Article pubs.acs.org/est
Dual Roles of Dissolved Organic Matter as Sensitizer and Quencher in the Photooxidation of Tryptophan Elisabeth M.-L. Janssen, Paul R. Erickson, and Kristopher McNeill* Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, CH-8092, Zurich, Switzerland S Supporting Information *
ABSTRACT: The photooxidation processes of tryptophan (Trp) in the presence of dissolved organic matter (DOM) were identified and quantified by steady-state photolysis experiments, laser spectroscopy and kinetic modeling. In sunlight, Trp photooxidation is dominated by the reaction with excited triplet DOM (3DOM), accounting for approximately 50−70% of the total degradation, depending on the DOM concentration and source. Reaction with singlet oxygen and direct photolysis are secondary processes that are both still more important than the reaction with hydroxyl radical. Both direct photolysis and reaction with 3DOM form Trp radical cation (Trp•+) via Trp photoionization and direct oxidation, respectively. The Trp•+ can be converted back to Trp by suitable electron or hydrogen atom donors. Transient absorption spectroscopy shows that DOM itself and low-molecular-weight analogues of redox-active moieties can reduce the lifetime of photochemically produced Trp•+ and thus quench Trp degradation. This study demonstrates that DOM plays dual roles in the photodegradation of Trp acting as a sensitizer and quencher. The photochemistry of Trp and the participation of DOM have direct implications for photochemical reactions in extracellular proteins as well as for organic compounds in aquatic systems with similar photoionization processes.
■
species (e.g., 1O2 and •OH). Triplet DOM reactivity is mainly attributed to excitation of aromatic ketones, aldehydes or quinone moieties.12 While the influence of DOM on photooxidation of pollutants has been more intensely studied, the interaction of DOM with amino acids has received less attention thus far. Boreen et al. showed enhanced degradation of amino acids in the presence DOM, emphasizing the photosensitizing properties of DOM without exciting the amino acids themselves.13 Effects of DOM on the direct photolysis of Trp remain to be investigated. The quenching properties of DOM have been investigated for some aquatic pollutants. Aromatic amines degraded slower during irradiation experiments with DOM present, leading to the assumption that DOM can reduce photochemically produced radical intermediates.14 Further studies investigated the triplet-sensitized photolysis rates of aniline and sulfonamide antibiotics, which decreased in the presence of DOM.15,16 The quenching ability of DOM is thus likely linked to its redox properties, which are mainly attributed to hydroquinone and phenol moieties.17,18 Aeschbacher et al. have shown that the titrated phenol content in particular correlates with electron donating capacities of DOM isolates.18 These redox-active DOM moieties may be represented by mono- and poly-
INTRODUCTION Sunlight photolysis is one major abiotic degradation pathway of biomolecules in aquatic systems and a source of dissolved organic nitrogen.1,2 For proteins, only some amino acids (e.g., tryptophan and tyrosine) are capable of absorbing sunlight, and the indole side chain of tryptophan (Trp) is the dominant amino acid−based chromophore in proteins. Photooxidation processes of Trp have been investigated in detail, including direct photolysis and reactions with singlet oxygen (1O2) and hydroxyl radical (•OH).3−6 Upon absorption of light, photoionization of Trp, to give Trp radical cation (Trp•+), competes with thermal relaxation from the excited state and represents a major photochemical process with high quantum yield.7 The Trp radical cation deprotonates to the neutral radical (Trp•), which can