Metal-binding properties of Hpn from Helicobacter pylori and implications for the therapeutic activity of bismuth

Abstract

Nickel is of particular importance to Helicobacter pylori in part because it acts as a cofactor of urease, which is critical to the survival of H. pylori. In this study the nickel storage, histidine-rich protein Hpn from H. pylori was converted into a Ni²⁺ probe by inserting it between two fluorescence resonance energy transfer (FRET) partners, cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP). The resulting construct, Hpn-FRET, exhibited a change in FRET upon the binding of Ni²⁺. Hpn-FRET has a moderate selectivity for Ni²⁺; it also responds to Zn²⁺ and Co²⁺ but not to other biometals. Competition experiments between Ni²⁺ and other metals plus the measured K_d values for Zn²⁺ and Ni²⁺ establish the selectivity order for Hpn-FRET as Zn²⁺ > Ni²⁺ > Co²⁺ ≫ other biometals. Bismuth is widely used as a therapeutic agent against H. pylori, and Hpn has been suggested as one of the possible targets. The dissociation constant of Bi³⁺ to Hpn-FRET was measured to be 6.19 × 10⁻⁵ M. Further experiments using Hpn-FRET in E. coli indicate that Hpn-FRET responds to Bi³⁺ but not to Ni²⁺ and Zn²⁺ inside E. coli. The result shows that unlike Ni²⁺ and Zn²⁺, which are tightly regulated in most bacteria, available Bi³⁺ can reach high micromolar levels inside E. coli.

---

Introduction

Helicobacter pylori is a gram-negative pathogenic bacterium which colonizes the digestive track and causes chronic gastritis, ulcers, and even some gastric cancers.¹ This bacterium has the remarkable ability to survive in the hostile environment of the stomach and is found in about half the world's human population.² Its ability to adapt to the acidic living conditions of the stomach lies in its high activity of urease, which catalyzes the hydrolysis of urea to produce carbon dioxide and ammonia that neutralize the internal pH.³Urease, accounting for about 10% of the total protein content in the bacterium, uses two nickel ions for catalysis in its holo form.⁴ Another important enzyme for H. pylori that also uses a nickel cofactor is hydrogenase which provides energy to this bacterium.⁵ Therefore, nickel is an essential metal for the survival of H. pylori, and its metabolism is globally regulated by the transcriptional regulatorHpNikR.⁶ An excess of nickel has been shown to be harmful, however, and the molecular mechanism of this toxicity is not understood in detail.

One line of treatment used against H. pylori are bismuth-containing therapeutics such as bismuth subsalicylate, bismuth subcitrate, and ranitidine bismuth citrate.⁷ While the exact mechanism of action for bismuth remains unknown, all current hypotheses suggest binding of bismuth to the metal-binding sites in various metalloproteins.⁸In vitro studies have shown that Bi³⁺ can bind to the active sites of urease⁹ and alcohol dehydrogenase,¹⁰ resulting in the inhibition of the catalytic activity of these enzymes. Other proposed target proteins of Bi³⁺ are metallothioneins¹¹ and the histidine-rich protein Hpn.¹² While Bi³⁺ has a low toxicity in humans, it is known to interact with serum proteins such as transferrin,¹³ lactoferrin,¹⁴ and serum albumin.¹⁵

In H. pylori as well as many other organisms, transition metals such as nickel play a dual role: they are at once essential as cofactors to enzymes that catalyze important transformations, but are also toxic at elevated levels. Substantial work has been done to understand the elaborate homeostatic networks for these metals, the individual regulation, uptake, distribution, delivery, storage, efflux, and the role these metals play in various diseases; however, the picture is still far from complete.^16,17 As part of the effort to elucidate the function of metals in living systems, fluorescent small-molecule and protein-based sensors have been developed to visualize metals.¹⁸ These two classes of sensors have complementary applications and advantages. While both small molecule and protein-based sensors for Ca²⁺,^19,20Zn²⁺,^21,22 and Cu^+23,24 have been developed and successfully used in vivo, small molecule and protein-based probes for Ni²⁺ and other transition metals are still lacking.²⁵

