Substitution of the Sole Tryptophan of the Cupredoxin, Amicyanin, with 5-Hydroxytryptophan Alters Fluorescence Properties and Energy Transfer to the Type 1 Copper Site

Abstract

Amicyanin is a type 1 copper protein with a single tryptophan residue. Using genetic code expansion, the tryptophan was selectively replaced with the unnatural amino acid, 5-hydroxytryptophan (5-HTP). The 5-HTP substituted amicyanin exhibited absorbance at 300-320 nm, characteristic of 5-HTP and not seen in native amicyanin. The fluorescence emission maximum in 5-HTP substituted amicyanin is redshifted from 318 nm in native amicyanin to 331 nm and to 348 nm in the unfolded protein. The fluorescence quantum yield of 5-HTP substituted amicyanin mutant was much less than that of native amicyanin. Differences in intrinsic fluorescence are explained by differences in the excited states of tryptophan versus 5-HTP and the intraprotein environment. The substitution of tryptophan with 5-HTP did not affect the visible absorbance and redox potential of the copper, which is 10 Å away. In amicyanin and other cupredoxins, an unexplained quenching of the intrinsic fluorescence by the bound copper is observed. However, the fluorescence of 5-HTP substituted amicyanin is not quenched by the copper. It is shown that the mechanism of quenching in native amicyanin is Förster, or fluorescence, resonance energy transfer (FRET). This does not occur in 5-HTP substituted amicyanin because the fluorescence quantum yield is significantly lower and the red-shift of fluorescence emission maximum decreases overlap with the near UV absorbance of copper. Characterization of the distinct fluorescence properties of 5-HTP relative to tryptophan in amicyanin provides a basis for spectroscopic interrogation of the protein microenvironment using 5-HTP, and long-distance interactions with transition metals.

1. INTRODUCTION

Amicyanin is a type 1 copper protein than mediates electron transfer from the tryptophan tryptophylquinone cofactor of methylamine dehydrogenase to c-type cytochromes in several methylotrophic and autotrophic bacteria [1]. The amicyanin used in this study is the well-characterized protein from Paracoccus denitrificans [2–4]. Amicyanin was chosen as this 12 kDa protein possesses a single tryptophan (Trp) residue, Trp45, that resides in the interior of the protein as shown in Figure 1 [5, 6]. Trp45 and the copper are separated by approximately 10 Å [4, 5]. A curious feature of amicyanin is that the fluorescence of Trp45 is quenched in the holoprotein with copper bound, relative to its fluorescence emission in the apoprotein, from which the copper had been removed [3]. The goal of this study is to modify Trp45 to examine the mechanism of this phenomenon and probe the influence of the protein microenvironment of this sole Trp residue on its spectroscopic and fluorescence properties. This was accomplished by substituting 5-hydroxytryptophan (5-HTP) for this Trp.

Figure 1.

Structure of amicyanin from Paracoccus denitrificans. A. The overall structure is shown. The copper and four amino acids that provide ligands and the single internal Trp residue are parallel with the z-axis. B. A focused representation of Trp45 and surrounding amino acids within the region highlighted in orange in A. C. The area shown in B rotated by 90° along the z-axis. Nitrogen is blue, oxygen is red, sulfur is yellow, and copper is brown. Carbon is green in Trp45 and cyan in the other residues. Figures were constructed using the coordinates in PDB entry 1AAN [5].

Genetic code expansion technology allows selective incorporation of unnatural amino acid analogs in proteins through codon suppression [7]. This technology has potential applications for studying post-translational modification of proteins and producing novel conjugated proteins for basic research and applications in biotechnology [8, 9]. Genetic code expansion can be accomplished through the suppression of a nonsense or a quadruplet codon [9]. Briefly, nonsense codon suppression uses an evolved aminoacyl-tRNA synthetase (aaRS) and tRNA pair to direct the incorporation of an unnatural amino acid, typically in response to stop codon (e.g., TAG), which has been inserted into the gene sequence in place of the codon for the amino acid residue to be replaced with the unnatural amino acid. This yields an additional orthogonal translation mechanism occurring in tandem with the natural translation mechanism of the host cell [7, 10]. Three categories of aaRS/tRNA pairs have been largely successful for the incorporation of analogs for tyrosine (Tyr), pyrrolysine, and Trp [10–16]. Incorporation of analogs of Tyr [13, 15, 17–19] and pyrrolysine [19, 20] using this approach are well documented, but applications for incorporation of Trp analogs are less well-developed.

