- Email: [email protected]
Protein carbonyl formation on mucosal proteins in vitro and in dextran sulfate-induced colitis
Free Radical Biology & Medicine, Vol. 27, Nos. 3/4, pp. 262–270, 1999 Copyright © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/99/$–see front matter
PII S0891-5849(99)00065-9
Original Contribution PROTEIN CARBONYL FORMATION ON MUCOSAL PROTEINS IN VITRO AND IN DEXTRAN SULFATE-INDUCED COLITIS ANNEKE C. BLACKBURN,* WILLIAM F. DOE,*
and
GARY D. BUFFINTON†
*Division of Molecular Medicine, John Curtin School of Medical Research Australian National University, Canberra, Australia and †Inflammatory Bowel Disease Research Unit The Canberra Hospital, Canberra, Australia (Received 5 October 1998; Revised 22 December 1998; Accepted 12 February 1999)
Abstract—Reactive oxygen and nitrogen species have been implicated as mediators of mucosal injury in inflammatory bowel disease, but few studies have investigated protein oxidation in the inflamed mucosa. In this study, protein carbonyl formation on colonic mucosal proteins from mice was investigated following in vitro exposure of homogenates to iron/ascorbate, hydrogen peroxide, hypochloric acid (HOCl), or nitric oxide (•NO). Total carbonyl content was measured spectrophotometrically by derivatization with dinitrophenylhydrazine (DNPH), and oxidation of component proteins within the tissue was examined by Western blotting for DNPH-derivatized proteins using anti-dinitrophenyl DNP antibodies. These results were compared with protein carbonyl formation found in the acutely inflamed mucosa from mice with colitis induced by dextran sulfate sodium (DSS) administered at 5% w/v in the drinking water for 7 d. In vitro, carbonyl formation was observed after exposure to iron/ascorbate, HOCl and •NO. Iron/ascorbate (20 M/20 mM) exposure for 5 h increased carbonyl groups by 80%, particularly on proteins of 48, 75–100, 116, 131, and 142 kDa. Oxidation by 0.1 and 0.5 mM HOCl did not increase total carbonyl levels, but Western blotting revealed carbonyl formation on many proteins, particularly in the 49 –95 kDa region. After exposure to 1–10 mM HOCl, total carbonyl levels were increased by 0.5 to 12 times control levels with extensive cross-linking and fragmentation of proteins rich in carbonyl groups observed by Western blotting. In mice with acute colitis induced by DSS, protein carbonyl content of the inflamed mucosa was not significantly different from control mucosa, (7.80 ⫾ 1.05 vs. 8.43 ⫾ 0.59 nmol/mg protein respectively, p ⫽ .16 n ⫽ 8, 10); however, Western blotting analysis indicated several proteins of molecular weight 48, 79, 95, and 131 kDa that exhibited increased carbonyl content in the inflamed mucosa. These proteins corresponded to those observed after in vitro oxidation of normal intestinal mucosa with iron/ ascorbate and HOCl, suggesting that both HOCl and metal ions may be involved in protein oxidation in DSS-induced colitis. Identification and further analysis of the mucosal proteins susceptible to carbonyl modification may lead to a better understanding of the contribution of oxidants to the colonic mucosa tissue injury in inflammatory bowel disease. © 1999 Elsevier Science Inc. Keywords—Inflammatory bowel disease, Protein carbonyls, Western blotting, Dextran sulfate, Inflammation, Hypochlorous acid, Iron, Free radical
INTRODUCTION
scopically by an infiltrate that includes neutrophils, macrophages, and eosinophils. Elevated levels of reactive oxygen and nitrogen species (RONS) attributed to these inflammatory cells have been found in inflamed mucosa [1–3], and the possibility of cell-generated oxidants interacting with iron sources in the inflamed intestine to result in hydroxyl radical (•OH) formation has been hypothesized [4,5]. Further, mucosal antioxidants are substantially depleted in IBD [6,7], suggesting that increased oxidative stress may be contributing to the mucosal injury of IBD. Although oxidants have been implicated in the tissue injury of IBD, only one report contains direct evidence of
Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of unknown cause, characterized microAddress correspondence to: G. D. Buffinton, Department of Paediatric Gastroenterology, St. Bartholomew’s and the Royal London School of Medicine and Dentistry Suite 31, 3rd Floor, Dominion House, 59 Bartholomew Close, London, EC1A 7BE, United Kingdom; Tel: ⫹44 (171) 601 8172; Fax: ⫹44 (171) 600 5901; E-Mail: [email protected]. William F. Doe’s present address is The Medical School, University of Birmingham Edgbaston; and Gary D. Buffinton’s present address is Department of Paediatric Gastroenterology, St. Bartholomew’s and the Royal London School of Medicine & Dentistry, London; United Kingdom. 262
Protein carbonyls in colitis
protein oxidation in the inflamed tissue. Oxidized thiol groups were detected on glyceraldehyde-3-phosphate dehydrogenase in colonic epithelial cells from inflamed but not noninflamed mucosa from IBD patients [8]. The potential for further protein oxidation in colitis is significant and warrants investigation of a general indicator of protein oxidation such as protein carbonyls. Protein carbonyl formation can occur via a variety of mechanisms, including oxidation by hypochloric acid (HOCl) [9], aldehydes [10], and lipid peroxidation products [11], and metal-catalyzed oxidation (MCO) of proteins [12], such that many amino acids are susceptible to carbonyl oxidation. Increases in protein carbonyl levels have been reported in the synovial fluid of patients with rheumatoid arthritis as compared with patients with osteoarthritis [13], experimental acute pancreatitis [14], and many other conditions where RONS are implicated [15–17]. In this study, the dextran sulfate sodium (DSS) model of acute colitis was used to examine mucosal protein oxidation. Ingestion of DSS by mice can cause acute or chronic inflammation of the colonic mucosa, depending on the pattern of DSS exposure, with similarities to human ulcerative colitis [18,19]. The importance of neutrophils to the development of severe mucosal damage [20] and the significant depletion of several mucosal antioxidants [21] suggest that, as with human disease, RONS may be involved in the tissue injury of DSSinduced colitis. Protein carbonyl formation on colonic mucosal proteins from mice after in vitro exposure to oxidants was investigated and compared with protein carbonyl formation found in acutely inflamed mucosa. This allowed assessment of the potential involvement of various inflammatory oxidants in carbonyl formation in colitis. MATERIALS AND METHODS
Colitis Colitis was induced in 42 to 49-day-old male CBA/H mice that had been raised in a specific pathogen-free facility, by administering distilled water supplemented with 5% DSS (TdB Consultancy AB, Uppsala, Sweden) for 7 d ad libitum as described previously [5]. Mice were observed daily for fluid intake and for the major symptoms of diarrhea and rectal bleeding. The disease observed in these mice was an acute, neutrophilic inflammation, as indicated by myeloperoxidase levels [5,21] and histology (not presented), which was most severe in the distal colon. On the last day of DSS administration, mucosa was collected from the cleaned and opened transverse and descending colon by scraping the tissue with a microscope slide, and the mucosa was frozen on dry ice. Mucosa was homogenized using a Teflon/glass
263
homogenizer in phosphate-buffered saline (PBS) at 50 mg tissue/ml and centrifuged at 15,000 ⫻ g for 5 min to remove debris. The resulting homogenate supernatants were used in analyses. In vitro oxidation of mucosal homogenates Mucosa was collected from healthy male 7-week-old CBA/H mice and the homogenates from three to six mice were pooled. Triplicate aliquots of homogenate (diluted to 20 mg tissue/ml, approximately 1 mg protein/ml) were exposed to oxidants at 37°C in incubation buffer consisting of PBS containing 25 mM N-[2-hydroxyethyl]piperazine-N⬘-[2-ethanesulfonic acid] (HEPES), pH 7.2, and 1 mM phenylmethylsulfonyl fluoride to inhibit protein degradation during exposure to oxidants. Homogenates were oxidized by the iron/ascorbate method as used by Shacter et al. [22]. Then twenty millimolar ascorbate and 20 –500 M FeSO4 were added to the incubation buffer (pH adjusted to 7.2 after addition of ascorbate and iron) and homogenates were incubated for 5 h. Homogenates were also exposed to hydrogen peroxide (H2O2), HOCl (British Drug House), or diethylamine NONOate (Cayman Chemical Company, Ann Arbor, MI, USA), a nitric oxide (•NO) donor, at concentrations of 0.1 to 10 mM for 30 min. After incubation, samples were cooled on ice, centrifuged at 15,000 ⫻ g, 4°C for 5 min, then aliquotted, frozen on dry ice, and stored at ⫺70°C until analysis. The concentrations of stock solutions of H2O2, HOCl, and NONOate were determined as described previously [5]. Spectrophotometric determination of carbonyl groups Carbonyl groups were determined according to the method of Levine et al. [23]. Homogenate supernatants were divided into four aliquots of 200 l (⬃0.2 mg protein). Proteins were precipitated by the addition of 100 l 20% trichloroacetic acid (TCA) for 5 min on ice, and centrifuged at 4000 ⫻ g, 5 min. The pellet was redissolved in 100 l 0.2 M NaOH, and 100 l of 2 M HCl or 10 mM dinitrophenylhydrazine (DNPH) (Fluka, Switzerland) in 2 M HCl added to duplicate aliquots for sample blanks or the derivatizing of carbonyl groups, respectively. Samples were reacted for 30 min at room temperature. Proteins were reprecipitated with TCA, and washed three times with 500 l 1:1 ethanol:ethyl acetate with 15-min standing periods to remove excess DNPH. Samples were redissolved in 200 l 6 M guanidine buffer containing 20 mM KH2PO4, pH 2.3 and the absorbance read at 370 nm in microquartz cuvettes requiring a volume of 75 l (Hellma Mu¨llheim/Baden, Germany) in a Cary 1 UV/visible spectrophotometer. The
