Detection of protein oxidation in endothelial cells by fluorescently labelled tyramine.

BACKGROUND: Endothelial cell injury mediated by activated polymorphonuclear leucocytes (PMN) occurs during inflammation or reperfusion after brain ischemia. Protein oxidation caused by activated PMN may lead to functional disturbances, degeneration and death of the endothelial cells. The aim of this study was to detect protein oxidation in endothelial cells induced by activated neutrophils by using a novel fluorescent probe. MATERIAL AND METHODS: Protein oxidation of Human Umbilical Vein Endothelial Cells (HUVEC) in culture was investigated by a 15-min incubation with human neutrophils activated by phorbol myristate acetate (PMA) in the presence of tyramine coupled to the succinimidyl ester of (fluorescein -5 (and-6)-carboxamido) hexanoic acid. Dityrosine bond formation as reflected by the linkage of the fluorescent tyramine to proteins was determined by Western-blotting. RESULTS: The oxidative burst generated by activated neutrophils induced dityrosine formation in the extracellular proteins (ECP) of HUVEC. Similar results were obtained, when horseradish peroxidase (HRP) was used for the induction of oxidative stress. However, when hydrogen peroxide (0.1 mM) was used, dityrosine formation was not detected. CONCLUSIONS: Fluorescently labelled tyramine is a powerful tool for the detection of ECP oxidation in endothelial cells. As long as the oxidation by the activated neutrophils is limited to ECP, the endothelial cells may be protected by antioxidants.

Detection of protein oxidation in rat-1 fibroblasts by fluorescently labeled tyramine.

Oxidative damage to proteins has been postulated as a major cause of various degenerative diseases including the loss of functional capacity during aging. A prominent target for oxidation by reactive oxygen species (ROS) is the tyrosine residue. Here we present a highly sensitive method for the detection of tyrosyl radical formation in cells. The method is based on the fluorescein-labeled tyrosine analogue, tyramine, which upon oxidation may couple to proteins carrying a tyrosyl radical. Coupling of the probe (denoted TyrFluo) to standard proteins could be induced by generating ROS with horseradish peroxidase/hydrogen peroxide, SIN-1 or with peroxides (cumene or hydrogen peroxide) in combination with a transition metal. TyrFluo added to rat-1 fibroblasts remained outside the cell, whereas the acetylated form (acetylTyrFluo) was membrane-permeable and accumulated in the cell. Exposure of the cells to oxidative stress in the presence of either TyrFluo or acetylTyrFluo gave a cellular labeling characteristic for each probe. Western blot analysis confirmed that each probe labeled a specific set of proteins. This new method for the detection of ROS-induced oxidation of proteins may mimic the tendency of oxidized proteins to form dityrosine bonds.

Peptide-based targeting of fluorophores to organelles in living cells.

Peptides carrying organelle-specific import or retention sequences can target the fluorophore BODIPY(581/591) to the nucleus, peroxisomes, endoplasmic reticulum (ER), or the trans-Golgi network (TGN). The peroxisomal peptide contains the PTS1 sequence AKL. For targeting to the ER or TGN, the peptides carry the retention sequences KDEL and SDYQRL, respectively. A peptide carrying the nuclear leader sequence of the simian virus SV40 large tumor antigen, KKKRK, was used to direct the fluorophore to the nucleus. The fluorescent peptides for peroxisomes, ER, and the TGN spontaneously incorporate into living fibroblasts at 37 degrees C and accumulate in their target organelles within minutes. The uptake is still significant at 4 degrees C, indicating that endocytosis is not required for internalization. The highly charged nuclear peptide (net charge +4) does not spontaneously internalize. However, by transient permeabilization of the plasma membrane, this fluorescent peptide was found to rapidly accumulate in the nucleus. These fluorescent peptides open new opportunities to follow various aspects of specific organelles such as their morphology, biogenesis, dynamics, degradation, and their internal parameters (pH, redox).