Interest in 5-hydroxytryptophan (5-HTP) is due to its different electronic properties from Trp. Early studies of substitution of Trp with 5-HTP in proteins were not accomplished selectively using the approach described above. Instead, 5-HTP incorporation in proteins was previously accomplished by growing cells in media lacking Trp but instead containing 5-HTP [21, 22]. A disadvantage of this approach is that the high abundance of 5-HTP in the media allowed translation of 5-HTP for every TGG codon of all genes in the organism. More recently, selective site-specific incorporation of Trp analogs by genetic code expansion was achieved [11, 16, 23]. Site-selective replacement of 5-HTP has recently been used to study post-translational modifications, and to enable coupling organic reactions to a specific site on the protein [24–26]. As such, it is an ideal approach to study the properties of Trp45 in amicyanin.

Trp45 in amicyanin was previously converted to Tyr by site-directed mutagenesis [27]. The crystal structure of this variant showed no significant perturbation of the structure, and the mutation did not affect the redox potential of the copper or its electron transfer properties. Thus, it seemed likely that substitution of Trp45 with 5-HTP would be tolerated by the protein. The 5-hydroxyl group of 5-HTP influences dipole moments in excited states [28]. Consequently, the fluorescence emission of 5-HTP is significantly different from that of Trp. Furthermore, 5-HTP has a broader UV absorption than does Trp. Thus, substitution of Trp45 with 5-HTP, allows exploitation of their distinct absorbance and fluorescence properties to study the effect of the protein environment on 5-HTP, as well as its effects on the interaction between Trp and copper in amicyanin. The data also indicate the mechanism of quenching in native amicyanin is Förster, or fluorescence, resonance energy transfer (FRET).

2. MATERIALS AND METHODS

2.1. Site-Directed Mutagenesis.

The template for his-tagged amicyanin was previously generated [29]. Primers were ordered from IDT DNA (Coralville, IA, USA) and site directed mutagenesis was accomplished with a Lightning Quikchange kit from Agilent Technologies (Santa Clara, CA). The forward and reverse primers are ACCGTCACCTGAATCAACCGC and GCGGTTGATTCAGGTGACGGT, respectively. Plasmids were transformed into XL-10 Gold Ultracompetent cells from Agilent Technologies. The miniprep kit was purchased from Zymo Research (Irvine, CA, USA) to purify the mutated plasmid and Sanger sequence verification was performed by Genewiz (South Plainfield, NJ, USA).

2.2. Expression and Purification of 5-HTP Amicyanin.

ATMW1-BL21(DE3) cells were transformed with the TGA codon substitution for Trp45, along with the pEvoltac-EcW-TGA-h14 plasmid for orthogonal translation of 5-HTP via codon suppression [11, 16]. The genotype for the strain is BL21(DE3) pUltraG_ScW40cca trpS::ZeoR trpT::GentR. In other words, the natural tryptophan tRNA/aminoacyl synthetase of the strain was initially evolved from Saccharomyces cerevisiae, and the orthogonal translation for 5-HTP was evolved from the discarded, previous E. coli tRNA/aminoacyl synthetase for Trp. Cells were grown in the presence of 10 μg/mL gentamycin, 15 μg/mL zeocin, 25 μg/mL chloramphenicol, 50 μg/mL ampicillin, 100 μg/mL spectinomycin, and 125 μM CuSO₄ in LB media at 37 °C. Copper was added to ensure that enough is present for incorporation into the expressed amicyanin. Cells were grown to an optical density (600 nm) of 0.8 before addition of 5-HTP and L-arabinose to final concentrations of 1 mM and 3.33 mM, respectively. 0.6 mM IPTG was added after 30 min of incubation with 5-HTP and L-arabinose. The temperature was lowered to 30 °C and induction lasted for 4-5 h.

