Research article

Proteins containing oxidized amino acids induce apoptosis in human monocytes

Rachael A. Dunlop, Ulf T. Brunk, Kenneth J. Rodgers


Cellular deposits of oxidized and aggregated proteins are hallmarks of a variety of age-related disorders, but whether such proteins contribute to pathology is not well understood. We previously reported that oxidized proteins form lipofuscin/ceroid-like bodies with a lysosomal-type distribution and up-regulate the transcription and translation of proteolytic lysosomal enzymes in cultured J774 mouse macrophages. Given the recently identified role of lysosomes in the induction of apoptosis, we have extended our studies to explore a role for oxidized proteins in apoptosis. Oxidized proteins were biosynthetically generated in situ by substituting oxidized analogues for parent amino acids. Apoptosis was measured with Annexin-V/PI (propidium iodide), TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling), MMP (mitochondrial membrane permeabilization), caspase activation and cytochrome c release, and related to lysosomal membrane permeabilization. Synthesized proteins containing the tyrosine oxidation product L-DOPA (L-3,4-dihydroxyphenylalanine) were more potent inducers of apoptosis than proteins containing the phenylalanine oxidation product o-tyrosine. Apoptosis was dependent upon incorporation of oxidized residues, as indicated by complete abrogation in cultures incubated with the non-incorporation control D-DOPA (D-3,4-dihydroxyphenylalanine) or when incorporation was competed out by parent amino acids. The findings of the present study suggest that certain oxidized proteins could play an active role in the progression of age-related disorders by contributing to LMP (lysosomal membrane permeabilization)-initiated apoptosis and may have important implications for the long-term use of L-DOPA as a therapeutic agent in Parkinson's disease.

  • aging
  • apoptosis
  • DOPA
  • lysosome
  • oxidized protein


The deposition and accumulation of oxidized proteins has been implicated in a wide variety of age-related pathologies including neurodegenerative diseases such as Alzheimer's and Parkinson's diseases (for a review see [1]), atherosclerosis [24], cataract [5], rheumatoid arthritis [6], and a variety of chronic inflammatory disorders, including diabetes [7]. Additionally, increased levels of oxidized proteins are observable in tissues from older animals, whereas caloric restriction, designed to increase life-span, results in a reduction in the extent of protein oxidation [8]. Although the primary mechanism for detoxification of oxidized proteins is their complete degradation, this process can be incomplete, as is evidenced by their accumulation [911].

Throughout their lifetime, cells are constantly exposed to ROS (reactive oxygen species), which may be formed by any one of a large number of physiological and non-physiological processes. The oxidative modification of proteins by ROS, usually catalysed by redox-active transition metals, results in a multitude of events, including dissociation of subunits, local or global unfolding, exposure of hydrophobic residues, aggregation and backbone fragmentation, and the modification of exposed amino acid side chains [12]. The susceptibility of amino acids in the peptide chain varies, although aromatic amino acids, such as tyrosine and phenylalanine, are particularly susceptible to modification by ROS. For example, in human carotid plaques [4] and cataractic lenses [13] elevated levels of oxidized phenylalanine (meta- and ortho-tyrosine) and tyrosine [DOPA (3,4-dihydroxyphenylalanine)] have been reported to be present in proteins. Despite the finding of elevated levels of oxidized proteins in a wide range of age-related diseases, their possible roles in the progression of these diseases has not been elucidated.

Unlike some other oxidative modifications to amino acids, DOPA is not a stable end-product, but a reactive species in its own right. Non-enzymatic generation of DOPA occurs via covalent addition of a hydroxyl radical (HO) to carbon three on the phenolic ring of tyrosine [14]. In the presence of reductants, such as glutathione and cysteine, the protein-bound DOPA moiety can redox-cycle to generate superoxide and its dimerization product hydrogen peroxide [15]. In the presence of redox-active transition metals, e.g. Fe(II), the highly aggressive and peroxidatively active HO can then be formed by Fenton-type reactions [12,16].

We previously described a novel approach to generate aggregates of proteins containing oxidized amino acids in cultured cells by the (mis)incorporation of oxidized amino acids via protein synthesis [17,18] and this approach has since been adopted by others [19]. Using this model, we reported that DOPA-containing proteins were inefficiently removed, formed autofluorescent, SDS-stable, high molecular mass aggregates, and increased the activity of the lysosomal proteases cathepsin B and cathepsin L more than 3-fold [20]. Autofluorescent DOPA-containing aggregates exhibited a distinct perinuclear punctuate pattern, suggesting a lysosomal localization. This was consistent with the observed up-regulation of lysosomal enzymes, which supposedly is a response to the accumulation of non-degradable materials within the lysosomal compartment. Conversely, o-tyrosine-containing proteins were rapidly removed, were not detected in lysosomes, and did not affect the activity of proteasomes or the activity of cathepsins B or L [20].

