I. INTRODUCTION

Developing nuclear energy is of social and economic importance, while this arises a major concern over environmental pollution by long-lived radioactive wastes, such as uranyl ion (UO₂²⁺), the most stable form of uranium under physiological conditions [1]. In addition to its radiation hazard, UO₂²⁺ is of high toxicity because it interacts with both DNA [2] and proteins [3-6] and disrupts their biological functions. To date, plentiful proteins have been found to be the targets of UO₂²⁺, such as transferrin, ferritin and albumin [3-6]. A protein data bank (PDB) survey shows that UO₂²⁺ binds to proteins mainly through carboxylic acid groups such as those of aspartate (Asp) and glutamate (Glu), as well as other coordinating amino acids such as histidine (His) and tyrosine (Tyr) [7]. In previous papers [8, 9], we studied the structural and functional consequences of UO₂²⁺ binding to cytochrome b₅ (cyt b₅), cytochrome c (cyt c), and the protein-protein complex of cyt b₅-cyt c, by both experimental and theoretical approaches.

Cyt b₅ is a small membrane heme protein, which is characterized by a highly negatively charged surface in presence of a series of acidic residues surrounding the heme group, known as an acidic cluster (Fig. 1) [10]. It has been shown that the acidic cluster participates in formation of cyt b₅-cyt c complex [11-17], and is crucial in mediating cyt c signaling in apoptosis [18]. With a positive charge, UO₂²⁺ has a strong tendency to be absorbed at negatively charged surface of membrane [19], where it has a large possibility of interacting with membrane proteins such as cyt b₅. In previous study [8], we proposed a UO₂²⁺ binding site for cyt b₅ at surface residues of Glu37 and Glu43 (Fig. 1), based on molecular modeling and dynamics simulation.

Fig. 1.Full-screen View

Modeling structure of UO₂²⁺-cyt b₅ complex, showing the UO₂²⁺ ion binding to Glu37 and Glu43, the residues in acidic cluster, the heme group and Trp22 in the hydrophobic pocket.

To further probe the role of acidic cluster in uranyl-cyt b₅ interaction and to reveal the effect of surface mutations of cyt b₅ on UO₂²⁺ binding, we herein choose a quadruple mutant of cyt b₅ as a target, in which four acidic residues were replaced with alanine, E44/48/56A/D60A cyt b₅, namely 4A cyt b₅. This mutant was designed previously for studying the interface between cyt b₅ and cyt c, where the replaced residues were considered to be interactive sites [16]. It keeps the uranyl-binding site of Glu37 and Glu43 but has a low binding affinity for cyt c, and therefore is an ideal model protein for investigating the regulation effect of cyt c by a comparison with that of cyt b₅-cyt c complex.

II. MATERIALS AND METHODS

4A cyt b₅ was expressed and purified as described previously [16]. Horse heart cyt c (Type VI) was purchased from Sigma Chemical Co. Uranyl nitrate and other chemicals were commercial products and of analytical grade. Double distilled water was used throughout the experiments.

Fluorescence spectra of 4A cyt b₅ and 4A cyt b₅-cyt c complex (4 μm in 50 mM Tris.HCl buffer, pH 7.0) with titration of UO₂²⁺ up to 10 equivalents, were collected at 25 ℃ on a LS45 fluorescence spectrometer (Perkin Elmer, USA). The excitation wavelength was 295 nm and emission spectra were recorded from 310 nm to 410 nm. The apparent dissociation constant (K_D) was calculated from double reciprocal plot by fitting to Eq. (1) [20],

where ΔF is the difference between the maximum fluorescence in the absence and presence of UO₂²⁺ and ΔF_inf is the fluorescence change for the complete binding of UO₂²⁺. The number of UO₂²⁺ binding site was calculated from ΔF of fluorescence spectra using the Hill plot [21], i.e. the yielded slope.

Circular dichroism (CD) spectra of 4A cyt b₅ and 4A cyt b₅-cyt c (20 μm in 50 mM Tris.HCl buffer, pH 7.0), in the absence or presence of 5 equivalents of UO₂²⁺, were collected at 25 ℃ from 300 nm to 600 nm (1.0 cm path length), with a Jasco J720 spectrometer (Japan). Dilution effect of addition of UO₂²⁺ was corrected by adding the same volume of buffer solution into the protein solution.

III. RESULTS AND DISCUSSION

The single tryptophan residue (Trp22) in cyt b₅ (Fig. 1) serves as a convenient and sensitive reporter for studying the interaction of UO₂²⁺-cyt b₅ by fluorescence spectroscopy. Figure 2(a) shows that the fluorescence intensity of 4A cyt b₅ decreases gradually upon titration of UO₂²⁺ ions, suggesting a fluorescence quenching as observed for uranyl titration of transferrin, ferritin and albumin [3-6]. This observation is similar to that of wild-type (WT) cyt b₅ [9]. When the changes of the maximum fluorescence were fitted to a double reciprocal plot (Fig. 2(a), right up) and a Hill plot (Fig. 2(a), right down), its binding affinity (K_D = 40 μm) is 4 times lower than that of cyt b₅ (Table 1), though 4A cyt b₅ remains a single binding site of UO₂²⁺ (slope = 1.1). It is likely due to a conformational change of residues Glu37 and Glu43 as a result of the four mutations in the acidic cluster, especially for Glu44 close to the proposed uranyl-binding site, as observed in the crystal structure of 4A cyt b₅ (PDB entry 1M2M) [16]. These observations indicate that the negative charges on the acidic cluster of cyt b₅ play important roles in uranyl-protein interaction.

