2022 Volume 69 Issue 12 Pages 1387-1394
Protein carbonylation is an irreversible and degenerative modification that can be used to evaluate oxidative stress caused by glucocorticoids. In this study, we focused on protein carbonylation in dogs with hypercortisolism (HC). Sera samples were collected from 14 dogs diagnosed with HC and treated with trilostane, 12 dogs with inflammatory diseases (disease control group), and eight clinically healthy dogs. When the carbonylated protein levels were detected by the immunoblot analysis, one band of approximately 40 kDa was predominantly increased in the dogs with HC. The band was identified as haptoglobin using the liquid chromatography tandem mass spectrometry method. Furthermore, haptoglobin immune reactivity was higher in the dogs with HC. Although the average protein carbonylation level of the HC group was not significantly different from that of the other groups, the carbonylation level was significantly higher for the poorly controlled HC cases than for the well-controlled HC group. Additionally, the primary culture of canine hepatocytes was used to clarify the direct effect of glucocorticoids on protein carbonylation in dog livers. Both the carbonylated protein and haptoglobin clearly increased after 72 h. These findings suggest that haptoglobin and its carbonylated form are increased with canine HC, and that the protein carbonylation ratio and/or haptoglobin level could be related to disease management. These factors could be useful as biomarkers for an oxidative stress reaction, at least in the liver, and for treatment monitoring of HC.
HYPERCORTISOLISM (HC), also known as Cushing’s syndrome and hyperadrenocorticism, has various clinical signs that are caused by chronic exposure to glucocorticoids [1-3]. Excess glucocorticoids are closely related to the induction of superoxide production in various tissues [4]. The serum contents of several oxidative stress markers, such as lipid peroxidation and F2-isoprostanes, are enhanced in human patients with HC [4-6]. Additionally, a chronic overdose of corticosterone increases the concentrations of carbonylated proteins in the brain, liver, and heart of rats [7]. Plasma protein carbonylation is an irreversible and degenerative modification [8-10] that can be used to evaluate protein degradation caused by excess free radicals. Plasma malondialdehyde, an indicator of oxidative stress, has been reported to be high in human patients with overt HC compared with cured HC patients and healthy individuals [6]. Therefore, protein carbonylation is considered a marker of pathophysiological malfunction in various tissues related to HC in dogs and humans.
HC is a common endocrine disease in dogs. The annual prevalence of HC is approximately 1–2 cases per 1,000 dogs, while the occurrence of HC in humans is rare [11]. As the clinical presentation of HD is highly similar between dogs and humans, canine HC often serves as an animal model for human HC [11]. During a recent study, lipid peroxidation was increased in canine HC, suggesting that controlling the disease would be beneficial for normalizing the state of oxidative stress [12]. However, there have been only a few studies related to protein carbonylation in dogs. Transportation stress experienced by dogs increases serum oxidative stress markers, including protein carbonylation [13]. The administration of antioxidants, such as polyunsaturated fatty acids, has been reported to suppress oxidative stress markers and protein carbonylation in aging dogs [14]. To the best of our knowledge, no reports have investigated a direct connection between carbonylated proteins and HC in dogs.
During this study, we focused on protein carbonylation with canine HC. One prevalent carbonylated protein band was found to be increased in the serum of dogs with HC; it was identified using the liquid chromatography tandem mass spectrometry (LC-MS/MS) method. Because it was predicted to be haptoglobin, its immune intensity and the ratio of carbonylation were also investigated. Moreover, to clarify the direct effect of glucocorticoids on protein carbonylation in the dog livers, a primary culture of canine hepatocytes was used. Therefore, we evaluated whether protein carbonylation and/or haptoglobin intensity could be a useful oxidative stress marker for the medical management of HC in dogs.