In this study Hpn, which is a histidine-rich protein, was converted into a FRET-based fluorescent sensor, Hpn-FRET, by inserting Hpn between the two FRET partners CFP (cyan fluorescent protein) and YFP (yellow fluorescent protein) (Fig. 1b). This approach to construct a protein based FRET sensor by introducing a metal binding domain which undergoes a conformational change between two FRET partners has been widely used^19,21,26 and we have previously converted metallothioneins which are small cysteine-rich proteins into effective FRET probes that demonstrate high selectivity to copper(I) due to the assembly of a tetracopper(I) cluster.²³ Hpn was chosen as a sensory domain because it has been reported to function as a nickel storage protein that provides nickel to the active site of urease in H. pylori.^27,28 Similarly to metallothioneins Hpn bears a large number of coordinating amino acids in a short sequence and changes conformation upon metal binding as has been observed by CD.^29,30 Overexpression of Hpn protects H. pylori from high concentrations of metal, especially nickel, and its expression is upregulated by HpNikR in the presence of excess Ni²⁺.^29,30 Hpn has also been proposed as a target for bismuth because deletion of the hpngene decreases the survival of H. pylori with bismuth treatment.¹² Hpn, is a histidine-rich small protein that contains 28 histidines out of a total of 60 residues (Fig. 1a) and has been shown to bind approximately 5 equiv. of Ni²⁺ (K_d = 7.1 × 10⁻⁶ M), 5 equiv. of Zn²⁺ (K_d = 19.28 × 10⁻⁶ M), 8.5 equiv. of Cu²⁺ (K_d = 2.16 × 10⁻⁶ M), and 3.81 equiv. of Bi³⁺ (K_d = 11.1 × 10⁻⁶ M), in a study in which the dissociation constants were determined by equilibrium dialysis.³¹

Fig. 1 Design of Hpn-FRET. (a) Sequence of Hpn, which is a histidine-rich metallothionein. (b) Hpn sequence is inserted between the FRET partners CFP and YFP. The binding of metals to Hpn results in a conformational change which leads to a change of FRET signal. (c) Fluorescence response of Hpn-FRET to Ni²⁺. Apo Hpn-FRET is shown in blue, and Hpn-FRET with 6 equiv. Ni²⁺ is shown in pink.

In the presented work, the metal binding properties of the resulting probe, Hpn-FRET, for different biometals and Bi³⁺ were characterized in vitro and the selectivity and affinity of Hpn-FRET were established. When Hpn-FRET was expressed in E. coli which is used as a model system, FRET change was only observed with the addition of Bi³⁺, but not Ni²⁺ and Zn²⁺. These measurements show that Hpn is a likely target for Bi³⁺in vivo. Unlike Ni²⁺ and Zn²⁺, which are tightly controlled in bacterial cells, the free concentration of Bi³⁺ can reach high micromolar levels.

Results

Design of Hpn-FRET and the response to Ni²⁺

Hpn is a histidine-rich protein and has been shown to bind to transition metals, resulting in a change in secondary structure (Fig. 1a). Taking advantage of this conformational change upon metal binding, we cloned Hpn between the two FRET partners CFP and YFP to yield Hpn-FRET (Fig. 1b). Hpn-FRET was overexpressed in E. coli, purified by Ni-NTA and gel filtration columns, and its purity was confirmed by a denaturing SDS-PAGE protein gel (Fig. S1, ESI†). Fluorescence spectra of Hpn-FRET (250 nM) were recorded in the absence and presence of 1.5 μM Ni²⁺ by exciting CFP (433 nm) and scanning from 450 to 600 nm. The addition of Ni²⁺ to Hpn-FRET resulted in a decrease in the fluorescence intensity of CFP (477 nm) and an increase in fluorescence intensity of YFP (527 nm) due to a change in the FRET efficiency upon metal binding (Fig. 1c). The conformational change of Hpn-FRET upon metal binding was also observed in the CD spectrum (Fig. S2, Table S1, ESI†).