Owing to their high concentration of low mass iron, a result of autophagic degradation of ferruginous materials such as ferritin and mitochondrial complexes, lysosomes are particularly sensitive to oxidative stress [16]. A role for lysosomes in the induction of apoptosis (e.g. following oxidative stress) has previously been characterized [21]. Intralysosomal iron redox-cycling during periods of oxidative stress leads to the formation of HO that damage lysosomal membrane phospholipids, resulting in LMP (lysosomal membrane permeabilization) and the release of redox-active iron as well as lysosomal cathepsins [22]. Cathepsin B can activate Bax/Bid with consequent creation of pores in the outer mitochondrial membrane resulting in MMP (mitochondrial membrane permeabilization) and release of cytochrome c with activation of the intrinsic mitochondrial apoptotic pathway (reviewed in [23]).

Based on our previous findings concerning the inefficient cellular catabolism of DOPA-containing proteins [18,20,24], we hypothesized that their propensity to lysosomal accumulation might contribute to the onset of apoptosis by inducing LMP, thus implicating oxidized proteins in the progression of diseases where deposition of oxidized proteins is a feature.



D- and L-DOPA, D-L-o-tyrosine, z-VAD-fmk (z-Val-Ala-Aspfluoromethylketone) and AO (Acridine Orange) were purchased from Sigma. The InnoCyte™ flow cytometric cytochrome c release kit (catalogue number CBA077) was purchased from Calbiochem, the Annexin V–FITC apoptosis detection kit I (catalogue number 556547) was purchased from BD Pharmingen™, and the MitoProbe™ JC-1 assay kit for flow cytometry (catalogue number M34152) and APO-BrdU™ (BrdU is bromodeoxyuridine) TUNEL [TdT (terminal deoxynucleotidyltransferase)-mediated dUTP nick-end labelling] assay kit (catalogue number A23210) were from Invitrogen. Antibodies against pro- and active caspase 3 were purchased from Abcam. Ac-DEVD-AMC (N-acetyl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin) to measure caspase activity was purchased from BD Biosciences. All aqueous solutions and buffers were prepared using water filtered through a four-stage Milli-Q system (Millipore). EMEM (Eagle's minimal essential medium) deficient in tyrosine, phenylalanine and Phenol Red was from JRH Biosciences. All other chemicals, solvents and chromatographic materials were of analytical reagent or cell-culture grade.

Cell cultures

The following cell lines were used in this study: THP1 cells, a human monocytic line; SH-SY5Y, a human neuroblastoma cell line; and MRC5s, human lung fibroblasts. Cells were cultured in DMEM (Dulbecco's modified Eagle's medium; Sigma) supplemented with 10% FCS (fetal calf serum; Invitrogen) under standard conditions. For experiments, cells were washed three times in PBS by centrifugation (2500 g for 5 min) and re-suspended in Phenol-Red-free EMEM deficient in phenylalanine and tyrosine at a density of 5×105 cells/ml. Adherent cells (SH-SY5Y and MRC5s) were allowed to adhere overnight. THP1 cells were treated immediately.

The in situ synthesis of proteins containing L-DOPA has been described previously [17,20]. Briefly, cultures were incubated for up to 24 h in EMEM deficient in phenylalanine and tyrosine and supplemented with 500 μM oxidized amino acids (L-DOPA, D-DOPA and D,L-o-tyrosine). Control cells were grown in the same medium in the absence of oxidized amino acids. The medium used for cultures incubated in the presence of parent amino acids was EMEM supplemented with 390 μM phenylalanine and 400 μM tyrosine.

Binding of Annexin V to PS (phosphatidylserine) exposed on the plasma membrane

PS externalization, an early-stage apoptosis event, was assayed by the binding of fluorescently labelled (FITC) Annexin V, a 35 kDa phospholipid-binding protein that has a high affinity for PS. Late-stage apoptosis or necrosis was measured by simultaneous staining with PI (propidium iodide) using the BD Pharminigen™ Annexin V–FITC apoptosis detection kit according to the manufacturer's instructions. Cells were harvested by centrifugation (2500 g at 4 °C for 5 min) after being put on ice and washed three times in ice-cold PBS, after which pellets were resuspended in 500 μl of 1×Binding Buffer (0.01 M Hepes, 0.14 M NaCl and 2.5 mM CaCl2, pH 7.4).

A 100 μl fraction of the cell suspension was aliquoted into flow cytometry tubes and 5 μl of both PI and Annexin V–FITC were added. The tubes were then briefly vortex-mixed. The cell suspension was incubated for 15 min at room temperature (22 °C) min the dark. The percentage of cells undergoing early-stage apoptosis (Annexin V–FITC positive) or late-stage apoptosis/necrosis (Annexin V–FITC and PI positive) was measured with excitation at 488 nm and emission in FL1 (525 nm) for FITC and FL3 (610 nm) for PI in a Becton FC500 flow cytometer. Data were collected using BD FACScan software (version 10.2) and 10 000 cells were analysed per sample.


