2-3-1.The catalytic activity of Art_KSI and the binding affinities toward steroids.
The Art_KSI protein was successfully purified by anion exchange column and its quality confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Figure 2-2). The precise molecular weight determined by LC/MS analysis showed the m/z= 13402 (M+H+), which was consistent with the molecular weight calculated from the amino acid composition of Art_KSI. Art_KSI enzyme possesses substantially activity with the value of kcat/Km = 1.12x107 (M-1s-1) (Table 2-2).
Figure 2-2. SDS-PAGE analysis of protein purity. Land 1 shows the protein marker and Lane 2 is Art_ KSI protein produced in E. coli, and purified on an anion exchange column.
Table 2-2. Michaelis-Menten parameters of different KSI mutants
Since the Art_KSI contains no tryptophan and only one tyrosine, i.e. Tyr-14, present in the steroid binding site, Art_KSI exhibits unique fluorescence emission to describe the protein-steroid binding interaction. The dissociation constants (Kd) of
KSI-steroid complex was determined by the quenching yield of tyrosine fluorescence at λem= 307 nm (λex= 278 nm). Several Kd values were listed in Table 2-3. Among those steroids, 19-norandrostendione presents the strongest interaction with Art-KSI (Kd= 10.9 μM) with a binding energy of 6.75 Kcal/mol. On the other hand, 1, 5-EDANS showed much weak binding to Art_KSI with a Kd value of 350 μM.
Table 2-3. Dissociation constant of steroid derivatives.
2-3-2. Identification and sequencing of the peptide labeled with I-14.
α-Halo acid or amide reagents (such as iodoacetic acid or iodoacetamide) are among the most frequently used reagents for thiol-modification in protein chemistry.
In most proteins, the site of reaction is the cysteine residues. Note that, there is no cysteine residue present in the wild type KSI. However, interestingly, when Y14only was treated with I-14, a stoichiometric and covalent label was observed by LC/MS analysis (Figure 2-3A).
Figure 2-3A. LC/MS analyses on the chemical modification of KSI. The species with a molecular mass of 13978 Da corresponds to the Y14only (13672 Da.) plus the moiety of 5-[2-(acetamido)ethylamino]naphthalene-1-sulfonic acid (306 Da).
The molecular weights of Y14only before and after chemical modification are 13672 Da and 13978 Da. The 306 amu increase confirmed the successful attachment of the moiety of 5-[2-(acetamido)ethylamino] naphthalene-1-sulfonic acid on KSI. It is reasonable to assume that a specific nucleophile in KSI is accessible in order for the substitution reaction to take place. Further, proteolytic digestion combining with an extensive LC/MS/MS analysis on sequencing the (I-14)-labeled peptide were employed to resolve the target residue of labeling. As shown in Figure 2-3B, the total ion chromatogram of the digest of the (I-14)-labeled Y14only exhibited a large number of peaks. Each peak corresponds to at least one peptide in the digest mixture.
The labeled peptides were located within the chromatogram by comparison of the peptides present within digests of labeled and unlabeled Y14only, using 335 nm absorption wavelength of I-14 to confirm the locations of the peptides of interest. The shaded region (retention time within 17-19 min) contains the (I-14)-labeled fragment.
Figure 2-3B. LC/MS analyses on the chemical modification of KSI. Y14only labeled with I-14 followed by pepsin digestion and HPLC separation. Through comparison with the control digest, the shaded portion was found to contain the labeled fragment (retention time 17-19 min).
The doubly-charged peptides of m/z = 975.8 and 822.6 were obtained by RP-HPLC separation of the labeled and unlabelled samples, respectively. The mass difference of the two fragments at the singly-charged state is 307 amu, which is consistent with the labeling moiety. Both fragments were subjected to MS/MS analysis and yielded the spectra shown in Figure 2-4.
Figure 2-4A. MS/MS analysis of the peptide labeled with I-14. MS/MS daughter-ion spectrum of the unlabelled peptide (m/z 822.6 in the doubly-charged state).
Figure 2-4B. MS/MS analysis of the peptide labeled with I-14. MS/MS daughter-ion spectrum of the labeled peptide (m/z 975.8 in the doubly-charged state).
The presence of b-ion and y-ion unequivocally confirmed the sequence of the unlabelled peptide as ATVEDPVTGSEPRSGTAA, corresponding to residues 34-51 of Y14only. Identification of the point of attachment of the probe moiety was achieved through inspection of the daughter ions of (I-14)-labeled peptide (m/z = 975.8, doubly charged). By comparing a series of b-ion and y-ion derived from both fragments (Figure 2-4C), we were able to conclude unequivocally that the moiety of 5-[2-(acetamido)ethylamino] naphthalene-1-sulfonic acid was covalently labeled on the Asp-38 residue. Although other effects may assist the labeling reaction of I-14, the hydrophobic nature existing in the active site of KSI was considered to play a major role, for which it not only promoted the driving force for I-14 binding but also enhanced the nucleophilicity of Asp-38 to react with I-14.