Peptide-based targeting of fluorophores to peroxisomes in living cells.

Peptides carrying different fluorophores can be designed to incorporate spontaneously into living cells when added to the medium. By incorporating the peroxisome-targeting sequence PTS1, the peptide is recognized by the protein-import machinery of peroxisomes and, as a result, can accumulate in these organelles. Depending on the cell type, an inhibitor of the multidrug-resistance protein might be required to ensure strong accumulation. In this update, we discuss the potential of these peptide-linked fluorophores in solving issues related to organelle function and dynamics.

Ratio-fluorescence microscopy of lipid oxidation in living cells using C11-BODIPY(581/591).

A ratio-fluorescence assay was developed for on-line localization and quantification of lipid oxidation in living cells. The assay explores the oxidative sensitivity of C11-BODIPY(581/591). Upon oxidation, the fluorescence of this fluorophore shifts from red to green. The probe incorporates readily into cellular membranes and is about twice as sensitive to oxidation as arachidonic acid. Using confocal microscopy, the cumene hydroperoxide-induced oxidation of C11-BODIPY(581/591) was visualized at the sub-cellular level in rat-1 fibroblasts. Preloading of the cells with tocopherol retarded this oxidation. The data demonstrate that C11-BODIPY(581/591) is a valuable tool to quantify lipid oxidation and anti-oxidant efficacy in single cells.

Outer membrane phospholipase A is dimeric in phospholipid bilayers: a cross-linking and fluorescence resonance energy transfer study.

In the cell, the activity of outer membrane phospholipase A (OMPLA) is strictly regulated to prevent uncontrolled breakdown of the membrane lipids. Previously, it has been shown that the enzymatic activity is modulated by reversible dimerization. The current studies were carried out to define the oligomeric state of OMPLA in a membrane and to investigate the activation process. Three single-cysteine variant proteins H26C, H234C, and S144C were produced and purified to homogeneity. Using maleimido-based homo-bifunctional cross-linking reagents, H26C could be efficiently cross-linked as assessed by SDS-PAGE, whereas S144C and H234C could not be cross-linked. These data suggest that residue 26 is located close to the dimer symmetry axis. H26C was specifically labeled with 5-({[(2-iodoacetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid and N,N'-dimethyl-N-(iodoacetyl)-N-(7-nitrobenz-2-oxa-1, 3-diazol-4-yl)ethylenediamine as the fluorescent energy donor and acceptor, respectively, and dimerization was investigated using fluorescence resonance energy transfer (FRET). Quenching of the donor in the presence of the acceptor demonstrated the dimeric nature of OMPLA, in agreement with cross-linking data. The observed FRET effect was dependent on the cofactor calcium, and the presence of substrate, indicating the specificity of the dimerization process. The labeled protein was reconstituted in phospholipid vesicles. In bilayers, OMPLA exhibited low activity and was dimeric as assessed by FRET. Addition of detergent resulted in a 70-fold increase in activity, while the protein remained dimeric. The results are discussed in terms of the activation of dimeric OMPLA due to changes in the physical state of the bilayer which occur upon perturbation of the membrane integrity.

Reorientational properties of fluorescent analogues of the protein kinase C cofactors diacylglycerol and phorbol ester.

