FeON-FeOFF: the Helicobacter pylori Fur regulator commutates iron-responsive transcription by discriminative readout of opposed DNA grooves

Abstract

Most transcriptional regulators bind nucleotide motifs in the major groove, although some are able to recognize molecular determinants conferred by the minor groove of DNA. Here we report a transcriptional commutator switch that exploits the alternative readout of grooves to mediate opposite output regulation for the same input signal. This mechanism accounts for the ability of the Helicobacter pylori Fur regulator to repress the expression of both iron-inducible and iron-repressible genes. When iron is scarce, Fur binds to DNA as a dimer, through the readout of thymine pairs in the major groove, repressing iron-inducible transcription (FeON). Conversely, on iron-repressible elements the metal ion acts as corepressor, inducing Fur multimerization with consequent minor groove readout of AT-rich inverted repeats (FeOFF). Our results provide first evidence for a novel regulatory paradigm, in which the discriminative readout of DNA grooves enables to toggle between the repression of genes in a mutually exclusive manner.

Figure 1.

Different Fur binding stoichiometry and operator consensus motifs involved in FeOFF-FeON regulation. (A) HpFur can directly regulate the transcription of both the iron-repressible frpB1 (FeOFF) and the iron-inducible pfr (FeON) genes, according to the metallation state of the protein, through affinity variations for specific operators in their promoters. In both cases, the promoters are derepressed in fur mutants. The sites with highest affinity span the −10 and/or −35 core promoter elements, whereas the lower affinity Fur binding sites are located further upstream from the primary Fur box. This fosters the current model of Fur competing with the RNA polymerase for binding to target promoters. (B) Gel shifts with OPI_frpB and OPI_pfr probes represent archetypal examples of Fur-DNA complexes with holo- or apo-Fur operator elements. Increasing amounts of Fur protein were incubated with end-labelled probes in presence of 150 μM 2,2 dipyridyl (Dipy) and 150 μM soluble Fe²⁺. Lanes 1–7; 0, 1.2, 2.4, 4.8, 9.6, 18, 39 nM Fur added, respectively. Single asterisks indicate free probe, double asterisks indicate free radiolabeled pBS vector used as non-specific competitor in binding reactions. HMC and LMC denote the high and low mobility complexes formed in a metal-dependent manner, respectively. (C) Sequence alignment for holo-Fur (FeOFF) and apo-Fur (FeON)-specific operators. Asterisks above the sequence alignments indicate conserved bases (grey boxes). Alignments were used to identify a consensus motif, sketched with WebLOGO. Coloured figure available in online version.

Figure 2.

Nucleotide sequences and representative EMSA experiments for wild-type OPI_frpB (A) or mutated loss of function OPI_{frpB rep}⁻ (B), gain of function OPI_{frpB ind}⁺ (C) and swapped OPI_{frpB rep}⁻_ind⁺ (D) probes. Nucleotides protected from •OH cleavage (data from Figure 4) on the wild-type OPI_frpB operator are shaded in grey. Convergent arrows above the sequence indicate the holo-operator inverted repeat. Mutated nucleotides are shaded in black. Nucleotides forming a TCATT_n10TT consensus motif within the OPI_frpB holo-operator element are underscored in bold letters.

Figure 3.

Reconstitution of a TCATT_n10TT motif within a holo-operator confers apo-regulation in vivo. Representative lacZ primer extensions on total RNA extracted at mid-log growth phase from G27 wild-type (wt) or fur knock out (Δfur) strains encompassing the vacA::P_OPIfrpB-lacZ transcriptional fusions, with wild-type OPI_frpB or the gain-of-function OPI_{frpB ind}⁺ operator in response to iron repletion [1mM (NH₄)₂Fe(SO₄)₂; Fe+] or iron chelation (150 μM 2,2’-dipyridyl; Fe-). Arrows above the OPI_frpB nucleotide sequence indicate the inverted repeat of the holo-Fur binding consensus motif. Bent arrows mark the transcriptional start site; the mapped −10 box of the promoter is shaded in black. The two mutagenized bases in OPI_frpBind⁺ are indicated in bold lowercase letters. The FeOFF/FeON transcript ratio is reported in the graph; grey bars: G27 wt genetic background; black bars: Δfur genetic background; vertical bars indicate the standard deviation of three independent replicates.

Figure 4.

