory activity of the native and mutant proteins over distinct promoters (wild-type and mutated) fused to the lacZ reporter gene by determination of the levels of accumulated β-galactosidase. In previous studies we have shown that in these conditions the cellular level of both mutant proteins AraR E30A and AraR Y5F is comparable to that seen with wild-type AraR, ruling out the possibility of deregulation originated by degradation of the repressor [21]. The results of the confrontation of the different araR alleles and the various promoters in the series of strains constructed are summarized in Table 4.
The circle illustrates the B. subtilis chromosome and the location of the amyE, araE/araR, and thrC loci indicated in degrees. The construction containing the wild-type or mutant araR alleles placed at the amyE locus is represented in the top left. The araR-null genetic background is depicted in middle left. The regulatory activity exerted by the araR alleles over the wild-type or mutant araA promoter sequences is measured by a promoter lacZ fusion placed at the thrC locus (bottom left).
The analysis of repression index of the wild-type AraR with the different promoter fragments showed a decrease in the regulatory activity when a mutated box ORA1 was used, compared to the wild-type ORA1A2. The mutation ORA1 T16→G displayed the higher deregulation, while ORA1 A1→C and T6→G exhibited similar less drastic effects. These results are comparable to those obtained in the in vitro assays (Table 3). The dissociation constant of the mutant Y5F suggested that this amino acid might interact with two nucleotides in the operator sequence, T16 and A1 (Table 3). However, the in vivo analysis does not corroborate the hypothesis (Table 4), as mutations at position T16 and A1 have a drastic effect in the regulatory activity of mutant Y5F (IQB792 and IQB793; Table 4). The in vivo results are in agreement with the results of the crystal structure of the AraR-DNA binding domain bound to ORA1 [14] that revealed Y5 interacting with the DNA backbone near nucleotide T6, thus this residue is not involved in direct or indirect contact with T16 and A1 (discussed below).
The EMSA assays indicated that residue E30 could be relevant for the interaction of the AraR protein with the T6 nucleotide (Table 3), although both the N-terminal AraR model [21] and the N-terminal AraR-ORA1 structure [14] suggest non-specific contacts of E30 to the DNA backbone (discussed below). This observation was supported by the in vivo data because the regulatory activity of mutant AraR E30A over the mutant ORA1 T6→G-lacZ promoter fusion is 2.7-fold higher (strain IQB798, Table 4) than that observed for the wild type promoter ORA1A2WT-lacZ (strain IQB779, Table 4). Furthermore, the lower level of expression observed in the strain bearing the mutant AraR E30A and the mutant ORA1 T6→G-lacZ promoter fusion (strain IQB798, Table 4), both in the presence and absence of inducer, compared to that obtained in the strain harbouring the wild-type AraR regulator and the mutant ORA1 T6→G-lacZ promoter fusion (strain IQB790, Table 4) suggests a stronger interaction of the E30A protein towards the mutated DNA operator.
Overall the in vivo results highlight the importance of amino acid E30 in the regulatory activity AraR and in the contact of the protein with the nucleotide T6 in ORA1.
Residue E30 is important for the AraR regulatory activity in distinct promoters
As T6 is a well-conserved nucleotide in the consensus signature of the AraR DNA binding site, present in all AraR operators characterized so far (Figure 1), to establish that E30 is an important amino acid for the AraR contact to the thymine at position 6 we assayed this effect in the context of a different promoter. The abf2 gene is regulated by cooperative binding of AraR to two in-phase operators ORX1X2 similarly to that observed in the arabinose metabolic operon promoter ([18]; Figure 1). Thus, using the same strategy the wild-type ORX1 (ATACATACGTACAAAT) and mutant ORX1T6→G abf2′-lacZ fusions were constructed and introduced at the B. subtilis thrC locus.