Selectivity of Hpn-FRET

The metal-binding properties of Hpn-FRET were investigated in detail by titrating Hpn-FRET with different biologically relevant metal ions. Titration of Hpn-FRET with Ni²⁺, the proposed metal that binds Hpn, showed 5 equiv. binding with a quite noticeable change in FRET (∼ 40%). Beside Ni²⁺, the addition of Zn²⁺ or Co²⁺ also resulted in a change in FRET. The FRET change with Zn²⁺ is larger than that observed with Ni²⁺ (∼55%), and the FRET change with Co²⁺ is smaller than that with Ni²⁺ (∼22%) (Fig. 2, Fig. S3, ESI†). Additionally when the CD spectra of Hpn-FRET in the presence of these metals were taken an increase in α-helix and a decrease in β-sheet content were observed where Zn²⁺ resulted in the biggest change and Co²⁺ in the least showing the same trend as the FRET response (Fig. S2, Table S1, ESI†). The titration also indicated that Hpn-FRET binds 5 equiv. of Zn²⁺, but only 4 equiv. of Co²⁺. A small change in FRET was observed upon addition of Cu²⁺ or Bi³⁺. Almost no FRET change was observed with Mg²⁺, Ca²⁺, Cr³⁺, Mn²⁺, Fe²⁺ or Fe³⁺ (Fig. 2).

Fig. 2 Hpn-FRET titration with different metal ions. Hpn-FRET (250 nM) was titrated with Zn²⁺, Ni²⁺, Co²⁺, Cu²⁺, Mn²⁺, Ca²⁺, Mg²⁺, Fe²⁺, Fe³⁺, Cr³⁺ and Bi³⁺ up to 8 equiv.

The selectivity of Hpn-FRET for Ni²⁺ was further evaluated in competition experiments in which the response of Hpn-FRET to Ni²⁺ was measured in the presence of an equal amount of the competing metal ions Mg²⁺, Ca²⁺, Cr³⁺, Mn²⁺, Fe²⁺, Co²⁺, Cu²⁺, Zn²⁺, Fe³⁺ and Bi³⁺. In these experiments, first the response of Hpn-FRET was measured in the presence of 6 equiv. metal ions. Then, an equal amount of Ni²⁺ was added to the sample, and the response was measured again (Fig. 3). In all cases except for Zn²⁺, the full response to Ni²⁺ was observed, indicating that Ni²⁺ binds tighter than other metal ions and the presence of other metal ions does not disturb the Ni²⁺ binding. When the same experiments were done with Zn²⁺ or Co²⁺, again none of the other metal ions could interfere with the binding. Zn²⁺ was able to displace Ni²⁺ and Co²⁺, while Co²⁺ displaced neither Zn²⁺ nor Ni²⁺ (Fig. S4a,b, ESI†). Furthermore, the binding stoichiometry of 5 equiv. Ni²⁺, 5 equiv. Zn²⁺ and 4 equiv. Co²⁺ to Hpn-FRET was confirmed by submitting Hpn-FRET samples, incubated with 10 equiv. of the respective metal and separated from the unbound metal, for ICP-MS analysis (Table S2, ESI†). In agreement with the selectivity observed in the fluorescence measurements, when Hpn-FRET was incubated with two metals at the same time Hpn-FRET preferentially bound to Zn²⁺ over Ni²⁺ and Co²⁺, and to Ni²⁺ over Co²⁺ (Table S2, ESI†). From these competition experiments, a relative binding order of Zn²⁺ > Ni²⁺ > Co²⁺ ≫ other metals, can be obtained.

Fig. 3 Competition of Ni²⁺ with other metal ions for binding to Hpn-FRET. Hpn-FRET (250 nM) with 6 equiv. of the respective metal ions is shown in black, and the addition of 6 equiv. Ni²⁺ to the sample is shown in white. All measurements were done in triplicates.

K _d measurements of Hpn-FRET for Zn²⁺ and Ni²⁺

To quantify the absolute binding affinity of Hpn-FRET to Ni²⁺ and Zn²⁺, binding constants were determined. Free Ni²⁺ concentrations were buffered from 1.37 × 10⁻⁹ to 6.32 × 10⁻⁶ M using various amounts of Ni²⁺ and citrate.^32,33 The percentage of metal loading to Hpn-FRET was measured in these buffers and the binding curve was fitted using the Hill equation (Fig. 4a, Table S3, ESI†). A dissociation constant of Hpn-FRET to Ni²⁺ was computed to be 7.89 × 10⁻⁸ ± 8.9 × 10⁻⁹ M with no cooperativity observed (n = 1.21 ± 0.16). The dissociation constant for Zn²⁺ was measured similarly by buffering free Zn²⁺ from 3.94 × 10⁻¹¹ to 6.64 × 10⁻⁷ M using Zn²⁺ and cyanide.^32,33 From fitting the binding curve, a dissociation constant of 1.03 × 10⁻⁹ ± 9.3 × 10⁻¹¹ M was obtained with no cooperativity observed (n = 1.42 ± 0.22) (Fig. 4b, Table S3, ESI†). These experiments indicate that Hpn-FRET has a high affinity for Ni²⁺ and an even higher (75-fold as compared to Ni²⁺) affinity for Zn²⁺.