2.3. Preparation of Proteins.

Purification of the His-tagged recombinant proteins was as previously described using a Ni-NTA resin [29]. Protein was purified from the periplasmic fraction of harvested cells, as the protein possesses a signal sequence that transports it to the periplasm. Cells were grown in the presence of 125 μM CuSO₄ and periplasmic fractions were treated with 1 mM CuSO₄ to ensure full occupancy of copper. The periplasmic fraction was prepared using a lysozyme-osmotic shock method [30]. Protein was eluted from the Ni-NTA resin at 150 mM imidazole. The copper-free apoproteins were generated as previously described [3] with minor modification. Briefly, the protein was reduced with ascorbate and the Cu⁺ was removed by dialysis against 50 mM potassium phosphate, pH 7.0, with 50 mM thiourea or EDTA for 16 h. Unfolded proteins were prepared by incubation in 50 mM potassium phosphate, pH 7.5 with 6 M guanidinium chloride. Concentrations of native and 5-HTP substituted amicyanin were determined from absorption at 595 nm from the bound Cu(II), using a previously determined extinction coefficient of 4,610 M⁻¹ cm⁻¹ [3]. The concentrations of the copper-free apoproteins were determined from absorption at 280 nm using extinction coefficients calculated from the amino acid sequences using Protparam from Expasy (https://www.expasy.org/). HPLC-coupled ESI-MS analysis of intact proteins was performed using a 1260 Agilent Infinity Series HPLC/6230 Agilent TOF Mass Spectrometer.

2.4. Fluorescence measurements.

Fluorescence was measured with a Perkin Elmer LS55 Fluorescence Spectrometer (Waltham, Massachusetts, USA). Samples were incubated in Quartz SUPRASIL Microcells for 600 μL from Perkin Elmer while measuring fluorescence. Unless otherwise stated measurements were made using a bandpass of 4 nm and scan rate of 50 nm per min.

2.5. Measurement of oxidation-reduction midpoint potential (E_m) values.

Measurement of E_m values for native and 5-HTP substituted amicyanins were performed as previously described [27]. Measurements were performed using an oxidation-reduction potential electrode from Microelectrodes (Bedford, NH, USA). The electrode was calibrated with quinhydrone (a 1:1 mixture of hydroquinone and benzoquinone) as a standard with an E_m, which has a known value of 286 mV at pH 7.0 [31]. The titrations were performed in 50 mM phosphate buffer at pH 7.5, with 400 μM ferricyanide and 200 μM quinhydrone present as mediators. The ambient potential value was incrementally decreased by additions of ascorbate. The relative concentrations of oxidized and reduced amicyanin were determined during the titration from the absorbance at 595 nm, which decreases as amicyanin is reduced and is absent in the fully reduced state. The data were fit by Equation 1, where R is the gas constant, T is temperature in kelvin, n is the number of electrons required for full reduction, and F is Faraday’s constant.

$$E=E_m+(2.3RT/nF)\log {[\text{amicyanin}]}_{\text{oxidized}}/{[\text{amicyanin}]}_{\text{reduced}}$$ (1)

2.6. Calculation of fluorescence quantum yield and Förster radius.

Quantum yield (Φ) was calculated from Equation 2 where A is absorbance, F is fluorescence intensity and RI is refractive index [32]. For measurements in aqueous solution, the RI value for water of 1.333 was used. For measurements in guanidine hydrochloride an RI value of 1.436 was used [33]. The subscripts are s for sample and ref for reference, for either Trp or 5-HTP. The reference quantum yields for Trp and 5-HTP in aqueous solution at neutral pH are 0.14 and 0.27, respectively [34].