The analysis of the regulatory index exerted by the native AraR in the strain IQB927 showed no effect of ORX1T6→G mutation when compared to the wild-type promoter (strain IQB926, Table 5). On the other hand, mutant AraR E30A leads to a complete loss of the regulation of the wild-type abf2′-lacZ promoter fusion abf2, showing once again the importance of this amino acid in the regulatory mechanism of this transcription factor. Nevertheless, the confrontation of the mutant E30A with mutation T6→G (strain IQB929, Table 5) leads to an increase in the regulatory activity when compared to the wild-type promoter (strain IQB928, Table 5). Therefore, the T6→G single nucleotide change partially suppresses the loss of regulation caused by the E30A amino acid substitution pointing out that E30 is an important amino acid for the AraR contact to the thymine at position 6 of both operator sequences ORA1 and ORX1.
Discussion
The sequence-specificity of DNA recognition by proteins should be viewed in a complete framework. At the atomic level the specificity of DNA-binding proteins is mainly accomplished through direct hydrogen bond and hydrophobic interactions between specific amino acid side chains and functional groups of nucleotide bases in the major and minor groove [1], [2], [3], [4], [30]. Nevertheless these direct or water-mediated hydrogen bonds are insufficient to completely explain the specificity of many DNA-binding proteins. In addition to the chemical complementarity between protein and DNA atoms, it is required a structural complementarity along the networking surfaces of the protein and DNA molecules [31]. The use of genetic methods to identify amino acid base pair contacts in a specific protein-DNA complex is a complementary approach to the X-ray diffraction and to two-dimensional nuclear magnetic resonance spectroscopic (2D NMR) analyses. Furthermore, the construction and analysis of single amino acid substitutions is the only method to determine the apparent binding free energy contribution and the apparent specificity free energy contribution of an amino acid-base pair contact [27 and references therein].
The GntR family members, in general, possess a DNA binding at the N-terminus of the protein and an effector-binding and/or oligomerisation domain at the C-terminus (Pfam family: PF00392; Prosite Family PS50949; [15]). The DNA-binding domain is conserved throughout the GntR family, consisting of a 3-helical bundle core with a small beta-sheet (wing), winged-HTH motif. Despite the vast number of GntR family members sequences deposited in databases there are only a few crystal structures available to examine in detail structure/function relationships. AraR is a transcription factor that typifies one of the sub-families of the GntR group, and recently the three-dimensional crystal structure of the AraR C-terminal domain [13] and the DNA-binding domain [14] were separately and independently determined. In this work, AraR was used to characterize specific interactions with the DNA by an in vitro missing-contact probing and posterior validation in vivo. In the in vitro a fragment
The results obtained in vitro with the AraR wild-type protein correlate well with those previously obtained in in vivo experiments [19], except for the mutation G5→T that showed a more accentuated decrease in the affinity measured in vitro than the loss of regulation observed in vivo [21]. Moreover, the data obtained in vivo in this study with the AraR wild-type protein are consistent with those previously observed in vivo using a different genetic system [21]. Although, The in vitro EMSA analysis using AraR mutant Y5F and the different DNA fragments bearing point mutations in the ORA1 operator suggested that residue Y5 could be important for protein contacts with two nucleotides in opposite sites of the operator palindromic sequence, T16 and A1 (Table 3), however the in vivo results do not corroborate this hypothesis (Table 4). The in vivo results validate the data of the crystal structure of the AraR DNA-binding domain in complex with two different operators, ORA1 and ORR3, showing specific contacts with DNA [14]. In fact, Y5 is not involved in direct or indirect contact with these nucleotides because it interacts with the DNA backbone near nucleotide T6. The analysis of the in vitro interaction between mutant AraR E30A with the mutant DNA fragments A1→C, T16→G and G5→T revealed a decrease in affinity when compared to the wild-type DNA indicating that residue E30 is not indirectly involved in contacts with the mutated bases. These mutated nucleotides are highly conserved across all AraR operators characterized so far [21], and accordingly to the AraR-ORA1 structure involved in the interaction with the protein. The opposite nucleotides of A1 and T16 are contacted by the same amino acid, G62, through an acetated or water-mediated interaction, respectively, but from different monomers, while G5 establishes a direct contact with amino acid R41 [14]. Surprisingly, the in vitro interaction studies with mutant T6→G displayed no decrease in the affinity of the mutant AraR E30A suggesting that residue E30 could be indirectly involved in contacts with T6. Furthermore, in vivo analysis performed with two distinct promoters showed that mutation T6→G partially suppresses the effect of substitution E30A in AraR improving its regulatory activity. In both strains bearing a lacZ fusion to different promoters an increase in the regulatory activity of the mutant E30A is observed (IQB798 Table 4 and IQB929 Table 5). Thus, the presence of an alanine at position 30 seems to have positive contribution to the interaction of the mutant ORA1 T6→G with the protein.