Figure 2-4C. MS/MS analysis of the peptide labeled with I-14. Rationalization of the observed singly-charged y- and b-ions with the pattern of amino acid sequence of the expected peptide.
2-3-3. The comparison of the fluorescence polarization factors of the labeled KSI mutants
In order to explore the fluorescence polarization factors of I-14 in KSI, we utilized two mutant proteins: Y14only and mKSI_126C. Depended on the specific labeling residues designed, we could easily control the fluorophore to label in active site or on the surface of protein. For instance, the mKSI_126C was designed for conducting the I-14 labeling on the surface of protein. This labeling reaction is presumably through the nucleophilic attack of Cys-126, which is the C-terminus of the protein and is structurally predicted to exposes on the surface. The steady state emission spectra of I-14 in various environments have been performed to mimic the various environmental conditions. The results are shown in Figure 2-5. The samples of I-14 in Tris buffer and in isopropyl alcohol (IPA) were used to simulate the hydrophilic and hydrophobic environment, respectively.
Figure 2-5. The steady-state emission spectra of I-14 in various environments.
Spectra were indicated as follows: I-14 in 100% IPA solution (▼), I-14 in 60% IPA solution (▲), I-14 in 20% IPA solution (●), and I-14 in 0% IPA solution (■). Note that H2O was used to compensate the percentage of IPA.
In all cases, the absorption band was centered at ~335 nm. However, the emission spectra of I-14 in Tris buffer and in IPA were significantly different. In hydrophilic environment the emission was centered at ~500 nm, whereas the emission spectra blue-shifted to ~450 nm in hydrophobic condition (IPA). Therefore, this spectral shift can be used as an indicator for the local environment in which the fluorophore exists. In mKSI_126C/I-14, the only accessible reaction site was Cys-126 of the protein, and it was expected that the emission spectra should similar to that in Tris buffer. As shown in Figure 2-6, the emission spectrum of mKSI_126C/I-14 was centered at ~500 nm with a shoulder at ~475 nm. This result supports that the moiety of 5-[2-(acetamido)ethylamino] naphthalene-1-sulfonic acid is surrounded by a hydrophilic environment; however the shoulder at ~475 nm indicated that the local environment of the 5-[2-(acetamido)ethylamino] naphthalene-1-sulfonic acid moiety is somewhat less hydrophilic due to the hydrophobic shell of the entire protein. The emission spectrum of Y14only/I-14 is centered at ~450 nm, which is similar to the emission spectrum of I-14 in IPA. This result strongly supported that the fluorophore was buried in a hydrophobic environment.
On the other hand, the steady-state fluorescence spectrum of Art_KSI-mA51 was shown to center at ~475 nm (Figure 2-6). As we added the 30 μM 19-norandrostendione into this system, the fluorescence emission became red-shifted and centered at ~500 nm. It indicated that the ligand and steroid binding domain existed on thermodynamic equilibrium. But, the equilibrium of
association/dissociation was disrupted by competition inhibition because of 19-norandrostendione contains the bigger absolute association energy than mA51, which causes the more population of mA51 to be exposed to the polar environment and results in the increasing intensity of fluorescence emission at ~500 nm.
Figure 2-6. The emission spectra of KSI mutants labeled with I-14. Spectra were indicated as follows: Y14only/I-14 (■), mKSI_126C/I-14 (●), Art_KSI/mA51, and Art_KSI/mA51 treated 30 μM 19-norandrostendione (○). All emission spectra obtained in Tris HCl buffer (50 mM, pH 7.5).
2-3-4. Fluorescence anisotropy decays of I-14 in KSI mutants.
According to the steady-state results abovementioned, the binding site of I-14 in Y14only_KSI was determined to be inside the hydrophobic cavity. In order to confirm
350 400 450 500 550 600
this result, we measured time-resolved fluorescence anisotropy to obtain the depolarization kinetics for the rotational Brownian motion of the fluorophore in various biological environments. In a TCSPC experiment, the fluorescence anisotropy decay can be determined with the following equation 46-47:
VV VH
where the G factor is the compensation for the polarization dependence of the monochromator grating, optics, and detector. By definition, G = IHV/IHH, where IHV
and IHH represent the intensities of fluorescence excitation with horizontal polarized light and monitoring at vertical and horizontal polarizations, respectively. Similarly, IVV and IVH represent the intensities of fluorescence excitation with vertically polarized light monitored at vertical and horizontal polarizations, respectively.
The fluorescence anisotropy decay curves of I-14 only, I-14 in Y14only_KSI and I-14 in mKSI_126C are shown in Figure 2-7, respectively. The observed time-dependent anisotropy curves all decay to zero background at longer times and they can be well described by a single-exponential (free isotropic rotation) or bi-exponential (free anisotropic rotation) decay function with the rotational time coefficients as indicated in the figures. For I-14 only in Tris buffer, the rotational time coefficient is 0.15 ns, while for I-14 in Y14only_KSI, the rotational time coefficient increases by two orders of magnitude (14.8 ns). The single-exponential decay feature of the system indicates a purely isotropic orientation relaxation that reflects the size of the system. According to Perrin’s equation, the rotational correlation time coefficient (τ) for a spherical molecule is proportional to the hydrodynamic molecular volume of the fluorophore (V) and can be represented using the following equation46:
V
= RTη
τ (2)
where η is the viscosity of the medium, T is the absolute temperature (fixed at 298 K) and R is the gas constant. As a result, the hydrodynamic radius of I14 only and I14 in Y14only_KSI can be estimated to be 5.5 and 25.4 Å, respectively.
