Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 DOI 10.1186/s12917-016-0839-9 RESEARCH ARTICLE Open Access Obese dogs with and without obesity- related metabolic dysfunction – a proteomic approach Asta Tvarijonaviciute1,2*, Jose J. Ceron2, Carlos de Torre3, Blanka B. Ljubić4, Shelley L. Holden5, Yann Queau6, Penelope J. Morris7, Josep Pastor1 and Alexander J. German5 Abstract Background: Approximately 20 % of obese dogs have metabolic disturbances similar to those observed in human metabolic syndrome, a condition known as obesity-related metabolic dysfunction. This condition is associated with insulin resistance and decreased circulating adiponectin concentrations, but clinical consequences have not been reported. In order to define better the metabolic changes associated with obesity-related metabolic dysfunction (ORMD), we compared the plasma proteomes of obese dogs with and without ORMD. A proteomic analysis was conducted on plasma samples from 8 obese male dogs, 4 with ORMD and 4 without ORMD. The samples were first treated for the depletion of high-abundance proteins and subsequently analysed by using 2-DE DIGE methodology. Results: Using mass spectrometry, 12 proteins were identified: albumin, apoliprotein A-I, C2, C3, C5, C4BPA, A2M, Uncharacterised protein (Fragment) OS = Canis familiaris, fibrinogen, IGJ, ITIH2, and glutathione peroxidase. In obese dogs with ORMD, the relative amounts of ten proteins (albumin, apoliprotein A-I, C2, C3, C5, C4BPA, A2M, Uncharacterised protein (Fragment) OS = Canis familiaris, fibrinogen, and ITIH2) were increased and two proteins (IGJ and glutathione peroxidase) were decreased, compared with obese dogs without ORMD. Specific assays were then used to confirm differences in serum albumin, apoliprotein A-I and glutathione peroxidase in a separate group of 20 overweight dogs, 8 with ORMD and 12 without ORMD. Conclusions: The current study provides evidence that, in obese dogs with ORMD, there are changes in expression of proteins involved in lipid metabolism, immune response, and antioxidant status. The clinical significance of these changes remains to be defined. Keywords: Antioxidants, Complement system, Dog, Immune response, Lipid metabolism, Metabolic syndrome, Obesity Background been estimated that one-quarter of the world’s adult In humans metabolic syndrome (MetS) is an extensively population has the MetS [1]. Recently, when human studied entity that comprises obesity, hypertension, dys- MetS criteria were adapted for dogs to define the condi- lipidaemia, and glucose intolerance [1], and is associated tion of obesity-related metabolic dysfunction (ORMD), with the development of cardiovascular disease and type approximately 20 % of dogs with naturally occurring 2 diabetes [2]. The prevalence of metabolic syndrome in obesity were described as presenting with concurrent humans varies with age and ethnic groups, and it has ORMD [3]. The presence of ORMD was associated with increased circulating insulin and decreased adiponectin * Correspondence: asta@um.es concentrations, suggesting that ORMD might be associ- 1 Departament de Medicina i Cirugia Animals, Universitat Autónoma de ated with a range of metabolic disturbances. However, to Barcelona, 08193 Barcelona, Spain 2 Interdisciplinary Laboratory of Clinical Pathology, Iterlab-UMU, Campus of date, no study has identified any direct association between Excellence Mare Nostrum, University of Murcia, Murcia, Spain these disturbances and clinical diseases of obese dogs, such Full list of author information is available at the end of the article © 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 2 of 9 that the pathophysiological significance of the changes that These dogs were otherwise healthy as recorded on occur in ORMD remains unknown. physical examination, CBC and biochemistry analyses. Proteomic analysis has been increasingly used in All procedures were approved by the Ethics Committees human medicine and experimental animals in order to of University of Murcia and Murcia Region; in addition, evaluate possible changes in blood, adipose tissue, or adi- an informed consent was obtained from the owner before pocyte proteomes in obese subjects [4–7], with the aim of inclusion of the dog in the study. identifying and clarifying the underlying mechanisms re- sponsible for development of secondary diseases. To the Clinical measurements authors’ knowledge, the only proteomic study to date in Body condition score was determined using a validated canine obesity was performed under experimental condi- 9-point body condition score system in both groups tions, and identified three proteins (retinol binding protein [13, 14]. In addition for group 1 dogs, body composition 4, clusterin, and α-1 antitrypsin) that differed between was analysed using fan-beam DEXA (Lunar Prodigy obese and lean dogs [3]. Whilst this study demonstrated Advance; GE Lunar) [15–17]. Blood pressure in all dogs proof of principle, it is possible to refine the technique re- was measured non-invasively using an oscillometric ported and improve detection sensitivity by first depleting method. All dogs were fully conscious. A cuff of the ap- high-abundance protein species in the biological samples propriate size (e.g. the cuff chosen had a width of ~40 % examined [8]. Such a depletion technique has been circumference of the leg) was placed on the right forelimb. performed in some human proteomic studies of obesity Once the dog was