264
A. C. BLACKBURN et al.
carbonyl content in nanomoles per milligram protein was calculated using a molar extinction coefficient of 22,000 M⫺1 cm⫺1 at 370 nm [23] after subtraction of the blank absorbance. Protein concentration of the homogenates was determined by the BioRad (Hercules, CA, USA) protein assay on an aliquot of homogenate taken before derivitization. Although some protein was lost in the wash steps, the loss was consistent across control and inflamed samples (data not shown). Data are expressed as mean ⫾ SD, and statistical differences were determined using the unpaired Student’s t-test, with significance considered as p ⬍ .05. Assessment of deoxyribonucleic acid (DNA) content of samples Deoxyribonucleic acid, a common contaminant influencing the measurement of protein carbonyls, was removed by precipitation with 1% streptomycin sulfate. To nine volumes of mucosal homogenate supernatant, one volume of 10% streptomycin sulfate in 50 mM HEPES was added. Samples were allowed to stand on ice for 15 min before being centrifuged at 15,000 ⫻ g, 4°C for 10 min. To confirm the successful removal of DNA, the DNA and protein content of the supernatants before and after streptomycin sulfate precipitation were compared. Aliquots of supernatant (treated and untreated) were precipitated with TCA, and washed with 4% TCA to remove residual streptomycin sulfate, which can interfere with the assays, before being redissolved in NaOH, neutralized, diluted in appropriate buffers, and used in the assays as described. For comparison of control mice and mice exposed to 5% DSS, carbonyl content was determined with and without streptomycin sulfate precipitation. Carbonyl assays on in vitro oxidized homogenates were performed without DNA removal; however, the DNA concentration of the supernatants was determined to ensure that any changes in carbonyl content were not caused by altered amounts of DNA present in the supernatant. At high concentrations of oxidant, increased amounts of DNA were present; however, they were not sufficient to account for the increases in carbonyl content observed (data not shown). The DNA content of homogenates was determined by fluorescence using the Hoechst 33258 stain according to Brunk et al. [24]. To aliquots of supernatant or whole homogenate, an equal volume of assay buffer containing 50 mM sodium phosphate, 2 M NaCl, and 2 mM ethylenediaminetetraacetate (EDTA) was added. Whole homogenates underwent three cycles of rapid freezing and thawing, followed by sonication to release DNA from the nuclei. Samples were centrifuged, and 20 to 40 l of supernatant was added to a cuvette containing 2 ml of 2
g/ml Hoechst 33258 in assay buffer. The DNA content was determined by comparing the increase in fluorescence (Ex 356 nm, Em 460 nm, Hitachi F-3000 fluorospectrophotometer, Tokyo, Japan) caused by the addition of supernatant to the increase produced by addition of 0.5 to 5 g calf thymus DNA. Western blotting for carbonyl groups Immunoblot analysis for protein carbonyls was performed according to Shacter et al. [22]. Homogenate supernatant (50 l) was precipitated with 25 l 20% TCA on ice 5 min and centrifuged at 4000 ⫻ g. The pellet was dissolved in 15 l 6% sodium dodecyl sulfate (SDS) and mixed with an equal volume of 10 mM DNPH in 10% trifluoroacetic acid. Derivatization was allowed for 15–30 min at room temperature, at which time 20 l sample buffer (2 M Tris buffer containing 30% glycerol and 1 M -mercaptoethanol) was added and 10 g protein loaded per well. A negative control of mucosal proteins reacted with 10% trifluoroacetic acid only was also loaded. Samples were subjected to 9% SDS polyacrylamide gel electrophoresis (SDS-PAGE) (BioRad Mini-Protean II gel system) followed by electroblotting (BioRad Mini Trans-blott transfer cell) onto nitrocellulose (Hybond, Amersham, Sydney, Australia). Replicate blots were stained for protein with amido black or were assayed immunologically for the presence of DNPH-derivatized proteins. Blots were blocked with gelatin and incubated with polyclonal rabbit antidinitrophenyl (DNP) antibody (Sigma, St. Louis, MO, USA; diluted 1/2000) followed by a horseradish peroxidase-conjugated sheep antirabbit antibody (Silenus, diluted 1/10,000). Bands containing DNPH-derivatized proteins were visualized by enhanced chemiluminescence (Amersham ECL kit, Amersham). A 48-kDa protein band was subjected to N-terminal amino acid sequence analysis. Freshly homogenized mouse colonic mucosa was subject to SDS-PAGE as described above and stained for protein with Coomassie blue. The 48-kDa band was excised from the gel; the protein was eluted and adsorbed onto polyvinyl difluoride membrane. The N-terminal amino acid sequence analysis, determined using ABI Procise 494 automated peptide sequencer (Applied Biosystems, Inc, Foster City, CA, USA), was performed by the Biomolecular Resource Facility of John Curtin School of Medical Research (Canberra, Australia). Total reduced thiols The total reduced thiol content of the mucosal homogenate supernatants after in vitro exposure to oxidants was
Protein carbonyls in colitis Table 1. Total Carbonyl Content (nmol/mg) and Total Reduced Thiol Content (nmol/mg) of Mouse Colonic Mucosal Homogenate Supernatants Exposed to Oxidants In Vitro Iron/ascorbate (M/mM) Carbonyl H2O2 (mM) Carbonyl Thiols •
NO (mM) Carbonyl Thiols
0/0
20/20
100/20
500/20
3.7 ⫾ 0.4
6.7 ⫾ 0.1
7.4 ⫾ 0.8
10.3 ⫾ 0.8
0 4.5 ⫾ 1.4 113.8 ⫾ 6.6
0.1 4.4 ⫾ 0.5 105.1 ⫾ 2.2
1 4.7 ⫾ 1.1 99.2 ⫾ 1.2
10 3.9 ⫾ 0.7 85.9 ⫾ 7.2
0 3.3 ⫾ 0.5 99.0 ⫾ 1.6
0.1 2.6 ⫾ 0.2 97.8 ⫾ 1.8
1 2.1 ⫾ 0.3 57.1 ⫾ 0.4
10 5.6 ⫾ 1.7 10.5 ⫾ 1.3
Mean ⫾ SD, n ⫽ 3.
determined as described previously [21]. In brief, homogenates containing 0.1% SDS were mixed with an equal volume of 2 mM 5,5⬘-dithiobis(2-nitrobenzoic acid) in 0.5 M potassium phosphate buffer pH 7.0 [25]. After 15 min, the supernatants were transferred to a 96-well plate for determination of the absorbance at 410 nm, and the thiol content was calculated from a standard curve using reduced glutathione. This method could not be applied to the samples containing iron and ascorbate, as iron/ascorbate can hydrolyze 5,5⬘-dithiobis(2-nitrobenzoic acid) to release the chromophore, thus giving artificially high results. RESULTS
In vitro oxidation of mucosal proteins Total carbonyl analysis. Iron/ascorbate-mediated oxidation of normal mouse colonic mucosal homogenates resulted in increased carbonyl content with increasing concentrations of iron. Exposure to 20 M iron/20 mM ascorbate for 5 h increased carbonyl groups by 80% (Table 1). Further, H2O2 exposure (10 mM) of mucosal homogenates did not result in protein carbonyl formation, although it did result in a 25% decrease in reduced thiol content (Table 1). Exposure to 10 mM •NO resulted in a 70% increase in protein carbonyl groups, although no increase was observed at the lower concentrations of • NO, and 1 and 10 mM •NO resulted in depletion of reduced thiols by 42% and 90%, respectively (Table 1). In contrast to the small increases in carbonyls from H2O2 and •NO exposure, 10 mM HOCl exposure resulted in a 12-fold increase in carbonyl groups with loss of all
265
detectable reduced thiols (Table 2). The carbonyl formation by HOCl occurred in a dose-dependent manner, and was preceded by depletion of reduced thiols. (The reduced thiol content in these samples may be underestimated, as chloramines are known to react with the chromophore, 5-thio-2-nitrobenzoic acid [26]).
Western blotting analysis with anti-DNP detection Normal mouse colonic mucosa exposed to iron/ascorbate displayed many protein bands that had a more intense anti-DNP signal, indicating increased carbonyl content, compared with the unoxidized mucosal homogenate (Fig. 1A). Exposure to iron/ascorbate resulted in no changes to the protein staining pattern of the blots, but generated particularly strong anti-DNP signals from proteins within the molecular weight range of 75 to 100 kDa, and also from bands of 29, 39, 48, and 52 kDa. Furthermore, proteins of molecular weight 116, 131, and 142 kDa were easily observed by anti-DNP detection, although barely visible by protein staining, which indicated a very high susceptibility to carbonyl modification by MCO (Fig. 1A). Despite the high sensitivity of this technique, Western blotting of homogenates exposed to 0.1 or 1.0 mM •NO did not reveal carbonyl formation. The increased carbonyl content after exposure to 10 mM • NO was visible as general darkening of protein bands, with no change in the pattern of protein staining (Fig. 1B). Minimal carbonyl formation was observed in the homogenates exposed to H2O2 (Fig. 1C). Oxidation by low concentrations of HOCl (0.1 and 0.5 mM) generated carbonyl groups on many different proteins particularly in the 45- to 95-kDa molecular weight region (Fig. 2). At these concentrations of HOCl, no significant increase in total carbonyl content was detected spectrophotometrically (Table 2). At 0.5 mM HOCl, high molecular weight material rich in carbonyl groups was generated that was not visible by protein staining. At higher HOCl concentrations (1–10 mM), protein aggregation or cross-linking was clearly evident by protein staining, to the extent that some material did not penetrate the 4% polyacrylamide stacking gel (not shown). This was accompanied by fragmentation of proteins with smearing of both protein staining and particularly of anti-DNP signal (Fig. 2).
Table 2. Total Carbonyl Content (nmol/mg) and Total Reduced Thiol Content (nmol/mg) of Mouse Colonic Mucosal Homogenate Supernatants Exposed to HOCl In Vitro HOCl (mM) Carbonyl Thiols
0 3.2 ⫾ 0.3 82.2 ⫾ 3.0
Mean ⫾ SD, n ⫽ 3.
0.1 2.6 ⫾ 0.7 54.3 ⫾ 1.9
0.5 3.4 ⫾ 0.7 8.8 ⫾ 0.6
1 4.9 ⫾ 0.4 ⫺6.1 ⫾ 0.7
2.5 11.0 ⫾ 0.5 ⫺1.7 ⫾ 3.7
5 20.9 ⫾ 0.5 ⫺1.4 ⫾ 5.3
10 41.8 ⫾ 3.7 ⫺1.4 ⫾ 1.2
266
A. C. BLACKBURN et al.
Fig. 2. In vitro oxidation of mouse colonic mucosal proteins by HOCl. (A) Detection of carbonyl groups by immunoblotting using anti-(DNP) antibody with sizes (kDa) indicating carbonyl modified proteins. (B) Duplicate blot stained for protein with amido black with the positions of molecular weight markers (kDa) indicated. The -DNP lane contains underivatized proteins for negative control.
Protein carbonyl formation in DSS colitis
Fig. 1. In vitro oxidation of mouse colonic mucosal proteins by iron/ ascorbate (A), •NO (B), and H2O2 (C). The left panels show the detection of carbonyl groups by immunoblotting using anti-(DNP) antibody, and the right panels show the corresponding duplicate blots stained for protein with amido black. The sizes (kDa) of noted carbonyl modified proteins are indicated in panel A (anti-DNP) whereas the positions of molecular weight markers are indicated on other panels. The -DNP lane contains underivatized proteins for negative control.