be further oxidized.8 Photochemically produced triplet states of other organic molecules such as flavins or coumarin can oxidize Trp to also form the same radical intermediates as in direct photolysis.9,10 Quenching of both Trp•+ and Trp• has been observed by some antioxidants including alpha-tocopherol, ascorbic acid, and p-methoxyphenol using pulse radiolysis to monitor the short-lived Trp radicals.8,11 Under environmental conditions the contributions of each photodegradation pathway and the relevance of the quenching of radical intermediates have not been quantified yet. In the aquatic environment, dissolved organic matter (DOM) is ubiquitous and acts as a photosensitizer and also a quencher. The photosensitizing properties are due to production of triplet DOM (3DOM) and reactive oxygen © 2014 American Chemical Society
Received: Revised: Accepted: Published: 4916
January 30, 2014 April 4, 2014 April 7, 2014 April 7, 2014 dx.doi.org/10.1021/es500535a | Environ. Sci. Technol. 2014, 48, 4916−4924
Environmental Science & Technology
Article
tubes (10 cm, inner diameter 1.1 cm) were angled at 30°. For UV light exposure, cork-stoppered borosilicate tubes were used (Pyrex, 7.5 cm, inner diameter 1.0 cm) on a turntable inside a photoreactor (Rayonet; Southern New England Ultraviolet Co, six RPR-3000 Å bulbs). All test solutions contained FFA (40 μM) as a singlet oxygen probe as discussed below. Experiments were performed in the presence of DOM (5−20 mgC L−1) or cinnamic acid and coumarin derivatives (50 μM). To probe whether the reaction of Trp with 3DOM is quenchable, samples were irradiated in open borosilicate tubes (Pyrex, 7.5 cm, inner diameter 1.0 cm) with a Xe lamp (Newport, at 300 W) and a 320 nm cutoff filter to prevent direct photolysis of Trp. These samples contained SRFA(II) (10 mgC L−1) and degradation rates of Trp were assessed in the presence and absence hydroxylated cinnamic acid derivatives (50 μM) as model quenchers. Reaction with Singlet Oxygen. The reaction rate constant of Trp with 1O2 was measured by producing different concentrations of 1O2. The reaction solutions contained sensitizer (Rose Bengal, 1−3 μM), FFA (40 μM) and Trp (10 μM) at pH 7.5, irradiated (Xe lamp, Newport, at 300 W, 455 nm cutoff filter) in cork-stoppered, stirred, borosilicate tubes. The reaction rate constant of Trp with 1O2 determined relative to the previously measured FFA rate constant (8.3 × 107 M−1 s−1)22 was 3.43 × 107 (±0.02 × 107) M−1 s−1 and agrees with literature values.6 In the photodegradation experiments with sunlight and UVB light, the steady-state 1O2 concentrations, [1O2]ss, were determined by dividing the observed FFA degradation rate by the reaction rate constant of FFA with 1O2. The apparent degradation rate constants of Trp by reaction with 1O2 was calculated by multiplying these [1O2]ss with the Trp reaction rate constant for each experiment. Sample and Data Analysis. Tryptophan and FFA concentrations were measured by ultra high-pressure liquid chromatography (UPLC) separation with fluorescence detection (excitation 275 nm, emission at 350 nm) and UV detection (219 nm), respectively. The mobile phase was 75% acetate buffer (pH 5.9) and 25% acetonitrile with 0.15 mL min−1 flow rate on a C18 column (Aquity, BEH130 C18, 1.7 μm; 2.1 × 150 mm). The first-order degradation rate constant of Trp, kobs (s−1), was assessed by linear regression of ln(A/A0) versus irradiation time. All data reported herein was corrected for light screening by DOM, coumarin or cinnamic acid derivatives and details about screening factors are provided in the Supporting Information (SI) (Table S1−S2, Figure S1). Laser Spectroscopy. Measurements of lifetimes of Trp radical intermediates were performed with laser flash photolysis by adopting previous methods.4,7 