calm and still, at least five systolic blood [9, 10]. In addition, when comparing protein differences pressure (SBP) readings were taken and averaged. Finally, between two samples, fluorescence two-dimensional dif- blood samples were collected by jugular venepuncture ferential gel electrophoresis (2-DE DIGE) has shown bene- after a fast of at least 16 h. fits over conventional 2-DE methodology, since it enables two different samples to be run on the same gel in both Definition of obesity-related metabolic dysfunction dimensions thus minimising gel-to-gel variation [11, 12]. Dogs were considered as having ORMD if the following The main objective of the current study was to compare criteria were met: (a) BCS 7-9/9; AND (b) any two of the plasma proteomes of obese dogs with and without the following: 1) plasma triglycerides > 2.3 mmol/L; 2) ORMD in order to evaluate if the presence of ORMD plasma cholesterol > 7.8 mmol/L; 3) SBP > 160 mmHg; 4) imply changes in proteome of obese dogs. fasting plasma glucose > 5.6 mmol/L, or previously diag- nosed diabetes mellitus [3]. Methods Animals Dog plasma pre-treatment (ProteoMiner) The study protocol adhered to the University of Liverpool Plasma samples from group 1 dogs were treated with Animal Ethics Guidelines, and was approved by both the ProteoMiner treatment according to manufacturer’s rec- University of Liverpool Research Ethics Committee and ommendation (BioRad Laboratories, Hercules, CA.). the WALTHAM Animal Welfare & Ethical Review Board. Briefly, 20 μL aliquots of ProteoMiner beads (small cap- All clinical procedures were performed for the direct acity kit) were equilibrated with washed buffer (phosphate benefit of the animals, and their owners gave informed buffer saline (PBS): 150 mM NaCl, 10 mM NaH2PO4, written consent. pH 7.4). Subsequently, 200 μL plasma samples were Plasma samples from eight neutered male obese dogs applied and incubated for 2 h at room temperature (group 1; 4 with ORMD, and 4 without ORMD) were under constant rotation. After incubation, the beads used in the proteomic study. All dogs were originally re- were washed three times with 200 μL wash buffer to ferred to the Royal Canin Weight Management Clinic, remove unbound proteins. Bound proteins were eluted University of Liverpool UK for investigation and manage- with three washes of 20 μL elution reagent (8 M urea, 2 % ment of obesity. Any surplus plasma, after that required CHAPS). Eluted proteins were cleaned using 2D clean-up for use in routine clinical biochemical analyses, was frozen kit (GE Healthcare Europe GmbH, Freiburg, Germany) at −20 °C, usually within 30 min of sample collection. and resuspended in rehydration buffer (30 mmol/L All samples remained frozen until shipment on dry ice to Tris basic, 2 mol/L thiourea, 7 mol/L urea, 4 % wt/vol the Interlab-UMU, Murcia, Spain, where all proteomic CHAPS). analyses were conducted. In order to corroborate the proteomic analysis findings Protein sample preparation and fluorescence labelling serum samples from a further 20 dogs (group 2) from Protein concentration was determined with 2Dquant kit different private veterinary clinics of Murcia Region, (GE Healthcare Europe GmbH, Freiburg, Germany). The Spain, were used. Twelve dogs were overweight/obese DIGE fluors –N-hydroxysuccinimidyl ester derivatives of without ORMD and eight overweight/obese with ORMD. the cyanine dyes, Cy2, Cy3 and Cy5 (GE Healthcare Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 3 of 9 Europe GmbH, Freiburg, Germany) –were prepared using respective Cy3 and Cy5 signals of the internal stand- using freshly opened N,N-dimethylformamide (Sigma- ard from the same gel, and the normalised Cy2 signal was Aldrich). Protein samples (50 μg) were labelled with 400 then used for spot intensity comparison and analysis. The pmol of Cy2, Cy3 and Cy5 DIGE fluors in separate tubes protein abundance for each spot in each sample was then on ice for 30 min. An internal standard was created by expressed as a ratio relative to the internal standard. Stat- mixing equal amounts of proteins of all samples and label- istical analysis was performed for every matched spot set, ling it with Cy2, whereas individual samples were labelled comparing the average ratio for a given spot between with Cy3 and Cy5. The labelling reaction was quenched groups. Only those spots with a p < 0.05 and a fold change by adding 1 μL of 10 mmol/L lysine. greater than 1.5 and lower than −1.5 were considered. Two-Dimensional DIGE Preparatives 2D gel The immobilised pH gradient strips (pH 4–7; GE Preparative gel, containing 275 μg of protein, were pre- Healthcare Europe GmbH, Freiburg, Germany) were ac- pared for a mix used as internal standard using PlusOne™ tively rehydrated for 16 h according to the instructions Bind-Silane following manufacturer recommendations (GE of the manufacturer. The Cy2-labelled internal standard Healthcare Europe GmbH, Freiburg, Germany). These gels of pooled samples was used to minimise experimental were fixed and stained with Colloidal Coomassie G-250 gel-to-gel variation and to facilitate cross-gel quantitative following the previously described protocol [18]. The im- analysis. Equal amounts (50 μg protein per dye per gel) ages were then captured using the Image Scanner III Sys- of Cy3- and Cy5-labelled individual samples and the tem (GE Healthcare Europe GmbH, Freiburg, Germany). Cy2-labeled internal standard were mixed together and The desired protein