Fig. 4 Binding curve of Hpn-FRET to (a) Ni²⁺ and (b) Zn²⁺. The fitting curve is shown in red and the blue points represent the minimum (apo-Hpn-FRET) and maximum (Hpn-FRET + 6 equiv. Ni²⁺ or + 6 equiv. Zn²⁺) signals, respectively. All measurements were done in triplicates.

K _d measurements of Hpn-FRET for Bi³⁺

We were puzzled by the low FRET change observed for Bi³⁺ shown in Fig. 2, which seems to be inconsistent with the proposal that Hpn in H. pylori is a target of Bi³⁺. Therefore, we set out to measure the binding constant for Bi³⁺ by adding increasing concentrations of Bi³⁺ in 10 μM increments to Hpn-FRET (250 nM). To our surprise, we observed a significant change of the FRET ratio from 2.0 to 4.0 at higher concentrations of Bi³⁺, and this increase in FRET was saturated at ∼90 μM Bi³⁺ (Fig. 5). The binding curve was fitted to the Hill equation and the dissociation constant was found to be 6.19 × 10⁻⁵ ± 1.0 × 10⁻⁶ M with cooperativity (n = 6.23 ± 0.71) (Table S3, ESI†). The Hill equation will only give an accurate number of binding sites when very strong cooperativity is present between the first and subsequent binding sites. It gives a lower limit to the number of binding sites in the case of moderate cooperativity.³⁴ For the binding curves of Ni²⁺ and Zn²⁺, where no cooperativity is observed (n ∼ 1), the binding equivalence was determined separately and found to be 5 equiv. of the respective metal. In the case of Bi³⁺, n = 6.23 was obtained from the fitting of the Hill equation, and therefore, there are at least six metal binding sites with some cooperativity.

Fig. 5 Binding curve of Hpn-FRET to Bi³⁺. The fitting curve is shown in red. All measurements were done in triplicate.

Response of Hpn-FRET to Bi³⁺, Zn²⁺ and Ni²⁺ in E. coli

To evaluate the ability of Hpn-FRET to function as a probein vivo, Hpn-FRET was expressed in E. coli and FRET changes were measured when cells were incubated with Ni²⁺, Zn²⁺ or Bi³⁺.³⁵Bismuth subsalicylate was used as a bismuth source because it is the active ingredient in Pepto Bismol, a common treatment for H. pylori.Pepto Bismol was used as well for comparison. E. colicells in MOPS minimal medium expressing Hpn-FRET were incubated with 1 to 400 μM metal complexes after protein expression was induced with 0.1 mM IPTG and cells grown overnight. To ensure that the metals did not inhibit growth, the OD₆₀₀ was measured and no effect was observed. The fluorescence spectra of these cells were taken and the fluorescent energy transfer was quantified by taking the intensity ratio of the YFP emission (527 nm) over the CFP emission (477 nm). We did not observe a change of the FRET ratio when cells were incubated with Zn²⁺ or Ni²⁺, as compared to the control cells with no metal added (Fig. 6). On the other hand, addition of bismuth subsalicylate or Pepto Bismol resulted in a significant increase of the FRET ratio starting at 10–20 μM and saturating at ∼100 μM Bi³⁺ (Fig. 6, Fig. S5, ESI†). These results demonstrate that Hpn-FRET can efficiently bind to Bi³⁺ in the bacterial cellular environment, and that free Ni²⁺ and Zn²⁺ concentrations inside E. coli are not high enough for Hpn-FRET to respond to them.

Fig. 6 Response of Hpn-FRET inside E. coli to Ni²⁺ (orange), Zn²⁺ (green), bismuth subsalicylate (black) and Pepto Bismol (pink) (the total concentration of bismuth is shown in the x-axis). Cells were excited at 420 nm and the ratio of the intensities at 527 nm over 477 nm was plotted as response. Error bars represent the standard deviation from four independent cell cultures.

