Soluble Cyanobacterial Carotenoprotein as a Robust Antioxidant Nanocarrier and Delivery Module

Abstract

To counteract oxidative stress, antioxidants including carotenoids are highly promising, yet their exploitation is drastically limited by the poor bioavailability and fast photodestruction, whereas current delivery systems are far from being efficient. Here we demonstrate that the recently discovered nanometer-sized water-soluble carotenoprotein from Anabaena sp. PCC 7120 (termed AnaCTDH) transiently interacts with liposomes to efficiently extract carotenoids via carotenoid-mediated homodimerization, yielding violet-purple protein samples. We characterize the spectroscopic properties of the obtained pigment-protein complexes and the thermodynamics of liposome-protein carotenoid transfer and demonstrate the delivery of carotenoid echinenone from AnaCTDH into liposomes with an efficiency of up to 70 ± 3%. Most importantly, we show efficient carotenoid delivery to membranes of mammalian cells, which provides protection from reactive oxygen species (ROS). Incubation of neuroblastoma cell line Tet21N in the presence of 1 μM AnaCTDH binding echinenone decreased antimycin A ROS production by 25% (p < 0.05). The described carotenoprotein may be considered as part of modular systems for the targeted antioxidant delivery.

Keywords: carotenoid; carotenoid delivery; liposomes; oxidative stress; protein-membrane interactions; protein-protein interactions.

Figure A1

Spectral properties of ECN and CAN in different environments. (A) Normalized absorption of AnaCTDH containing CAN (violet) and ECN (purple). Orange and red lines show the absorption of ECN and CAN in liposomes, respectively. Dotted and dashed lines show the absorption of COCP and the W288A sequence variant thereof, respectively, both containing CAN. Spectra were shifted along the y-axis for the better presentation. (B,C) Resonance Raman spectra of CTD-like carotenoid proteins recorded with 532 nm excitation. (B) AnaCTDH (solid violet line) and COCP (dotted line), both containing CAN. (C) AnaCTDH containing CAN (violet) and ECN (purple). (D) Raman spectra of ECN in liposomes. The Raman signal in the 900–1060 cm⁻¹ region was magnified 2.5 times for better presentation. Note the x-axis break from 1350 to 1475 cm⁻¹. (E) Dependency of the Raman shift of the -C=C- stretching mode (ν₁) on the position of the S₀–S₂ maximum (λ_max) in absorption spectra of ECN and CAN embedded in liposomes and proteins. Arrows indicate, how the effective conjugation length and the average number of double bonds changes among samples. The average number of conjugated double bonds was calculated using the empirical dependency denoted in the text [48]. (F) Chemical structures of cis-cis and trans-trans isomers of CAN.

Figure A2

Analysis of ECN and CAN by thin layer chromatography and absorbance spectroscopy. (A) thin layer chromatogram in a 80:20 kerosene:acetone (v/v) system showing the carotenoid content in the membranes of the ECN/CAN-producing E. coli strain [29,34] (1) and the samples containing AnaCTDH(CAN) (2), CAN-loaded liposomes (3), AnaCTDH(ECN) (4), or ECN-loaded liposomes (5) used in the study. (B) absorbance spectrum of ECN in liposomes obtained by running SEC with full absorbance spectrum detection. The spectrum was corrected for the high scattering from liposomes by subtracting the scattering profile measured on empty liposomes. (C) Absorbance spectrum of CAN in liposomes obtained by running SEC with full absorbance spectrum detection. The spectrum was corrected for the high scattering from liposomes by subtracting the scattering profile measured on empty liposomes. Note the difference of the peak maximum position for ECN and CAN.

Figure A3

ECN uptake from liposomes by AnaCTDH apoprotein. (A) Absorption spectra of ECN in liposomes before (orange) and after (pink) four hours of incubation with 300 µM of AnaCTDH at 30 °C. Resulting spectrum (pink) was decomposed into components which correspond to absorption of ECN remaining in liposomes (dashed line) and in AnaCTDH (dotted line). Percentiles indicate the ECN content in different fractions comparing to the indicial concentration in liposomes. Some fraction of carotenoid (~7%) was lost due to photodegradation upon the continuous absorption measurements. (B) Time-course of accumulation of AnaCTDH(ECN) due to carotenoid uptake from liposomes monitored as increase of optical density at 550 nm. The inset shows cuvettes with solutions of ECN-containing liposomes before (left) and after incubation with AnaCTDH apoprotein.