Initial total carbonyl measurements on mucosa from control mice and mice with DSS-induced colitis showed a 4-fold increase in carbonyl content in the inflamed mucosa (Table 3). However, specific measurement of the DNA content of the supernatants using Hoechst 33258 revealed that DNA levels from DSS-induced colitis supernatants were eight-fold the DNA content of the control supernatants. After removal of DNA by streptomycin sulfate precipitation, no difference was observed in the protein carbonyl content of control mucosa and mucosa from mice with DSS-induced colitis (Table 3). The total reduced thiol content of the inflamed mucosal superna-
Protein carbonyls in colitis
267
Table 3. Analysis of Mucosal Homogenates from Control and DSS-Exposed Mice
Carbonyl (nmol/mg) ⫺SS (without DNA removal) ⫹SS* (soluble DNA removed) Total thiols (nmol/mg)† DNA (g/mg) Homogenate supernatant Homogenate, whole
Control (n)
DSS colitis (n)
5.9 ⫾ 0.9 (10)
24.4 ⫾ 3.0 (10)
8.4 ⫾ 0.6 (8)
7.8 ⫾ 1.0 (10)
133 ⫾ 15 (10)
115 ⫾ 23 (8)
8.8 ⫾ 1.5 (10) 175 ⫾ 51 (9)
71 ⫾ 15 (9) 169 ⫾ 18 (9)
Mean ⫾ standard deviation. Carbonyl content of supernatants with and without removal of DNA by streptomycin sulfate (SS) precipitation; supernatant total reduced thiol content; DNA content of homogenate supernatants (10,000 ⫻ g) and whole homogenates, as determined using the fluorescent dye Hoechst 33258. * p ⫽ .16 † p ⫽ .097 for control vs. DSS colitis.
tants was decreased by 13.1%, although this was not statistically significant (Table 3). The A280:A260 ratio has been used as an indicator of the necessity for removal of nucleic acids [23]; however, in these samples it was not informative. The ratio A280: A260 was very similar in control and DSS-exposed mice and, although streptomycin sulfate precipitation reduced the DNA content of the DSS samples to the same level as control samples, it did not alter the A280:A260 ratio (data not shown). The elevated supernatant DNA content of the inflamed mucosa contrasted markedly with the unchanged DNA content found in the whole homogenates (Table 3), which may be caused by necrosis or tissue degradation within inflamed sites, or increased fragility of the nuclear and chromosome structures that are disrupted during tissue homogenizing. Western blotting analysis of mucosa from control and DSS-exposed mice revealed marked differences in antiDNP signal intensity from several protein bands, indicating altered carbonyl content. In Fig. 3, proteins of size 48, 79, 95, and 131 kDa showed stronger anti-DNP signal in samples from mice with acute colitis, although the 79-, 95-, and 131-kDa proteins were not always as prominent compared with control mucosa as in the samples displayed. The prominence and relative abundance of the 48-kDa band allowed it to be excised and subjected to N-terminal amino acid sequence analysis; however, it was composed of at least three proteins, thus confounding identification of the oxidized protein(s). DISCUSSION
Reactive oxygen and nitrogen species have been implicated in the pathogenesis of IBD; however, very few reports have been made of protein oxidation in IBD or
Fig. 3. Carbonyl formation on colonic mucosal proteins from control mice and mice exposed to 5% DSS. (A) Detection of carbonyl groups by immunoblotting using anti-(DNP) antibody with sizes (kDa) indicating carbonyl modified proteins. (B) Duplicate blot stained for protein with amido black. The DNP lane contains underivatized proteins for negative control.
experimental colitis [8]. In this study, protein oxidation in the form of carbonyl groups was investigated in the mucosa of mice with DSS-induced colitis. Although increased total protein carbonyl content was not observed in the inflamed mucosa, Western blotting with anti-DNP detection of derivatized proteins revealed differences in the carbonyl content of specific proteins, with approximate molecular weights of 48, 79, 95, and 131 kDa (Fig. 3), indicating that carbonyl oxidation of mucosal proteins had occurred in DSS-induced colitis. Similar analysis of mucosal homogenates exposed in vitro to a range of inflammatory oxidants provided insights into which oxidants may be contributing to carbonyl oxidation in the inflamed mucosa. In vitro oxidation of mucosa by iron/ascorbate, HOCl, and, at high concentrations,
268 •
A. C. BLACKBURN et al.
NO resulted in increased levels of protein carbonyl groups on colonic mucosal proteins (Tables 1 and 2), whereas Western blotting analysis of these proteins revealed different patterns of carbonyl formation for each oxidant (Figs. 1 and 2). Proteins susceptible to iron/ ascorbate-mediated carbonyl formation included 48- and 131-kDa proteins, and carbonyl formation by HOCl was observed on many proteins between 45 and 100 kDa, including 48, 79, and 95 kDa, corresponding to the oxidized proteins observed in the DSS– colitis mucosa. These results suggest that protein oxidation by both HOCl and MCO is occurring in the mucosa of DSSinduced colitis, supporting proposed mechanisms of RONS-mediated tissue injury in colitis. Evidence of HOCl involvement in colitis includes the finding of elevated myeloperoxidase levels as a result of neutrophil infiltration in both the DSS-induced colitis model [21] and in ulcerative colitis [27], whereas elevated chemiluminescence of whole biopsies taken from inflamed mucosa of IBD patients was reduced by azide, which inhibits myeloperoxidase, and the HOCl scavenger taurine [1,2], suggesting that neutrophil-derived HOCl is generated in the mucosa in active IBD. Aminosalicylates, used in the treatment of IBD, may inhibit this injury process, as they are potent scavengers of HOCl [28,29] and can inhibit myeloperoxidase [30]. Catalytic iron is potentially abundant in the inflamed intestinal mucosa, and its involvement in tissue injury in IBD has been hypothesized, although not thoroughly investigated. Luminal iron, which can be as high as 320 M in the human colon [31], and catalytic heme iron from mucosal bleeding may interact with inflammatory cell-derived oxidants to mediate MCO of the mucosa [4]. Evidence supporting this mechanism includes the detection of a metabolite produced by •OH attack on 5-aminosalicylate in fecal extracts from IBD patients on aminosalicylate therapy [32], the ability of 5-aminosalicylate to scavenge •OH in vitro [33,34], and the ability of iron chelation by desferrioxamine and •OH scavenging by dimethylsulfoxide to significantly attenuate formylmethionyl-leucyl-phenylalanine–induced mucosal damage in rats [35]. The involvement of •OH in DSS-induced colitis has been investigated using salicylate hydroxylation; however, the susceptibility of the hydroxylated salicylate derivatives to oxidation by HOCl and the severity of other metabolic changes during disease confounded the interpretation of these results [5]. The current study has provided further evidence that these oxidative processes may be involved in tissue injury in colitis, and has helped in identifying specific targets that can be investigated with respect to their role in the pathogenesis of disease. These proteins are targets for oxidation for several possible reasons. The relative abundance of the 48-kDa protein may account for its
susceptibility to oxidation by both of these oxidants, whereas the 95- and 131-kDa proteins are present in only small amounts and thus may be expected to possess features making them particularly susceptible to oxidation. Proteins with metal binding sites are vulnerable to MCO, especially of the residues in close proximity to the metal binding site [36,37], whereas carbonyl formation by HOCl can result from decomposition of primary chloramines that may form on exposed lysine residues or terminal amino groups [9,38,39]. Identification of the 48-kDa modified protein was attempted; however, the identity remains elusive, as this band contains at least three comigrating proteins. The overall changes to tissue proteins observed by applying the Western blotting technique to mucosa oxidized in vitro are consistent with previous studies of protein oxidation. In the mucosa, carbonyl formation by HOCl was preceded by depletion of thiols (Table 2), an order of events also observed in cultured cells exposed to HOCl [40]. Carbonyl formation was initially observed on selected mucosal proteins and was followed by extensive cross-linking (Fig. 2), a well-recognized phenomenon of HOCl-mediated protein oxidation that has been suggested to play a role in inflammatory processes such as arthritis and atherosclerosis [26,39,41]. In contrast to HOCl, carbonyl formation via a MCO mechanism may not require thiol depletion. Iron-mediated carbonyl formation is a site-specific reaction, where iron can bind to metal-ion binding sites on proteins and result in oxidation of nearby residues [36], thus reduced thiols can be spared even in the event of carbonyl formation [12]. The lack of protein cross-linking and fragmentation induced in mucosal homogenates exposed to iron/ascorbate contrasted with the pattern of oxidation induced by HOCl, a finding consistent with the changes observed in plasma [22] and in studies of enzyme inactivation after MCO, where aggregation or fragmentation have not been reported [36]. Hydrogen peroxide can generate protein carbonyls by participating in iron-mediated oxidation and •OH generation, a model used by many investigators for in vitro studies of MCO [36]. In the current investigation, mucosal catalase may have consumed the H2O2, thus protecting the proteins from both carbonyl and thiol oxidation (Table 1). On bovine serum albumin and in rat lung homogenates, peroxynitrite (ONOO⫺) exposure was shown to generate carbonyl groups; this was preceded by depletion of reduced thiols [42]. The formation of ONOO⫺ or other reactive nitrogen species is possible at the highest concentration of •NO used in the present study [43], which may account for the carbonyl formation observed. A recent study of oxidative protein modification in experimental acute pancreatitis applied the Western blotting technique [14]. Unlike our findings in the DSS-
Protein carbonyls in colitis
induced colitis model, increased carbonyl content in pancreatitis tissue compared with control tissue was evident in the blots, but selective oxidation of individual protein bands was not observed [14]. The model of pancreatitis used in that study was extremely acute, with a time course of only 24 h, during which carbonyl levels increased to three times control levels. This is more akin to the outcome produced by high concentrations of oxidant exposure in vitro in our studies, where the selective oxidation of proteins is masked by the high level of overall protein oxidation. The contrast of these two disease models illustrates that this technique is most useful at identifying susceptible proteins where overall protein oxidation levels are relatively low, thus potentially identifying molecules that are targets early in the oxidative injury process. One report of carbonyl levels in colonic biopsies from IBD patients, measured by derivitization with DNPH and spectrophotometric measurement of the derivatized proteins [44], found elevated carbonyl content in mucosal biopsies from patients with IBD compared with control patients, but not compared with noninflamed tissue from the same patient. In that paper, however, soluble nucleic acids were not removed from the biopsy homogenates, a step that we have shown to be important when dealing with inflamed mucosa. Thus, the relationship of the carbonyl levels measured to protein oxidation in that study remains to be confirmed. The recently described enzymelinked immunosorbent assay method [45] may be a more reliable method for quantitation of protein carbonyl groups in complex tissue samples of very limited size, such as colonic biopsies. The application of Western blotting analysis of protein carbonyls to DSS-induced colitis indicated several proteins that may be particular targets of carbonyl modification, with the pattern of oxidation seen in the inflamed mucosa bearing similarity to that obtained by in vitro oxidation of mucosa. This suggested that both HOCl and MCO may be involved in protein oxidation in DSS-induced colitis. Identification and further analysis of the mucosal proteins susceptible to carbonyl modification may lead to a better understanding of the contribution of oxidants to the tissue injury in the colonic mucosa in IBD. Acknowledgements — G.D.B. acknowledges the support of the NHMRC.