$${\mathrm{\Phi }}_{\mathrm{s}}=({\mathrm{A}}_{\text{ref}}/{\mathrm{A}}_{\mathrm{s}})({\mathrm{F}}_{\mathrm{s}}/{\mathrm{F}}_{\mathrm{ref}}){({\mathrm{RI}}_{\mathrm{s}}/{\mathrm{RI}}_{\mathrm{ref}})}^2{\mathrm{\Phi }}_{\mathrm{ref}}$$ (2)

The Förster radius was calculated using Equation 3 [35], which is appropriate since the fluorophore and acceptor are fixed in distance within amicyanin. The fixed distance of R is 10.1 Å between the donor (Trp residue) and acceptor (bound copper) as determined from the crystal structure. The FRET efficiency, or quenched fluorescence intensity, normalized to unity is E. The Förster radius is R₀.

$$R=R_0\sqrt[6]{\frac{1-E}{E}}$$ (3)

3. RESULTS

3.1. Expression and purification of 5-HTP substituted amicyanin.

After expression and purification, the 5-HTP substituted amicyanin was pure as judged by a single band from SDS-PAGE. The substituted protein also had blue color, consistent with the formation of the type 1 copper site. The predicted molecular weight of amicyanin, including the 6-histidine tag and linker region is 12,313 Da. The 5-HTP substituted amicyanin is expected to have a 16 Da greater mass. Analysis of native and 5-HTP substituted amicyanins by ESI-MS yielded peaks corresponding to 12,313 Da and 12,329 Da, respectively, confirming the complete substitution of 5-HTP for Trp45.

3.2. Effect of 5-HTP substitution on the absorbance spectrum of amicyanin.

The major distinction in the spectra of the natural and unnatural amino acids is that 5-HTP has a broader absorbance peak in the 280 nm region and an additional absorbance in the 300-320 nm range (Figure 2A). The notable difference in the absorbance spectrum of the 5-HTP substituted amicyanin relative to native amicyanin is the clear presence of an absorbance feature in the 300 nm to 320 nm region (Figure 2B). This is consistent with the presence of 5-HTP in the protein. A characteristic of amicyanin and other type 1 copper proteins is a broad absorbance at 595 nm [1]. This is why they are also referred to as blue copper proteins. This absorbance at 595 nm is correlated to the typical charge transfer from the 3p sulfur in the Cys ligand to the 3d_x²_-y² copper observed in type 1 copper proteins [36]. This absorbance feature is sensitive to the copper coordination by amino acid side chains in the copper binding site. The absorbance feature from 500-700 nm that is centered at 595 nm in the 5-HTP substituted amicyanin is essentially identical to that of native amicyanin (Figure 2B). This indicates that the electronic transitions that give rise to this absorbance feature were not perturbed by the substitution of 5-HTP for Trp.

Figure 2.

Absorbance spectra of Trp and 5-HTP in solution and in amicyanin. A. Spectra of Trp (blue) and 5-HTP (red). Concentrations were 70 μM. B. Spectra of native amicyanin (blue) and 5-HTP substituted amicyanin (red). Protein concentrations were 60 μM.

While not typically discussed, the bound copper in amicyanin also contributes to the UV absorbance at 280 nm in amicyanin. While this contribution would not be significant in a protein with multiple Trp residues, since amicyanin has only one Trp, the effect is noticeable. As seen in Figure 3A, the 280 nm absorbances with Cu²⁺ in oxidized amicyanin and Cu¹⁺ in reduced amicyanin are similar. As expected, the 595 nm absorbance is lost in the reduced protein [3]. However, when copper is removed to yield apoamicyanin, there is also a significant decrease in 280 nm absorbance due to the loss of copper. It should be noted that this is not due to structural changes in the apoprotein. The crystal structure of the apoprotein is essentially identical to that of the holoprotein, except for the absence of copper [5]. Furthermore, previous study of the holo and apo forms of amicyanin by circular dichroism concluded that the absence of copper decreased the thermal stability of the protein, but did not significantly alter the overall structure of electronic structure of the copper site [27]. It can also be seen that copper-dependent absorbance is absent in the 300-350 nm range. The same spectral features of copper are lost then it is removed from the 5-HTP substituted amicyanin (Figure 3B). Without the background from copper absorbance, the features of the substituted 5-HTP absorbance are more evident.