TUNEL is a common method to detect the DNA fragmentation that is typical of apoptosis. Nucleases degrade DNA into fragments of approx. 200 bp in length and these ‘nicks’ are identified by TdT, an enzyme that catalyses the addition of dUTPs that are secondarily labelled with a fluorescent marker. The level of fluorescence can be measured via flow cytometry using the APO-BrdU™ TUNEL assay kit. Cells were cultured as described above, washed in PBS, fixed for 15 min in ice-cold 1% formaldehyde freshly prepared from paraformaldehyde, washed again in PBS, resuspended in a 1:1 mix of PBS and 70% ethanol, and stored at −20 °C for 16 h. Cells were then incubated in the DNA-labelling solution mix (10 μl of reaction buffer, 0.75 μl of TdT enzyme, 8.0 μl of BrdUTP and 31.25 μl of distilled H2O per sample) for 60 min in a 37 °C water bath. Cells were then rinsed, concentrated by centrifugation (300 g), and exposed to Alexa Fluor® 488 dye-labelled anti-BrdU antibodies for 30 min, exposed to PI/RNAse A buffer, incubated for a further 30 min and finally analysed via flow cytometry as described above.

Measurement of the potential over the inner mitochondrial membrane (Ψm) by JC-1

The ΔΨm is the result of asymmetric distribution of protons and other ions on either side of the inner mitochondrial membrane, giving rise to both a chemical (pH) and an electrical gradient, both of which are essential for mitochondrial function. JC-1 is a metachromatic concentration-dependent fluorescent probe that exhibits potential-dependent accumulation in mitochondria as indicated by the emission of red fluorescence from healthy mitochondria with normal potential over their inner membranes, whereas organelles with reduced potential emit green fluorescence. The ΔΨm was measured using the MitoProbe™ JC-1 assay kit for flow cytometry (Invitrogen) according to the manufacturer's instructions. Briefly, cells were incubated (or not) for 4–26 h in EMEM with 500 μM oxidized amino acids as described previously, washed three times in PBS, exposed to 20 μM JC-1 for 30 min (37 °C; 5% CO2) and analysed by flow cytometry as described above.

Release of cytochrome c from mitochondria

Cells were incubated in EMEM deficient in phenylalanine and tyrosine with added oxidized amino acids as described above, washed three times in PBS and concentrated by centrifugation (2500 g for 5 min). The amount of cytochrome c remaining in the mitochondria was determined by fluorescent microscopy using the InnoCyte™ flow cytometric cytochrome c release kit from Calbiochem® according to the manufacturer's instructions. Briefly, cell pellets were permeabilized to enable the release of cytosolic cytochrome c, incubated on ice for 10 min, then fixed in 4% formaldehyde freshly prepared from paraformaldehyde. The pellets were washed, exposed to anti-cytochrome c antibody (for 60 min) and finally labelled with FITC-conjugated anti-IgG antibody (for 60 min). The cells were collected by centrifugation (2500 g for 5 min) and resuspended in 100 μl of wash buffer supplemented with 1 μl of 100 μg/ml DAPI (4′,6-diamidino-2-phenylindole). Cells were incubated on ice in the dark for 10 min. A drop of cell suspension was then placed on a microscope slide and viewed under UV and blue filter cubes in an Olympus IX71 inverted fluorescent microscope. The intensity of the FITC fluorescence was quantified using ImageJ software (version 1.42) and expressed as a function of total cell number, as determined by DAPI staining.

Determination of LMP using AO

AO is a metachromatic fluorescent cationic dye that emits red and green fluorescence at high and low concentrations respectively. It is also a lysosomotrophic agent, and in living cells it accumulates in lysosomes where it gives rise to a distinct red fluorescence when excited with blue or green light, whereas the lower AO concentrations in the cytosol and nuclei give rise to green fluorescence following blue light excitation [25]. Cells were cultured for 4–24 h as described above, and then stained for 15 min with 5 μg/ml AO in complete culture medium (DMEM with 10% FCS, 2 mM L-glutamine and 1% penicillin/streptomycin) at 37 °C. The AO-containing medium was removed and replaced with complete medium for a further 25 min incubation period (37 °C; 5% CO2). ‘Pale’ cells with decreased numbers of red lysosomes, indicating various degrees of lysosomal rupture, were quantified by flow cytometry, as described above, or fluorescent microscopy using a using a super wideband pass blue filter (λex, 420–480 nm; λem, 520 nm).