Distinctive binding architecture of Fur to holo- and apo-operator elements. (A and B) Specific DNA probes for P_frpB (A) and P_pfr (B) fragments, end labelled on the coding strand, were incubated with increasing amounts of recombinant Fur protein in presence of 150 μM 2,2’dipyridyl (Dipy, left panel) or in presence of 150 μM MnCl₂ (Mn²⁺, right panel). Lanes 1–7; 0, 29, 61, 122, 244, 490, 980 nM Fur (monomer) added, respectively. Fur binding sites, protected by DNase I, with highest affinity for either holo- (light grey boxes) or apo-Fur (dark boxes). The open boxes on the right indicate the extended region of •OH protection, whereas the arrowheads indicate short protected areas from •OH cleavage. Bent arrows mark transcriptional start sites, the position of the −10 and −35 hexamers are marked by open rectangles; open reading frames are indicated by vertical open arrows to the left of each gel. (C and D) Summary of protection data on operators from P_frpB (C) and P_pfr (D). For each operator, the numbers are referred to the respective transcriptional start site (+1). Open circles indicate bases protected by Fur on the coding or the non-coding strand. Strongly protected bases are shaded in grey. (E and F) Representative helical projections of the •OH protected residues (open circles) on the OPI_frpB (E) and OPI_pfr (F) DNA backbone. Shaded and black bars in the DNA helix represent adenine and thymine bases, respectively.

Figure 5.

A swap in the iron-dependent binding affinity changes the binding architecture of Fur to the operator element. (A) The •OH footprinting assays on OPI_frpB, OPI_{frpB rep}⁻_ind⁺ and OPI_pfr probes. Symbols are described in the legend of Figure 4. (B) Sequence alignments of the OPI_frpB, OPI_{frpB rep}⁻_ind⁺ and OPI_pfr operators. The four point mutations reconstituting a TCATT_n10TT element in a holo-Fur loss-of-function operator are shaded in black. The nucleotides protected from •OH cleavage in panel A are shaded in grey. Arrows above the OP_frpB nucleotide sequence indicate the inverted repeat of the holo-Fur binding consensus motif. Horizontal bars mark the TCATT_n10TT elements.

Figure 6.

Minor groove readout in iron-repressible Fur regulation. (A) Distamycin A interference assays in vitro with OPI_frpB, OPI_pfr and OPI_{frpB ind}⁺ operator probes. Lane 1, free probe. Lane 2, Fur-DNA complexes formed in the absence of distamycin A. Lanes 3–5, Fur-DNA complexes formed in the presence of 1.2, 2.4 and 4.8 nM distamycin A, respectively. Symbols are as in Figure 2. (B) I-C box substitutions of OPI_frpB in OPI_{frpB I=C} are shaded in black. Nucleotides protected from •OH cleavage are shaded in grey. Arrows above the nucleotide sequences indicate the inverted repeat of the holo-operator consensus motif. (C) EMSA of OPI_frpB (left panel) and I-C-substituted OPI_{frpB I=C} probes with increasing concentrations of Fur in the presence of either 150 μM iron (Fe²⁺) or 150 μM 2,2 dipyridyl (Dipy). Lane 1: free probe, 0 nM Fur (*); lanes 2–4 and 5–7, 19, 38 and 190 nM Fur dimer, respectively. Open diamond, circle and square indicate HMC, LMC and LMC2 Fur-DNA complexes, respectively. (D) Effects of distamycin A on iron-dependent Fur repression of frpB and pfr. Transcript levels were quantified by qRT-PCR on RNA extracted from mid-log H. pylori G27 cultures treated 15 min with 1 mM soluble Fe²⁺ or 150 µM 2,2 dipyridyl (Dipy), after 20 min preincubation with ddH₂O (cntrl) or 20 µM distamycin A (Dist A). The housekeeping gene ppk was used as control. To take in account only Fur-dependent responses, results were normalized to frpB and pfr transcript levels observed in the G27Δfur strain under the same conditions. Statistically significant differences were assessed by Student’s t-test. Error bars indicate the standard deviation deriving from two independent biological duplicates, each analysed twice in independent qRT-PCR runs, in triplicate technical replicates for each sample.

Figure 7.

Fur-DNA interaction models. Best docking models resulting for Fur-OPI_pfr (A) and Fur-OPI_frpB (B) complexes (coloured in the online version). The protein is reported as ribbon diagram coloured from deep blue in the proximity of the N-terminal to red at the C-terminus. Zn²⁺ ions are reported as purple spheres. The DNA is reported as ribbon coloured in blue, with exception of the active residues, reported in red.