REFERENCES [1] Simmonds, N. J.; Allen, R. E.; Stevens, T. R.; van Someren, R. N.; Blake, D. R.; Rampton, D. S. Chemiluminescence assay of mucosal reactive oxygen metabolites in inflammatory bowel disease. Gastroenterology 103:186 –196; 1992. [2] Sedghi, S.; Fields, J. Z.; Klamut, M.; Urban, G.; Durkin, M.; Winship, D.; Fretland, D.; Olyaee, M.; Keshavarzian, A. In-
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
269
creased production of luminol enhanced chemiluminescence by the inflamed colonic mucosa in patients with ulcerative colitis. Gut 34:1191–1197; 1993. Middleton, S. J.; Shorthouse, M.; Hunter, J. O. Increased nitric oxide synthesis in ulcerative colitis. Lancet 341:465– 466; 1993. Babbs, C. F. Oxygen radicals in ulcerative colitis. Free Radic. Biol. Med. 13:169 –181; 1992. Blackburn, A. C.; Doe, W. F.; Buffinton, G. D. Salicylate hydroxylation as an indicator of hydroxyl radical generation in dextran sulfate-induced colitis. Free Radic. Biol. Med. 25:305– 313; 1998. Buffinton, G. D.; Doe, W. F. Altered ascorbic acid status in the mucosa from inflammatory bowel disease patients. Free Radic. Res. 22:131–143; 1995. Buffinton, G. D.; Doe, W. F. Depleted mucosal antioxidant defences in inflammatory bowel disease. Free Radic. Biol. Med. 19:911–918; 1995. McKenzie, S. J.; Baker, M. S.; Buffinton, G. D.; Doe, W. F. Evidence of oxidant-induced injury to epithelial cells during inflammatory bowel disease. J. Clin. Invest. 98:136 –141; 1996. Pereira, W. E.; Hoyano, Y.; Summons, R. E.; Bacon, V. A.; Duffield, A. M. Chlorination studies. II. The reaction of aqueous hypochlorous acid with ␣-amino acids and dipeptides. Biochim. Biophys. Acta. 313:170 –180; 1973. O’Neill, C. A.; Halliwell, B.; van der Vliet, A.; Davis, P. A.; Packer, L.; Tritschler, H.; Strohman, W. J.; Rieland, T.; Cross, C. E.; Reznick, A. Z. Aldehyde-induced protein modifications in human plasma: protection by glutathione and dihydrolipoic acid. J. Lab. Clin. Med. 124:359 –370; 1994. Blakeman, D. P.; Ryan, T. P.; Jolly, R. A.; Petry, T. W. Diquatdependent protein carbonyl formation. Identification of lipiddependent and lipid-independent pathways. Biochem. Pharmacol. 50:929 –935; 1995. Levine, R. L. Oxidative modification of glutamine synthetase. I. Inactivation is due to loss of one histidine residue. J. Biol. Chem. 258:11823–11827; 1983. Chapman, M. L.; Rubin, B. R.; Gracy, R. W. Increased carbonyl content of proteins in synovial fluid from patients with rheumatoid arthritis. J. Rheumatol. 16:15–18; 1989. Reinheckel, T.; Nedelev, B.; Prause, J.; Augustin, W.; Schulz, H. U.; Lippert, H.; Halangk, W. Occurrence of oxidatively modified proteins: an early event in experimental acute pancreatitis. Free Radic. Biol. Med. 24:393– 400; 1998. Starke-Reed, P. E.; Oliver, C. N. Protein oxidation and proteolysis during aging and oxidative stress. Arch. Biochem. Biophys. 275: 559 –567; 1989. Serbinova, E.; Khwaja, S.; Reznick, A. Z.; Packer, L. Thioctic acid protects against ischemia-reperfusion injury in the isolated perfused langendorff heart. Free Radic. Res. Commun. 17:49 –58; 1992. Liu, Y.; Rosenthal, R. E.; Starke-Reed, P.; Fiskum, G. Inhibition of postcardiac arrest brain protein oxidation by acetyl-L-carnitine. Free Radic. Biol. Med. 15:667– 670; 1993. Okayasu, I.; Hatakeyama, S.; Yamada, M.; Ohkusa, T.; Inagaki, Y.; Nakaya, R. A novel method in the induction of reliable experimental acute and chronic ulcerative colitis in mice. Gastroenterology 98:694 –702; 1990. Cooper, H. S.; Murthy, S. N.; Shah, R. S.; Sedergran, D. J. Clinicopathologic study of dextran sulfate sodium experimental murine colitis. Lab. Invest. 69:238 –249; 1993. Wozniak, A.; Sharma, D. P.; van de Pol, E.; McColl, S. R.; Doe, W. F. Depletion of neutrophils decreases disease severity and mortality in a mouse model of ulcerative colitis. Gastroenterology 108(Suppl.):A944; 1995. Blackburn, A. C.; Doe, W. F.; Buffinton, G. D. Colonic antioxidant status in dextran sulfate-induced colitis in mice. Inflammatory Bowel Dis. 3:198 –203; 1997. Shacter, E.; Williams, J. A.; Lim, M.; Levine, R. L. Differential susceptibility of plasma proteins to oxidative modification: examination by Western blot immunoassay. Free Radic. Biol. Med. 17:429 – 437; 1994.
270
A. C. BLACKBURN et al.
[23] Levine, R. L.; Garland, D.; Oliver, C. N.; Amici, A.; Climent, I.; Lenz, A. G.; Ahn, B. W.; Shaltiel, S.; Stadtman, E. R. Determination of carbonyl content in oxidatively modified proteins. Methods Enzymol. 186:464 – 478; 1990. [24] Brunk, C.; Jones, K.; James, T. Assay for nanogram quantities of DNA in cellular homogenates. Anal. Biochem. 92:497–500; 1979. [25] Jocelyn, P. C. Spectrophotometric assay of thiols. Methods Enzymol. 143:44 – 67; 1987. [26] Vissers, M. C.; Winterbourn, C. C. Oxidative damage to fibronectin. I. The effects of the neutrophil myeloperoxidase system and HOCl. Arch. Biochem. Biophys. 285:53–59; 1991. [27] Izzo, R. S.; Witkon, K.; Chen, A. I.; Hadjiyane, C.; Weinstein, M. I.; Pellecchia, C. Interleukin-8 and neutrophil markers in colonic mucosa from patients with ulcerative colitis. Am. J. Gastroenterol. 87:1447–1452; 1992. [28] Williams, J. G.; Hallett, M. B. The reaction of 5-amino-salicylic acid with hypochlorite. Implications for its mode of action in inflammatory bowel disease. Biochem. Pharmacol. 38:149 –154; 1989. [29] Aruoma, O. I.; Wasil, M.; Halliwell, B.; Hoey, B. M.; Butler, J. The scavenging of oxidants by sulphasalazine and its metabolites. A possible contribution to their anti-inflammatory effects? Biochem. Pharmacol. 36:3739 –3742; 1987. [30] Kettle, A. J.; Winterbourn, C. C. Mechanism of inhibition of myeloperoxidase by anti-inflammatory drugs. Biochem. Pharmacol. 41:1485–1492; 1991. [31] Babbs, C. F. Free radicals and the etiology of colon cancer. Free Radic. Biol. Med. 8:191–200; 1990. [32] Ahnfelt-Ronne, I.; Nielsen, O. H.; Christensen, A.; Langholz, E.; Binder, V.; Riis, P. Clinical evidence supporting the radical scavenger mechanism of 5-aminosalicylic acid. Gastroenterology 98:1162–1169; 1990. [33] Betts, W. H.; Whitehouse, M. W.; Cleland, L. G.; VernonRoberts, B. In vitro antioxidant properties of potential biotransformation products of salicylate, sulphasalazine and amidopyrine. Free Radic. Biol. Med. 1:273–280; 1985. [34] Allgayer, H.; Hofer, P.; Schmidt, M.; Bohne, P.; Kruis, W.; Gugler, R. Superoxide, hydroxyl and fatty acid radical scavenging by aminosalicylates. Direct evaluation with electron spin resonance spectroscopy. Biochem. Pharmacol. 43:259 –262; 1992. [35] von Ritter, C.; Grisham, M. B.; Hollwarth, M.; Inauen, W.; Granger, D. N. Neutrophil-derived oxidants mediate formyl-methionyl-leucyl-phenylalanine-induced increases in mucosal permeability in rats. Gastroenterology 97:778 –780; 1989. [36] Stadtman, E. R. Metal ion-catalyzed oxidation of proteins: biochemical mechanism and biological consequences. Free Radic. Biol. Med. 9:315–325; 1990. [37] Amici, A.; Levine, R. L.; Tsai, L.; Stadtman, E. R. Conversion of
[38]
[39]
[40]
[41]
[42] [43]
[44]
[45]
amino acid residues in proteins and amino acid homopolymers to carbonyl derivatives by metal-catalyzed oxidation reactions. J. Biol. Chem. 264:3341–3346; 1989. Stelmaszynska, T.; Zgliczynski, J. M. N-(2-Oxoacyl)amino acids and nitriles as final products of dipeptide chlorination mediated by the myeloperoxidase/H2O2/CI⫺ system. Eur. J. Biochem. 92: 301–308; 1978. Hazell, L. J.; van-den Berg, J. J.; Stocker, R. Oxidation of lowdensity lipoprotein by hypochlorite causes aggregation that is mediated by modification of lysine residues rather than lipid oxidation. Biochem. J. 302:297–304; 1994. Schraufstatter, I. U.; Browne, K.; Harris, A.; Hyslop, P. A.; Jackson, J. H.; Quehenberger, O.; Cochrane, C. G. Mechanisms of hypochlorite injury of target cells. J. Clin. Invest. 85:554 –562; 1990. Jasin, H. E. Oxidative cross-linking of immune complexes by human polymorphonuclear leukocytes. J. Clin. Invest. 81:6 –15; 1988. Ischiropoulos, H.; Al-Mehdi, A. B. Peroxynitrite-mediated oxidative protein modifications. FEBS Lett. 364:279 –282; 1995. Crow, J. P.; Beckman, J. S. Reactions between nitric oxide, superoxide, and peroxynitrite: footprints of peroxynitrite in vivo. Adv. Pharmacol. 34:17– 43; 1995. Lih-Brody, L.; Powell, S. R.; Collier, K. P.; Reddy, G. M.; Cerchia, R.; Kahn, E.; Weissman, G. S.; Katz, S.; Floyd, R. A.; McKinley, M. J.; Fisher, S. E.; Mullin, G. E. Increased oxidative stress and decreased antioxidant defenses in mucosa of inflammatory bowel disease. Dig. Dis. Sci. 41:2078 –2086; 1996. Buss, H.; Chan, T. P.; Sluis, K. B.; Domigan, N. M.; Winterbourn, C. C. Protein carbonyl measurement by a sensitive ELISA method. Free Radic. Biol. Med. 23:361–366; 1997.