Western blots for pro- and active caspase

The conversion of pro-caspase 3 to its active form was measured by Western blotting. Protein (50 μg) was resolved via SDS/PAGE using Invitrogen™ NuPAGE gels and transferred to nitrocellulose with the iBlot™ gel transfer stacks system (Invitrogen) according to the manufacturer's instructions. The membrane was blocked with 4% skimmed milk in 0.1% TBS-T [100% TBS-T (Tris-buffered saline containing Tween) is 8.8 g of NaCl, 0.2 g of KCl, 3 g of Tris base and 500 μl of Tween 20, dissolved into 1 litre of MilliQ water and adjusted to pH 7.4 (working solution is 0.1% TBS-T in MilliQ water)] with 1% goat serum overnight at 4 °C. The primary antibodies, polyclonal antibodies against the pro- and active forms of caspase 3 (Abcam), were incubated with the membrane overnight at 4 °C at a dilution of 1:1000 in 0.1% TBS-T with 1% goat serum and 4% skimmed milk. Following three washes with 4% skimmed milk in 0.1% TBS-T, the membrane was incubated with the secondary antibody [goat anti-rabbit IgG peroxidase heavy and light chain (Vector Laboratories)] at a dilution of 1:333 for 1 h at room temperature. Finally, the membranes were washed three times in 0.1% TBS-T and visualized with DAB (diaminobenzidine). Images were acquired on a ChemiDoc using Quantity One, 1D Analysis Software, version 4.6.6 (Bio-Rad).

Caspase 3 activity

Caspase 3 is a key effector caspase that is activated by several independent pro-apoptotic mechanisms and its activity serves as a good integral indicator of apoptosis. Cultures were washed once in PBS, pelleted by centrifugation (2500 g for 5 min), then snap-frozen in liquid nitrogen and stored at −80 °C until required. The pellets were lysed in lysis buffer [consisting of 100 mM Hepes (pH 7.25), 10% sucrose, 0.1% CHAPS, 0.4% Nonidet P40 and 2 mM DTT (dithiothreitol)] with three freeze–thaw cycles in liquid nitrogen and a water bath at 37 °C. Lysates were centrifuged at 10000 g for 10 min at 4 °C to remove particulates, the supernatant was collected, and the protein concentration determined using the Bio-Rad Coomassie protein assay method [26]. Lysates were diluted in lysis buffer to allow for a final concentration of 5 μg of protein per well in a 96-well plate. For every part lysate (loaded in triplicate), three parts lysis buffer were added, followed by 5 μl of 1 mM Ac-DEVD-AMC. Changes in fluorescence were read on a CytoFluor fluorescent plate reader (λex, 360/40; λem, 460/40; gain 60). Caspase activity was determined by the change in FU (fluorescence units)/min per mg of protein and expressed as a percentage of the values from control lysates.

Statistical analysis

Statistical comparisons were made using one-way ANOVA followed by a Tukey's multiple comparison test and Student's unpaired two tailed t tests in GraphPad Prism (version 4.0c). P<0.05 was considered significant.


L-DOPA- and o-tyrosine-containing proteins induce apoptosis in THP1 cells

THP1 cells were incubated for up to 24 h in medium containing L-DOPA, D-DOPA or o-tyrosine, and apoptosis was measured using the Annexin V–FITC plus PI binding assay. No evidence of increased apoptosis was observed in cells incubated in complete medium (Figure 1a). However, when THP1 cells were exposed to L-DOPA or o-tyrosine in medium deficient in tyrosine and phenylalanine, to allow for the incorporation of oxidized amino acids into newly synthesized proteins [17], there was a significant increase in apoptosis as measured by Annexin V–FITC/PI binding (Figures 1b and 1c) and also by the TUNEL assay to identify DNA fragmentation (Figure 1d). The effect of L-DOPA on apoptosis was time-dependent as shown in Figure 1(c), becoming significant at t=18 h. In all experiments, there was no increase in nuclear PI staining, indicating that oxidized amino acids did not induce post-apoptotic necrosis in THP1 cells within 24 h (results not shown).

Figure 1 Proteins containing oxidized amino acids induce apoptosis in THP1 cells

Cultures were incubated for 24 h in complete medium (a) or in medium deficient in tyrosine and phenylalanine in the presence or absence of 500 μM oxidized amino acids (bd). Cells were also incubated in medium deficient in tyrosine and phenylalanine with 1 mM of the lysosomotropic detergent LeuLeuOMe for 4 h and 24 h and L-DOPA for 24 h (e). The cells were assayed for apoptosis by Annexin V–FITC/PI binding (a, b, c and e) and the TUNEL assay (d) using flow cytometry (10 000 events). Values are means+S.D. from a minimum of three independent experiments. Statistical significance was calculated compared with control cells and D-DOPA; *P<0.05; **P<0.01; ***P<0.001. In (e), L-DOPA was significantly different to control, D-DOPA, and LeuLeuOMe t=4 and t=24 (P<0.001). LeuLeuOMe t=4 was significantly different from LeuLeuOMe t=24 (P<0.001).

Under identical conditions D-DOPA, which can enter cells [20] but is not a substrate for mammalian protein synthesis, did not affect the Annexin V–FITC/PI-binding or TUNEL staining (Figures 1b, 1c and 1d), relative to the untreated control cells. Thus incorporation of oxidized amino acid into proteins was essential for the induction of apoptosis.

For comparison with a commonly used lysosomotrophic and apoptogenic compound [27], THP1 cells were incubated with L-DOPA for 24 h and LeuLeuOMe (L-leucine-L-leucine methyl ester) for up to 24 h. As shown in Figure 1(e), 1 mM LeuLeuOMe was significantly (P<0.001) more apoptogenic than L-DOPA at both t=4 and t=24 h. Apoptosis induced by LeuLeuOMe increased significantly after 24 h when compared with 4 h.