The E30 residue is highly conserved in the GntR-family proteins, and the corresponding residue in FadR, E34, was shown to contact the DNA backbone [32], [33]. The FadR-DNA structure indicates that E34 also contacts nearby amino acids, contributing presumably to the stabilization of residues that interact specifically with the DNA bases. Similarly, both the N-terminal AraR model and the N-terminal AraR-ORA1 structure suggest non-specific contacts of E30 to the DNA backbone [14], [21], and indicate possible interactions with R41 and R45 [14], [21]; and Figure 5A). The core of HTH motif is comprised by two α-helices, H2 and H3, spaced by a short four-residues turn (T) in between. In AraR E30 belongs to H2, the stabilizing helix, while R41 and 45 to H3, the recognition helix. The angle between H2 and H3 is typically of 120°, however it can vary between 100° and 150° [34]. Since E30 interacts with R41 and R45, this interaction is crucial to settle the geometry and spatial arrangement of H2 and H3, and protein docking on DNA by the recognition helix, H3 (Figure 5A and B). The role of the E30 is not only the interaction with the DNA but is also to limit the rotation of the recognition helix. In the E30A mutant, R41 and R45 are no longer interacting with E30, moreover this alanine substitution impairs the contacts of this residue with the DNA backbone (Figure 5C). As a result, the regulatory activity of the mutant protein decreases in the presence of the wild-type ara operon promoter, which does not occur in the presence of mutant ORA1 T6→G promoter as a consequence of a spatial orientation of H2 and H3 (Table 4). On the other hand, enrichment of the operator DNA with another guanine, T6→G, could lead to a significant alteration in DNA conformation. In fact, the exocyclic 2-amino groups of the guanines are crucial elements in DNA structure and recognition, as they are known to exert a substantial influence on DNA bending, flexibility and intrinsic curvature [35], [36], [37], [38]. Therefore if the functional groups in the protein do not correctly juxtapose with those in the DNA, protein-DNA complex stability is impaired, which seems to be the case of the wild-type AraR interaction with the mutated operator T6→G. An amino acid not directly involved in contacts with bases, such as E30, placed within or adjacent to the DNA binding domain can therefore indirectly affect the affinity of the protein to DNA by properly modulating the protein conformation, allowing a correct alignment between the functional groups of the protein and the DNA.
Native AraR structure displaying interactions between: A) residues E30, R41 and R45 and B) residues (E30, R41 and R45), and R41 and guanine 5 (ORA1). Mutant AraR E30A displaying possible interactions between: C) residues E30A, R41 and R45 and D) residues (R41, R45) and guanine 5 and 6 (ORA1 T6→G). E30 (yellow), A30 (grey), R41 (light blue), R45 (purple) and direct hydrogen bonds (side chain or protein-DNA) are shown in dashed lines. The structures were drawn using PyMOL (http://pymol.sourceforge.net/) and the data of the structure of the AraR N-terminal domain in complex with ORA1 (PDB access no. 4EGY; [14]).
Although there is no ‘recognition code’ between amino acids and nucleotides, they possess some preferential interactions, for instance arginines are known to interact favourably with guanines [4], [5], [6]. Thus, we propose that the effect observed in vivo of the recovery of regulation in the double mutant E30A ORA1 T6→G is due to the loss of interaction between E30, and R41 or R45, which results in a conformational change that allows a proper arrangement between the functional groups of the protein and the new operator DNA composition. R41 and R45 became free to establish new interactions with the nucleotides, not only the G at position 5, but also with the new G at position 6 (Figure 5D). Thus, the E30A mutation results in a better contact of the latter residues (R41 or R45) with G5 and the mutated G6 adjusting to the new DNA sequence, as observed by the increased regulatory activity of the mutant protein in the presence of the mutated operators (ORA1 and ORX2) when compared to the native protein (Table 4 and Table 5).