spots were picked from the Coomassie loaded onto an 18-cm immobilised pH gradient strip stained gels with Ettan Spot Picker (GE Healthcare Europe (pH 4–7). The proteins were focused in the first dimen- GmbH, Freiburg, Germany). sion for a total of 60000 Vh in the Ettan IPGphor 3 sys- tem (GE Healthcare Europe GmbH, Freiburg, Germany). Protein identification The immobilised pH gradient strips were equilibrated Proteins within gel spots were first reduced and alky- first in buffer I [6 mol/L urea, 2 % SDS, 75 mmol/L Tris lated using DTT and iodoacetamide, respectively, and (pH 8.8), 30 % glycerol, 1 % DTT] for 15 min and then then digested to peptides by trypsin proteomics grade in buffer II (same as buffer I plus 2.5 % iodoactamide) (Sigma-Aldrich, St Louis, MO, USA) [19].The tryptic for 15 min and subsequently transferred onto 12.5 % peptides were analysed by capillary reverse-phase liquid gels. Proteins were then separated in the second dimen- chromatography coupled online with MS/MS. The column, sion based on their molecular weight in the Ettan DALT BioBasic-18, 5 μm particles, 300 Å pore size, 0.18 mm ID- six (GE Healthcare Europe GmbH, Freiburg, Germany). A 100 mm L (Thermo, San Jose, CA), was connected to a total of six SDS gels were run together at 5 mA per gel for Surveyor MS Pump Plus (Thermo, San Jose, CA) and then 2 h and 2 W per gel for 12 h with the temperature of the coupled with an ion trap mass spectrometer (LXQ, running buffer maintained at 23 °C. Thermo, San Jose, CA). The flow rate was set at 180 μl/min but split to a flow rate of 2 μl/min. Mobile phase A was Image acquisition and analysis 0.1 % formic acid/2 % methanol in water and B was 0.1 % Gels were scanned directly between glass plates at 488/ formic acid in methanol. The peptide samples were injected 520 nm, 532/580 nm and 633/670 nm (excitation/emis- and gradient elution was done under the following condi- sion wavelengths) for the Cy2, Cy3 and Cy5 signals, re- tions: 5 % B in 2 min; a linear increase of 5 to 70 % B in spectively, using a Typhoon 9410 fluorescent scanner 20 min; 70 % for 10 min; 5 % B for 10 min. The ion trap (GE Healthcare Europe GmbH, Freiburg, Germany). MS was operated in a data-dependent MS/MS mode where Raw data of images were analysed with DeCyder 7.0 the five most abundant peptide molecular ions in every MS software (GE Healthcare Europe GmbH, Freiburg, scan were sequentially selected for collision-induced Germany). Briefly, original images of Cy2, Cy3 and Cy5 dissociation with a normalised collision energy of 34 %. images for each gel were cropped and smoothed to clar- Dynamic exclusion was applied to minimise repeated ify spots. The software was used following manufacturer selection of peptides previously selected for collision in- recommendations, and the estimated number of spots duced dissociation. The capillary temperature and electro- was set to 1500. The internal standard image with the spray voltage were set to 200 °C and 3.5 kV, respectively. most detected spots was assigned as the master and it The resulting mass spectra were searched against the was selected to create a MatchSet image. A MatchSet Canis familiaris Uniprot database (26456 sequences, image was created for comparing and analysing spots released at September 19, 2013) with the Proteome from all 11 images and a master image. The Cy2 spot data Discoverer 1.3 software (Thermo, San Jose, CA). The from each gel were normalised on a spot-to-spot basis following search parameters were applied: default charge Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 4 of 9 states of 2+, 3+, and 4+ were used; a maximum of one Table 1 The baseline characteristics of dogs included in missed cleavage was allowed with an average peptide mass proteomic study with and without obesity-related metabolic tolerance of 1.5 Da. A fragment ion search tolerance of dysfunction (ORMD) 0.8 Da was permitted. Fixed modification on cysteine was Variable ORMD No ORMD carbamidomethylation and a variable modification was Sex 4 males (all neutered) 4 males oxidation of methionine. All data were searched against a (all neutered) decoy database. The peptide concentration score cut-off Age, months 60 (25–120) 70 (43–120) for each of the runs was automatically adjusted to ensure Breed Labrador Retriever (2), Border Terrier a 1 % false discovery rate throughout the experiments. A Golden Retriever; Patterdale terrier Labrador positive identification was assigned when two or more Mixed Breed unique peptides were identified. Newfoundland Verification of proteomic analysis by specific assays Weight, kg 36.7 (13.8–100.0) 49.6 (25.8–60.8) Serum samples from group 2 dogs were analysed for Total BodyFat, % 52.4 (50.8–60.8) 52.1 (44.4–55.1) albumin (Albumin, Beckman Coulter Ireland Inc., BCS 9 (9–9) 9 (7–9) Lismeehan, Ireland), apoliprotein A-1 (Apoliprotein A-I SBP, mmHg 174 (166–177) 141 (123–159) (APO A-I), BioSystems, Barcelona, Spain), and glutathione Glucose, mmol/L 5.6 (5.1–6.0) 5.3 (4,8–5,5) peroxidase (Ransel, Randox Laboratories Limited, Crumlin, Cholesterol, mmol/L 6.9 (4.6–8.0) 5.9 (4,9–6,1) United Kingdom). All the methods were performed in automated analyzer (Olympus AU600, Beckman Coulter, TG, mmol/L 1.3 (0.8–2.3) 0.8 (0.5–1.2) Brea, USA). All assays showed an inter- and intra-assay Data are presented as median (interquartile range) BCS body condition score, SBP systolic blood pressure, TG triglycerides imprecision less than 15 % and a linearity under