The present study was designed as a retrospective study. Serum samples from dogs that visited the Veterinary Medical Center of the University of Tokyo for follow-up of HC and other inflammatory diseases between June 2012 and January 2014 were used. Theses samples were the residual blood samples acquired during the follow-up examination. Written informed consent for secondary blood use for research was received from all dog owners during this study. Fourteen dogs with HC were included: six males, five castrated males, two females, and one spayed female. The age range of the dogs was 6 to 15 years (mean, 11.2 years). The breeds were Bichon Frise (one), Bull Terrier (one), Chihuahua (two), French Bull Dog (two), Maltese (one), Miniature Dachshund (four), Pomeranian (one), Shetland Sheepdog (one), and Shih Tzu (one). These 14 dogs with HC included 12 with pituitary-dependent hypercortisolism and two with functional adrenal tumors. They had been diagnosed with HC, such as pituitary hypercortisolism or an adrenal tumor, according to their history, clinical signs, adrenocorticotropic hormone stimulation test results, and abdominal ultrasound test results at least 2 months before inclusion in this study [15]. All dogs were treated with an appropriate dose of trilostane (range, 0.5–3.0 mg/kg once or twice daily [15]). Their conditions were evaluated based on their medical records, including the general physical condition, polydipsia/polyuria, integumentary issues (such as hair loss and/or dermatosis), adrenal tumors, and other issues related to HC. Dogs with no problems or one problem were classified as the well-controlled group; those with more than one problem were classified as poorly controlled group. Of the 14 dogs with HC, there were six in the well-controlled group and eight in the poorly controlled group. As the disease control group, 12 dogs that did not have HC but did have inflammatory diseases were used as the inflammatory disease group. There was one male, four castrated males, three females, and four spayed females. Their diagnoses were allergic dermatitis (two), arthritis (two), bacterial rhinitis (one), cholangiohepatitis (two), Malassezia infection (one), pancreatitis (two), and superficial pyoderma (two). The breeds were American Eskimo (one), Chihuahua (one), Chin (one), Kaninchen Dachshund (one), Miniature Dachshund (two), mixed breed (one), Shih Tzu (one), Toy poodle (one), and Yorkshire terrier (three). The age range was 2 to 13 years (mean, 8.6 years). Addition, the stocked sera of eight clinically healthy spayed female Beagle dogs that had been maintained by our laboratory (9–13 years old) were used as the healthy control group.
Collecting blood samplesWhole blood was collected from the dogs via jugular vein puncture. After 30 min of incubation at room temperature for blood coagulation, the serum was separated by centrifugation at room temperature at 1,500 × g for 5 min. The obtained serum was transferred to a micro tube and stored frozen at –30°C until use.
Detection of carbonylated protein in dog serumA common method of detecting carbonylated proteins in a biological sample involves generating a reaction between the carbonyl groups with 2,4-Dinitrophenylhydrazine to generate 2,4-Dinitrophenylhydrazone [16-18]. During this study, a carbonylated protein in dog serum was detected using a commercially available kit based on the aforementioned method (OxyBlot kit, S7150; Millipore, Darmstadt, Germany). According to the manufacturer’s instructions, this kit provides the reagents needed to perform immunoblot detection of carbonyl groups that are introduced into proteins via oxidative reactions with ozone or oxides of nitrogen or by metal-catalyzed oxidation. First, the total serum protein concentration was measured using a protein assay kit (BioRad, Hercules, CA, USA). Then, the serum was diluted with ultra-pure water to a protein concentration of 5.5 μg/μL. After dilution, 5 μL of diluted serum, 5 μL of 12% SDS solution (Wako, Osaka, Japan), and 10 μL of 2,4-Dinitrophenylhydrazine solution were mixed in a micro tube for 15 min at room temperature. The reaction was stopped by adding 7.5 μL of neutralization solution. This conjugated solution was electrophoresed using a 10% polyacrylamide gel and transferred to a nitrocellulose membrane (Hybond ECL Nitrocellulose Blotting Membrane; Amersham, Munich, Germany). The membrane was blocked with 1% bovine serum albumin (Sigma, St. Louis, MO, USA) and reacted with the primary antibody (rabbit anti-DNP polyclonal antibody) in the kit. After this reaction, the membrane was washed using washing solution (Tris-buffered saline, 0.1% Tween 20) and reacted with horseradish peroxidase-labeled secondary antibody (goat anti-rabbit IgG polyclonal antibody) from the kit. The membrane was illuminated using Pierce Western Blotting Substrate Plus (Thermo Scientific, Rockford, IL, USA) and sensitized to the film (Hyperfilm ECL, high-performance chemiluminescence film; Amersham). The obtained image was digitized using gel imager (Gel Doc EZ Imager, BioRad) and analyzed using image analysis software (Image Lab version 5.0; BioRad). The relative index values of carbonylated protein and haptoglobin were normalized to the value of healthy control number 1.