Fig. 2.Full-screen View

Fluorescence titration of 4A cyt b₅ (4 μm) (a) and 4A cyt b₅-cyt c complex (4 μm) (b), with UO₂²⁺ in 50 mM Tris.HCl buffer (pH 7.0) at 25 ℃. Double reciprocal plot and Hill plot are shown in right top and right down, respectively.

On the other hand, in titration of 4A cyt b₅-cyt c complex with UO₂²⁺ (Fig. 2(b)), the single uranyl-binding site on 4A cyt b₅ surface in the complex has a higher affinity of UO₂²⁺ (K_D = 15 μm) than that on the surface of isolated 4A cyt b₅ (K_D = 40 μm). This observation suggests that the conformation of surface residues, Glu37 and Glu43, are tuned to be suitable for uranyl-binding as a result of cyt c interacting with 4A cyt b₅, though the binding constant decreases from cyt b₅-cyt c (2.2×10⁴ M^-1) to 4 A cyt b₅-cyt c (5.5×10³ M^-1), as determined by NMR technique in a previous study [15].

In Table 1, the uranyl-binding factor (K_D =15 μm) of 4A cyt b₅-cyt c complex is two times lower than that of cyt b₅-cyt c complex (30 μm), suggesting that a different protein-protein interface was adopted for 4 A cyt b₅-cyt c complex as a result of mutation of four acidic residues on the same heme-exposed edge side of cyt b₅ (Fig. 1). In an earlier study, Huang and co-workers [17] showed that the charge neutralization of 4 A cyt b₅ greatly increased the relative contribution of the heme propionate and other charged residues surrounding the heme edge to contact with cyt c. Thus, our observations indicate that UO₂²⁺-cyt b₅ interaction is regulated by both surface mutations of cyt b₅ and its partner cyt c.

The structural perturbations of 4A cyt b₅ and its complex with cyt c upon uranyl binding were studied by circular dichroism (CD) spectroscopy (Fig. 3). It shows that UO₂²⁺-4A cyt b₅ complex has a slightly decreased negative Cotton effect at 419 nm, whereas the Cotton effect in 300–350 nm is more positive than that of 4A cyt b₅ (Fig. 3(a)), which indicates that, similar to the observation for UO₂²⁺-cyt b₅ [9], UO₂²⁺ binding to the surface of 4A cyt b₅ slightly alters the conformation of heme-binding domain and aromatic amino acids such as Phe35 and Trp22 in the protein hydrophobic core. The CD spectra also show that upon uranyl binding, the 4A cyt b₅-cyt c complex has an altered positive Cotton effect and a decreased negative Cotton effect compared to that in the absence of UO₂²⁺ ions, with decreased intensity in 300–350 nm (Fig. 3(b)). These spectral changes are similar to the case of UO₂²⁺ binding to cyt b₅-cyt c complex, whereas both cases are less obvious than that for UO₂²⁺ binding to cyt c [9]. These observations suggest that UO₂²⁺ ions alter the heme active site in both 4A cyt b₅-cyt c and cyt b₅-cyt c complexes, which are regulated by cyt c, likely through forming dynamic protein-protein complexes and competing with UO₂²⁺ ions.

Fig. 3.Full-screen View

CD spectra of 4A cyt b₅ (20 μm) (a) and 4A cyt b₅-cyt c complex (20 μm) (b) in the absence (black lines) and presence (grey lines) of UO₂²⁺ (100 μm) in 50 mM Tris.HCl buffer (pH 7.0) at 25 ℃.

IV. CONCLUSION

In summary, we investigated the interactions of UO₂²⁺-4A cyt b₅ and UO₂²⁺-4A cyt b₅-cyt c complex, and compared to that of UO₂²⁺-cyt b₅, UO₂²⁺-cyt c and UO₂²⁺-cyt b₅-cyt c from our previous studies [8, 9]. With four acidic residues on protein surface replaced by alanine, 4A cyt b₅ exhibits a lower affinity for UO₂²⁺ compared to that for WT cyt b₅, though it remains the proposed uranyl-binding site at Glu37 and Glu43, which suggests a key role of the acidic cluster of cyt b₅ in tuning uranyl-protein interaction. On the other hand, UO₂²⁺ binds more tightly to 4A cyt b₅-cyt c complex than to the cyt b₅-cyt c complex, presumably due to the conformational changes of Glu37 and Glu43 by a protein-protein interface different from that of cyt b₅-cyt c complex. These observations indicate that UO₂²⁺-cyt b₅ interaction is regulated by both the surface acidic cluster of cyt b₅ and its interacting protein partner cyt c, which provides valuable information on metal-protein-protein interactions, and clues for understanding the mechanism of uranyl toxicity.