dilution resulted in coefficient of correlation close to 1 with canine serum samples. Statistical analysis Statistical analysis was performed using routine descrip- tive statistical procedures and software (Graph Pad Prism, Table 2 The baseline characteristics of dogs included in Version 5). In order to compare different variables be- proteomic study verification with and without obesity-related tween dogs with and without ORMD of the group 2, the metabolic dysfunction (ORMD) data were first evaluated for normality of distribution Variable ORMD No ORMD P using a D’Agostino & Pearson omnibus normality test, giv- Number 8 12 ing not parametric distribution. For this, Mann–Whitney Sex 3 males and 5 females 5 males and 7 females test was performed. Values of P < 0.05 were considered (all neutered) (all neutered) significant. Age, months 100 (36–139) 81 (24–132) NS Results Breed Mixed Breed (2), Mixed Breed (7) Comparison of obese dogs with and without ORMD Boxer, Yorkshire terrier (2) The baseline characteristics of group 1 dogs with and Beagle, Bichon Maltese without ORMD (participating in proteomic analyses) are Brittany Spaniel Collie shown in Table 1, with no marked differences between Cocker Spaniel, Shih tzu the groups, in terms of age, breed, weight, and body fat French Bulldog, percentage. All 4 dogs with ORMD had increased SBP, whilst 3, 3, and 2 dogs, respectively, had increased glucose, Yorkshire terrier, cholesterol, and triglyceride concentrations. Weight, kg 12.4 (11.4–18.7) 20.3 (5.2–44.5) NS The baseline characteristics of group 2 dogs the animals BCS 9 (8–9) 9 (7–9) NS with and without ORMD (used for assay verification) are SBP, mmHg 140 (123–150) 130 (110–158) NS shown in Table 2. Again, there were no marked differ- Glucose, mmol/L 5.8 (5.7–7.6) 4.9 (4.7–5.3) 0.047 ences between the groups, in terms of age, breed, and Cholesterol, 9.1 (7.6–11.0) 5.8 (5.2–7.6) 0.007 weight. Of the 8 dogs with ORMD, five had hypercholes- mmol/L terolaemia, four had hypertriglyceridaemia, and six had TG, mmol/L 2.8 (0.8–3.8) 0.9 (0.6–1.4) NS hyperglycaemia. Of the 12 dogs without ORMD, two Data are presented as median (interquartile range) had hypercholesterolaemia, one had hypertriglyceridae- BCS body condition score, SBP systolic blood pressure, TG triglycerides, NS not mia, and two had hyperglycaemia. statistically significant Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 5 of 9 Proteomic analysis (P = 0.044) and apoliprotein A-I (P < 0.001) concentra- More than 1200 individual spots, ranging from 15 to tions were greater in dogs with ORMD, and there 250 kDa masses over pH 4–7 were detected in gels of was also decreased activity of glutathione peroxidase protein mapping of canine plasma pre-treated with Pro- (P = 0.039) (Fig. 3). teoMiner. Using image analysis and further statistical analysis, 8 spots were detected and subsequently identi- fied with differential protein concentrations between Discussion dogs with and without ORMD (Fig. 1). Among these 8 The aim of the present study was to evaluate differ- protein spots, protein concentrations were increased in ences in the plasma proteome of obese dogs with and 3 and decreased in 5 obese dogs with ORMD compared without ORMD. For this, we used 2D DIGE and mass with obese dogs without ORMD (Table 3, Fig. 2). spectrometry to identify a set of proteins that differed significantly in abundance in the plasma of obese dogs Mass spectrometry with ORMD relative to obese dogs without ORMD. In Mass spectrometry allowed the identification of 12 pro- total, 12 differentially expressed proteins were identified. teins from seven out of eight spots detected (Table 4), These were proteins associated with lipid metabolism which were identified as: alpha-2 macroglobulin (A2M) (albumin, apoliprotein A-I), the complement system and, apoliprotein A-I (spot number 268), C2 and C5(spot (C2, C3, C5, C4BPA, A2M, Uncharacterised protein 410), fibrinogen and Uncharacterised protein (Fragment) (Fragment) OS = Canis familiaris), coagulation (fibrinogen), OS = Canis familiaris (spot 624), albumin (spot 730), C3 immune response (IGJ), hyaluronan metabolic process (spots 730, 736 and 1190), ITIH2 (spots 730. 736 and (ITIH2), and antioxidants (glutathione peroxidase). These 1190), C4BPA (spot 1190), IGJ(spot 1199), and glutathione findings provide further evidence that ORMD in dogs is as- peroxidase (spot 1201). Unfortunately, the proteins in spot sociated with a range of metabolic changes, although their 1205 could not be identified. Among these 12 proteins, clinical significance remains unknown. concentrations were increased in 10 and decreased in 2 Increased albumin abundance was observed in the in obese dogs with ORMD compared with obese dogs plasma of obese dogs with ORMD compared with dogs without ORMD. The proteins with increased expression without ORMD by both proteomic and specific analysis. in ORMD were: albumin, apoliprotein A-I, C2, C3, Significantly greater albumin concentrations have previ- C5, C4BPA, A2M, Uncharacterised protein (Fragment) ously been reported in obese compared with healthy OS = Canis familiaris, fibrinogen, and ITIH2, whilst dogs [20], and circulating concentrations of albumin are the two proteins with decreased expression in ORMD known to decrease after weight loss [17, 20]. However, were: IGJ and glutathione peroxidase. not all studies have identified such differences [21, 22]. That said, increased albumin expression has been