Proteins containing o-tyrosine and L-DOPA induce LMP

Having established that proteins containing oxidized amino acids can induce apoptosis in human THP1 cells, we investigated the mechanism(s) by which this may occur.

Recent evidence has implicated the lysosomes in apoptosis by a mechanism involving destabilization of the membrane (LMP) and the release of pro-apoptotic proteases (for a review see [23]).

We have previously reported that L-DOPA- (but not o-tyrosine-) containing proteins accumulate as proteolytically resistant, SDS-stable, high-molecular-mass aggregates, evidently within lysosomes, and increase the activity of the lysosomal proteases cathepsins B and L [20]. The pattern of autofluorescence that resulted from L-DOPA incorporation suggested localization of the aggregates to the lysosomal compartment. Considering the propensity for the DOPA moiety (both free and protein-bound) to redox-cycle [15], and the capacity of lysosomal redox-active iron to catalyse the formation of aggressive radicals [16], we proposed that DOPA-containing proteins were capable of inducing LMP. To test this, THP1 human monocytes were cultured with L-DOPA, D-DOPA or o-tyrosine for up to 24 h in a medium free of tyrosine and phenylalanine. The presence of intact stable lysosomes was determined using the metachromatic fluorescent probe AO and the AO uptake method [28].

As shown in Figure 2(a), there was an increased number of ‘pale’ cells in cultures synthesizing L-DOPA-containing proteins compared with those incubated in medium containing D-DOPA or medium without added oxidized amino acids (control). ‘Pale’ cells are indicative of cells with lysosomes that have become damaged and ‘leaky’ hence, instead of concentrating within the lysosomes, AO distributes throughout the cytosol, resulting in cells with a reduced number of red lysosomes and enhanced green cytosolic fluorescence. For L-DOPA, there was a significant increase in LMP at 4 h, which reached over 40% of the total number of cells after 24 h (Figure 2a). Conversely, for o-tyrosine LMP did not become significant until 20 h (Figure 2b). No increase in LMP was found in cells exposed to D-DOPA (Figure 2a) indicating a requirement for the incorporation of DOPA into proteins in order for LMP to occur and, further, that free DOPA did not contribute significantly to apoptosis. Representative flow cytometry histograms for 24 h are shown in Figures 2(c) (D-DOPA), 2(d) (o-tyrosine) and 2(e) (L-DOPA).

Figure 2 Proteins containing oxidized amino acids induce LMP

THP1 cells were incubated in the presence or absence of oxidized amino acids for up to 24 h, then stained with 5 μg/ml AO at otherwise standard culture conditions, rinsed in PBS and kept for a further 25 min in complete culture medium (DMEM). Increases in the number of ‘pale’ cells, indicative of ‘leaky’ lysosomes, were measured with the AO uptake method and quantified by flow cytometry (10000 events). When THP1 cells were incubated with L-DOPA, LMP increased over time and was significantly enhanced after 4 h (a; P<0.01 compared with control cells). For o-tyrosine, significant increases in LMP did not occur until 20 h (b; P<0.001 compared with 4 h o-tyrosine). Values are expressed as means+S.D. from a minimum of three independent experiments. Samples that show a significant increase in ‘pale’ cells are indicated with asterisks; *P<0.05; **P<0.01; ***P<0.001). Representative histograms and scatter plots from flow cytometric analyses following exposure for 24 h to (c) D-DOPA, (d) o-tyrosine and (e) L-DOPA. The emergent peaks of cells with reduced red fluorescence, appearing to the left of the major fluorescence peaks, represent cells with damaged lysosomes. (c) Shows a small population of ‘pale’ apoptotic cells 24 h after exposure to D-DOPA. (d) Shows a moderate number of ‘pale’ apoptotic cells after exposure to o-tyrosine. (e) Shows a major number of ‘pale’ cells following exposure to L-DOPA.

It should be pointed out that ‘pale’ cells are broadly defined as cells with any degree of increased LMP and, as such, some of these cells are only slightly affected, whereas others show extensive lysosomal leakage, similar to those treated with L-DOPA. To determine whether LMP induced by L-DOPA occurred in other cell types apart from THP1 cells, we repeated these experiments in both human lung fibroblasts (MRC5 cells) and a human neuroblastoma cell line (SH-SY5Ycells; see Supplementary Material at We report that these cell types were also susceptible to LMP to differing degrees when cultured in the presence of L-DOPA.

L-DOPA-containing proteins induce loss of potential over the inner mitochondrial membrane (Δψm)

Following a 24 h incubation of THP1 cells with D-DOPA, L-DOPA or o-tyrosine in medium deficient in both tyrosine and phenylalanine, we observed a 7-fold increase in cells with depolarized (green) mitochondria in cultures exposed to L-DOPA (Figures 3a and 3h), and a 4-fold increase in those exposed to o-tyrosine (Figures 3a and 3g) compared with untreated controls. These effects were time-dependent, with a significant loss of MMP at 8 h for L-DOPA (Figure 3b) which peaked by 18 h.