Our results provide information beyond the pairwise analysis, the data highlight and demonstrate that residues that are not involved in specific interactions with nucleotides, but act as linker residues by positioning other amino acids in the correct 3D context of a nucleoprotein complex, can be as important for the protein-DNA interaction as residues making direct contact with DNA bases, and have a crucial role in the modulation of DNA recognition. Furthermore, we show that by manipulating these residues it is possible to redesign the specificity of protein–DNA interactions.
Acknowledgments
We are grateful to Charles Moran Jr. for the helpful discussions and suggestions and Jaime Mota for critically reading this manuscript. The assistance of Augusta Correia in the statistic analysis of data is greatly appreciated.
Author Contributions
Conceived and designed the experiments: ILC ISF IS-N. Performed the experiments: ILC ISF. Analyzed the data: ILC IS-N. Wrote the paper: ILC ISF IS-N.
References
- 1. Schleif R (1988) DNA binding by proteins. Science 241: 1182–1187.
- 2. Pabo CO, Sauer RT (1984) Protein-DNA recognition. Annu Rev Biochem 53: 293–321.
- 3. Pabo CO, Sauer RT (1992) Transcriptions factors: structural families and principles of DNA recognition. Annu Rev Biochem 61: 1053–1095.
- 4. Luscombe NM, Laskowski RA, Thornton JM (2001) Amino acid–base interactions: a three-dimensional analysis of protein–DNA interactions at an atomic level. Nucleic Acids Res 29: 2860–2874.
- 5. Luscombe NM, Thornton JM (2002) Protein–DNA interactions: Amino acid conservation and the effects of mutations on binding specificity. J Mol Biol 320: 991–1009.
- 6. Marabotti A, Spyrakis F, Facchiano A, Cozzini P, Alberti S, et al. (2008) Energy-based prediction of amino acid-nucleotide base recognition. J Comput Chem 12: 1955–1969.
- 7. Urnov FD, Rebar EJ, Holmes MC, Zhang HS, Gregory PD (2010) Genome editing with engineered zinc finger nucleases. Nat Rev Genet 11 (9) 636–646.
- 8. Bogdanove AJ, Voytas DF (2011) TAL Effectors: Customizable Proteins for DNA Targeting. Science 333: 1843–1846.
- 9. Sá-Nogueira I, Mota LJ (1997) Negative regulation of L-arabinose metabolism in Bacillus subtilis: characterization of the araR (araC) gene. J Bacteriol 179: 1598–1608.
- 10. Mota LJ, Tavares P, de Sá-Nogueira I (1999) Mode of action of AraR, the key regulator of L-arabinose metabolism in Bacillus subtilis. Mol Microbiol 33: 476–489.
- 11. Haydon DJ, Guest JR (1991) A new family of bacterial regulatory proteins. FEMS Microbiol Lett 63: 291–295.
- 12. Weickert MJ, Adhya S (1992) A family of bacterial regulators homologous to Gal and Lac repressors. J Biol Chem 267: 15869–15874.
- 13. Procházková K, Čermáková K, Pachl P, Sieglová I, Fabry M, et al. (2012) Structure of the effector-binding domain of the arabinose repressor AraR from Bacillus subtilis. Acta Crystallogr D Biol Crystallogr 68: 176–185.
- 14. Jain D, Nair DT (2013) Spacing between core recognition motifs determines relative orientation of AraR monomers on bipartite operators. Nucleic Acids Res 41: 639–647.
- 15. Hoskisson PA, Rigali S (2009) Chapter 1. Variation in form and function the helix-turn-helix regulators of the GntR superfamily. Adv Appl Microbiol 69: 1–22.
- 16. Sá-Nogueira I, Nogueira TV, Soares S, de Lencastre H (1997) The Bacillus subtilis L-arabinose (ara) operon: nucleotide sequence, genetic organization and expression. Microbiology 143: 957–969.