Figure A4

Left: fluorescence decay kinetics of the TagRFP-AnaCTDH chimera in the absence of carotenoid (black), after transfer of CAN from COCP (red) and after addition of liposomes (blue) which were previously enriched by ECN via the AnaCTDH(ECN) holoprotein. Right: FLIM image of HeLa cells with the TagRFP-AnaCTDH chimera showing no signs of colocalization of the protein at the membranes.

Figure 1

Interactions of the AnaCTDH apoprotein with membranes. (A) Threshold voltages (V_bd) that caused electrical breakdown of DOPC and DPhPC membranes at different concentrations of the AnaCTDH apoprotein in solution. All data are reported as mean values ± SD from three independent experiments. The means accompanied by different letters (a, b, and c) are significantly different at p < 0.05. (B) Fluorescence micrographs of giant unilamellar vesicle membranes made from 50 mol% DOPC and 50 mol% TMCL (top row) and 67 mol% DOPC and 33 mol% CHOL (bottom row) (stained with fluorescent lipid marker) in the absence and in the presence of AnaCTDH apoprotein. The lipid:protein ratio was 100:1. The liquid, disordered (l_d) phase appears red, while the solid, ordered (s_o) phase remains dark. Image size is 15 × 15 μm. (C) DSC thermograms of DMPC liposomes in the absence (control) and in the presence of the AnaCTDH-apoprotein in a lipid to protein molar ratios indicated using color coding. (D) The increase of phase transition temperature of liposomes composed of DMPC and DPPC in response to the increasing AnaCTDH concentration. The results were averaged based on three independent experiments (mean ± SEM). (E) Model of the interaction of the AnaCTDH apoprotein with PC. Glu-133 is shown in red (numbering from 6FEJ structure), positively charged residues are shown in blue, leucines are shown in cyan. The structure of the AnaCTDH apoprotein was drawn in Pymol using PDB ID 6FEJ chain A [28].

Figure 2

CAN uptake by the AnaCTDH apoprotein from liposomes in PBS (pH 7.4). (A) Absorption spectra of CAN in liposomes and after incorporation into AnaCTDH. Rayleigh scattering (RS) was subtracted. (B) Increase of O.D. at 550 nm upon addition of different amounts of the AnaCTDH apoprotein to CAN-containing liposomes. (C) Time-courses of AnaCTDH holoprotein accumulation due to ECN uptake from liposomes at different temperatures. (D) The corresponding temperature dependency of rate constants (Arrhenius plot). All values are presented as mean values ± SD from three independent experiments.

Figure 3

ECN delivery by AnaCTDH(ECN) holoprotein into liposomes. (A) The absorption spectrum of AnaCTDH(ECN) before (purple) and after addition of liposomes (orange). The percentage of carotenoid remaining in AnaCTDH after addition of liposomes was determined by decomposing the resulting spectrum into AnaCTDH(ECN) absorption and scattering of liposomes, considering also the fact that ECN absorption in liposomes, peaked at ~465 nm, is significantly blue-shifted and does not overlap with protein-bound ECN in the 550–700 nm region. (B) The percentage of carotenoid remaining in AnaCTDH upon increasing the concentration of liposomes in solution. Values are presented as mean ± SD from three independent experiments. The dependency was approximated by an exponential function in order to determine the maximum yield of carotenoid delivery. The inset shows cuvettes with the AnaCTDH(ECN) holoprotein solution before (left) and after (right) incubation with liposomes. (C) Time-courses of optical density at 530 nm corresponding to the disappearance of the AnaCTDH(ECN) holoprotein due to carotenoid delivery into liposomes at different temperatures (shown in °C). (D) The corresponding temperature dependence of rate constants (Arrhenius plot). All values are presented as mean ± SD from three independent experiments.