Figure 3.

Absorbance spectra of native and 5-HTP substituted amicyanin and free 5-HTP. A. Overlay of absorbance spectra of amicyanin. Oxidized protein is blue, reduced protein is black and copper-free apoprotein is red. Protein concentrations are 11 μM. B. Overlay of absorbance spectra of 5-HTP substituted amicyanin. Oxidized protein is blue, reduced protein is black and copper-free apoprotein is red. Protein concentrations are 11 μM. C. Comparison of spectra of 5-HTP free and within the protein. The HTP substituted apoprotein is red. 5-HTP in aqueous solution is purple. 5-HTP solvated by 2-propanol is green. The concetration of each is 54 μM.

Since the environment of 5-HTP in the protein interior is not the same as in aqueous solution, the absorbance spectrum of the free 5-HTP was recorded in 2-propanol, in addition to aqueous solution, to attempt to mimic the internal protein environment and compared to that in aqueous solution (Figure 3C). The absorbance in the 300-330 range is extended to higher wavelength in 2-propanol, compared to aqueous solution. When overlaid with the spectrum of the 5-HTP substituted apoamicyanin, the absorbance features and intensity in this range are now nearly identical (Figure 3C). This is consistent with the crystal structure of amicyanin, which showed this Trp to be shielded from solvent.

3.3. Effect of 5-HTP substitution on the redox potential value of amicyanin.

The E_m values for the Cu²⁺/Cu⁺ redox couple of type-1 copper sites in proteins, including amicyanin, vary by hundreds of mV. The factors that influence the E_m values include copper ligand interactions, hydrogen bonding to the sulfur of the cysteine ligand, and protein constraint on the orientations of the copper ligand orientations [37, 38]. This can include subtle interactions outside the primary coordination sphere of the copper [38]. In order to determine whether the substitution of 5-HTP for Trp45 affected the electronic properties of the copper site, oxidation-reduction potential titrations were performed with the native and 5-HTP substituted amicyanins for comparison. As can be seen in Figure 4, the results of the titrations were essentially identical. Fits of these data to Eq. 1 yielded E_m values of 251 ± 3 mV and 255 ± 4 mV, respectively, for native and 5-HTP substituted amicyanin at pH 7.5. The fitted N values (number of electrons transferred) were 1.21 ± 0.19 and 1.05 ± 0.13, respectively. The results are comparable to previous determinations for native amicyanin [27, 39] and demonstrate that the substitution of 5-HTP for Trp45 has no significant effect on the E_m value of the copper site of amicyanin. This provides further evidence that the substitution of 5-HTP does not effect the electronic structure of the type 1copper site.

Figure 4.

Redox titrations of amicyanin. A. Titration of native amicyanin. B. Titration of 5-HTP substituted amicyanin. Measurements were performed as described under Experimental Procedures. The curve is a fit to Eq. 1. The R² value for each fit is 0.99.

3.4. Effect of 5-HTP substitution on the fluorescence properties of amicyanin.

Copper-dependent fluorescence quenching was previously described in amicyanin [3]. It was shown that the copper-free apoamicyanin exhibited fluorescence that was quenched by the re-addition of copper to reform the holoprotein. Previous to this, copper-dependent fluorescence quenching had also been reported for other type 1 copper proteins; azurin [40], stellacyanin [41], and plastocyanin [42]. The mechanism for this phenomenon was not determined. As such, it is of interest not only to describe the fluorescence spectrum of the 5-HTP in the substituted amicyanin, but to also determine whether this phenomenon for fluorescence quenching is operative in the 5-HTP substituted amicyanin. Since the absorbance spectrum of 5-HTP in the substituted amicyanin extends to higher wavelengths than that of Trp, excitation wavelengths of 295 nm and 310 nm were used to study fluorescence emission (Figure 5).