Discussion

Hpn-FRET was designed by inserting Hpn between the FRET partners CFP and YFP, taking advantage of the previously reported change in Hpn conformation upon Ni²⁺ binding.²⁹ The resulting Hpn-FRET serves as an excellent probe to study metal binding to Hpn while avoiding potential problems of protein aggregation and denaturation. Hpn-FRET shows a moderate selectivity for Ni²⁺; it also responds to Zn²⁺ and Co²⁺ but not other biometals, such as Mg²⁺, Ca²⁺, Cr³⁺, Mn²⁺, Fe²⁺, Fe³⁺ and Cu²⁺. Competition experiments give a relative binding order of Zn²⁺ > Ni²⁺ > Co²⁺ for Hpn. The measured binding constant of Hpn-FRET to Ni²⁺ is 7.89 × 10⁻⁸ M while the binding constant of Hpn-FRET to Zn²⁺ is 1.03 × 10⁻⁹ M, which is 75-fold tighter than that to Ni²⁺.

The metal-binding properties of Hpn-FRET show that it binds 5 equiv. Ni²⁺ with a K_d = 7.89 × 10⁻⁸ M, 5 equiv. Zn²⁺ with a K_d = 1.03 × 10⁻⁹ M, 4 equiv. of Co²⁺ with an affinity lower than that of Ni²⁺ or Zn²⁺, and no appreciable binding to the other bioavailable metals Mg²⁺, Ca²⁺, Cr³⁺, Mn²⁺, Fe²⁺, Fe³⁺ and Cu²⁺. In addition, binding to the therapeutic metal Bi³⁺ is observed at a higher concentration with K_d = 6.19 × 10⁻⁵ M. Thus, the affinity order of Hpn-FRET to biometals is Zn²⁺ > Ni²⁺ > Co²⁺ > other metals tested and Bi³⁺. The metal-binding behavior of Hpn-FRET is different from previous reports that used the isolated Hpn peptide. While the binding stoichiometry of Hpn-FRET for Ni²⁺ and Zn²⁺ are in agreement with the Hpn peptide, Hpn-FRET binds much tighter to these two metal ions than previously reported using equilibrium dialysis and analyzing metal content by ICP-MS.^29,31 Furthermore, binding of ∼8 equiv. of Cu²⁺ and ∼4 equiv. Bi³⁺ were not observed for Hpn-FRET. One potential explanation for this difference in the metal-binding behavior could be the difference in oligomerization of the Hpn peptide and Hpn-FRET. The Hpn peptide can form different large multimers depending on the buffer conditions.²⁹ Hpn-FRET is less prone to aggregation as it elutes as one peak on a size-exclusion column in the presence and absence of Ni²⁺ (Fig. S6, ESI†). In a different study reporting the relative binding affinity of Hpn peptide to various metals using western blots in combination with radioisotopes, an affinity order of Ni²⁺ and Zn²⁺ ≫ Co²⁺ > Cu²⁺ > Mn²⁺ was reported,²⁸ which is in agreement with what we observed with Hpn-FRET. Because the FRET technique is sensitive, and therefore less concentrated samples were used, as well as the presence of CFP and YFP at the two termini, multimerization of Hpn can be avoided.

Hpn has been suggested as a target for the therapeutic metal Bi³⁺ based on the higher susceptibility of the hpn knock-out strain of H. pylori to bismuth treatment.¹² The much higher binding affinities of Hpn to Ni²⁺ and Zn²⁺ over Bi³⁺ pose the question of how Bi³⁺ can target Hpn in H. pylori. Therefore, the FRET response of Hpn-FRET inside a model bacterium, E. coli, towards Ni²⁺, Zn²⁺, Bi³⁺ and Pepto Bismol was evaluated. In these experiments, up to 400 μM metals were added to bacterial cells and no toxicity was observed. While the addition of Ni²⁺ or Zn²⁺ resulted in no increase in FRET compared to control cells, an increase of FRET was observed following incubation with either Pepto Bismol or bismuth subsalicylate, beginning with 10–20 μM Bi³⁺ and saturating at ∼100 μM Bi³⁺. This result shows that Hpn-FRET binds to Bi³⁺ efficiently inside E. coli, and that Hpn-FRET does not bind to Ni²⁺ or Zn²⁺ tightly enough to compete for these metals inside E. coli. While metal homeostasis in H. pylori and E. coli is clearly different it is intriguing to observe such a high accumulation of Bi³⁺ inside E. coli considering the therapeutic activity of Bi³⁺ against H. pylori. However, one should be cautious to extrapolate results obtained from experiments performed in E. coli to H. pylori. The nickel content in H. pylori could be significantly higher than that in E. coli, which could make Hpn predominantly occupied with nickel in H. pylori. To test the hypothesis experiments with Hpn-FRET in H. pylori would be very interesting for future studies.