ABBREVIATIONS
DNP— dinitrophenyl DNPH— dinitrophenylhydrazine DSS— dextran sulfate sodium IBD—inflammatory bowel disease MCO—metal-catalyzed oxidation PAGE—polyacrylamide gel electrophoresis RONS—reactive oxygen and nitrogen species SDS—sodium dodecyl sulfate TCA—trichloroacetic acid
PII S0891-5849(99)00065-9
Original Contribution PROTEIN CARBONYL FORMATION ON MUCOSAL PROTEINS IN VITRO AND IN DEXTRAN SULFATE-INDUCED COLITIS ANNEKE C. BLACKBURN,* WILLIAM F. DOE,*
and
GARY D. BUFFINTON†
*Division of Molecular Medicine, John Curtin School of Medical Research Australian National University, Canberra, Australia and †Inflammatory Bowel Disease Research Unit The Canberra Hospital, Canberra, Australia (Received 5 October 1998; Revised 22 December 1998; Accepted 12 February 1999)
Abstract—Reactive oxygen and nitrogen species have been implicated as mediators of mucosal injury in inflammatory bowel disease, but few studies have investigated protein oxidation in the inflamed mucosa. In this study, protein carbonyl formation on colonic mucosal proteins from mice was investigated following in vitro exposure of homogenates to iron/ascorbate, hydrogen peroxide, hypochloric acid (HOCl), or nitric oxide (•NO). Total carbonyl content was measured spectrophotometrically by derivatization with dinitrophenylhydrazine (DNPH), and oxidation of component proteins within the tissue was examined by Western blotting for DNPH-derivatized proteins using anti-dinitrophenyl DNP antibodies. These results were compared with protein carbonyl formation found in the acutely inflamed mucosa from mice with colitis induced by dextran sulfate sodium (DSS) administered at 5% w/v in the drinking water for 7 d. In vitro, carbonyl formation was observed after exposure to iron/ascorbate, HOCl and •NO. Iron/ascorbate (20 M/20 mM) exposure for 5 h increased carbonyl groups by 80%, particularly on proteins of 48, 75–100, 116, 131, and 142 kDa. Oxidation by 0.1 and 0.5 mM HOCl did not increase total carbonyl levels, but Western blotting revealed carbonyl formation on many proteins, particularly in the 49 –95 kDa region. After exposure to 1–10 mM HOCl, total carbonyl levels were increased by 0.5 to 12 times control levels with extensive cross-linking and fragmentation of proteins rich in carbonyl groups observed by Western blotting. In mice with acute colitis induced by DSS, protein carbonyl content of the inflamed mucosa was not significantly different from control mucosa, (7.80 ⫾ 1.05 vs. 8.43 ⫾ 0.59 nmol/mg protein respectively, p ⫽ .16 n ⫽ 8, 10); however, Western blotting analysis indicated several proteins of molecular weight 48, 79, 95, and 131 kDa that exhibited increased carbonyl content in the inflamed mucosa. These proteins corresponded to those observed after in vitro oxidation of normal intestinal mucosa with iron/ ascorbate and HOCl, suggesting that both HOCl and metal ions may be involved in protein oxidation in DSS-induced colitis. Identification and further analysis of the mucosal proteins susceptible to carbonyl modification may lead to a better understanding of the contribution of oxidants to the colonic mucosa tissue injury in inflammatory bowel disease. © 1999 Elsevier Science Inc. Keywords—Inflammatory bowel disease, Protein carbonyls, Western blotting, Dextran sulfate, Inflammation, Hypochlorous acid, Iron, Free radical
INTRODUCTION
scopically by an infiltrate that includes neutrophils, macrophages, and eosinophils. Elevated levels of reactive oxygen and nitrogen species (RONS) attributed to these inflammatory cells have been found in inflamed mucosa [1–3], and the possibility of cell-generated oxidants interacting with iron sources in the inflamed intestine to result in hydroxyl radical (•OH) formation has been hypothesized [4,5]. Further, mucosal antioxidants are substantially depleted in IBD [6,7], suggesting that increased oxidative stress may be contributing to the mucosal injury of IBD. Although oxidants have been implicated in the tissue injury of IBD, only one report contains direct evidence of
Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of unknown cause, characterized microAddress correspondence to: G. D. Buffinton, Department of Paediatric Gastroenterology, St. Bartholomew’s and the Royal London School of Medicine and Dentistry Suite 31, 3rd Floor, Dominion House, 59 Bartholomew Close, London, EC1A 7BE, United Kingdom; Tel: ⫹44 (171) 601 8172; Fax: ⫹44 (171) 600 5901; E-Mail: [email protected]. William F. Doe’s present address is The Medical School, University of Birmingham Edgbaston; and Gary D. Buffinton’s present address is Department of Paediatric Gastroenterology, St. Bartholomew’s and the Royal London School of Medicine & Dentistry, London; United Kingdom. 262
Protein carbonyls in colitis
protein oxidation in the inflamed tissue. Oxidized thiol groups were detected on glyceraldehyde-3-phosphate dehydrogenase in colonic epithelial cells from inflamed but not noninflamed mucosa from IBD patients [8]. The potential for further protein oxidation in colitis is significant and warrants investigation of a general indicator of protein oxidation such as protein carbonyls. Protein carbonyl formation can occur via a variety of mechanisms, including oxidation by hypochloric acid (HOCl) [9], aldehydes [10], and lipid peroxidation products [11], and metal-catalyzed oxidation (MCO) of proteins [12], such that many amino acids are susceptible to carbonyl oxidation. Increases in protein carbonyl levels have been reported in the synovial fluid of patients with rheumatoid arthritis as compared with patients with osteoarthritis [13], experimental acute pancreatitis [14], and many other conditions where RONS are implicated [15–17]. In this study, the dextran sulfate sodium (DSS) model of acute colitis was used to examine mucosal protein oxidation. Ingestion of DSS by mice can cause acute or chronic inflammation of the colonic mucosa, depending on the pattern of DSS exposure, with similarities to human ulcerative colitis [18,19]. The importance of neutrophils to the development of severe mucosal damage [20] and the significant depletion of several mucosal antioxidants [21] suggest that, as with human disease, RONS may be involved in the tissue injury of DSSinduced colitis. Protein carbonyl formation on colonic mucosal proteins from mice after in vitro exposure to oxidants was investigated and compared with protein carbonyl formation found in acutely inflamed mucosa. This allowed assessment of the potential involvement of various inflammatory oxidants in carbonyl formation in colitis. MATERIALS AND METHODS
Colitis Colitis was induced in 42 to 49-day-old male CBA/H mice that had been raised in a specific pathogen-free facility, by administering distilled water supplemented with 5% DSS (TdB Consultancy AB, Uppsala, Sweden) for 7 d ad libitum as described previously [5]. Mice were observed daily for fluid intake and for the major symptoms of diarrhea and rectal bleeding. The disease observed in these mice was an acute, neutrophilic inflammation, as indicated by myeloperoxidase levels [5,21] and histology (not presented), which was most severe in the distal colon. On the last day of DSS administration, mucosa was collected from the cleaned and opened transverse and descending colon by scraping the tissue with a microscope slide, and the mucosa was frozen on dry ice. Mucosa was homogenized using a Teflon/glass
263
homogenizer in phosphate-buffered saline (PBS) at 50 mg tissue/ml and centrifuged at 15,000 ⫻ g for 5 min to remove debris. The resulting homogenate supernatants were used in analyses. In vitro oxidation of mucosal homogenates Mucosa was collected from healthy male 7-week-old CBA/H mice and the homogenates from three to six mice were pooled. Triplicate aliquots of homogenate (diluted to 20 mg tissue/ml, approximately 1 mg protein/ml) were exposed to oxidants at 37°C in incubation buffer consisting of PBS containing 25 mM N-[2-hydroxyethyl]piperazine-N⬘-[2-ethanesulfonic acid] (HEPES), pH 7.2, and 1 mM phenylmethylsulfonyl fluoride to inhibit protein degradation during exposure to oxidants. Homogenates were oxidized by the iron/ascorbate method as used by Shacter et al. [22]. Then twenty millimolar ascorbate and 20 –500 M FeSO4 were added to the incubation buffer (pH adjusted to 7.2 after addition of ascorbate and iron) and homogenates were incubated for 5 h. Homogenates were also exposed to hydrogen peroxide (H2O2), HOCl (British Drug House), or diethylamine NONOate (Cayman Chemical Company, Ann Arbor, MI, USA), a nitric oxide (•NO) donor, at concentrations of 0.1 to 10 mM for 30 min. After incubation, samples were cooled on ice, centrifuged at 15,000 ⫻ g, 4°C for 5 min, then aliquotted, frozen on dry ice, and stored at ⫺70°C until analysis. The concentrations of stock solutions of H2O2, HOCl, and NONOate were determined as described previously [5]. Spectrophotometric determination of carbonyl groups Carbonyl groups were determined according to the method of Levine et al. [23]. Homogenate supernatants were divided into four aliquots of 200 l (⬃0.2 mg protein). Proteins were precipitated by the addition of 100 l 20% trichloroacetic acid (TCA) for 5 min on ice, and centrifuged at 4000 ⫻ g, 5 min. The pellet was redissolved in 100 l 0.2 M NaOH, and 100 l of 2 M HCl or 10 mM dinitrophenylhydrazine (DNPH) (Fluka, Switzerland) in 2 M HCl added to duplicate aliquots for sample blanks or the derivatizing of carbonyl groups, respectively. Samples were reacted for 30 min at room temperature. Proteins were reprecipitated with TCA, and washed three times with 500 l 1:1 ethanol:ethyl acetate with 15-min standing periods to remove excess DNPH. Samples were redissolved in 200 l 6 M guanidine buffer containing 20 mM KH2PO4, pH 2.3 and the absorbance read at 370 nm in microquartz cuvettes requiring a volume of 75 l (Hellma Mu¨llheim/Baden, Germany) in a Cary 1 UV/visible spectrophotometer. The