Figure 5.

Fluorescence properties of native and 5-HTP substituted amicyanins. In each set of spectra, the holoprotein is blue, the apoprotein is red, and the unfolded protein is green. A. Emission spectrum of native proteins with excitation at 295 nm. B. Emission spectrum of 5-HTP substituted proteins with excitation at 295 nm. C. Emission spectrum of native proteins with excitation at 310 nm. D. Emission spectrum of 5-HTP substituted proteins with excitation at 310 nm. Samples contained 4 μM protein. Fluorescence is plotted as arbitrary units that are the same for each panel so that they can be directly compared.

Excitation at 295 nm is specific for Trp fluorescence, whereas other amino acid residues can also absorb at 280 nm. For native amicyanin (Figure 5A), the fluorescence emission maxima of the holo and apo proteins is at 318 nm. The presence of copper in the holoprotein causes a decrease in fluorescence intensity relative to the apoprotein. When the protein is unfolded, a further decrease in intensity is observed with a shift in emission maximum to 360 nm. For the 5-HTP substituted amicyanin (Figure 5B), the fluorescence emission maxima of the holo and apo proteins were at 331 nm. The fluorescence intensities of each are less than for the native protein. Furthermore, the fluorescence of the holoprotein is not quenched relative to the apoprotein in 5-HTP substituted amicyanin. Another distinction is that the intensity of fluorescence of the unfolded 5-HTP substituted amicyanin is increased compared to the folded holo and apo proteins. This is the opposite of what was seen with native amicyanin where unfolding decreased the fluorescence.

The excitation wavelength of 310 nm was chosen to selectively excite 5-HTP, as neither Trp nor any other amino acid residue absorbs at this wavelength. This exploits the shoulder of its UV absorbance relative to Trp (Figure 2A). As expected, essentially no fluorescence was seen for the native amicyanin (Figure 5C). For the 5-HTP substituted amicyanin (Figure 5D), the fluorescence intensities of the holo and apo proteins are approximately equal and of relatively low intensity. It should be noted that it is not possible to observe fluorescence below 320 nm as the excitation wavelength is 310 nm. As seen at 295 nm excitation, the unfolded protein has increased fluorescence. The increase in fluorescence in the unfolded 5-HTP-substituted amicyanin is likely due to solvent now being able to interact with the exposed 5-HTP residue.

These results clearly show that 5-HTP fluorescence is not quenched by the bound copper, as Trp fluorescence is in the native protein. Furthermore, the fluorescence intensity of 5-HTP within the protein is much less than that of Trp in the unquenched apoprotein. Shifts in the position of the fluorescence maxima of the 5-HTP substituted amicyanin relative to native amicyanin are also observed.

3.5. Fluorescence quantum yield and Förster radius.

In order to obtain a quantitative comparison of the fluorescence properties of the different forms of amicyanin, fluorescence quantum yields for the native and 5-HTP substituted amicyanins were calculated. An excitation at 295 nm that selectively excites only Trp and 5-HTP residues was used and data were fit using Eq 2. The excitation did not include contributions to fluorescence intensity from Phe and Tyr residues that also absorb at 280 nm. At this wavelength, the quantum yields for copper-free apoamicyanin and oxidized holoamicyanin (Figure 5C) are 0.32 and 0.17, respectively. For comparison, the quantum yield of free Trp in aqueous solution is 0.14 [34]. Thus, the hydrophobic environment of the Trp enhances its fluorescence relative to its environment in aqueous solution. This also explains why fluorescence intensity decreases when the protein is unfolded and the Trp is exposed to solvent. For the 5-HTP substituted amicyanin at this wavelength (Figure 5D), the quantum yields for the apo and holo forms are each much lower and both are 0.03. The quantum yield of free 5-HTP in aqueous solution is 0.27 [34]. Thus, the presence of the hydroxyl substituent of 5-HTP significantly affects the excited state relative to Trp such that the hydrophobic environment depresses fluorescence intensity. This also explains why fluorescence intensity increases when the protein is unfolded, the opposite of what is observed for native amicyanin. The Förster radius is the distance between the fluorescence donor and acceptor for FRET at which energy transfer 50% efficient. The Förster radius for this process in amicyanin was calculated using Eq 3 to be 9.9 Å.