Figure 3 Proteins containing oxidized amino acids interfere with the potential over the inner mitochondrial membrane (ΔΨm)

THP1 cells were incubated in the presence or absence of 500 μM oxidized amino acids in medium deficient in phenylalanine and tyrosine (a, b and d) or in complete medium (c) for 4–26 h. (eh) Representative scatter plots for flow cytometric analyses; (e) control 24 h, (f) 24 h D-DOPA, (g) 24 h o-tyrosine and (h) 24 h L-DOPA. Cells were washed in PBS, then incubated with 20 μM JC-1 for 30 min (37 °C; 5% CO2) and analysed by flow cytometry (argon laser; FL1 and FL2 channels) using a Becton flow cytometer FC500 (10 000 events, data collected using the CXP software). Values (means+S.D.) are from a minimum of three independent experiments. Statistical significance was calculated compared with control cells and D-DOPA-treated cells for (a) and (c) using one-way ANOVA and Student's unpaired two-tailed t test, and one-way ANOVA compared with 4 h for (b) and (d); **P<0.01; ***P<0.001.

The percentage of cells with depolarized mitochondria in cells treated with o-tyrosine reached significance much later (at 20 h) and remained high until measurements were stopped at 26 h (Figure 3d). These effects were dependent upon protein synthesis as indicated by no measurable MMP when the ‘non-incorporating’ control D-DOPA was present (Figures 3a and 3b) or when the parent amino acids (Figure 3c) were present in the medium.

Representative scatter plots for control and D-DOPA at 24 h, are shown in Figures 3(e) and 3(f) respectively.

L-DOPA-induced LMP was independent of caspase activity

To confirm that the LMP observed in cells treated with L-DOPA was not a consequence of an early release of caspases, we measured LMP in the presence of a pan-caspase inhibitor. We selected 5 h as a time point when there was significant LMP but before any significant MMP and well before we had observed cytochrome c release (see Table 1). If any stray caspases were contributing to LMP, the presence of a pan-caspase inhibitor would abrogate the effect. When we co-incubated cells with 50 μM z-VAD-fmk and 500 μM L-DOPA and then measured LMP, we saw no significant reduction in LMP, indicating that caspases were not contributing to LMP at this time (Figure 4).

View this table:
Table 1 Time course of events for apoptosis induced by L-DOPA-containing proteins

THP1 cells were incubated with 500 μM L-DOPA for up to 24 h and measures for apoptosis were conducted. LMP (measured with AO) preceded MMP (JC-1), which in turn preceded apoptosis (Annexin V–FITC and TUNEL). + denotes a significant change above time-matched untreated controls and D-DOPA treated cells. +P<0.05, ++P<0.01, +++P<0.001.

Figure 4 LMP in cultures treated with L-DOPA is independent of caspase activity

Cells were treated as described in Figure 3, but incubated in the presence and absence of 50 μM z-VAD-fmk for 5 h, then analysed for LMP. No significant difference between L-DOPA and L-DOPA co-incubated with z-VAD-fmk indicates no contribution from caspases to LMP at this time point, thus LMP precedes MMP and the activation of caspases. L-DOPA (***P<0.001) and L-DOPA plus z-VAD-fmk (**P<0.01) were significantly different from all other treatments. NS, not significant.

L-DOPA-containing proteins induce MMP and release of cytochrome c from the mitochondria

Given the propensity of DOPA-containing proteins to interfere with Δψm, we next assessed whether this was sufficient to cause the release of cytochrome c. Depending on the degree of insult, changes in Δψm may or may not, result in the release of the pro-apoptotic protein cytochrome c from the mitochondrial intermembrane space. However, sufficient imbalance can result in cytochrome c liberation, through pores in the outer membrane, and activation of the effector caspases 3 and 7 via the apoptosome (for a review see [29]). Thus we measured the amount of cytochrome c released from THP1 cells incubated with oxidized amino acids.

In cells treated with L-DOPA, a loss of mitochondrial cytochrome c was indicated by a reduction in green staining compared with cells exposed to D-DOPA (Figure 5). Quantification of fluorescent staining indicated a 30% decrease in the amount of cytochrome c remaining in the mitochondria of cells incubated with L-DOPA compared with those incubated with D-DOPA (Figure 5). See Supplementary Figure S1 at for fluorescent images.

Figure 5 Proteins containing L-DOPA induce the release of cytochrome c from mitochondria

THP1 cells were fixed and permeabilized to remove cytosolic cytochrome c, then stained with a FITC-tagged anti-cytochrome c antibody according to the manufacturer's protocol. Cell nuclei were stained with DAPI. Fluorescence was visualized with an Olympus IX71 inverted microscope at 400× magnification and intensity was quantified using ImageJ software. Values are expressed as the percentage of green fluorescence per cell. Data are expressed as means+S.D. from at least three independent experiments. Statistical significance was calculated relative to cells treated with D-DOPA; ***P<0.001.