A pump−probe transient absorbance system (EOS, Ultrafast Systems) was used for timeresolved absorbance experiments. Excitation pulses were generated by a regeneratively amplified Ti:sapphire ultrafast laser (Solstice, Newport Spectra-Physics). The tunable primary output was set to 795 nm (pulse width 18 MΩ cm, Barnstead nanopure System), experiments at pH 3.0 were carried out in nanopure water acidified with hydrochloric acid (Fluka), and experiments at pH 7.5 were carried out in 10 mM phosphate buffer from sodium phosphate dibasic dihydrate (Sigma-Aldrich, ≥ 99%) adjusted to 30 mM ionic strength with sodium chloride (Merck, ACD reagent grade). The following reagents were used as received: Tryptophan (Sigma-Aldrich, ≥ 98% reagent grade); caffeic acid, coumarin (Sigma, ≥ 98%); 6,7-dihydroxycoumarin, p-coumaric acid, 3-(4-hydroxyphenyl)propionic acid, 7-hydroxycoumarin, 3-methoxy-4-hydroxycinnamic (ferulic acid), 3,5-dimethoxy-4hydroxycinnamic acid (sinapinic acid) (Aldrich ≥98%); mcoumaric acid, trans-cinnamic acid (ACROS organics, ≥ 98%); 6-hydroxycoumarin, 6-hydroxy-7-methoxycoumarin (TCI, ≥ 98%); ammonium acetate (Prolabo), acetic acid, 2-propanol (Fluka); acetonitrile, methanol (Merck KGaA, ≥ 99.9), synthetic air (80% N2, 20% ± 1% O2), N2O (AlphaGaz). Furfuryl alcohol (Aldrich) was distilled prior to use. Tryptophan stock solutions were prepared daily in nanopure water. Stock solutions of the coumarin and cinnamic acid derivatives were prepared in methanol and the amount spiked into aqueous photolysis solutions resulted in 320 nm light to excite DOM but not Trp. If Trp•+ were formed in these experiments, the effectiveness of suitable quenchers should mirror observations in the laser spectroscopy experiments, in which Trp•+ was directly monitored. Tryptophan degraded in the presence of excited SRFA(II) and its degradation was quenched upon addition of hydroxylated cinnamic acids by up to 70%. Dihydroxy cinnamic acid (caffeic acid) and 6-hydroxy-7-methoxy cinnamic acid (ferulic acid) had a greater quenching effect than monohydroxylated cinnamic acid (p-cinnamic acid) (SI, Figure S12− S13). This trend of quenching efficiency of the reaction of Trp with 3DOM parallels observations for quenching of direct photolysis during steady-state experiments in UVB light and quenching rates, kq, Trp•+, assessed by laser spectroscopy. Hydroxylated cinnamic acids may also react with 3DOM and reduce Trp degradation by competition rather than quenching the Trp•+. The decay rate constant of 3DOM can be estimated as 5 × 104 s−1 based on lifetime estimates.37 Under aerated conditions it decays ten times faster, 5 × 105 s−1, due to quenching by oxygen (k(O2) ∼ 2 × 109 M−1 s−1 and 240 μM O2 concentration, see Table S3 in the SI). At a quencher concentration of 5 × 10−5 M (50 μM) and a diffusioncontrolled reaction rate constant of hydroxylated cinnamic acids with 3DOM, approximately 109 M−1 s−1, the steady-state 3 DOM concentration may be reduced by up to 10%. This competition effect alone cannot explain the decrease in Trp degradation by up to 70% achieved here. These observations strongly suggest that photooxidation of Trp by reaction with 3 DOM is indeed quenchable by redox-active compounds. Further, these quenching experiments support the assignment of Trp•+ as the key intermediate in the reaction of 3DOM with Trp.
model for SRFA(II) and is comparable to the values measured for model compounds discussed below. Data in Figure 5 show kq, Trp•+ of Trp•+ obtained by laser spectroscopy for all hydroxylated cinnamic acid and coumarin
Figure 5. Quenching rate constants, kq,Trp•+, of tryptophan radical cation, Trp•+, measured by time-resolved laser spectroscopy for cinnamic acid (A) and coumarin (B) derivatives. Error bars represent one standard deviation.