The efficiency of FRET is also influenced by the degree of spectral overlap between the wavelength of fluorescence emission of the donor and absorbance of the energy acceptor. It can be seen in Figure 3A that the type 1 copper present in oxidized amicyanin exhibits absorbance at 300 nm which extends and decreases to 380 nm. This absorbance is not present in the apoprotein spectrum (Figure 3A). The near UV absorbance that can be attributed to the bound copper can be clearly seen by subtracting the spectrum of the apoprotein from the holoprotein. The spectral overlap of the fluorescence emission of Trp and the absorbance of copper can be seen in Figure 6. This overlap allows the observed FRET. The fluorescence emission maxima for the 5-HTP substituted amicyanin is shifted from that of the native protein from 318 nm and 331 nm. Since the emission maximum of the FRET acceptor must be greater than that of the FRET donor, the spectral overlap with copper absorbance will be much less for the 5-HTP substituted amicyanin fluorescence. This shift coupled with the relatively low fluorescence quantum yield of the5-HTP substituted amicyanin accounts for the lack of FRET.

Figure 6.

Overlap of Trp fluorescence with copper absorbance in amicyanin. The near UV absorbance spectrum of the type 1 copper in amicyanin (black) was obtained by subtracting the spectrum of apoamicyanin from the spectrum of holoamicyanin. The fluorescence spectra of apoamicyanin (red) taken from the data shown in Figure 5. Since the units for absorbance and fluorescence are different, the spectra cannot be quantitatively compared. The spectra presented here are presented on a comparable scale to allow qualitative evaluation of the wavelength range of potential overlap.

3.6. Effect of the redox state of copper in amicyanin on FRET.

The visible absorbance feature of the oxidized type 1 copper of amicyanin is lost upon reduction to Cu⁺. However, most of the UV and near UV absorbance is retained (see Figure 3A). This retention of near UV absorbance was previously noted in reductive titrations of amicyanin using dithionite as a reductant [3] and using enzymatic reduction [2]. This region of the spectrum is expanded in Figure 7A for comparison. The reduction to Cu⁺ results in a small decrease in near UV absorbance in the region of potential spectral overlap with Trp fluorescence. As a consequence of this, a corresponding decrease in the level of fluorescence quenching of the reduced Cu⁺ amicyanin is observe, compared to that of the oxidized Cu²⁺ amicyanin (Figure 7B). The calculated quantum yield for the reduced amicyanin is 0.21 compared to 0.17 for the oxidized amicyanin, and the calculated Förster radius is 9.1Å compared to 9.9 Å. This correlation of the extent of spectral overlap with level of fluorescence quenching provide additional support for the conclusion that the mechanism of copper-dependent quenching of Trp fluorescence in amicyanin is FRET.

Figure 7.

Effect of the redox state of copper on the quenching of Trp fluorescence in amicyanin. A. Effect of the oxidation state on the near UV absorbance of copper. Absorbance spectra of oxidized Cu²⁺ amicyanin (blue) and dithionite-reduced Cu⁺ amicyanin (black). B. Effect of the copper oxidation state on the fluorescence of of Trp in amicyanin. Fluorescence spectra using an excitation wavelength of 295 nm are shown for apoamicyanin (red), oxidized Cu²⁺ amicyanin (blue) and dithionite-reduced Cu⁺ amicyanin (black). Fluorescence is plotted as arbitrary units.