It has been well established that most bacterial cells tightly control metal availability.¹⁶ In E. coli, nickel availability is limited to 10⁻¹² M, which has been determined by the study of the Ni²⁺ -dependent transcriptional regulator NikR.³⁶ The free zinc level in E. coli is likewise regulated to ∼10⁻¹⁵ M as was determined by the study of the zinc regulatory proteins Zur and ZntR.³⁷ Since the K_d values of Hpn-FRET to Ni²⁺ and Zn²⁺ are 7.89 × 10⁻⁸ M and 1.03 × 10⁻⁹ M, respectively, they are unable to compete for these metals inside E. coli, which agrees with the absence of any FRET change in our experiments with the use of Hpn-FRET.

In H. pylori the Ni²⁺ level is controlled by HpNikR,³⁸ which has been reported to bind Ni²⁺ at ∼1.2 × 10⁻⁸ M (tighter binding was also suggested depending on techniques used to determine the K_d value).^39,40 This K_d value of HpNikR agrees very well with our measured Hpn-FRET K_d to Ni²⁺ of 7.89 × 10⁻⁸ M, supporting the hypothesis that Hpn acts as a storage protein to scavenge excess Ni²⁺ inside H. pylori. The previously reported Hpn peptideK_d of 7.1 × 10⁻⁶ M to Ni²⁺ is too high to be consistent with Hpn acting as a be a nickel storage and scavenger protein considering the response threshold of HpNikR. On the other hand, Bi³⁺ is not regulated in bacterial cells and no specific detoxification mechanism for Bi³⁺ is known. Bismuth thus can accumulate to a high cellular concentration and occupy available metal-binding sites, interfering with biological functions. In fact, the suggested single dose for Pepto Bismol is high enough to result in ∼700 μM Bi³⁺ in the stomach, which is higher than the concentration needed to saturate Hpn-FRET inside E. coli. Our measurements in E. coli show that free, available Bi³⁺ can accumulate to high micromolar levels. If Bi³⁺ accumulates to the similarly high levels inside H. pylori it may target Hpn or related proteins, eventually leading to bactericidal effects. The exact consequence of bismuth binding to Hpn still remains to be evaluated, but it has been found that the hpn deficient H. pylori cannot adapt to decreases in pH of the growth medium and has lower urease activity when Ni²⁺ in the growth media was chelated. Thus, if binding of Bi³⁺ prevents Hpn function, the bacteria would be more susceptible to changing environmental conditions.

Conclusions

In closing, the histidine-rich protein Hpn from H. pylori, which functions as a Ni²⁺ storage protein and is a target for Bi³⁺ therapy, is converted into a FRET-based metal probe, Hpn-FRET. Taking advantage of the metal-binding-induced FRET change of this construct, we determined the metal-binding properties of this protein to be Zn²⁺ (K_d = 1.03 × 10⁻⁹ M) > Ni²⁺ (K_d = 7.89 × 10⁻⁸ M) > Co²⁺ > other biometals. In addition, the dissociation constant for the therapeutic metal Bi³⁺ to Hpn-FRET was measured to be 6.19 × 10⁻⁵ M. Experiments in E. coli showed that Ni²⁺ and Zn²⁺ can not bind to Hpn-FRET due to their tight regulation inside the cell; however, high micromolar levels of Bi³⁺ accumulate and result in the FRET change of Hpn-FRET inside E. coli. The FRET-based probe is an excellent tool to study sensitive binding of metal ions to metalloproteins, and can be used to probe metal availability in living cells.

Acknowledgements

This work was supported by supported by a CAREER award from the National Science Foundation (NSF0544546 to C. H.) and Office of Basic Energy Sciences, U. S. Department of Energy under Contract No. DE-FG02-07ER15865 (to C. H.)

