Figure 1.
Molecular mechanisms of dityrosine cross-linking (Cortho-Cortho). (
A) Reaction mechanism for the formation of dityrosine. (B) Intermolecular dityrosine cross-linking in proteins.
Figure 2.
Tertiary and primary structures of insulin, dimer and 2Zn hexamer forms and receptor binding sites.
(A) 3D structures of insulin monomer and dimer extracted from the crystallized structured of the 2Zn pig insulin hexamer (4INS.pdb) [37]. Insulin A chains of insulin are displayed in gray (two different shades in the insulin dimer) and B chains are displayed in black. The 4 Tyr residues of insulin (Y, orange) and the 6 Cys (C, green) involved in 3 disulphide bridges (SS) are displayed in orange and green, respectively (left-side structure for insulin monomer and top structure for insulin dimer). The two insulin binding sites are displayed in CPK (0.5 Å radius) in blue (binding site 1) and red (binding site 2) (right side structure for insulin monomer and bottom structure for insulin dimer). (B) Primary structure of human insulin. The amino acid residues involved in the association of 2 insulin molecules into a dimer (blue residues) and in the molecular assembly of 3 dimers into 2Zn insulin hexamer (red residues) are displayed. The amino acid residues belonging to insulin receptor binding sites 1 and 2 are indicated by wide blue and red arrows, respectively. Amino acid residues important for insulin receptor recognition of the binding site 1 but not belonging to this site are displayed with thin blue arrows (adapted from [36], information regarding dimer association, hexamer formation, and insulin receptor binding sites extracted from [35], [36], [38], [48]–[51]).
Figure 3.
Absorbance spectra of human insulin after prolonged 276 nm UV-excitation.
(A) Absorbance spectra obtained before and after 276 nm light continuous exc. (0.25 h, 0.5 h, 0.75 h, 1 h, 2 h, and 2.5 h) of human insulin in solution. In the insert, are plotted the absorbance values at 220, 276 and 290 nm vs 276 nm exc. time. Absorbance at 276 and 290 nm increase linearly with 276-nm excitation time (linear fitting, R2 = 96.78% for Abs276 nm and R2 = 99.28% for Abs290 nm). (B) Fitting of the absorbance spectrum (obtained after 2 h exc. with 276 nm light) with a 2 gaussian peak function. The peak at 276 nm was fixed. The original experimental spectrum, the two individual fitting curves obtained for each peak and the cumulative curve of the two fittings are displayed.
Table 1.
Parameter values obtained upon fitting of the insulin absorbance spectra with a 2 Gaussian peak function.
Figure 4.
Fluorescence emission of human insulin (276 nm excitation) upon prolonged 276 nm UV-excitation.
(A) Fluorescence emission spectra (276 nm exc.) obtained before and after 276 nm light continuous exc. (0.5 h, 1 h, 1.5 h, 2.5 h, 3.5 h, and 7 h) of human insulin in solution. There is a continuous decrease in emission intensity at 303 nm with 276 nm exc. time. The insert shows a zoom of the emission spectra between 350 and 550 nm. Emission intensity at ∼405 nm increases progressively with 276 nm excitation time. (B) Fluorescence emission intensity kinetic traces obtained at 330 and 405 nm (exc. fixed at 276 nm) upon continuous of human insulin with 276 nm light. Fitting of the experimental traces was carried out using an exponential function F(t) = C1– C2.e−kt. Fitted parameter values and corresponding errors, and root mean square error values were obtained after fitting each kinetic trace (Table 2).
Table 2.
Parameter values obtained upon fitting the insulin fluorescence emission kinetic traces recorded with emission fixed at 303 nm (exc. 276 nm) and 405 nm (exc. 276 or 320 nm).
Figure 5.
Fluorescence emission of human insulin (320 nm excitation) upon prolonged 276 nm UV-excitation.