Proteins that contain L-DOPA or o-tyrosine activate caspase 3 in THP1 cells

Among the irreversible steps in apoptosis is the conversion of pro-caspase 3 into the active effector enzyme caspase 3, signalling the cells commitment to programmed cell death. After 24 h, the conversion of pro-caspase 3 into active caspase 3 was demonstrated in THP-1 cells that had incorporated L-DOPA and, to a lesser extent, o-tyrosine (Figure 6a) and this correlated with increased activity (Figure 6b). Incubation with D-DOPA under identical conditions did not result in caspase 3 activation (Figures 6a and 6b). Prevention of incorporation of L-DOPA and o-tyrosine into proteins, by competition with the parent amino acids phenylalanine and tyrosine, prevented caspase 3 activation (Figure 6c), providing further evidence for apoptosis being dependent upon incorporation of the oxidized amino acids into the protein chain.

Figure 6 Proteins containing oxidized amino acids activate caspase 3

Western blot analyses showed conversion of pro-caspase 3 (33 kDa) into active caspase 3 (19 kDa) in lysates from cells treated with oxidized amino acids. Capsase 3 was activated by o-tyrosine and L-DOPA, but not D-DOPA, after 24 h; (a) shows a representative blot. Caspase activity was measured as hydrolysis of a fluorigenic substrate in cultures incubated with oxidized amino acids in medium deficient in phenylalanine and tyrosine (b) or in complete culture medium (c). Values are expressed as the percentage of controls and reported as means+S.D. from a minimum of three independent experiments. Statistical significance was calculated compared with control cells and cells treated with D-DOPA; ***P<0.001.


In 1956, Harman [30] postulated that oxygen-derived free radicals cause cumulative damage to the cellular system, resulting in aging. More than 50 years later, the basic concept of a progressive accumulation of free-radical-damaged cellular components in aging cells and tissues remains valid [31]. In addition, it is now recognized that many age-related diseases in humans are characterized by the accumulation of oxidatively damaged proteins [32]. Although some focus has been directed towards lipofuscin/ceroid [33,34], as yet, a precise role for oxidized proteins in advancing the disease process has not been thoroughly characterized.

In the present study, we used a novel approach that allowed us to selectively generate oxidized proteins in situ, thereby minimizing peripheral oxidative stress, and for the first time demonstrate that these proteins can be potent inducers of apoptosis. We provide clear evidence that oxidized amino acids themselves do not induce apoptosis, but require incorporation into the polypeptide chain of a protein. We support this with studies in which we demonstrate that the D-isomer of DOPA does not induce apoptosis since, unlike the L-isomer, it cannot be incorporated into proteins by mammalian cells.

DOPA is one of the major oxidized species found in proteins in cells and tissues in age-related diseases such as atherosclerosis [4] and cataract [13]. We have previously demonstrated that DOPA-containing proteins accumulate in cells in vitro and generate autofluorescent aggregates with a pattern indicative of localization to the lysosomal compartment [20]. When isolated, the aggregates were found to be SDS-stable and of a high molecular mass [20]. Using the same culture conditions for the formation of DOPA-containing proteins in the present study, we showed that these proteins are potent inducers of apoptosis. This is the first study to report a role for oxidized proteins in apoptosis and to characterize the pathway(s) by which it proceeds.

We report that the presence of DOPA-containing proteins resulted in LMP in a human monocytic line (THP1), a human neuroblastoma cell line (SH-SY5Y) and human lung fibroblasts (MRC5). LMP-mediated MMP, with release of cytochrome c and subsequent caspase activation, appears to be a conserved mechanism for programmed cell death since it has now been demonstrated using different stimuli and in multiple cellular models [27]. LMP can be induced by a variety of distinct stimuli, including ROS generated intralysosomally during oxidative stress and membrane-damaging lysosomotropic compounds, such as 3-amino propanal and the detergents sphingosine, MSDH and LeuLeuOMe (reviewed in [22]). LeuLeuOMe accumulates in lysosomes via receptor-mediated endocytosis where it is converted into a membranolytic compound by dipeptidyl peptidase I, causing rapid loss of lysosomal membrane integrity [35]. Cell death is then catalysed by the translocation of redox-active iron and cathepsins into the cytosol and cleavage of the pro-apoptotic Bcl-2 family member Bid which, in turn, activates Bax with resulting pore formation in the outer mitochondrial membrane and ensuing release of cytochrome c [27,36]. We also observed rapid cell death with LeuLeuOMe (1 mM) in THP1 cells with 64% of the cells binding Annexin V–FITC after only a 4 h incubation period, which reached 98% by 24 h (Figure 1e). By contrast, proteins containing L-DOPA also induced apoptosis, but to a much smaller degree; at 24 h, only 39% of cells treated with L-DOPA were positive for Annexin V–FITC (Figure 1b).