- 17. Sá-Nogueira I, Ramos SS (1997) Cloning, functional analysis, and transcriptional regulation of the Bacillus subtilis araE gene involved in L-arabinose utilization. J Bacteriol 179: 7705–7711.
- 18. Raposo MP, Inácio JM, Mota LJ, de Sá–Nogueira I (2004) Transcriptional regulation of genes encoding arabinan-degrading enzymes in Bacillus subtilis. J Bacteriol 186: 1287–1296.
- 19. Mota LJ, Sarmento LM, de Sá -Nogueira I (2001) Control of the arabinose regulon in Bacillus subtilis by AraR in vivo: crucial roles of operators, cooperativity, and DNA looping. J Bacteriol 183: 4190–4201.
- 20. Franco IS, Mota LJ, Soares CM, de Sá-Nogueira I (2006) Functional domains of Bacillus subtilis transcription factor AraR and identification of amino acids important for nucleoprotein complex assembly and effector-binding. J Bacteriol 188: 3024–3036.
- 21. Franco IS, Mota LJ, Soares CM, de Sá-Nogueira I (2007) Probing key DNA contacts in AraR-mediated transcriptional repression of the Bacillus subtilis arabinose regulon. Nucleic Acids Res 35: 4755–4766.
- 22.
Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor Laboratory. Cold Spring Harbor. NY.
- 23. Pascal M, Kunst F, Lepesant JA, Dedonder R (1971) Characterization of two sucrase activities in Bacillus subtilis Marburg. Biochimie 53: 1059–1066.
- 24. Spizizen J (1958) Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleotide. Proc Natl Acad Sci USA 22: 1072–1078.
- 25.
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
- 26. Guérout-Fleury AM, Frandsen N, Stragier P (1996) Plasmids for ectopic integration in Bacillus subtilis. Gene 180 (1–2) 57–61.
- 27. Ebright RH (1991) Identification of amino acid-base pair contacts by genetic methods. Methods Enzymol 208: 620–640.
- 28. Brunelle A, Schleif R (1987) Missing contact probing of DNA-protein interactions. Proc Natl Acad Sci USA 84: 6673–6676.
- 29. Brunelle A, Schleif R (1989) Determining residue-base interactions between AraC protein and araI DNA. J Mol Biol 209: 607–622.
- 30. Martin AM, Sam MD, Reich NO, Perona JJ (1999) Structural and energetic origins of indirect readout in site-specific DNA cleavage by a restriction endonuclease. Nat Struct Biol 6: 269–377.
- 31. Hilchey SP, Koudelka GB (1997) DNA-based loss of specificity mutations. J Biol Chem 272: 1646–1653.
- 32. van Aalten DM, DiRusso CC, Knudsen J (2001) The structural basis of acyl coenzyme A-dependent regulation of the transcription factor FadR. EMBO J 20: 2041–2050.
- 33. Xu Y, Heath RJ, Li Z, Rock CO, White SW (2001) The FadR-DNA complex. Transcriptional control of fatty acid metabolism in Escherichia coli. J Biol Chem 276: 17373–17379.
- 34. Gajiwala KS, Burley SK (2000) Winged helix proteins. Curr Opin Struct Biol 10: 110–116.
- 35. Bailly C, Møllegaard NE, Nielsen PE, Waring MJ (1995) The influence of the 2-amino group of guanine on DNA conformation. Uranyl and DNase I probing of inosine/diaminopurine substituted DNA. EMBO J 14: 2121–2131.
- 36. Bailly C, Waring MJ, Travers AA (1995) Effects of base substitutions on the binding of a DNA-bending protein. J Mol Biol 253: 1–7.
- 37. Møllegaard NE, Bailly C, Waring MJ, Nielsen PE (1997) Effects of diaminopurine and inosine substitutions on A-tract induced DNA curvature. Importance of the 3′-A-tract junction. Nucleic Acids Res 25: 3497–3502.
- 38. Lindemose S, Nielsen PE, Møllegaard NE (2008) Dissecting direct and indirect readout of cAMP receptor protein DNA binding using an inosine and 2,6-diaminopurine in vitro selection system. Nucleic Acids Res 36 (14) 4797–807.