de- Verification of proteomic analysis by specific assays tected in the proteome of mature adipocytes collected Specific assays were used to verify the results of proteomic from overweight and morbidly obese persons [7]. Fur- analyses in a separate group of dogs. Plasma albumin ther studies would be indicated in order to determine Fig. 1 Representative 2-DE gel images proteins of ProteoMiner pre-treated plasma samples obtained from dogs with and without obesity-related metabolic dysfunction (ORMD). Differentially expressed protein spots between the dogs with and without ORMD are circled and numbered Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 6 of 9 Table 3 Fold change of differentially expressed spot matches in In the present study, six different proteins (C2, C3, C5, dogs with (n = 4) and without (n = 4) obesity-related metabolic C4BPA, A2M, Uncharacterised protein (Fragment) OS = dysfunction (ORMD) Canis familiaris) associated with the complement arm Spota Fold change noORMD dogs/ORMD dogs P of the immune system were higher in abundance in dogs 1205 42.68 0.017 with ORMD. In two human proteomics studies, plasma 1201 2.55 0.005 C3 was identified as a key marker for differences in body 1199 1.53 0.006 fat and body fat changes [10, 27]. Further, in humans C3 was associated with the prevalence of MetS and diabetes 410 −1.76 0.040 mellitus type 2 development [25]. It was speculated that 268 −2.08 0.041 increased concentrations of proteins of complement sys- 730 −2.28 0.036 tem could be due to decreased activity of the comple- 624 −2.39 0.009 ment system or due to changes in production in adipose 736 −3.09 0.031 tissue or at other sites [10, 26]. Similar mechanisms a Spot label number from annotated gel image (see Figs. 1 and 2) might be responsible for the findings in obese dogs with ORMD, but more work would be required to clarify this. the mechanisms involved for the increased circulating Dogs with ORMD also had greater fibrinogen concen- albumin in obese dogs with ORMD. trations than dogs without ORMD. It has been suggested Apolipoprotein A1 (APO-A1) participates in the re- fibrinogen to be an additional defining component of verse transport of cholesterol from tissues to the liver, metabolic syndrome in humans [28], as fibrinogen con- ready for its subsequent excretion. APO-A1 has also centrations were increased in subjects with metabolic been reported to have acyl coenzyme A cholesterol syndrome compared with those without it [28, 29]. It acyltransferase-like activity; this may synergistically regu- has been hypothesised that fibrinogen acts as a link be- late adipogenesis and lipid metabolism [23]. In the tween prothrombotic and proinflammatory states in current study, greater APO-A1 concentrations were de- metabolic syndrome in humans [28]. tected in dogs with ORMD as observed in proteomic J-chain of immunoglobulin (IGJ) is a protein of about analysis and verified by a specific assay. These results are 15 kDa polypeptide, expressed by mucosal and glandular consistent with those reported by Briand et al. [24] who, plasma cells. It regulates polymer formation of immuno- in a stable isotope study, observed increased APO-A1 globulin (Ig) A and IgM. In humans increased IgA con- production in insulin resistant dogs. centration is associated with components of metabolic Fig. 2 Example of the 3-dimensional images of differentially regulated spot proteins (one increased and one decreased) in dogs with (left) and without (right) obesity-related metabolic dysfunction (ORMD). Graphics represent the changes in Standardised Abundance of specific spots after logarithmic transformation obtained directly from DeCyder 7.0 software analysis: dotted lines represent individual data, entire lines represent mean Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 7 of 9 Table 4 Mass spectrometry identification of proteins Spot Noa Protein name Accession numberb Score Unique peptides Sequence MW (kDa)/pI coverage (%)c 268 Uncharacterised protein Canis familiaris (A2M) F6UME0 99.79 10 8.14 165.1/6.71 Apolipoprotein A-I Canis familiaris F1PDJ5 11.42 2 9.40 30.2/5.39 410 Uncharacterised protein Canis familiaris (C2) E2RS75 132.97 9 16.56 86.3/7.37 Uncharacterised protein Canis familiaris (C5) F1P7J4 18.61 2 1.36 188.5/6.60 624 Fibrinogen alpha chain OS = Canis familiaris J9NRV7 112.52 5 7.05 88.9/6.16 Uncharacterised protein (Fragment) OS = Canis familiaris F1PWR2 12.25 3 2.45 192.8/6.89 730 Uncharacterised protein Canis familiaris (C4BPA) F1PGM9 390.27 11 30.93 57.9/6.83 Albumin Canis familiaris F2Z4Q6 174.32 8 15.30 68.6/5.69 Uncharacterised protein Canis familiaris (C3) F1PGM1 64.87 5 2.81 176.0/7.12 Uncharacterised protein Canis familiaris (ITIH2) F1PG39 4.96 3 4.96 107.2/7.53 736 Uncharacterised protein Canis familiaris (C4BPA) F1PGM9 456.80 8 23.93 57.9/6.83 Uncharacterised protein Canis familiaris (ITIH2) F1PG39 237.71 6 11.39 107.2/7.53 Uncharacterised protein Canis familiaris (C3) F1PGM1 128.64 11 8.62 176.0/7.12 Albumin Canis familiaris F2Z4Q6 38.31 2 5.59 68.6/5.69 1199 Uncharacterised protein Canis familiaris (IGJ) J9JHH5 43.95 2 12.58 18.2/4.82 1201 Glutathione peroxidase (Fragment) Canis familiaris J9P028 322.78 4 37.91 17.3/7.97 MW molecular weight a Spot label number from annotated gel image (see Figs. 1 and 2). Spot No 1205 Not identified b UniProtKB database. Canis familiaris(26456 sequences, 15304077 residues, 09/19/2013) c Sequence coverage: percentage of identified sequence to the