The reorientational properties of the fluorescently labelled protein kinase C (PKC) cofactors diacylglycerol (DG) and phorbol ester (PMA) in vesicles and mixed micelles have been investigated using time-resolved polarised fluorescence. The sn-2 acyl chain of DG was replaced by diphenylhexatriene- (DPH) propionic acid, while a dansyl labelled analogue of phorbol ester was used. The extent of ordering of DPH-DG in vesicles turned out to be slightly different from that of the control choline lipid DPH-PC. Addition of PKC to vesicles containing 30 mole% brain PS considerably slowed down the DPH-DG anisotropy decay. This was not observed when DPH-DG was replaced by DPH-PC. Analysis of the fluorescence anisotropy decays of these DPH-lipids in micelles polyoxyethylene-9-laurylether mixed with 10 mole% of the essential phosphatidylserine allowed estimation of their lateral diffusion, orientation distribution and reorientational dynamics within the micelles. Addition of PKC resulted in a significantly slower decay of the fluorescence anisotropy of both DPH-DG and DPH-PC even in the absence of calcium, indicating a calcium independent complexation of PKC with the PS containing micelles. Addition of calcium resulted in a further reduction of the decay of anisotropy of DPH-DG but not of DPH-PC indicating that the Ca2+ dependent immobilisation is cofactor-specific. Similar specific interactions with PKC resulted in a slower decay of dansylated PMA when calcium and PS were present.

Monoclonal antibodies against two electron reduced riboflavin and a quantification of affinity constants for this oxygen-sensitive molecule.

In order to create a protein environment that binds preferentially to the two-electron reduced form of flavin, monoclonal antibodies have been raised against a reduced flavin derivative. Due to the low fluorescence quantum yield and visible light absorption and to the instability of reduced flavin in an aerobic environment, it is not possible to determine the affinities of these antibodies for two-electron-reduced flavin using standard techniques. Because of its sensitivity, time-resolved fluorescence can be used to overcome this problem. This technique has been applied to study the binding of two antibodies, an IgG1 and an IgM, to reduced riboflavin (1,5-dihydroriboflavin) and oxidized riboflavin (riboflavin). The affinity of the IgG1 is more than 80 times larger for 1,5-dihydroriboflavin than for riboflavin. From analysis of the dynamical parameters of the system it is apparent that the internal motion of 1,5-dihydroriboflavin bound to IgG1 is much more restricted than that of riboflavin. In contrast, the affinity of the IgM is only slightly higher for 1,5-dihydroriboflavin than for riboflavin and the flexibility of binding of both flavin redox states in the antigen binding site is almost similar.

Quantitative analysis of lipid-lipid and lipid-protein interactions in membranes by use of pyrene-labeled phosphoinositides.

The lateral and rotational dynamics of phosphoinositides and their interactions with proteins were characterized using pyrene-labeled lipid analogues. In these systems, the collision frequency of pyrene-labeled lipids was studied by monitoring the monomeric pyrene fluorescence yield as a function of their mole fraction in the membranes. From this dependence, the lateral diffusion coefficient and a repulsion factor between two pyrene phosphoinositides could be estimated by applying an extended form of the Milling Crowd model [Eisinger, J., Flores, J., & Petersen, W. P. (1986) Biophys. J. 49, 987-1001]. The repulsion appeared to be highly dependent on the amount of negative charge of the lipid headgroups. From experiments with dioleoylphosphatidylcholine vesicles containing band 3 protein, the fraction of lipid molecules bound to this protein and the minimum number of sites possessing affinity for phosphatidylinositol-4-phosphate could be approximately estimated. The results of this study indicate that phosphoinositides are located preferentially adjacent to band 3. Intramolecular excimer formation of dipyrene-labeled phosphatidylcholine, phosphatidylinositol, and phosphatidylinositol-4-phosphate yielded information about the acyl chain dynamics of lipids surrounding the protein and of lipids in the bulk membrane. Time-resolved measurements of the pyrene fluorescence anisotropy showed that in membranes of resealed erythrocyte ghost cells the rotational freedom of pyrene-labeled phosphatidylinositol-4,5-bisphosphate is smaller than that of pyrene-labeled phosphatidylcholine. In contrast, no significant differences could be detected when these pyrene lipids were dispersed in dioleoylphosphatidylcholine membranes. It is proposed that the nonrandom distribution of the phosphoinositides induced by lipid-lipid repulsion and protein-lipid attraction will have a profound effect on the phospholipase C-catalyzed hydrolysis of the phosphoinositides into second-messenger molecules.