4. DISCUSSION

Copper-dependent fluorescence quenching was described for amicyanin [3] and for other type 1 copper proteins including azurin [40], stellacyanin [41], and plastocyanin [42]. However, a mechanism to explain this phenomenon was not described. The data presented indicate that FRET is responsible. The shifted fluorescence emission maximum for the 5-HTP substituted proteins, as well as the calculated quantum yields and Förster radius support our conclusion. Given the very similar structures of the other type 1 copper proteins, this is likely that FRET is responsible for the observed copper-dependent fluorescence in the other cupredoxins as well.

To further investigate the interaction of the sole Trp residue of amicyanin with the protein environment in which it resides, genetic code expansion technology was used to replace it with 5-HTP. The substitution of 5-HTP for Trp45 did not affect the electronic properties of the type 1 copper site, as evidenced by no effect on the distinctive visible absorbance of the oxidized protein or the E_m value of the copper. However, the fluorescence properties of the 5-HTP relative to Trp were significantly affected. The intensity of the fluorescence emission from 5-HTP in apoamicyanin is much less than that of Trp in native apoamicyanin, and the fluorescence of 5-HTP in the substituted amicyanin is not quenched by the copper.

The excited state of 5-HTP is significantly different from that of Trp. Both residues have two low energy excited states, referred to as ¹L_a and ¹L_b [22, 28, 43]. The two excited states have different characteristics for the separation between the absorption of light and the resulting emission of fluorescence, otherwise known as a Stokes-shift. The ¹L_a excited state has a Stokes-shift significantly dependent on the local polarity around the fluorophore and is the lower of the two excited states for Trp. In contrast, the ¹L_b excited state is less dependent on the nearby polarity around the residue and is the lower of the two excited states for 5-HTP. The intensity of the fluorescence emission for 5-HTP is influenced by interactions with the 5-hydroxyl group which influences the dipole observed from the corresponding excited state required for fluorescence emission [21, 22, 28]. The hydrophobic environment of the 5-hydroxl of the 5-HTP residue in amicyanin makes the excited state dipole less favorable, suppressing fluorescence relative to what is observed for Trp in the native protein.

The calculated Förster radii of 9.9 Å and 9.1 Å between Trp and copper for oxidized and reduced amicyanin, respectively, are within the range of other intrinsic FRET pairs formed by aromatic amino acid residues in proteins, which range from 4 Å to 16 Å [35]. Engineered FRET pairs, which are used to study protein motions and protein-protein interactions in cells, are designed to have significantly higher spectral overlap and quantum yields. These pairs have Förster radii ranging from 30 Å to 70 Å [44]. Given the relatively low quantum yield of 5-HTP in amicyanin and weak spectral overlap, fluorescence quenching of the 5-HTP residue is not possible. However, if 5-HTP could be paired with an engineered FRET acceptor with a larger spectral overlap, this would increase the distance over which FRET could be observed. UV light-absorbing chromophores for labeling proteins are currently being used and could potentially be partnered with 5-HTP as a FRET pair without influence from other Trp residues in the protein [45]. A benefit to this approach is that it only requires labeling with one dye or partner instead of two synthetic molecules to form a FRET pair. Since 5-HTP can be excited at a wavelength that is not absorbed by Trp of other amino acid residues, there will be no interference from the protein and depending on the nature of the study, the 5-HTP could be inserted in an interior section of the protein or on the surface.

HIGHLIGHTS.

Substitution of Trp in amicyanin with 5-HTP was achieved by genetic code expansion

The mechanism of long-distance Cu-dependent fluorescence quenching of Trp is FRET

5-HTP substitution allows spectroscopic interrogation of the protein microenvironment

The protein environment affects excited states of 5-HTP and its fluorescence properties

ABBREVIATIONS

5-HTP

5-hydroxytryptophan

aaRs

aminoacyl-tRNA synthetase

E_m

oxidation-reduction midpoint

FRET

fluorescence (Förster) resonance energy transfer