(A) Fluorescence emission spectra (320 nm exc.) recorded before and after 276 nm light continuous exc. (0.5 h, 1 h, 1.5 h, 2.5 h, 3.5 h, and 7 h) of human insulin in solution. There is a continuous increase in emission intensity at 405 nm with 276 nm exc. time. (B) Fluorescence emission intensity kinetic trace obtained at 405 nm (exc. at 320 nm) upon continuous of human insulin with 276 nm light. Fitting of the experimental traces was carried out using an exponential function F(t) = C1– C2.e−kt. Fitted parameter values and corresponding errors, and root mean square error values were obtained after fitting each kinetic trace (Table 2).
Figure 6.
Fluorescence excitation spectra of human insulin (emission fixed at 303 nm) upon prolonged 276 nm UV-excitation.
The excitation spectra (em. 303 nm) were obtained before and after 276 nm light continuous exc. (0.5 h, 1 h, 1.5 h, 2.5 h, 3.5 h, and 7 h) of human insulin in solution. There is a continuous decrease in excitation intensity at 276 nm with 276 nm exc. time.
Figure 7.
Fluorescence excitation spectra of human insulin (emission fixed at 405 nm) upon prolonged 276 nm UV-excitation.
Fluorescence excitation spectra (emission fixed at 405 nm) obtained before and after 276 nm light continuous exc. (0.5 h, 1 h, 1.5 h, 2.5 h, 3.5 h, and 7 h) of human insulin in solution. There is a continuous increase in excitation intensity at 320 nm with 276 nm exc. time.
Figure 8.
Concentration of detected free thiol groups (open circles) in human insulin vs 276 nm exc. time.
Detection of free thiol groups was carried out using the Ellman’s assay before and after 276 nm light continuous exc. (0.25 h, 0.5 h, 0.75 h, 1 h, 2 h, and 2.5 h) of human insulin in solution. The concentration of free thiol groups was estimated from the absorbance of the Ellman’s assay reaction product, TNB2−, at 412 nm (ε412 nm = 14150 M−1.cm−1 [44]). The experimental values were fitted using an exponential function y = y0– A.e−R0t (fitted curve in red), where y is the concentration of thiol groups (µM) at the 276 nm excitation time t (h), y0 and A are constants and R0 is the rate of thiol group formation (h−1). Fitted experimental parameters were: y0 = 5.04±0.24 µM, A = 5.27±0.22 µM, R0 = 0.87±0.09 h−1. Root mean square error was 99.41%.
Figure 9.
Far-UV CD spectra of human insulin recorded before upon prolonged 276 nm UV-excitation.
The far-UV CD spectra were obtained before and after 276 nm light continuous exc. (0.5 h, 1 h, 1.5 h, 2.5 h, 3.5 h, and 7 h) of insulin in solution. There is a progressive loss of ellipticity signal with 276 nm exc. time at 195, 209 and 222 nm.
Figure 10.
Effect of continuous 276 nm light exposure of human insulin on its recognition by guinea-pig anti-insulin antibodies.
Human insulin concentration was detected using a radioimmunoassay. Insulin samples were previously excited with 276 nm light during 1.5 and 3.5 h. A positive control (PC) was carried out for each excitation duration, where insulin was left in the dark for the same time period (1.5 and 3.5 h). Uncertainty values (standard errors) for the detected insulin concentration are displayed with error bars.
Figure 11.
Effect of continuous 276 nm light exposure of human insulin on its hormonal function in vitro.
The glucose uptake by human skeletal muscle cells was measured after 1 hour of incubation at three different conditions: no insulin (control), insulin kept in the dark (positive control, PC), and UV-illuminated insulin (1.5 h exc. at 276 nm). Uncertainty values (standard errors) for the concentration of glucose (uptake) are displayed with error bars. The students t-tests resulted in the following p-values: p (control vs PC) = 27.6.10−4; p (control vs 1.5 h exc. at 276 nm) = 0.002165; p(PC vs 1.5 h exc. at 276 nm) = 0.00179.