The interaction between protein kinase C and lipid cofactors studied by simultaneous observation of lipid and protein fluorescence.

The interaction of protein kinase C (PKC) with lipids was probed by a dual approach. Pyrene-labeled lipid analogues of diacylglycerol, phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylinositol 4-phosphate (PIP), and phosphatidylcholine (PC) were used both as acceptors of tryptophan excitation energy of PKC and as membrane probes for intra- and intermolecular lipid chain collisions by measuring the ratio of excimer-to-monomer fluorescence intensity (EM). Both in micelles of polyoxyethylene 9-lauryl ether and in dioleoyl-PC vesicles, interaction of PKC with monopyrenyl PS (pyr-PS) in the absence of calcium resulted in a relatively slow decrease of the EM value. This effect on the lipid dynamics was accompanied by quenching of the tryptophan fluorescence of PKC. Addition of calcium resulted in a rapid further decrease of the EM ratio of pyr-PS and in additional quenching of the tryptophan fluorescence. When 4 mol % of pyr-PS was replaced by 0.5 mol % of dipyrenyl-labeled diacylglycerol a decrease of the intramolecular excimer formation rate and tryptophan fluorescence could only be detected in the presence of calcium and PS. Strong binding was also observed with dipyrenyl-labeled PIP (dipyr-PIP), but not with the other dipyrenyl-labeled lipids: PI, PS, or PC. In addition, the EM ratios of dipyr-PIP were not affected by phorbol 12-myristate 13-acetate, indicating that phorbol 12-myristate 13-acetate and dipyr-PIP can bind simultaneously to PKC.

Fluorescence dynamics of diphenyl-1,3,5-hexatriene-labeled phospholipids in bilayer membranes.

A comparative study of the dynamical fluorescence properties of three phosphatidylcholines having a diphenyl-1,3,5-hexatriene (DPH) group attached at different depths from the head group incorporated into membrane vesicles has been carried out. The probes were covalently attached to the sn-2 position of the glycerol part of the phosphatidylcholine via either carboxyl, ethyl or propanoyl links. The vesicles were composed of either dimyristoylphosphatidylcholine or dipalmitoylphosphatidylcholine. The experimental time-resolved polarized fluorescence data of the probes were analysed by two different methods: maximum entropy and global analysis. Distributed fluorescence lifetimes and correlation times of the DPH derivatives were obtained with the maximum entropy method. All DPH derivatives exhibited a bimodal distribution of fluorescence lifetimes with a dependence of the lifetime peak positions on the lipid phase, confirming previous data in the literature. The anisotropic rotational dynamics of the DPH moieties in the membranes could be described by several distributed correlation times. In the fluid phase of the membrane the residual anisotropy of free DPH became very small in contrast with those of the other probes, indicating that restriction of probe rotation is mainly imposed by the molecular geometry of the lipid probes. A two-dimensional analysis using the maximum entropy method demonstrated that both rotational correlation times were associated with the same set of fluorescence lifetimes. Global analysis of the data sets according to the general rotational diffusion model yielded weighted orientational distributions. Unexpectedly, a component of the DPH moiety oriented parallel to the membrane surface was obtained in the orientational distributions of the DPH lipids (as was reported earlier for DPH and TMA-DPH), which seems at variance with the geometric constraints imposed by the headgroups.

Steady-state fluorescence studies on lipase-vesicle interactions.