complete sequence of the known protein syndrome, such as hyperglycaemia, hypertriglyceridaemia, homologous heavy chains (ITIHs) have been identified in and abdominal obesity [30]. It has been suggested that in- this family (ITIH1-ITIH5) [33]. The only currently known creased IgA and IgM concentrations could be attributed function of the heavy chains is to bolster the stability of to the presence of low-grade inflammation in obesity [30]. extracellular matrix through covalent linkage to hyalur- Taken together, it could be hypothesised that lower abun- onic acid [34], which is a major component of this matrix dance of IGJ observed in dogs with ORMD in present [33].A number of studies have reported on the biological study, could be attributed to its greater utilisation by IgA effects of the ITI molecules, proposing involvement in or/and IgM. However, future studies are required in order various acute-phase processes, such as inflammation or to clarify this. cancer [33, 35]. ITIH2 was shown to be down-regulated in The inter-alpha (globulin) inhibitor (ITI) family (more tumours of the breast, compared with other ITIH pro- commonly called the family of inter-alpha-trypsin inhibi- teins, suggesting its potential function as tumour suppres- tors) is composed of serine protease inhibitors that are sor or metastasis repressor [33]. In the present study, assembled from two precursor proteins: a light chain and increased abundance of ITIH2 was observed in obese dogs either one or two heavy chains [31, 32]. Whilst there is with ORMD and, to the authors’ knowledge, no previous only one type of light chain, to date, five different studies have linked this protein to metabolic syndrome. Fig. 3 Serum albumin, apoliprotein A-I, and gluthatione peroxidase in dogs with (ORMD; n = 8) and without (Non-ORMD; n = 12) obesity related metabolic dysfunction (ORMD) Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 8 of 9 Glutathione peroxidase is an 80 kDa protein, which is Funding composed of four identical subunits, and acts as a major The study was funded by a research grant from the WALTHAM® Foundation. Additional financial support was also provided by the Program “Juan de la antioxidant in the plasma [35]. Decreased glutathione Cierva” of ‘Ministerio de Economia y Competitividad’, Spain, through a peroxidase abundance was observed by proteomic postdoctoral grant. study in obese dogs that presented with ORMD. These findings were also confirmed by use of a specific assay. Availability of data and materials If the reader requires further information to help interpret the results, please During the development of obesity, glutathione perox- contact the corresponding author. idase precursor expression decreases in obesity-prone rats, but increases in both obesity-resistant and nor- Authors’ contributions mal rats [35]. It was hypothesised that these changes AT designed the study, performed laboratory assays, drafted manuscript, and reviewed the manuscript; JJC designed the study, performed laboratory could lead to augmented systemic oxidative stress and assays, drafted the manuscript, and reviewed the manuscript; CdT performed the onset of metabolic complications such as diabetes laboratory assays, and reviewed the manuscript; BBL performed laboratory mellitus [36]. assays, and reviewed the manuscript; SLH collected clinical data, and reviewed manuscript; YQ reviewed the results, and reviewed the manuscript; The main limitation of the present study was the rela- PJM reviewed the results, and reviewed the manuscript; VB reviewed the tively small number of animals studied by proteomic ana- results, and reviewed the manuscript; JP drafted the manuscript, and lysis. In light of this, only samples from male dogs were reviewed the manuscript; AJG designed the study, collected the clinical data, analysed the results, and drafted the manuscript. All authors have approved used to perform proteomic analysis, so there could be no the final article. possible confounding effect of sex or/and hormones. Pre- vious studies have suggested that three biological samples Competing interests are sufficient to detect induced biological variations in the The following conflicts of interest apply: AJG’s Readership is funded by Royal Canin; the diet used in this study is manufactured by Royal Canin; PJM is an proteome of animals [37, 38]. A second potential limita- employee of WALTHAM, whilst YQ is employed by Royal Canin. tion of the present study relates to storage of the samples used for proteomic analysis prior to their transportation Ethics approval and consent to participate to InterLab-UMU, Spain. It is possible that such storage The study protocol adhered to the University of Liverpool Animal Ethics Guidelines, and was approved by both the University of Liverpool Research may have affected the results, for instance by resulting in Ethics Committee and the WALTHAM Animal Welfare & Ethical Review Board. fewer spots/proteins being detected. Thirdly, the included All clinical procedures were performed for the direct benefit of the animals, animals were client-owned dogs with variable living condi- and their owners gave informed written consent. tions and different diets, what could have influenced the Author details results of the study. Nevertheless, the data are representa- 1 Departament de Medicina i Cirugia Animals, Universitat Autónoma de tive of a true clinic picture. Despite these limitations and, Barcelona, 08193 Barcelona, Spain. 