When incorporated into proteins, both L-DOPA and o-tyrosine were capable of inducing apoptosis. The course of events for L-DOPA was LMP, followed by MMP, followed by apoptosis (Table 1). The impact on LMP and MMP, and hence apoptosis, by proteins containing o-tyrosine was moderate compared with that of L-DOPA. No changes in LMP or MMP were detected before 20 h following exposure to o-tyrosine and at this time both measurements showed significant changes. We have previously reported that proteins containing o-tyrosine are efficiently degraded and, as such, do not form autofluorescent aggregates to the same extent as DOPA-containing proteins [20]. It is thus evident that proteins containing o-tyrosine induce much less damage to the lysosomal compartment, which could be indicative of less oxidative stress (the LMP value at 24 h was 18% for o-tyrosine compared with 43% for L-DOPA). Although it is clear that o-tyrosine-containing proteins can damage lysosomal membranes, as indicated by the observed moderate (but significant) increase in LMP, this is unlikely to be a primary route of metabolism for these proteins because their degradation profile suggests that they are probably not processed by lysosomes [20]. Furthermore, since both LMP and MMP were significantly increased at 20 h, it was not possible to ascertain which came first. We propose that alternative pathways to apoptosis may be contributing to the toxicity of o-tyrosine, and this is a hypothesis we are currently exploring.

Since we have previously shown that DOPA-containing proteins are generated in L-DOPA-treated patients with Parkinson's disease [24], the present observation that DOPA-containing proteins induce apoptosis may be of particular relevance to the long-term use of L-DOPA in Parkinson's disease. Although L-DOPA is effective at ameliorating the clinical symptoms of this disorder, particularly during early stages of the disease, there have been long-standing concerns that it may also contribute to the progressive nature of neurodegeneration in the Parkinson's disease brain. In vitro studies have shown that L-DOPA is toxic to dopaminergic neurons [37], although data from animal studies investigating the possible neurotoxicity of L-DOPA are conflicting [38,39]. Autopsy studies using immunohistochemistry on Parkinson's disease substantia nigra have begun to define the pre-mitochondrial apoptosis signalling pathways in this disease and implicated a p53–GAPDH (glyceraldehyde-3-phosphate dehydrogenase)–BAX pathway in combination with the Fas receptor–FADD (Fas-associated death domain)–caspase 8–BAX pathway [40]. The question of whether L-DOPA negatively effects cell function and/or survival has been investigated in three large-scale human trials [4143]. Nevertheless, despite extensive discussion in the scientific literature, this conundrum remains unresolved. Interestingly, the fact that L-DOPA, which is an amino acid and a close structural analogue of tyrosine, can be a substrate for protein synthesis [44] has not previously been considered of importance for L-DOPA toxicity. The results of the present study support the view that DOPA-containing proteins generated in cells in L-DOPA-treated Parkinson's disease patients might eventually reach levels that could trigger apoptosis through a mechanism mediated by LMP and MMP. One could also predict that cells such as neurons, which are not capable of division and, therefore, do not have the ability to dilute protein aggregates and prevent their accumulation, could be particularly vulnerable to apoptosis induced by DOPA-containing proteins.

The findings of the present study indicate that oxidized proteins should not be considered as ‘innocent bystanders’ in aging and age-related diseases, but may function as second messengers mediating effects downstream of the initial oxidative insult. This raises the possibility that certain oxidized proteins could play an active role in the progression of age-related disorders, where such proteins are often found at relatively high levels. Interventions designed to bolster the proficiency of cells in preventing the accumulation and aggregation of oxidized proteins, for example increasing levels of antioxidants and heat-shock proteins [45] to prevent their accumulation or improving the ability to chelate redox-active lysosomal iron, may provide an alternative approach to the treatment of neurodegenerative diseases including Parkinson's disease and Alzheimer's disease.


Rachael Dunlop conducted all experimental work. Kenneth Rodgers and Ulf Brunk contributed to project and experimental design, offered advice and guidance on execution, and contributed to the writing of the manuscript.


This work was supported by The Heart Research Institute and the National Health and Medical Research Council Grant in Aid [grant number 358330].

Abbreviations: Ac-DEVD-AMC, N-acetyl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin; AO, Acridine Orange; BrdU, bromodeoxyuridine; DAPI, 4',6-diamidino-2-phenylindole; DMEM, Dulbecco's modified Eagle's medium; DOPA, 3,4-dihydroxyphenylalanine; EMEM, Eagle's minimal essential medium; FCS, fetal calf serum; LeuLeuOMe, L-leucine-L-leucine methyl ester; LMP, lysosomal membrane permeabilization; MMP, mitochondrial membrane permeabilization; PI, propidium iodide; PS, phosphatidylserine; ROS, reactive oxygen species; TdT, terminal deoxynucleotidyltransferase; TUNEL, TdT-mediated dUTP nick-end labelling; z-VAD-fmk, z-Val-Ala-Asp-fluoromethylketone


View Abstract