Notes and references

1. M. J. Blaser and J. C. Atherton, J. Clin. Invest., 2004, 113, 321.
2. T. L. Cover and M. J. Blaser, Adv. Int. Med., 1996, 41, 85.
3. K. Stingl and H. De Reuse, Int. J. Med. Microbiol., 2005, 295, 307.
4. R. J. Maier, S. L. Benoit and S. Seshadri, Biometals, 2007, 20, 655.
5. S. L. Benoit and R. J. Maier, Ann. N. Y. Acad. Sci., 2008, 1125, 242.
6. (a) F. D. Ernst, E. J. Kuipers, A. Heijens, R. Sarwari, J. Stoof, C. W. Penn, J. G. Kusters and A. H. van Vliet, Infect. Immun., 2005, 73, 7252; (b) N. S. Dosanjh and S. L. Michel, Curr. Opin. Chem. Biol., 2006, 10, 123.
7. (a) U. Dittes, E. Vogel and B. K. Keppler, Coord. Chem. Rev., 1997, 163, 345; (b) P. J. Sadler, H. Li and H. Sun, Coord. Chem. Rev., 1999, 185–186, 689; (c) N. Yang and H. Sun, Coord. Chem. Rev., 2007, 251, 2354; (d) H. Sun, L. Zhang and K. Y. Szeto, Met. Ions Biol. Syst., 2004, 41, 333.
8. R. Ge and H. Sun, Acc. Chem. Res., 2007, 40, 267.
9. L. Zhang, S. B. Mulrooney, A. F. Leung, Y. Zeng, B. B. Ko, R. P. Hausinger and H. Sun, BioMetals, 2006, 19, 503.
10. L. Jin, K. Y. Szeto, L. Zhang, W. Du and H. Sun, J. Inorg. Biochem., 2004, 98, 1331.
11. (a) H. Sun, H. Li, I. Harvey and P. J. Sadler, J. Biol. Chem., 1999, 274, 29094; (b) J. Pannequin, S. Kovac, J. P. Tantiongco, R. S. Norton, A. Shulkes, K. J. Barnham and G. S. Baldwin, J. Biol. Chem., 2003, 279, 2453.
12. H. L. Mobley, R. M. Garner, G. R. Chippendale, J. V. Gilbert, A. V. Kane and A. G. Plaut, Helicobacter, 1999, 4, 162.
13. H. Li, P. J. Sadler and H. Sun, J. Biol. Chem., 1996, 271, 9483.
14. L. Zhang, K. Y. Szeto, W. B. Wong, T. T. Loh, P. J. Sadler and H. Sun, Biochemistry, 2001, 40, 13281.
15. H. Sun, H. Li, A. B. Mason, R. C. Woodworth and P. J. Sadler, J. Biol. Chem., 2001, 276, 8829.
16. K. J. Waldron and N. J. Robinson, Nat. Rev. Microbiol., 2009, 7, 25.
17. K. J. Waldron, J. C. Rutherford, D. Ford and N. J. Robinson, Nature, 2009, 460, 823.
18. (a) D. W. Domaille, E. L. Que and C. J. Chang, Nat. Chem. Biol., 2008, 4, 168; (b) E. L. Que, D. W. Domaille and C. J. Chang, Chem. Rev., 2008, 108, 1517.
19. A. Miyawaki, J. Llopis, R. Heim, J. M. McCaffery, J. A. Adams, M. Ikura and R. Y. Tsien, Nature, 1997, 388, 882.
20. M. Mank, A. F. Santos, S. Direnberger, T. D. Mrsic-Flogel, S. B. Hofer, V. Stein, T. Hendel, D. F. Reiff, C. Levelt, A. Borst, T. Bonhoeffer, M. Hubener and O. Griesbeck, Nat. Methods, 2008, 5, 805.
21. (a) J. L. Vinkenborg, T. J. Nicolson, E. A. Bellomo, M. S. Koay, G. A. Rutter and M. Merkx, Nat. Methods, 2009, 6, 737; (b) W. Qiao, M. Mooney, A. J. Bird, D. R. Winge and D. J. Eide, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 8674; (c) P. J. Dittmer, J. G. Miranda, J. A. Gorski and A. E. Palmer, J. Biol. Chem., 2009, 284, 16289.
22. (a) C. J. Chang, E. M. Nolan, J. Jaworski, S. C. Burdette, M. Sheng and S. J. Lippard, Chem. Biol., 2004, 11, 203; (b) E. M. Nolan, J. W. Ryu, J. Jaworski, R. P. Feazell, M. Sheng and S. J. Lippard, J. Am. Chem. Soc., 2006, 128, 15517.