Identification of carbonylated protein using LC-MS/MSThe serum protein from the dogs with HC that formed a characteristic carbonylated band at approximately 40 kDa was identified using the LC-MS/MS method. The protein band was separated using the linear SDS-PAGE and cut out (Fig. 1). LC-MS/MS identification was determined by APRO Science (Tokusima, Japan). The database was searched for the spectra detected by mass spectrometry using the MASCOT program (www.matrixscience.com). The proteins that were determined to be significant were listed as candidate proteins.
The specific protein band and LC-MS/MS analysis of hypercortisolism (HC). A) Representative SDS-PAGE image of the sera of the control group and that of one of the dogs in the poorly controlled HC group. The band surrounding a white square was identified using LC-MS/MS. B) Canine haptoglobin alpha amino acid sequence, which is the most likely candidate according to the LC-MS/MS analysis results. Gray letters indicate amino acids that were covered (44% of the whole protein). Dark gray letters indicate exclusive unique amino acids.
The membranes that were used to detect the carbonylated protein level were reused for Western blotting. The antibodies were removed from the aforementioned nitrocellulose membrane using an acidic washing solution. Then, the membrane was blocked with bovine serum albumin (3%) and reacted with horseradish peroxidase-labeled anti-human haptoglobin antibody (1:5,000 dilution, ab8968; Abcam, Cambridge, UK) for 1 h. After washing, the film was illuminated with ECL Western Blotting Detection Reagents (Amersham GE Healthcare, Amersham, UK). The obtained images were digitized and analyzed using BioRad devices as described. The intensities of each band were measured using densitometry analysis (ImageJ 1.42q; National Institutes of Health, Bethesda, MD, USA). The relative index values of carbonylated protein and haptoglobin were normalized to the value of healthy control number 1.
Dexamethasone treatment of cultured canine hepatocytesCommercially available canine primary hepatocytes and culture media were used for in vitro experiments during this study. Dog Beagle cryopreserved hepatocyte (batch no. HEP185042), including the cell seeding solution and culture media, were supplied by Biopredic International (Rennes, France). The viable cell rate of thawed canine hepatocytes was 80.6% as determined by trypan blue staining. These cells were suspended in the cell seeding solution and seeded in a 24-well plate at 70,000 cells/well; they had been previously coated with collagen I. After the cells were cultured for 24 h, the media were replaced by the culture media. Additionally, the culture media were mixed with the final concentrations of 0 or 5 μM dexamethasone phosphate (Corson P injection; Intervet, Tokyo, Japan) and incubated for 24 h and 72 h at 37°C with 5% CO2. After treatment, the culture supernatant of each well was collected and centrifuged at 3,000 × g for 5 min and stored at –30°C. The hepatocytes attached to the plate were washed twice with phosphate-buffered saline and then scraped and collected using the lysis buffer (lysis buffer; Biopredic International). The lysate solution was centrifuged at 3,000 × g for 10 min, and the supernatant was stored at –30°C until use. The culture media with 5 μL of supernatants and 0.5 μg of cell lysates were applied to detect the protein carbonylation and haptoglobin using Western blotting and the same aforementioned method used for the sera.
Statistical analysisAll statistical analyses were conducted using XLSTAT Life Science (version 2021.2.2.1141; Addinsoft, Paris, France) as an add-in for MS Excel (Microsoft Corporation, Redmond, WA, USA). The Kruskal-Wallis and Dunn-Bonferroni post hoc tests were used for comparisons among the three groups. The Mann–Whitney U test was used for comparisons between the well-controlled and poorly controlled groups. Significance was considered when p < 0.05 for all statistical comparisons.
When carbonylated proteins were detected, the carbonylation intensity of the proteins varied depending on the individual. However, a characteristic carbonylated protein band of approximately 40 kDa was strong in one of the dogs in the poorly controlled group compared with that in the control group. A protein corresponding to this band was cut out and identified using the LC-MS/MS method (Fig. 1A). Canine haptoglobin alpha chain (GI: 123,511, 36.457 kDa) was nominated as a likely candidate (Fig. 1B). Twelve exclusive unique peptides, 18 exclusive unique spectra, 42 total spectra, and 154 of 329 amino acids were covered. To confirm the existence of haptoglobin, the same membranes were reused for immunoblotting with a haptoglobin antibody. Haptoglobin immunoreactivity was observed at the same position as the characteristic carbonylated protein band (Fig. 2). Carbonylation of the other protein bands varied in individual cases (Fig. 2).
Representative image of the protein carbonylation signals and immunoblot of haptoglobin using the same membrane. The OxyBlot method was used to detect carbonylated proteins (see Materials and Methods).