264
A. C. BLACKBURN et al.
carbonyl content in nanomoles per milligram protein was calculated using a molar extinction coefficient of 22,000 M⫺1 cm⫺1 at 370 nm [23] after subtraction of the blank absorbance. Protein concentration of the homogenates was determined by the BioRad (Hercules, CA, USA) protein assay on an aliquot of homogenate taken before derivitization. Although some protein was lost in the wash steps, the loss was consistent across control and inflamed samples (data not shown). Data are expressed as mean ⫾ SD, and statistical differences were determined using the unpaired Student’s t-test, with significance considered as p ⬍ .05. Assessment of deoxyribonucleic acid (DNA) content of samples Deoxyribonucleic acid, a common contaminant influencing the measurement of protein carbonyls, was removed by precipitation with 1% streptomycin sulfate. To nine volumes of mucosal homogenate supernatant, one volume of 10% streptomycin sulfate in 50 mM HEPES was added. Samples were allowed to stand on ice for 15 min before being centrifuged at 15,000 ⫻ g, 4°C for 10 min. To confirm the successful removal of DNA, the DNA and protein content of the supernatants before and after streptomycin sulfate precipitation were compared. Aliquots of supernatant (treated and untreated) were precipitated with TCA, and washed with 4% TCA to remove residual streptomycin sulfate, which can interfere with the assays, before being redissolved in NaOH, neutralized, diluted in appropriate buffers, and used in the assays as described. For comparison of control mice and mice exposed to 5% DSS, carbonyl content was determined with and without streptomycin sulfate precipitation. Carbonyl assays on in vitro oxidized homogenates were performed without DNA removal; however, the DNA concentration of the supernatants was determined to ensure that any changes in carbonyl content were not caused by altered amounts of DNA present in the supernatant. At high concentrations of oxidant, increased amounts of DNA were present; however, they were not sufficient to account for the increases in carbonyl content observed (data not shown). The DNA content of homogenates was determined by fluorescence using the Hoechst 33258 stain according to Brunk et al. [24]. To aliquots of supernatant or whole homogenate, an equal volume of assay buffer containing 50 mM sodium phosphate, 2 M NaCl, and 2 mM ethylenediaminetetraacetate (EDTA) was added. Whole homogenates underwent three cycles of rapid freezing and thawing, followed by sonication to release DNA from the nuclei. Samples were centrifuged, and 20 to 40 l of supernatant was added to a cuvette containing 2 ml of 2
g/ml Hoechst 33258 in assay buffer. The DNA content was determined by comparing the increase in fluorescence (Ex 356 nm, Em 460 nm, Hitachi F-3000 fluorospectrophotometer, Tokyo, Japan) caused by the addition of supernatant to the increase produced by addition of 0.5 to 5 g calf thymus DNA. Western blotting for carbonyl groups Immunoblot analysis for protein carbonyls was performed according to Shacter et al. [22]. Homogenate supernatant (50 l) was precipitated with 25 l 20% TCA on ice 5 min and centrifuged at 4000 ⫻ g. The pellet was dissolved in 15 l 6% sodium dodecyl sulfate (SDS) and mixed with an equal volume of 10 mM DNPH in 10% trifluoroacetic acid. Derivatization was allowed for 15–30 min at room temperature, at which time 20 l sample buffer (2 M Tris buffer containing 30% glycerol and 1 M -mercaptoethanol) was added and 10 g protein loaded per well. A negative control of mucosal proteins reacted with 10% trifluoroacetic acid only was also loaded. Samples were subjected to 9% SDS polyacrylamide gel electrophoresis (SDS-PAGE) (BioRad Mini-Protean II gel system) followed by electroblotting (BioRad Mini Trans-blott transfer cell) onto nitrocellulose (Hybond, Amersham, Sydney, Australia). Replicate blots were stained for protein with amido black or were assayed immunologically for the presence of DNPH-derivatized proteins. Blots were blocked with gelatin and incubated with polyclonal rabbit antidinitrophenyl (DNP) antibody (Sigma, St. Louis, MO, USA; diluted 1/2000) followed by a horseradish peroxidase-conjugated sheep antirabbit antibody (Silenus, diluted 1/10,000). Bands containing DNPH-derivatized proteins were visualized by enhanced chemiluminescence (Amersham ECL kit, Amersham). A 48-kDa protein band was subjected to N-terminal amino acid sequence analysis. Freshly homogenized mouse colonic mucosa was subject to SDS-PAGE as described above and stained for protein with Coomassie blue. The 48-kDa band was excised from the gel; the protein was eluted and adsorbed onto polyvinyl difluoride membrane. The N-terminal amino acid sequence analysis, determined using ABI Procise 494 automated peptide sequencer (Applied Biosystems, Inc, Foster City, CA, USA), was performed by the Biomolecular Resource Facility of John Curtin School of Medical Research (Canberra, Australia). Total reduced thiols The total reduced thiol content of the mucosal homogenate supernatants after in vitro exposure to oxidants was
Protein carbonyls in colitis Table 1. Total Carbonyl Content (nmol/mg) and Total Reduced Thiol Content (nmol/mg) of Mouse Colonic Mucosal Homogenate Supernatants Exposed to Oxidants In Vitro Iron/ascorbate (M/mM) Carbonyl H2O2 (mM) Carbonyl Thiols •
NO (mM) Carbonyl Thiols
0/0
20/20
100/20
500/20
3.7 ⫾ 0.4
6.7 ⫾ 0.1
7.4 ⫾ 0.8
10.3 ⫾ 0.8
0 4.5 ⫾ 1.4 113.8 ⫾ 6.6
0.1 4.4 ⫾ 0.5 105.1 ⫾ 2.2
1 4.7 ⫾ 1.1 99.2 ⫾ 1.2
10 3.9 ⫾ 0.7 85.9 ⫾ 7.2
0 3.3 ⫾ 0.5 99.0 ⫾ 1.6
0.1 2.6 ⫾ 0.2 97.8 ⫾ 1.8
1 2.1 ⫾ 0.3 57.1 ⫾ 0.4
10 5.6 ⫾ 1.7 10.5 ⫾ 1.3
Mean ⫾ SD, n ⫽ 3.
determined as described previously [21]. In brief, homogenates containing 0.1% SDS were mixed with an equal volume of 2 mM 5,5⬘-dithiobis(2-nitrobenzoic acid) in 0.5 M potassium phosphate buffer pH 7.0 [25]. After 15 min, the supernatants were transferred to a 96-well plate for determination of the absorbance at 410 nm, and the thiol content was calculated from a standard curve using reduced glutathione. This method could not be applied to the samples containing iron and ascorbate, as iron/ascorbate can hydrolyze 5,5⬘-dithiobis(2-nitrobenzoic acid) to release the chromophore, thus giving artificially high results. RESULTS
In vitro oxidation of mucosal proteins Total carbonyl analysis. Iron/ascorbate-mediated oxidation of normal mouse colonic mucosal homogenates resulted in increased carbonyl content with increasing concentrations of iron. Exposure to 20 M iron/20 mM ascorbate for 5 h increased carbonyl groups by 80% (Table 1). Further, H2O2 exposure (10 mM) of mucosal homogenates did not result in protein carbonyl formation, although it did result in a 25% decrease in reduced thiol content (Table 1). Exposure to 10 mM •NO resulted in a 70% increase in protein carbonyl groups, although no increase was observed at the lower concentrations of • NO, and 1 and 10 mM •NO resulted in depletion of reduced thiols by 42% and 90%, respectively (Table 1). In contrast to the small increases in carbonyls from H2O2 and •NO exposure, 10 mM HOCl exposure resulted in a 12-fold increase in carbonyl groups with loss of all
265
detectable reduced thiols (Table 2). The carbonyl formation by HOCl occurred in a dose-dependent manner, and was preceded by depletion of reduced thiols. (The reduced thiol content in these samples may be underestimated, as chloramines are known to react with the chromophore, 5-thio-2-nitrobenzoic acid [26]).
Western blotting analysis with anti-DNP detection Normal mouse colonic mucosa exposed to iron/ascorbate displayed many protein bands that had a more intense anti-DNP signal, indicating increased carbonyl content, compared with the unoxidized mucosal homogenate (Fig. 1A). Exposure to iron/ascorbate resulted in no changes to the protein staining pattern of the blots, but generated particularly strong anti-DNP signals from proteins within the molecular weight range of 75 to 100 kDa, and also from bands of 29, 39, 48, and 52 kDa. Furthermore, proteins of molecular weight 116, 131, and 142 kDa were easily observed by anti-DNP detection, although barely visible by protein staining, which indicated a very high susceptibility to carbonyl modification by MCO (Fig. 1A). Despite the high sensitivity of this technique, Western blotting of homogenates exposed to 0.1 or 1.0 mM •NO did not reveal carbonyl formation. The increased carbonyl content after exposure to 10 mM • NO was visible as general darkening of protein bands, with no change in the pattern of protein staining (Fig. 1B). Minimal carbonyl formation was observed in the homogenates exposed to H2O2 (Fig. 1C). Oxidation by low concentrations of HOCl (0.1 and 0.5 mM) generated carbonyl groups on many different proteins particularly in the 45- to 95-kDa molecular weight region (Fig. 2). At these concentrations of HOCl, no significant increase in total carbonyl content was detected spectrophotometrically (Table 2). At 0.5 mM HOCl, high molecular weight material rich in carbonyl groups was generated that was not visible by protein staining. At higher HOCl concentrations (1–10 mM), protein aggregation or cross-linking was clearly evident by protein staining, to the extent that some material did not penetrate the 4% polyacrylamide stacking gel (not shown). This was accompanied by fragmentation of proteins with smearing of both protein staining and particularly of anti-DNP signal (Fig. 2).
Table 2. Total Carbonyl Content (nmol/mg) and Total Reduced Thiol Content (nmol/mg) of Mouse Colonic Mucosal Homogenate Supernatants Exposed to HOCl In Vitro HOCl (mM) Carbonyl Thiols
0 3.2 ⫾ 0.3 82.2 ⫾ 3.0
Mean ⫾ SD, n ⫽ 3.
0.1 2.6 ⫾ 0.7 54.3 ⫾ 1.9
0.5 3.4 ⫾ 0.7 8.8 ⫾ 0.6
1 4.9 ⫾ 0.4 ⫺6.1 ⫾ 0.7
2.5 11.0 ⫾ 0.5 ⫺1.7 ⫾ 3.7
5 20.9 ⫾ 0.5 ⫺1.4 ⫾ 5.3
10 41.8 ⫾ 3.7 ⫺1.4 ⫾ 1.2
266
A. C. BLACKBURN et al.
Fig. 2. In vitro oxidation of mouse colonic mucosal proteins by HOCl. (A) Detection of carbonyl groups by immunoblotting using anti-(DNP) antibody with sizes (kDa) indicating carbonyl modified proteins. (B) Duplicate blot stained for protein with amido black with the positions of molecular weight markers (kDa) indicated. The -DNP lane contains underivatized proteins for negative control.
Protein carbonyl formation in DSS colitis
Fig. 1. In vitro oxidation of mouse colonic mucosal proteins by iron/ ascorbate (A), •NO (B), and H2O2 (C). The left panels show the detection of carbonyl groups by immunoblotting using anti-(DNP) antibody, and the right panels show the corresponding duplicate blots stained for protein with amido black. The sizes (kDa) of noted carbonyl modified proteins are indicated in panel A (anti-DNP) whereas the positions of molecular weight markers are indicated on other panels. The -DNP lane contains underivatized proteins for negative control.