2Interdisciplinary Laboratory of Clinical Pathology, Iterlab-UMU, Campus of Excellence Mare Nostrum, University of although ideally all identified proteins should have been Murcia, Murcia, Spain. 3Unidad de Proteómica, Hospital Clínico Universitario validated, the verification of three proteins agrees with the Virgen de la Arrixaca (HCUVA), IMIB-Arrixaca, 30120 Murcia, Spain. 4Internal data observed by proteomic study. Diseases Clinic, Faculty of Veterinary Medicine, University of Zagreb, 10000 Zagreb, Croatia. 5Department of Obesity and Endocrinology, University of Liverpool, Leahurst Campus, Chester High Road, Neston, Wirral CH64 7TE, UK. Conclusions 6 Royal Canin Research Center, B.P.4-650 Avenue de la Petite Camargue, The results of this comparative plasma proteome ana- 30470 Aimargues, France. 7The WALTHAM Centre for Pet Nutrition, Freeby Lane, Waltham-on-the-Wolds, Melton Mowbray LE14 4RT, UK. lysis have demonstrated that the obese dogs with ORMD have alterations in the concentrations of proteins related Received: 7 July 2016 Accepted: 10 September 2016 to lipid metabolism, immune response, and antioxidant system. Further studies should be performed to evaluate References the clinical implications that changes in these proteins 1. International Diabetes Federation. The IDF consensus worldwide definition could have. of the metabolic syndrome. http://www.idf.org/metabolic-syndrome. 16 Sept 2016. Abbreviations 2. Grundy SM, Hansen B, Smith Jr SC, Cleeman JI, Kahn RA, American Heart 2-DE DIGE: Two-dimensional differential gel electrophoresis; A2M: Alpha-2- Association, National Heart, Lung, and Blood Institute, American Diabetes macroglobulin; APO-A1: Apoliprotein A-I; BCS: Body condition score; Association. Clinical management of metabolic syndrome: report of the C2: Complement component 2; C3: Complement component 3; American Heart Association/National Heart, Lung, and Blood Institute/ C4BPA: Complement component 4 binding protein alpha; C5: Complement American Diabetes Association conference on scientific issues related to component 5; DEXA: X-ray absorptiometry dual energy; IgA: Immunoglobulin management. Arterioscler Thromb Vasc Biol. 2004;24(2):e19–24. A; IGJ: J-chain of immunoglobulin; IgM: Immunoglobulin B; ITIH: Inter-alpha- 3. Tvarijonaviciute A, Ceron JJ, Holden SL, Cuthbertson DJ, Biourge V, Morris trypsin inhibitors; ORMD: Obesity related metabolic dysfunction; PJ, German AJ. Obesity-related metabolic dysfunction in dogs: a comparison PBS: Phosphate buffer saline; SBS: Systolic blood pressure with human metabolic syndrome. BMC Vet Res. 2012;8:147. 4. Brandacher G, Golderer G, Kienzl K, Werner ER, Margreiter R, Weiss HG. Acknowledgements Potential applications of global protein expression analysis (proteomics) in The authors wish to acknowledge the referring veterinarians for referring morbid obesity and bariatric surgery. Obes Surg. 2008;18(7):905–10. cases, the owners of all dogs for allowing them to participate, and the 5. Peral B, Camafeita E, Fernández-Real JM, López JA. Tackling the human clinical staff at the University of Liverpool for assistance with case adipose tissue proteome to gain insight into obesity and related management. pathologies. Expert Rev Proteomics. 2009;6(4):353–61. Tvarijonaviciute et al. BMC Veterinary Research (2016) 12:211 Page 9 of 9 6. Galata Z, Moschonis G, Makridakis M, Dimitraki P, Nicolaides NC, Manios Y, including Factor D/adipsin: studies in anorexia nervosa and obesity. Bartzeliotou A, Chrousos GP, Charmandari E. Plasma proteomic analysis Clin Exp Immunol. 1997;108(3):507–15. in obese and overweight prepubertal children. Eur J Clin Invest. 27. van Dijk SJ, Feskens EJ, Heidema AG, Bos MB, van de Rest O, Geleijnse JM, 2011;41(12):1275–83. de Groot LC, Muller M, Afman LA. Plasma protein profiling reveals protein 7. Benabdelkamel H, Masood A, Almidani GM, Alsadhan AA, Bassas AF, Duncan clusters related to BMI and insulin levels in middle-aged overweight MW, Alfadda AA. Mature adipocyte proteome reveals differentially altered subjects. PLoS One. 2010;5(12):e14422. protein abundances between lean, overweight and morbidly obese human 28. Grundy SM, Brewer Jr HB, Cleeman JI, Smith Jr SC, Lenfant C, National Heart, subjects. Mol Cell Endocrinol. 2015;401:142–54. Lung, and Blood Institute, American Heart Association. Definition of 8. Selvaraju S, Rassi ZE. Liquid-phase-based separation systems for depletion, metabolic syndrome: report of the National Heart, Lung, and Blood prefractionation and enrichment of proteins in biological fluids and Institute/American Heart Association conference on scientific issues related matrices for in-depth proteomics analysis–an update covering the period to definition. Arterioscler Thromb Vasc Biol. 2004;24(2):e13–8. 2008–2011. Electrophoresis. 2012;33(1):74–88. 29. Imperatore G, Riccardi G, Iovine C, Rivellese AA, Vaccaro O. Plasma 9. Dayarathna MK, Hancock WS, Hincapie M. A two step fractionation fibrinogen: a new factor of the metabolic syndrome. A population-based approach for plasma proteomics using immunodepletion of abundant study. Diabetes Care. 1998;21(4):649–54. proteins and multi-lectin affinity chromatography: Application to the 30. Gonzalez-Quintela A, Alende R, Gude F, Campos J, Rey J, Meijide LM, analysis of obesity, diabetes, and hypertension diseases. J Sep Sci. Fernandez-Merino C, Vidal C. Serum levels of immunoglobulins (IgG, IgA, 2008;31(6–7):1156–66. IgM) in a general adult population and their relationship with alcohol 10. Oberbach A, Bluher M, Wirth H, Till H, Kovacs P, Kullnick Y, Schlichting N, consumption, smoking and common metabolic abnormalities. Clin Exp Tomm JM, Rolle-Kampczyk U, Murugaiyan J, Binder H, Dietrich A, von Immunol. 