To clarify the alteration of the haptoglobin and its carbonylation caused by HC, the intensity of the approximately 40-kDa carbonylated protein and haptoglobin immunoreactivity were detected (Figs. 2 and 3). Although it was not significantly changed (p = 0.228) in the sera of the dogs with HC compared with that of the healthy control group and the inflammatory disease control group, the carbonylated protein level tended to be increased (Fig. 3A). All cases with index values more than 3.0 were observed in the HC group. The immune reactivity of haptoglobin was statistically stronger in the HC group than in the healthy control group. The ratio of carbonylated protein to haptoglobin was not changed among these groups.
Relative levels of carbonylated protein (CP), haptoglobin (HG), and the ratio of CP to HG in the sera. Individual plots and median bars are shown for the healthy control, HC, and inflammatory diseases (IDs) groups (A), and for the poorly controlled HC and well-controlled HC groups (B).
Fig. 3B shows the stratified analysis results based on the medical condition of dogs with HC. Both the carbonylated protein level and the ratio of carbonylated protein to haptoglobin were significantly lower in the well-controlled group than in the poorly controlled group. However, haptoglobin immunoreactivity was not changed between these groups.
To investigate the direct effect of glucocorticoids on protein carbonylation in the liver of dogs, the primary culture method of canine hepatocytes was used (Fig. 4). Although there were no clear changes 24 h after adding dexamethasone to the media, both the amount of 40-kDa carbonylated proteins and haptoglobin immune intensity in the cell lysate had clearly increased after 72 h. In the media, there was only one clear band of carbonylated protein that was considered to be haptoglobin.
Effects of glucocorticoids on protein carbonylation and haptoglobin immune intensity of the liver cells of dogs in vitro. Protein carbonylation signals and immunoblot of haptoglobin in the culture media (A) and cell lysate (B) of canine hepatocytes under the primary culture condition. DEX, dexamethasone. The incubation time after the addition of DEX or vehicle was measured in hours.
Oxidative stress is considered one of the aggravating factors of HC [4-6, 12]. During this study, the carbonylated protein of approximately 40 kDa was detected in most dogs with HC. However, there were no statistically significant changes between the HC group and the other groups. All HC patients included in this study had been previously treated with trilostane for at least 2 months, which might have affected the variability of these data. Nevertheless, all cases with index values more than 3.0 were observed in the HC group. Our findings during this study suggest the presence of oxidative stress and protein degeneration in dogs with HC and are coincident with those of previous studies that investigated other oxidative stress markers in humans and dogs with excessive exposure to glucocorticoids [5, 6, 12]. To the best of our knowledge, this is the first report of the protein carbonylation profile of the dogs with HC.
Oxidative stress markers are believed to be altered by the medical management of the disease. Previous studies indicated that lipid peroxidation was reduced by stabilizing HC using medical therapy [5, 12]. During this study, the carbonylated protein level of the well-controlled HC group was clearly lower than that of the poorly controlled HC group. Our data suggest that protein carbonylation, which is one of the oxidative stress reactions, could be influenced by the medical control of HC. Although various methods have been considered for monitoring the health status after starting trilostane therapy, the gold standard has not yet been established for dogs with HC [19-23]. Additionally, the present results suggest that the carbonylation level of a 40-kDa protein could be a new factor used to monitor the medical condition of dogs with HC.
During this study, the approximately 40-kD band was identified as canine haptoglobin using the LC-MS/MS method. It cannot be determined whether all the contents of this band originated from haptoglobin because spot detection methods, such as two-dimensional gel electrophoresis, have not been performed. However, considering the Western blot analysis results for haptoglobin, at least most of the approximately 40-kD carbonylated protein could be haptoglobin. Several studies have reported that the haptoglobin concentration is higher in dogs with HC [24-26], thus supporting our findings. However, the haptoglobin level has been reported to decrease during trilostane treatment [27-29]. Another recent study of dogs with HC reported that haptoglobin is significantly associated with clinical scores [21]. During this study, the immune intensity of haptoglobin did not clearly alter based on the HC conditions, but the carbonylated haptoglobin ratio was low in well-managed HC patients treated with trilostane. These data suggest that the carbonylation intensity of haptoglobin could be a more sensitive marker than haptoglobin itself.