Figure 4

Carotenoid delivery from AnaCTDH(ECN) holoprotein to liposomes studied by size-exclusion spectrochromatography. (A,B) Elution profiles of AnaCTDH(ECN), liposomes, or their mixture pre-incubated for 40 min at 30 °C to ensure carotenoid transfer, followed by either 460 nm (A) or 280 nm (B) absorbance, using a Superdex 200 Increase 5/150 column (GE Healthcare) operated at a 0.45 mL/min flow rate by a Varian ProStar 335 diode-array system with the full spectrum absorbance detection at a 0.45 mL/min flow rate. Apparent Mw for the protein peaks was estimated from column calibration using protein standards (indicated in kDa). Note that liposomes elute in the void volume and give significant light scattering. Arrows indicate the observed changes as the result of carotenoid transfer from AnaCTDH (ECN) to liposomes accompanied by AnaCTDH dissociation into apoprotein monomers. Note that protein does not migrate to the liposome fraction, indicating the absence of tight binding. The most typical result is shown. (C) Retrieving the ECN absorption spectrum taking into account light scattering from sample containing pure liposomes in the absence of ECN. (D) Normalized absorbance spectra of the initial AnaCTDH(ECN) preparation and ECN in liposomes at the end of the transfer (normalization to the peak absorbance as indicated).

Figure 5

Carotenoid delivery by AnaCTDH(ECN) to mammalian cells. (A) Raman spectra of ECN in AnaCTDH (ECN) (violet) and in liposome suspensions after incubation with AnaCTDH (ECN) (black). The spectrum was decomposed into a linear combination of Raman spectra of ECN in AnaCTDH (ECN) (30%) and in liposomes (70%) (see Figure A1). The orange line shows characteristic Raman spectra of ECN in HeLa cells. The grey line shows Raman spectrum of HeLa cells before incubation with AnaCTDH(ECN), with no bands in the carotenoid region observed. The inset shows a tube with a suspension of HeLa cells containing 5 μM of AnaCTDH(ECN) before (top) and after two hours of incubation (bottom) at 37 °C. (B) Overlay of microscopic images of a typical HeLa cell in transmitted light with the of ν₁ band intensity (at 1522 cm⁻¹ minus background at 1550 cm⁻¹) presented in pseudo colors. The dotted line shows a cross section through the Raman image, the corresponding distribution of the ν₁ band signature (intensity at 1522 cm⁻¹ minus background at 1550 cm⁻¹) is shown in (C). Effect of carotenoid delivery by AnaCTDH(ECN) into Tet21N cells on ROS production induced by antimycin A was determined by the fluorescence of the dyes DHE (D) and DCFDA (E) followed by flow cytometry. (F) The yield of ROS in Tet21N cells (purple, control sample), under oxidative stress induced by antimycin A treatment (red) and in cells pre-incubated with CTDH (ECN) and treated by antimycin A (green). Values are presented as mean and standard deviation. * indicates significance at the level of p < 0.05 compared with cells exposed to antimycin A only (statistical analysis performed with the t-Student’s test at significance level of 5%).

Figure 6

Proposed model for the AnaCTDH-mediated carotenoid uptake and delivery. Carotenoid uptake by AnaCTDH from the membrane is promoted by electrostatic interactions of the CTT and lipid head groups resulting in anchoring and formation of a transient complex between the membrane and the protein facing its carotenoid binding cavity towards the membrane. In such a complex, spontaneous translocation of the carotenoid into the hydrophobic part of the protein may be stabilized by the formation of the hydrogen bonds between the carotenoid keto group and the conserved Trp/Tyr residues of AnaCTDH. Due to a significant length of the carotenoid molecule, it requires two AnaCTDH subunits to isolate it from the solvent. The presence of two keto groups in CAN results in most efficient carotenoid binding in the AnaCTDH dimer, while ECN binding is apparently weaker. Since both types of AnaCTDH holoproteins can transfer carotenoids into other proteins, we postulate that intermediary, spontaneous monomerization of the protein dimer occurs regardless of the carotenoid type [24]. However, only AnaCTDH monomers in which keto group of ECN loses connection with the protein give the carotenoid an opportunity to escape another protein subunit and return to the membrane.