2008;151(1):42–50. Bergen M. Combined proteomic and metabolomic profiling of serum 31. Bost F, Diarra-Mehrpour M, Martin JP. Inter-alpha-trypsin inhibitor reveals association of the complement system with obesity and identifies proteoglycan family–a group of proteins binding and stabilizing the novel markers of body fat mass changes. J Proteome Res. 2011;10(10):4769–88. extracellular matrix. Eur J Biochem. 1998;252(3):339–46. 11. Unlü M, Morgan ME, Minden JS. Difference gel electrophoresis: a single 32. Zhuo L, Hascall VC, Kimata K. Inter-alpha-trypsin inhibitor, a covalent protein- gel method for detecting changes in protein extracts. Electrophoresis. glycosaminoglycan-protein complex. J Biol Chem. 2004;279(37):38079–82. 1997;18(11):2071–7. 33. Hamm A, Veeck J, Bektas N, Wild PJ, Hartmann A, Heindrichs U, Kristiansen 12. Tonge R, Shaw J, Middleton B, Rowlinson R, Rayner S, Young J, Pognan F, G, Werbowetski-Ogilvie T, Del Maestro R, Knuechel R, Dahl E. Frequent Hawkins E, Currie I, Davison M. Validation and development of fluorescence expression loss of Inter-alpha-trypsin inhibitor heavy chain (ITIH) genes in two-dimensional differential gel electrophoresis proteomics technology. multiple human solid tumors: a systematic expression analysis. BMC Cancer. Proteomics. 2001;1(3):377–96. 2008;8:25. 13. Laflamme D. Development and validation of a body condition score system 34. Huang L, Yoneda M, Kimata K. A serum-derived hyaluronan-associated for dogs. Can Pract. 1997;22(4):10–5. protein (SHAP) is the heavy chain of the inter alpha-trypsin inhibitor. 14. German AJ, Holden SL, Moxham GL, Holmes KL, Hackett RM, Rawlings JM. J Biol Chem. 1993;268(35):26725–30. A simple, reliable tool for owners to assess the body condition of their dog 35. Choi JW, Wang X, Joo JI, Kim DH, Oh TS, Choi DK, Yun JW. Plasma or cat. J Nutr. 2006;136(7 Suppl):2031S–3S. proteome analysis in diet-induced obesity-prone and obesity-resistant rats. 15. Raffan E, Holden SL, Cullingham F, Hackett RM, Rawlings JM, German AJ. Proteomics. 2010;10(24):4386–400. Standardized positioning is essential for precise determination of body 36. Lee YS, Kim AY, Choi JW, Kim M, Yasue S, Son HJ, Masuzaki H, Park KS, Kim composition using dual-energy x-ray absorptiometry in dogs. J Nutr. 2006; JB. Dysregulation of adipose glutathione peroxidase 3 in obesity contributes 136(7 Suppl):1976–8. to local and systemic oxidative stress. Mol Endocrinol. 2008;22(9):2176–89. 16. German AJ, Holden SL, Bissot T, Hackett RM, Biourge V. Dietary energy 37. Westermeier R, Naven T, Höpker HR. Proteomics in Practice — A Guide to restriction and successful weight loss in obese client-owned dogs. J Vet Int Successful Experimental Design. 2nd ed. Weinheim: Wiley-VCH; 2008. Med. 2007;21(6):1174–80. 38. Schmid GM, Converset V, Walter N, Sennitt MV, Leung KY, Byers H, Ward M, 17. German AJ, Hervera M, Hunter L, Holden SL, Morris PJ, Biourge V, Hochstrasser DF, Cawthorne MA, Sanchez JC. Effect of high-fat diet on the Trayhurn P. Improvement in insulin resistance and reduction in plasma expression of proteins in muscle, adipose tissues, and liver of C57BL/6mice. inflammatory adipokines after weight loss in obese dogs. Dom An Proteomics. 2004;4(8):2270–82. Endocr. 2009;37(4):214–22. 18. Dyballa N, Metzger S. Fast and Sensitive Colloidal Coomassie G-250 Staining for Proteins in Polyacrylamide Gels. J Vis Exp. 2009;30:1431. doi:10.3791/1431. 19. Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc. 2006;1(6):2856–60. 20. Yamka RM, Friesen KG, Frantz NZ. Identification of canine markers related to obesity and the effects of weight loss on the markers of interest. Intern J Appl Res Vet Med. 2006;4(4):282–92. 21. Chikamune T, Katamoto H, Ohashi F, Shimada Y. Serum lipid and lipoprotein concentrations in obese dogs. J Vet Med Sci. 1995;57(4):595–8. 22. Eirmann LA, Freeman LM, Laflamme DP, Michel KE, Satyaraj E. Comparison of adipokine concentrations and markers of inflammation in obese versus lean dogs. Intern J Appl Res Vet Med. 2009;7(4):196–205. Submit your next manuscript to BioMed Central 23. Cheng AM, Handa P, Tateya S, Schwartz J, Tang C, Mitra P, Oram JF, Chait A, and we will help you at every step: Kim F. Apolipoprotein A-I attenuates palmitate-mediated NF-kB activation by reducing Toll-like receptor-4 recruitment into lipid rafts. PLoS One. • We accept pre-submission inquiries 2012;7(3):e33917. • Our selector tool helps you to find the most relevant journal 24. Briand F, Bailhache E, Nguyen P, Krempf M, Magot T, Ouguerram K. • We provide round the clock customer support Metabolism of high density lipoprotein apolipoprotein A-I and cholesteryl ester in insulin resistant dog: a stable isotope study. Diabetes Obes Metab. • Convenient online submission 2007;9(1):139–42. • Thorough peer review 25. Hertle E, Stehouwer CD, van Greevenbroek MM. The complement system in • Inclusion in PubMed and all major indexing services human cardiometabolic disease. Mol Immunol. 2014;61(2):135–48. 26. Pomeroy C, Mitchell J, Eckert E, Raymond N, Crosby R, Dalmasso AP. Effect • Maximum visibility for your research of body weight and caloric restriction on serum complement proteins, Submit your manuscript at www.biomedcentral.com/submit