23. S. V. Wegner, H. Arslan, M. Sunbul, J. Yin and C. He, J. Am. Chem. Soc., 2010, 132, 2567.
24. (a) L. Zeng, E. W. Miller, A. Pralle, E. Y. Isacoff and C. J. Chang, J. Am. Chem. Soc., 2006, 128, 10; (b) L. Yang, R. McRae, M. M. Henary, R. Patel, B. Lai, S. Vogt and C. J. Fahrni, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 11179.
25. (a) S. C. Dodani, Q. He and C. J. Chang, J. Am. Chem. Soc., 2009, 131, 18020; (b) T. Mizuno, K. Murao, Y. Tanabe, M. Oda and T. Tanaka, J. Am. Chem. Soc., 2007, 129, 11378.
26. J. Zhang, Y. Ma, S. S. Taylor and R. Y. Tsien, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 14997.
27. S. Seshadri, S. L. Benoit and R. J. Maier, J. Bacteriol., 2007, 189, 4120.
28. J. V. Gilbert, J. Ramakrishna, F. W. Sunderman, Jr, A. Wright and A. G. Plaut, Infect. Immun., 1995, 63, 2682.
29. R. Ge, R. M. Watt, X. Sun, J. A. Tanner, Q. Y. He, J. D. Huang and H. Sun, Biochem. J., 2006, 393, 285.
30. Y. B. Zeng, D. M. Zhang, H. Li and H. Sun, JBIC, J. Biol. Inorg. Chem., 2008, 13, 1121.
31. R. Ge, Y. Zhang, X. Sun, R. M. Watt, Q. Y. He, J. D. Huang, D. E. Wilcox and H. Sun, J. Am. Chem. Soc., 2006, 128, 11330.
32. L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini and A. Vacca, Coord. Chem. Rev., 1999, 184, 311.
33. A. E. Martell and R. M. Smith, Standard Reference Database, 2001, vol. 46.
34. J. N. Weiss, FASEB J., 1997, 11, 835.
35. F. C. Neidhardt, P. L. Bloch and D. F. Smith, J. Bacteriol., 1974, 119, 736.
36. (a) S. L. Bloom and D. B. Zamble, Biochemistry, 2004, 43, 10029; (b) P. T. Chivers and R. T. Sauer, Chem. Biol., 2002, 9, 1141; (c) S. Leitch, M. J. Bradley, J. L. Rowe, P. T. Chivers and M. J. Maroney, J. Am. Chem. Soc., 2007, 129, 5085.
37. C. E. Outten and T. V. O'Halloran, Science, 2001, 292, 2488.
38. C. Dian, K. Schauer, U. Kapp, S. M. McSweeney, A. Labigne and L. Terradot, J. Mol. Biol., 2006, 361, 715.
39. B. Zambelli, M. Bellucci, A. Danielli, V. Scarlato and S. Ciurli, Chem. Commun., 2007, 43, 3649.
40. (a) L. O. Abraham, Y. Li and D. B. Zamble, J. Inorg. Biochem., 2006, 100, 1005; (b) N. S. Dosanjh, N. A. Hammerbacher and S. L. Michel, Biochemistry, 2007, 46, 2520.

---

Footnote

† Electronic supplementary information (ESI) available: Materials and methods. Table S1 shows the CD fitting results, Table S2 gives ICP-MS quantification of metal ion loading to Hpn-FRET, and Table S3 shows the fitting results for binding curves of Hpn-FRET to Ni²⁺, Zn²⁺, and Bi³⁺, Fig. S1 shows protein gel of purified Hpn-FRET, Fig. S2 is the CD spectra of Hpn-FRET in the absence and presence of metals, Fig. S3 is the fluorescent spectra for the response of Zn²⁺, Co²⁺, and Ni²⁺ with Hpn-FRET, Fig. S4 displays the competition experiments for Zn²⁺ and Co²⁺ selectivity, Fig. S5 shows the fluorescence spectra of E. coli in the absence of additional metal, Fig. S6 shows the gel filtration chromatogram for purified Hpn-FRET, and Fig. S7 compares FRET response between Hpn-FRET and Hpn-FRET without a N-terminal His6-tag. See DOI: 10.1039/c0sc00411a

---