The interaction of lipase from Candida cylindracea (CCL) with positively charged polymerizable surfactant vesicles was studied by the use of steady-state fluorescence techniques. The phase transition of vesicles composed of nonpolymerized and polymerized N-allylbis[2-(hexadecanoyloxy)ethyl]methylammonium bromide (ABHEMA Br) was determined in the absence of lipase, by measuring the change in fluorescence anisotropy of the membrane probe 1,6-diphenyl-1,3,5-hexatriene (DPH). The phase transition temperature for nonpolymerized vesicles is 49 degrees C and for the polymerized analogues 45 degrees C. Fluorescence anisotropy and resonance energy transfer measurements were used to illustrate the incorporation of the lipase in the vesicle membrane. These studies demonstrated that CCL is incorporated into the hydrophobic bilayer of the vesicle. By using an interfacial membrane probe 1-[4-(trimethylammonium)phenyl]-6-phenyl-1,3,5-hexatriene p-toluene sulphonate, TMA-DPH) and an internal membrane probe (DPH), it could be determined that the enzyme is incorporated more efficiently into nonpolymerized vesicles, and that the penetration of the enzyme into the bilayer is less deep in the case of the polymerized vesicles.

Fluorescence methods to study lipid-protein association: The interaction of protein kinase C with lipid-loaded mixed micelles.

The interaction of protein kinase C with lipids was studied in a mixed micellar system. Two fluorescence spectroscopic methods are presented with a different but complementary information content. Diffusion monitored by fluorescence correlation spectroscopy provides information on the interaction of the protein with the whole lipid aggregate. Resonance energy transfer from tryptophans to pyrene-labeled lipids monitored by time-correlated single-photon counting supplies information on the interaction of the protein with specific lipid cofactors within the micelle. The results can be extended to postulate new mechanisms for the activation of protein kinase C by the signal transduction cascades in the cell. Both fluorescence spectroscopic methods can be easily applied to other protein systems which interact with lipids.

Quantitation of the interaction of protein kinase C with diacylglycerol and phosphoinositides by time-resolved detection of resonance energy transfer.

Quantitative studies of the binding of protein kinase C (PKC) to lipid cofactors were performed by monitoring resonance energy transfer with time-resolved fluorescence techniques. For that purpose, diacylglycerol (DG), phosphatidylinositol 4,5-biphosphate (PIP2), phosphatidylinositol 4-phosphate (PIP), phosphatidylinositol (PI), phosphatidylcholine (PC), and phosphatidylserine (PS) were labeled with a pyrenyl decanoyl moiety at the sn-2 position of the lipid glycerol. These labeled lipids proved excellent energy acceptors of light-excited tryptophan residues in PKC. The quenching efficiency of the tryptophan fluorescence was determined as function of lipid probe concentration in mixed micelles consisting of poly(oxyethylene)-9-lauryl ether, PS, and various mole fractions of probe lipid. The experimental conditions and method of data analysis allowed the estimation of binding constants of single or multiple pyrene lipids to PKC. The affinity of PKC for inositide lipids increases in the order PI < PIP < PIP2. The affinity of PKC for PIP and PIP2 is higher than that for DG. Determination of PKC activity in the presence of labeled lipids and PS showed that only PIP2 and DG activate PKC. Double-labeling experiments suggest that PIP2 and DG are not able to bind simultaneously to PKC, indicating a reciprocal binding relationship of both cofactors. The results support the notion that, besides DG, PIP2 can be a primary activator of PKC.

The interaction of pyrene labeled diacylglycerol with protein kinase C in mixed micelles.

The binding of protein kinase C (PKC) to pyrene-labeled diacylglycerol (pDG) has been studied in a mixed micellar system by monitoring resonance energy transfer from excited tryptophans to pyrene with time-correlated single photon counting. The average lifetime of the excited state of the tryptophans in PKC showed a clear dependence on the mole percentage pDG in micelles in contrast with pyrene-labeled phosphatidylcholine (pPC). The binding data has been analyzed to a simple model which encompasses the size of the micelles and the binding constant of the pDG-PKC complex. From our data, though, these quantities cannot be determined independently. If we have no size information on the micelles we can determine a lower boundary of this quantity compatible with the data. When the micellar size is known, a binding constant for the DG-PKC complex can be extracted. The presented analytical approach can be applied to other systems in which lipid-protein interactions must be quantified.