Initial total carbonyl measurements on mucosa from control mice and mice with DSS-induced colitis showed a 4-fold increase in carbonyl content in the inflamed mucosa (Table 3). However, specific measurement of the DNA content of the supernatants using Hoechst 33258 revealed that DNA levels from DSS-induced colitis supernatants were eight-fold the DNA content of the control supernatants. After removal of DNA by streptomycin sulfate precipitation, no difference was observed in the protein carbonyl content of control mucosa and mucosa from mice with DSS-induced colitis (Table 3). The total reduced thiol content of the inflamed mucosal superna-
Protein carbonyls in colitis
267
Table 3. Analysis of Mucosal Homogenates from Control and DSS-Exposed Mice
Carbonyl (nmol/mg) ⫺SS (without DNA removal) ⫹SS* (soluble DNA removed) Total thiols (nmol/mg)† DNA (g/mg) Homogenate supernatant Homogenate, whole
Control (n)
DSS colitis (n)
5.9 ⫾ 0.9 (10)
24.4 ⫾ 3.0 (10)
8.4 ⫾ 0.6 (8)
7.8 ⫾ 1.0 (10)
133 ⫾ 15 (10)
115 ⫾ 23 (8)
8.8 ⫾ 1.5 (10) 175 ⫾ 51 (9)
71 ⫾ 15 (9) 169 ⫾ 18 (9)
Mean ⫾ standard deviation. Carbonyl content of supernatants with and without removal of DNA by streptomycin sulfate (SS) precipitation; supernatant total reduced thiol content; DNA content of homogenate supernatants (10,000 ⫻ g) and whole homogenates, as determined using the fluorescent dye Hoechst 33258. * p ⫽ .16 † p ⫽ .097 for control vs. DSS colitis.
tants was decreased by 13.1%, although this was not statistically significant (Table 3). The A280:A260 ratio has been used as an indicator of the necessity for removal of nucleic acids [23]; however, in these samples it was not informative. The ratio A280: A260 was very similar in control and DSS-exposed mice and, although streptomycin sulfate precipitation reduced the DNA content of the DSS samples to the same level as control samples, it did not alter the A280:A260 ratio (data not shown). The elevated supernatant DNA content of the inflamed mucosa contrasted markedly with the unchanged DNA content found in the whole homogenates (Table 3), which may be caused by necrosis or tissue degradation within inflamed sites, or increased fragility of the nuclear and chromosome structures that are disrupted during tissue homogenizing. Western blotting analysis of mucosa from control and DSS-exposed mice revealed marked differences in antiDNP signal intensity from several protein bands, indicating altered carbonyl content. In Fig. 3, proteins of size 48, 79, 95, and 131 kDa showed stronger anti-DNP signal in samples from mice with acute colitis, although the 79-, 95-, and 131-kDa proteins were not always as prominent compared with control mucosa as in the samples displayed. The prominence and relative abundance of the 48-kDa band allowed it to be excised and subjected to N-terminal amino acid sequence analysis; however, it was composed of at least three proteins, thus confounding identification of the oxidized protein(s). DISCUSSION
Reactive oxygen and nitrogen species have been implicated in the pathogenesis of IBD; however, very few reports have been made of protein oxidation in IBD or
Fig. 3. Carbonyl formation on colonic mucosal proteins from control mice and mice exposed to 5% DSS. (A) Detection of carbonyl groups by immunoblotting using anti-(DNP) antibody with sizes (kDa) indicating carbonyl modified proteins. (B) Duplicate blot stained for protein with amido black. The DNP lane contains underivatized proteins for negative control.
experimental colitis [8]. In this study, protein oxidation in the form of carbonyl groups was investigated in the mucosa of mice with DSS-induced colitis. Although increased total protein carbonyl content was not observed in the inflamed mucosa, Western blotting with anti-DNP detection of derivatized proteins revealed differences in the carbonyl content of specific proteins, with approximate molecular weights of 48, 79, 95, and 131 kDa (Fig. 3), indicating that carbonyl oxidation of mucosal proteins had occurred in DSS-induced colitis. Similar analysis of mucosal homogenates exposed in vitro to a range of inflammatory oxidants provided insights into which oxidants may be contributing to carbonyl oxidation in the inflamed mucosa. In vitro oxidation of mucosa by iron/ascorbate, HOCl, and, at high concentrations,
268 •
A. C. BLACKBURN et al.
NO resulted in increased levels of protein carbonyl groups on colonic mucosal proteins (Tables 1 and 2), whereas Western blotting analysis of these proteins revealed different patterns of carbonyl formation for each oxidant (Figs. 1 and 2). Proteins susceptible to iron/ ascorbate-mediated carbonyl formation included 48- and 131-kDa proteins, and carbonyl formation by HOCl was observed on many proteins between 45 and 100 kDa, including 48, 79, and 95 kDa, corresponding to the oxidized proteins observed in the DSS– colitis mucosa. These results suggest that protein oxidation by both HOCl and MCO is occurring in the mucosa of DSSinduced colitis, supporting proposed mechanisms of RONS-mediated tissue injury in colitis. Evidence of HOCl involvement in colitis includes the finding of elevated myeloperoxidase levels as a result of neutrophil infiltration in both the DSS-induced colitis model [21] and in ulcerative colitis [27], whereas elevated chemiluminescence of whole biopsies taken from inflamed mucosa of IBD patients was reduced by azide, which inhibits myeloperoxidase, and the HOCl scavenger taurine [1,2], suggesting that neutrophil-derived HOCl is generated in the mucosa in active IBD. Aminosalicylates, used in the treatment of IBD, may inhibit this injury process, as they are potent scavengers of HOCl [28,29] and can inhibit myeloperoxidase [30]. Catalytic iron is potentially abundant in the inflamed intestinal mucosa, and its involvement in tissue injury in IBD has been hypothesized, although not thoroughly investigated. Luminal iron, which can be as high as 320 M in the human colon [31], and catalytic heme iron from mucosal bleeding may interact with inflammatory cell-derived oxidants to mediate MCO of the mucosa [4]. Evidence supporting this mechanism includes the detection of a metabolite produced by •OH attack on 5-aminosalicylate in fecal extracts from IBD patients on aminosalicylate therapy [32], the ability of 5-aminosalicylate to scavenge •OH in vitro [33,34], and the ability of iron chelation by desferrioxamine and •OH scavenging by dimethylsulfoxide to significantly attenuate formylmethionyl-leucyl-phenylalanine–induced mucosal damage in rats [35]. The involvement of •OH in DSS-induced colitis has been investigated using salicylate hydroxylation; however, the susceptibility of the hydroxylated salicylate derivatives to oxidation by HOCl and the severity of other metabolic changes during disease confounded the interpretation of these results [5]. The current study has provided further evidence that these oxidative processes may be involved in tissue injury in colitis, and has helped in identifying specific targets that can be investigated with respect to their role in the pathogenesis of disease. These proteins are targets for oxidation for several possible reasons. The relative abundance of the 48-kDa protein may account for its
susceptibility to oxidation by both of these oxidants, whereas the 95- and 131-kDa proteins are present in only small amounts and thus may be expected to possess features making them particularly susceptible to oxidation. Proteins with metal binding sites are vulnerable to MCO, especially of the residues in close proximity to the metal binding site [36,37], whereas carbonyl formation by HOCl can result from decomposition of primary chloramines that may form on exposed lysine residues or terminal amino groups [9,38,39]. Identification of the 48-kDa modified protein was attempted; however, the identity remains elusive, as this band contains at least three comigrating proteins. The overall changes to tissue proteins observed by applying the Western blotting technique to mucosa oxidized in vitro are consistent with previous studies of protein oxidation. In the mucosa, carbonyl formation by HOCl was preceded by depletion of thiols (Table 2), an order of events also observed in cultured cells exposed to HOCl [40]. Carbonyl formation was initially observed on selected mucosal proteins and was followed by extensive cross-linking (Fig. 2), a well-recognized phenomenon of HOCl-mediated protein oxidation that has been suggested to play a role in inflammatory processes such as arthritis and atherosclerosis [26,39,41]. In contrast to HOCl, carbonyl formation via a MCO mechanism may not require thiol depletion. Iron-mediated carbonyl formation is a site-specific reaction, where iron can bind to metal-ion binding sites on proteins and result in oxidation of nearby residues [36], thus reduced thiols can be spared even in the event of carbonyl formation [12]. The lack of protein cross-linking and fragmentation induced in mucosal homogenates exposed to iron/ascorbate contrasted with the pattern of oxidation induced by HOCl, a finding consistent with the changes observed in plasma [22] and in studies of enzyme inactivation after MCO, where aggregation or fragmentation have not been reported [36]. Hydrogen peroxide can generate protein carbonyls by participating in iron-mediated oxidation and •OH generation, a model used by many investigators for in vitro studies of MCO [36]. In the current investigation, mucosal catalase may have consumed the H2O2, thus protecting the proteins from both carbonyl and thiol oxidation (Table 1). On bovine serum albumin and in rat lung homogenates, peroxynitrite (ONOO⫺) exposure was shown to generate carbonyl groups; this was preceded by depletion of reduced thiols [42]. The formation of ONOO⫺ or other reactive nitrogen species is possible at the highest concentration of •NO used in the present study [43], which may account for the carbonyl formation observed. A recent study of oxidative protein modification in experimental acute pancreatitis applied the Western blotting technique [14]. Unlike our findings in the DSS-
Protein carbonyls in colitis
induced colitis model, increased carbonyl content in pancreatitis tissue compared with control tissue was evident in the blots, but selective oxidation of individual protein bands was not observed [14]. The model of pancreatitis used in that study was extremely acute, with a time course of only 24 h, during which carbonyl levels increased to three times control levels. This is more akin to the outcome produced by high concentrations of oxidant exposure in vitro in our studies, where the selective oxidation of proteins is masked by the high level of overall protein oxidation. The contrast of these two disease models illustrates that this technique is most useful at identifying susceptible proteins where overall protein oxidation levels are relatively low, thus potentially identifying molecules that are targets early in the oxidative injury process. One report of carbonyl levels in colonic biopsies from IBD patients, measured by derivitization with DNPH and spectrophotometric measurement of the derivatized proteins [44], found elevated carbonyl content in mucosal biopsies from patients with IBD compared with control patients, but not compared with noninflamed tissue from the same patient. In that paper, however, soluble nucleic acids were not removed from the biopsy homogenates, a step that we have shown to be important when dealing with inflamed mucosa. Thus, the relationship of the carbonyl levels measured to protein oxidation in that study remains to be confirmed. The recently described enzymelinked immunosorbent assay method [45] may be a more reliable method for quantitation of protein carbonyl groups in complex tissue samples of very limited size, such as colonic biopsies. The application of Western blotting analysis of protein carbonyls to DSS-induced colitis indicated several proteins that may be particular targets of carbonyl modification, with the pattern of oxidation seen in the inflamed mucosa bearing similarity to that obtained by in vitro oxidation of mucosa. This suggested that both HOCl and MCO may be involved in protein oxidation in DSS-induced colitis. Identification and further analysis of the mucosal proteins susceptible to carbonyl modification may lead to a better understanding of the contribution of oxidants to the tissue injury in the colonic mucosa in IBD. Acknowledgements — G.D.B. acknowledges the support of the NHMRC.