Canine haptoglobin is similar in nature to human haptoglobin, and it is often used as an experimental model for human haptoglobin studies [30-32]. Haptoglobin is an acute phase protein produced in response to internal events such as inflammation [33]. To evaluate whether HC patients could be distinguished from patients with other diseases, protein carbonylation and haptoglobin in patients with several inflammatory diseases were measured during this study. Although the levels of protein carbonylation and haptoglobin were not statistically different between dogs with HC and the disease control group, no cases in the disease control group had carbonylated protein levels more than 3.0. It is suggested that the carbonylation level could be a biomarker for differential diagnoses as well. Moreover, because oxidative stress and carbonylated protein production are supposedly caused by HC in humans [4-6]. Because canine HC often serves as an animal model for human HC [11], our data obtained by studying dogs could be extrapolated to humans in the future.
The precise effect of excessive cortisol on oxidative stress and protein carbonylation remains unknown. It has been reported that glucocorticoids induce the overproduction of reactive oxygen species, thereby causing dysregulation of physiological processes [4]. According to several studies, antioxidant drugs have a hepatoprotective effect on dogs [34, 35], and such drugs are frequently prescribed for HC cases. In the rat cell line, glucocorticoids increase protein carbonylation via mitochondrial dysfunction in a dose-dependent manner [36]. Furthermore, lipid peroxidation is augmented along with corticosterone levels in rats with chronic stress [37]. Haptoglobin is synthesized and reactive oxygen species are generated by the liver [38, 39]. During this study, the long exposure of canine hepatocytes to dexamethasone resulted in increased protein carbonylation and haptoglobin levels in vitro, suggesting that glucocorticoids could initiate an oxidative stress reaction, at least in the liver of dogs. Additionally, the ratios of carbonyl protein to haptoglobin did not change among the healthy, HC, and inflammatory disease dogs. In experiments using hepatocytes, the addition of dexamethasone increased the amount of both haptoglobin and carbonylated proteins. These results could mean that glucocorticoids increase the amount of haptoglobin but do not induce carbonylation. Because protein carbonylation is facilitated by iron and haptoglobin binds to iron-containing hemoglobin, it could be carbonylated easily by nature. It is possible that the glucocorticoid-induced increase in the amount of carbonylated proteins observed in this study is due to an increase in the amount of haptoglobin and not because carbonylation was accelerated by glucocorticoids. However, there were no additional data related to iron metabolism or any other evidence to clarify these mechanisms. A more precise molecular mechanism and/or signal cascade should be clarified by future studies.
In the clinical cases, protein carbonylation, haptoglobins, and their ratios had individual characteristics when compared between the well-controlled HC group and poorly controlled HC group. Protein carbonylation and its ratio was lower in the well-controlled HC group than in the poorly controlled HC group. It was speculated that protein carbonylation could be improved more than haptoglobin production when HC is well-controlled. Additionally, this speculation could suggest why only the haptoglobin level was high in treated HC dogs compared to the dogs in the healthy and disease control groups. To prove this hypothesis, future studies comparing dogs with untreated HC and dogs with HC undergoing treatment are needed.
There were certain limitations to this study. With regard to limitations in the research design, the main concern of this research was that no dogs with HC were studied before trilostane treatment was administered. Time-course samples from the same patients during the treatment period were not collected. Despite not being able to bind glucocorticoid receptors, trilostane is a synthetic steroid. This study did not clarify whether trilostane would cause protein carbonylation. Only one method was used to measure protein carbonylation. The protein amount in each well was carefully adjusted according to the results of the protein assay, but reconfirmation through staining of housekeeping proteins or Ponceau S on the Western blotting membranes was not performed. Regarding limitations in the statistical analysis, the correlations may have been influenced by the small sample size. This study was retrospective. The in vitro experiment was singly performed and not statistically analyzed because of the high expense of commercial canine frozen hepatocytes.
In summary, the 40-kDa carbonylated protein, identified as haptoglobin, was strongly detected in dogs with HC. Although the carbonylation level of this protein was not significantly different among all dogs with HC and dogs in the other groups, it was significantly higher in dogs in the poorly controlled HC group compared with dogs in the well-controlled HC group. The 40-kDa protein carbonylation and haptoglobin were induced by the addition of a glucocorticoid to the canine primary hepatocytes. These findings suggest that protein carbonylation and haptoglobin are increased with canine HC, and that good disease management could reduce protein carbonylation ratio and/or haptoglobin levels. Our findings also suggest that these factors could be useful as biomarkers for the oxidative stress reaction, at least in the liver, and for monitoring the treatment results of dogs and humans with HC.
None of the authors has any potential conflicts of interest associated with this research.