derivatives. All mono- and dihydroxylated compounds tested were able to quench Trp•+ with relatively high rate constants between 9.7 × 107 M−1 s−1 and 2.6 × 109 M−1 s−1. While monohydroxylated cinnamic acids show moderate quenching rate constants, the poly-substituted derivatives; caffeic acid, ferulic acid, and sinapinic acid show more efficient quenching, approaching the diffusion-controlled limit. During steady-state photolysis, caffeic acid did quench Trp photooxidation more effectively than ferulic and sinapinic acid, differing from the quenching rate trends obtained by the laser spectroscopy. One possible explanation is that coupling reactions can increase the number of electrons released per mole of caffeic acid, leading to more potent quenching during steady-state photolysis than expected.34 However, in laser spectroscopy, effects of downstream coupling are not expected because only the initial events are observed on the time scale of the measurement. In addition, no coupling reactions of caffeic acid were observed at low pH (pH 3) investigated by cyclic voltammetry.35 The speciation of cinnamic acid derivatives was not significantly affected between steady-state and laser spectroscopy experiments because the fraction of phenolates are relatively low at pH 7.5 (≤1.6%) and pH 3 (≤0.00005%). Quenching by nonhydroxylated cinnamic acid was not observed (
Dual Roles of Dissolved Organic Matter as Sensitizer and Quencher in the Photooxidation of Tryptophan Elisabeth M.-L. Janssen, Paul R. Erickson, and Kristopher McNeill* Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, CH-8092, Zurich, Switzerland S Supporting Information *
ABSTRACT: The photooxidation processes of tryptophan (Trp) in the presence of dissolved organic matter (DOM) were identified and quantified by steady-state photolysis experiments, laser spectroscopy and kinetic modeling. In sunlight, Trp photooxidation is dominated by the reaction with excited triplet DOM (3DOM), accounting for approximately 50−70% of the total degradation, depending on the DOM concentration and source. Reaction with singlet oxygen and direct photolysis are secondary processes that are both still more important than the reaction with hydroxyl radical. Both direct photolysis and reaction with 3DOM form Trp radical cation (Trp•+) via Trp photoionization and direct oxidation, respectively. The Trp•+ can be converted back to Trp by suitable electron or hydrogen atom donors. Transient absorption spectroscopy shows that DOM itself and low-molecular-weight analogues of redox-active moieties can reduce the lifetime of photochemically produced Trp•+ and thus quench Trp degradation. This study demonstrates that DOM plays dual roles in the photodegradation of Trp acting as a sensitizer and quencher. The photochemistry of Trp and the participation of DOM have direct implications for photochemical reactions in extracellular proteins as well as for organic compounds in aquatic systems with similar photoionization processes.
■
species (e.g., 1O2 and •OH). Triplet DOM reactivity is mainly attributed to excitation of aromatic ketones, aldehydes or quinone moieties.12 While the influence of DOM on photooxidation of pollutants has been more intensely studied, the interaction of DOM with amino acids has received less attention thus far. Boreen et al. showed enhanced degradation of amino acids in the presence DOM, emphasizing the photosensitizing properties of DOM without exciting the amino acids themselves.13 Effects of DOM on the direct photolysis of Trp remain to be investigated. The quenching properties of DOM have been investigated for some aquatic pollutants. Aromatic amines degraded slower during irradiation experiments with DOM present, leading to the assumption that DOM can reduce photochemically produced radical intermediates.14 Further studies investigated the triplet-sensitized photolysis rates of aniline and sulfonamide antibiotics, which decreased in the presence of DOM.15,16 The quenching ability of DOM is thus likely linked to its redox properties, which are mainly attributed to hydroquinone and phenol moieties.17,18 Aeschbacher et al. have shown that the titrated phenol content in particular correlates with electron donating capacities of DOM isolates.18 These redox-active DOM moieties may be represented by mono- and poly-