REFERENCES [1] Simmonds, N. J.; Allen, R. E.; Stevens, T. R.; van Someren, R. N.; Blake, D. R.; Rampton, D. S. Chemiluminescence assay of mucosal reactive oxygen metabolites in inflammatory bowel disease. Gastroenterology 103:186 –196; 1992. [2] Sedghi, S.; Fields, J. Z.; Klamut, M.; Urban, G.; Durkin, M.; Winship, D.; Fretland, D.; Olyaee, M.; Keshavarzian, A. In-
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
269
creased production of luminol enhanced chemiluminescence by the inflamed colonic mucosa in patients with ulcerative colitis. Gut 34:1191–1197; 1993. Middleton, S. J.; Shorthouse, M.; Hunter, J. O. Increased nitric oxide synthesis in ulcerative colitis. Lancet 341:465– 466; 1993. Babbs, C. F. Oxygen radicals in ulcerative colitis. Free Radic. Biol. Med. 13:169 –181; 1992. Blackburn, A. C.; Doe, W. F.; Buffinton, G. D. Salicylate hydroxylation as an indicator of hydroxyl radical generation in dextran sulfate-induced colitis. Free Radic. Biol. Med. 25:305– 313; 1998. Buffinton, G. D.; Doe, W. F. Altered ascorbic acid status in the mucosa from inflammatory bowel disease patients. Free Radic. Res. 22:131–143; 1995. Buffinton, G. D.; Doe, W. F. Depleted mucosal antioxidant defences in inflammatory bowel disease. Free Radic. Biol. Med. 19:911–918; 1995. McKenzie, S. J.; Baker, M. S.; Buffinton, G. D.; Doe, W. F. Evidence of oxidant-induced injury to epithelial cells during inflammatory bowel disease. J. Clin. Invest. 98:136 –141; 1996. Pereira, W. E.; Hoyano, Y.; Summons, R. E.; Bacon, V. A.; Duffield, A. M. Chlorination studies. II. The reaction of aqueous hypochlorous acid with ␣-amino acids and dipeptides. Biochim. Biophys. Acta. 313:170 –180; 1973. O’Neill, C. A.; Halliwell, B.; van der Vliet, A.; Davis, P. A.; Packer, L.; Tritschler, H.; Strohman, W. J.; Rieland, T.; Cross, C. E.; Reznick, A. Z. Aldehyde-induced protein modifications in human plasma: protection by glutathione and dihydrolipoic acid. J. Lab. Clin. Med. 124:359 –370; 1994. Blakeman, D. P.; Ryan, T. P.; Jolly, R. A.; Petry, T. W. Diquatdependent protein carbonyl formation. Identification of lipiddependent and lipid-independent pathways. Biochem. Pharmacol. 50:929 –935; 1995. Levine, R. L. Oxidative modification of glutamine synthetase. I. Inactivation is due to loss of one histidine residue. J. Biol. Chem. 258:11823–11827; 1983. Chapman, M. L.; Rubin, B. R.; Gracy, R. W. Increased carbonyl content of proteins in synovial fluid from patients with rheumatoid arthritis. J. Rheumatol. 16:15–18; 1989. Reinheckel, T.; Nedelev, B.; Prause, J.; Augustin, W.; Schulz, H. U.; Lippert, H.; Halangk, W. Occurrence of oxidatively modified proteins: an early event in experimental acute pancreatitis. Free Radic. Biol. Med. 24:393– 400; 1998. Starke-Reed, P. E.; Oliver, C. N. Protein oxidation and proteolysis during aging and oxidative stress. Arch. Biochem. Biophys. 275: 559 –567; 1989. Serbinova, E.; Khwaja, S.; Reznick, A. Z.; Packer, L. Thioctic acid protects against ischemia-reperfusion injury in the isolated perfused langendorff heart. Free Radic. Res. Commun. 17:49 –58; 1992. Liu, Y.; Rosenthal, R. E.; Starke-Reed, P.; Fiskum, G. Inhibition of postcardiac arrest brain protein oxidation by acetyl-L-carnitine. Free Radic. Biol. Med. 15:667– 670; 1993. Okayasu, I.; Hatakeyama, S.; Yamada, M.; Ohkusa, T.; Inagaki, Y.; Nakaya, R. A novel method in the induction of reliable experimental acute and chronic ulcerative colitis in mice. Gastroenterology 98:694 –702; 1990. Cooper, H. S.; Murthy, S. N.; Shah, R. S.; Sedergran, D. J. Clinicopathologic study of dextran sulfate sodium experimental murine colitis. Lab. Invest. 69:238 –249; 1993. Wozniak, A.; Sharma, D. P.; van de Pol, E.; McColl, S. R.; Doe, W. F. Depletion of neutrophils decreases disease severity and mortality in a mouse model of ulcerative colitis. Gastroenterology 108(Suppl.):A944; 1995. Blackburn, A. C.; Doe, W. F.; Buffinton, G. D. Colonic antioxidant status in dextran sulfate-induced colitis in mice. Inflammatory Bowel Dis. 3:198 –203; 1997. Shacter, E.; Williams, J. A.; Lim, M.; Levine, R. L. Differential susceptibility of plasma proteins to oxidative modification: examination by Western blot immunoassay. Free Radic. Biol. Med. 17:429 – 437; 1994.
270
A. C. BLACKBURN et al.
[23] Levine, R. L.; Garland, D.; Oliver, C. N.; Amici, A.; Climent, I.; Lenz, A. G.; Ahn, B. W.; Shaltiel, S.; Stadtman, E. R. Determination of carbonyl content in oxidatively modified proteins. Methods Enzymol. 186:464 – 478; 1990. [24] Brunk, C.; Jones, K.; James, T. Assay for nanogram quantities of DNA in cellular homogenates. Anal. Biochem. 92:497–500; 1979. [25] Jocelyn, P. C. Spectrophotometric assay of thiols. Methods Enzymol. 143:44 – 67; 1987. [26] Vissers, M. C.; Winterbourn, C. C. Oxidative damage to fibronectin. I. The effects of the neutrophil myeloperoxidase system and HOCl. Arch. Biochem. Biophys. 285:53–59; 1991. [27] Izzo, R. S.; Witkon, K.; Chen, A. I.; Hadjiyane, C.; Weinstein, M. I.; Pellecchia, C. Interleukin-8 and neutrophil markers in colonic mucosa from patients with ulcerative colitis. Am. J. Gastroenterol. 87:1447–1452; 1992. [28] Williams, J. G.; Hallett, M. B. The reaction of 5-amino-salicylic acid with hypochlorite. Implications for its mode of action in inflammatory bowel disease. Biochem. Pharmacol. 38:149 –154; 1989. [29] Aruoma, O. I.; Wasil, M.; Halliwell, B.; Hoey, B. M.; Butler, J. The scavenging of oxidants by sulphasalazine and its metabolites. A possible contribution to their anti-inflammatory effects? Biochem. Pharmacol. 36:3739 –3742; 1987. [30] Kettle, A. J.; Winterbourn, C. C. Mechanism of inhibition of myeloperoxidase by anti-inflammatory drugs. Biochem. Pharmacol. 41:1485–1492; 1991. [31] Babbs, C. F. Free radicals and the etiology of colon cancer. Free Radic. Biol. Med. 8:191–200; 1990. [32] Ahnfelt-Ronne, I.; Nielsen, O. H.; Christensen, A.; Langholz, E.; Binder, V.; Riis, P. Clinical evidence supporting the radical scavenger mechanism of 5-aminosalicylic acid. Gastroenterology 98:1162–1169; 1990. [33] Betts, W. H.; Whitehouse, M. W.; Cleland, L. G.; VernonRoberts, B. In vitro antioxidant properties of potential biotransformation products of salicylate, sulphasalazine and amidopyrine. Free Radic. Biol. Med. 1:273–280; 1985. [34] Allgayer, H.; Hofer, P.; Schmidt, M.; Bohne, P.; Kruis, W.; Gugler, R. Superoxide, hydroxyl and fatty acid radical scavenging by aminosalicylates. Direct evaluation with electron spin resonance spectroscopy. Biochem. Pharmacol. 43:259 –262; 1992. [35] von Ritter, C.; Grisham, M. B.; Hollwarth, M.; Inauen, W.; Granger, D. N. Neutrophil-derived oxidants mediate formyl-methionyl-leucyl-phenylalanine-induced increases in mucosal permeability in rats. Gastroenterology 97:778 –780; 1989. [36] Stadtman, E. R. Metal ion-catalyzed oxidation of proteins: biochemical mechanism and biological consequences. Free Radic. Biol. Med. 9:315–325; 1990. [37] Amici, A.; Levine, R. L.; Tsai, L.; Stadtman, E. R. Conversion of
[38]
[39]
[40]
[41]
[42] [43]
[44]
[45]
amino acid residues in proteins and amino acid homopolymers to carbonyl derivatives by metal-catalyzed oxidation reactions. J. Biol. Chem. 264:3341–3346; 1989. Stelmaszynska, T.; Zgliczynski, J. M. N-(2-Oxoacyl)amino acids and nitriles as final products of dipeptide chlorination mediated by the myeloperoxidase/H2O2/CI⫺ system. Eur. J. Biochem. 92: 301–308; 1978. Hazell, L. J.; van-den Berg, J. J.; Stocker, R. Oxidation of lowdensity lipoprotein by hypochlorite causes aggregation that is mediated by modification of lysine residues rather than lipid oxidation. Biochem. J. 302:297–304; 1994. Schraufstatter, I. U.; Browne, K.; Harris, A.; Hyslop, P. A.; Jackson, J. H.; Quehenberger, O.; Cochrane, C. G. Mechanisms of hypochlorite injury of target cells. J. Clin. Invest. 85:554 –562; 1990. Jasin, H. E. Oxidative cross-linking of immune complexes by human polymorphonuclear leukocytes. J. Clin. Invest. 81:6 –15; 1988. Ischiropoulos, H.; Al-Mehdi, A. B. Peroxynitrite-mediated oxidative protein modifications. FEBS Lett. 364:279 –282; 1995. Crow, J. P.; Beckman, J. S. Reactions between nitric oxide, superoxide, and peroxynitrite: footprints of peroxynitrite in vivo. Adv. Pharmacol. 34:17– 43; 1995. Lih-Brody, L.; Powell, S. R.; Collier, K. P.; Reddy, G. M.; Cerchia, R.; Kahn, E.; Weissman, G. S.; Katz, S.; Floyd, R. A.; McKinley, M. J.; Fisher, S. E.; Mullin, G. E. Increased oxidative stress and decreased antioxidant defenses in mucosa of inflammatory bowel disease. Dig. Dis. Sci. 41:2078 –2086; 1996. Buss, H.; Chan, T. P.; Sluis, K. B.; Domigan, N. M.; Winterbourn, C. C. Protein carbonyl measurement by a sensitive ELISA method. Free Radic. Biol. Med. 23:361–366; 1997.
ABBREVIATIONS
DNP— dinitrophenyl DNPH— dinitrophenylhydrazine DSS— dextran sulfate sodium IBD—inflammatory bowel disease MCO—metal-catalyzed oxidation PAGE—polyacrylamide gel electrophoresis RONS—reactive oxygen and nitrogen species SDS—sodium dodecyl sulfate TCA—trichloroacetic acid