In the case of mKSI_126C/I-14, where the fluorophore I-14 is expected to be labeled on the surface of the protein molecules, the time-dependent fluorescence anisotropy was observed to feature a bi-exponential decay with rotational time coefficients of 0.28 ns (amp1) and 12.6 ns (amp2). The observed two-component anisotropic decay feature is consistent with the rotational relaxation being mainly due to the rotation of the mKSI_126C molecule and the segmental motion of the fluorophore outside the protein. However, the contribution of the latter is much larger than that of the former. Because I14 was labeled on the surface of the protein, it is reasonable to observe that the fast-decay component corresponding to the segmental motion of the fluorophore outside protein becomes dominant in the observed depolarization curve. Furthermore, the rotational relaxation time of I14 on the surface of mKSI_126C is substantially longer than that of free I14 in Tris buffer solution (0.28 vs. 0.15 ns), indicating the significance of the restricted motion of the fluorophore affecting the observed depolarization kinetics.
Figure 2-7. Time-resolved fluorescence anisotropy decay of I-14 in (A) I14 in Tris buffer , (B) Y14only/I-14 and (C) mKSI_126C/I14. The raw data are shown as open circles and the fitting results are represented by solid curves.
2-3-5. Fluorescence anisotropy decays of mA51 moiety in Art_KSI/mA51
According to the results of Figure 2-7, we understood the anisotropy behaviors of fluorophore in protein active-cavity or labeled on surface of KSI are different.
When the fluorophore bound with long-chain in the KSI, the fluorophore can changes its orientation between inside and outside. The relative population of a
fluorophore-KSI which exists in two different environments displays complex kinetics. The observed time-dependent anisotropy, r (t)obs, of Art_KSI/mA51 was fitted by the following equation:
r (t)obs = x×rinside
(t) + (1-x)×r
outside(t)
= r01×x× exp(-t/τ1
)+(1-x)×r
02×[f×exp(-t/τ2)+(1-f)×exp(-t/τ
3)]
(3) According to the extinction coefficients at λ=375 nm and lifetimes monitored at 500 nm of two type of orientations are the same, the observed time-dependent anisotropy is a sum of these two components weighted by their corresponding intensities. The physical meaning of equation 3 is obvious: the anisotropy component resulting from the rotational relaxation decay of mA51 molecule inside the Art_KSI, whereas the other component resulting from the rotational relaxation decays bi-exponentionally of mA51 outside the KSI as expressed in Figure 2-8. x and (1-x) are the relative corresponding intensity of mA51 inside and outside the KSI cavity, respectively. r 01 and r are fundamental anisotropy values of two type orientations. Based on Figure 02 2-7 results, τ1 is 16.2 ns fixed by a KSI-Y14 rotational diffusion time which is calculated by a hydrodynamic radius (25.4 Å), and τ2 and τ3 are also fixed by mKSI-126C rotational diffusion time which are 0.27 and 13.4 ns, respectively. The f and (1-f) are the fractions of species for the anisotropy with τ2 and τ3, respectively.
The Figure 2-8 shows the fundamental anisotropy values of Art_KSI/mA51. The different fundamental anisotropies demonstrate the geometries of the fluorophore molecules in two different conditions are slightly altered by the protein and solvents.
The ratio of mA51 molecule inside to outside Art_KSI protein is 0.33 (0.25/0.75), therefore the mA51 molecule interact more strongly with solvents than protein in Art_KSI/mA51 system without any steroid derivatives.
Figure 2-8. Time-resolved fluorescence anisotropy decay of mA51 in Art_KSI.
2-4. Conclusion
The hydrophobic interaction prompted I-14 into steroid-binding site in conjugation with Asp-38. This interaction between Y14only_KSI and I-14 as well as identification of the site of I-14 labeling was confirmed by steady-state fluorescence and MS measurements. The Art_KSI was constructed to evaluate the feasibility of bioaffinity system for steroid detection. This new steroid-recognition model was successfully constructed by specifically conjugating the switching ligand to the Cys-86 of Art_KSI through the disulfide bond formation. The comparison between the steady-state fluorescence of Y14only_KSI/I-14, mKSI_126C/I-14 and Art_KSI/mA51 indicated the levels of fluorescence polarization of the fluorophore in various environments. The time-dependent fluorescence anisotropy studies on mA51 were employed to probe the environment around it. The anisotropy revealed that the mA51 on Art_KSI/mA51 exists both inside and outside of the steroid-binding cavity with 25% of the mA51 inside the protein cavity. Furthermore, the steady-state
fluorescence spectra demonstrated that 19-NA competed with mA51 for the binding sites and the population of mA51 in the steroid-binding site perturbing by 19-NA was also manifested by a small red-shift.