INTRODUCTION Sunlight photolysis is one major abiotic degradation pathway of biomolecules in aquatic systems and a source of dissolved organic nitrogen.1,2 For proteins, only some amino acids (e.g., tryptophan and tyrosine) are capable of absorbing sunlight, and the indole side chain of tryptophan (Trp) is the dominant amino acid−based chromophore in proteins. Photooxidation processes of Trp have been investigated in detail, including direct photolysis and reactions with singlet oxygen (1O2) and hydroxyl radical (•OH).3−6 Upon absorption of light, photoionization of Trp, to give Trp radical cation (Trp•+), competes with thermal relaxation from the excited state and represents a major photochemical process with high quantum yield.7 The Trp radical cation deprotonates to the neutral radical (Trp•), which can be further oxidized.8 Photochemically produced triplet states of other organic molecules such as flavins or coumarin can oxidize Trp to also form the same radical intermediates as in direct photolysis.9,10 Quenching of both Trp•+ and Trp• has been observed by some antioxidants including alpha-tocopherol, ascorbic acid, and p-methoxyphenol using pulse radiolysis to monitor the short-lived Trp radicals.8,11 Under environmental conditions the contributions of each photodegradation pathway and the relevance of the quenching of radical intermediates have not been quantified yet. In the aquatic environment, dissolved organic matter (DOM) is ubiquitous and acts as a photosensitizer and also a quencher. The photosensitizing properties are due to production of triplet DOM (3DOM) and reactive oxygen © 2014 American Chemical Society
Received: Revised: Accepted: Published: 4916
January 30, 2014 April 4, 2014 April 7, 2014 April 7, 2014 dx.doi.org/10.1021/es500535a | Environ. Sci. Technol. 2014, 48, 4916−4924
Environmental Science & Technology
Article
tubes (10 cm, inner diameter 1.1 cm) were angled at 30°. For UV light exposure, cork-stoppered borosilicate tubes were used (Pyrex, 7.5 cm, inner diameter 1.0 cm) on a turntable inside a photoreactor (Rayonet; Southern New England Ultraviolet Co, six RPR-3000 Å bulbs). All test solutions contained FFA (40 μM) as a singlet oxygen probe as discussed below. Experiments were performed in the presence of DOM (5−20 mgC L−1) or cinnamic acid and coumarin derivatives (50 μM). To probe whether the reaction of Trp with 3DOM is quenchable, samples were irradiated in open borosilicate tubes (Pyrex, 7.5 cm, inner diameter 1.0 cm) with a Xe lamp (Newport, at 300 W) and a 320 nm cutoff filter to prevent direct photolysis of Trp. These samples contained SRFA(II) (10 mgC L−1) and degradation rates of Trp were assessed in the presence and absence hydroxylated cinnamic acid derivatives (50 μM) as model quenchers. Reaction with Singlet Oxygen. The reaction rate constant of Trp with 1O2 was measured by producing different concentrations of 1O2. The reaction solutions contained sensitizer (Rose Bengal, 1−3 μM), FFA (40 μM) and Trp (10 μM) at pH 7.5, irradiated (Xe lamp, Newport, at 300 W, 455 nm cutoff filter) in cork-stoppered, stirred, borosilicate tubes. The reaction rate constant of Trp with 1O2 determined relative to the previously measured FFA rate constant (8.3 × 107 M−1 s−1)22 was 3.43 × 107 (±0.02 × 107) M−1 s−1 and agrees with literature values.6 In the photodegradation experiments with sunlight and UVB light, the steady-state 1O2 concentrations, [1O2]ss, were determined by dividing the observed FFA degradation rate by the reaction rate constant of FFA with 1O2. The apparent degradation rate constants of Trp by reaction with 1O2 was calculated by multiplying these [1O2]ss with the Trp reaction rate constant for each experiment. Sample and Data Analysis. Tryptophan and FFA concentrations were measured by ultra high-pressure liquid chromatography (UPLC) separation with fluorescence detection (excitation 275 nm, emission at 350 nm) and UV detection (219 nm), respectively. The mobile phase was 75% acetate buffer (pH 5.9) and 25% acetonitrile with 0.15 mL min−1 flow rate on a C18 column (Aquity, BEH130 C18, 1.7 μm; 2.1 × 150 mm). The first-order degradation rate constant of Trp, kobs (s−1), was assessed by linear regression of ln(A/A0) versus irradiation time. All data reported herein was corrected for light screening by DOM, coumarin or cinnamic acid derivatives and details about screening factors are provided in the Supporting Information (SI) (Table S1−S2, Figure S1). Laser Spectroscopy. Measurements of lifetimes of Trp radical intermediates were performed with laser flash photolysis by adopting previous methods.4,7 A pump−probe transient absorbance system (EOS, Ultrafast Systems) was used for timeresolved absorbance experiments. Excitation pulses were generated by a regeneratively amplified Ti:sapphire ultrafast laser (Solstice, Newport Spectra-Physics). The tunable primary output was set to 795 nm (pulse width 18 MΩ cm, Barnstead nanopure System), experiments at pH 3.0 were carried out in nanopure water acidified with hydrochloric acid (Fluka), and experiments at pH 7.5 were carried out in 10 mM phosphate buffer from sodium phosphate dibasic dihydrate (Sigma-Aldrich, ≥ 99%) adjusted to 30 mM ionic strength with sodium chloride (Merck, ACD reagent grade). The following reagents were used as received: Tryptophan (Sigma-Aldrich, ≥ 98% reagent grade); caffeic acid, coumarin (Sigma, ≥ 98%); 6,7-dihydroxycoumarin, p-coumaric acid, 3-(4-hydroxyphenyl)propionic acid, 7-hydroxycoumarin, 3-methoxy-4-hydroxycinnamic (ferulic acid), 3,5-dimethoxy-4hydroxycinnamic acid (sinapinic acid) (Aldrich ≥98%); mcoumaric acid, trans-cinnamic acid (ACROS organics, ≥ 98%); 6-hydroxycoumarin, 6-hydroxy-7-methoxycoumarin (TCI, ≥ 98%); ammonium acetate (Prolabo), acetic acid, 2-propanol (Fluka); acetonitrile, methanol (Merck KGaA, ≥ 99.9), synthetic air (80% N2, 20% ± 1% O2), N2O (AlphaGaz). Furfuryl alcohol (Aldrich) was distilled prior to use. Tryptophan stock solutions were prepared daily in nanopure water. Stock solutions of the coumarin and cinnamic acid derivatives were prepared in methanol and the amount spiked into aqueous photolysis solutions resulted in 320 nm light to excite DOM but not Trp. If Trp•+ were formed in these experiments, the effectiveness of suitable quenchers should mirror observations in the laser spectroscopy experiments, in which Trp•+ was directly monitored. Tryptophan degraded in the presence of excited SRFA(II) and its degradation was quenched upon addition of hydroxylated cinnamic acids by up to 70%. Dihydroxy cinnamic acid (caffeic acid) and 6-hydroxy-7-methoxy cinnamic acid (ferulic acid) had a greater quenching effect than monohydroxylated cinnamic acid (p-cinnamic acid) (SI, Figure S12− S13). This trend of quenching efficiency of the reaction of Trp with 3DOM parallels observations for quenching of direct photolysis during steady-state experiments in UVB light and quenching rates, kq, Trp•+, assessed by laser spectroscopy. Hydroxylated cinnamic acids may also react with 3DOM and reduce Trp degradation by competition rather than quenching the Trp•+. The decay rate constant of 3DOM can be estimated as 5 × 104 s−1 based on lifetime estimates.37 Under aerated conditions it decays ten times faster, 5 × 105 s−1, due to quenching by oxygen (k(O2) ∼ 2 × 109 M−1 s−1 and 240 μM O2 concentration, see Table S3 in the SI). At a quencher concentration of 5 × 10−5 M (50 μM) and a diffusioncontrolled reaction rate constant of hydroxylated cinnamic acids with 3DOM, approximately 109 M−1 s−1, the steady-state 3 DOM concentration may be reduced by up to 10%. This competition effect alone cannot explain the decrease in Trp degradation by up to 70% achieved here. These observations strongly suggest that photooxidation of Trp by reaction with 3 DOM is indeed quenchable by redox-active compounds. Further, these quenching experiments support the assignment of Trp•+ as the key intermediate in the reaction of 3DOM with Trp.
model for SRFA(II) and is comparable to the values measured for model compounds discussed below. Data in Figure 5 show kq, Trp•+ of Trp•+ obtained by laser spectroscopy for all hydroxylated cinnamic acid and coumarin
Figure 5. Quenching rate constants, kq,Trp•+, of tryptophan radical cation, Trp•+, measured by time-resolved laser spectroscopy for cinnamic acid (A) and coumarin (B) derivatives. Error bars represent one standard deviation.
derivatives. All mono- and dihydroxylated compounds tested were able to quench Trp•+ with relatively high rate constants between 9.7 × 107 M−1 s−1 and 2.6 × 109 M−1 s−1. While monohydroxylated cinnamic acids show moderate quenching rate constants, the poly-substituted derivatives; caffeic acid, ferulic acid, and sinapinic acid show more efficient quenching, approaching the diffusion-controlled limit. During steady-state photolysis, caffeic acid did quench Trp photooxidation more effectively than ferulic and sinapinic acid, differing from the quenching rate trends obtained by the laser spectroscopy. One possible explanation is that coupling reactions can increase the number of electrons released per mole of caffeic acid, leading to more potent quenching during steady-state photolysis than expected.34 However, in laser spectroscopy, effects of downstream coupling are not expected because only the initial events are observed on the time scale of the measurement. In addition, no coupling reactions of caffeic acid were observed at low pH (pH 3) investigated by cyclic voltammetry.35 The speciation of cinnamic acid derivatives was not significantly affected between steady-state and laser spectroscopy experiments because the fraction of phenolates are relatively low at pH 7.5 (≤1.6%) and pH 3 (≤0.00005%). Quenching by nonhydroxylated cinnamic acid was not observed (
