Untargeted and Targeted Proteomic Analysis of the Impact of Diabetes Mellitus on Two Adipose Tissue Depots and Blood of Obese Patients

Methods

Study population and design

Subcutaneous (SAT) and visceral (VAT) adipose tissue samples from the same patient were obtained during bariatric surgery (N = 38). As standard of operating procedure, patients were subjected to a hypocaloric diet before surgery in order to have them loose around 5–8% of the total body mass weight. All patients fasted for at least 12 hours overnight and were taken to the operating room. Blood samples were obtained at the time of intervention in order to analyze metabolic markers and serum protein composition. Written informed consent was obtained from all subjects and the study was conducted according to the recommendations of the Declaration of Helsinki and approved by the Centro Médico Teknon Ethical Committee.

Untargeted proteomics were performed on both fat depots, VAT and SAT (from same patient), in six type 3 obese non-diabetic patients and six type 3 obese diabetic patients with fully characterized demographics (Table 1 and Fig. 1). Targeted plasma proteomic characterization was performed by the Bio-Plex Multiplex System, powered by Luminex xMAP technology, a multiplex assay system to simultaneously detected and quantified multiple target analyzed in qualified complex samples obtained from patients characterized by increasing cardiovascular risk: non-obese non-diabetic (N = 11); type 3 obese/non-diabetic (N = 13); and type 3 obese/diabetic (N = 14).

Table 1

Demographic description and clinical parameters of the patients included in the proteomic analysis
	Non Diabetics	Diabetics
Diabetes Mellitus	NO	YES
Age	50 ± 3	53 ± 3
BMI (Kg/m²)	43.6 ± 1.0	43.3 ± 1.3
Gender (M/W)	3/3	3/3
CV Risk factors	1	3–5
Glucose (mg/dl)	127.6 ± 19.5	149.4 ± 13.6
TG (mg/dl)	173.2 ± 37.8	170.8 ± 12.7
Cholesterol (mg/dl)	168.4 ± 12.2	197.8 ± 21.9
HDL (mg/dl)	33.5 ± 4.0	39.5 ± 8.4
LDL (mg/dl)	226.8 ± 57.1	170.2 ± 40.8
Urea (mg/dl)	28.2 ± 5.1	47.3 ± 18.6
Total proteins (g/dl)	6.9 ± 0.3	7.5 ± 0.4
GOT (AST) UL	25.0 ± 4.6	17.6 ± 2.8
GPT (ALT) UL	20.8 ± 4.6	15.4 ± 2.8
GPT/GOT	0.9 ± 0.2	1.0 ± 0.1
Creatinine	1.0 ± 0.1	1.1 ± 0.2
Data expressed as Mean Value ± SEM. BMI: Body mass Index; M: Men; W: Women; CVD: Cardiovascular diseases

Sample collection and preparation

Blood samples were collected in anticoagulant-free Vacutainer tubes for serum preparation. Serum fractions were separated by centrifugation at 3000 x g at 4ºC for 20 min, aliquoted and stored at -80ºC until used.

Adipose tissue was collected under sterile conditions at the operating room; samples were washed in phosphate-buffered saline and immediately frozen in liquid N₂. Samples were stored at -80ºC until protein extraction.

Adipose tissue protein isolation was performed following the previously described method (Cortón et al. 2008) with minor modifications. In brief, AT was mechanically homogenized, diluted in lysis buffer (8.4 mol/l urea, 2.4 mol/l thiourea, 50 g/l CHAPS, 50 mmol/l DTT) and sonicated (6 cycles of 15 seconds). The suspension was shaken for 1 hour at room temperature and centrifuged at 60.000 rpm for 1.5 hours.

Biochemical analysis

Blood levels of glucose; triglyceride; and total, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol were determined with a CLIMA MC-15 analyzer (RAL) (Table 1).

Two-Dimensional Gel Electrophoresis (2-DE)

For the 2-DE analysis, samples were desalted using the ReadyPrep 2D Cleanup Kit (Bio Rad), and proteins were diluted in urea/thiourea buffer (7 mol/l urea, 2 mol/l thiourea and 20 g/l CHAPS). The protein concentration was determined using the 2-D Quant Kit (GE Healthcare).

Protein extracts from SAT and VAT (100 µg) were diluted in urea/thiourea buffer supplemented with 1% (v/v) IPG buffer, and 50 mmol/l DTT and then applied to IPG strips (pH 4–7; Bio-Rad) by active rehydration (12 h; 50V). IPG strips were then equilibrated in equilibration buffer (0.1% DTT) to maintain the fully reduced state of proteins, followed by 2.5% iodoacetamide to prevent reoxidation of thiol groups during electrophoresis. The second dimension was performed in 12% acrilamide gels using an Ettan DALT II System (GEHealthcare). 2D-gels were stained with Flamingo™ Fluorescent Gel Stain (Bio Rad) and scanned with a Typhoon FLA9500 (GE Healthcare). For each independent experiment, 2-DE for protein extracts from SAT and VAT were processed in parallel to guarantee a maximum of comparability. Analysis for differences in protein patterns were performed with the PD-Quest 8.0 (BioRad), using a single master that included all gels of each independent experiment. Each spot was assigned a relative value that corresponded to the single spot volume compared to the volume of all spots in the gel, following background extraction and normalization between gels.

Mass spectrometry analysis

Proteins were identified after in-gel tryptic digestion and extraction of peptides from the gels pieces, as previously described[15], by matrix – assisted laser desorption/ionization time-of-flight (MALDI-TOF) using an AutoFlex III Smartbeam MALDI-TOF/TOF (Bruker Daltonics). Samples were applied to Prespotted AnchorChip plates (Bruker Daltonics) surrounding calibrates provided on the plates. Spectra were acquired with flexControl on reflector mode, (mass range 850–4000 m/z, reflector 1: 21.06 kV; reflector 2: 9.77 kV; ion source 1 voltage: 19 kV; ion source 2: 16.5 kV; detection gain 2.37x) with an average of 3500 added shots at a frequency of 200 Hz. Each sample was processed with flexAnalysis (version 3.0, Bruker Daltonics) considering a signal-to-noise ratio over 3, applying statistical calibration and eliminating background peaks. For identification, peaks between 850 and 1000 were not considered as in general only matrix peaks are visible on this mass range. After processing, spectra were sent to the interface BioTools (version 3.2, Bruker Daltonics) and MASCOT search on Swiss-Prot 57.15 database was done (Taxonomy: Homo Sapiens, Mass Tolerance 50 to 100, up to 2 miss cleavage, Global Modification: Carbamidomethyl (C), Variable Modification: Oxidation (M)). Identification was accepted with a score higher than 56.

Quantification of serum protein levels

Multiplexed bead-based immunoassay was utilized for the simultaneous measurement of different proteins. A 37-plex assay was performed to measure inflammation related proteins (171AL001M), a 10-plex assay for the measure of diabetes related proteins (171A7001M), and finally a 2-plex assay was used to measure Adipsin and Adiponectin due to the different dilution needed (171A7002M). Multiplex assay is based on an internal color-coded bead that is coupled to analyze specific antibodies allowing the simultaneous measurement of multiple analyzes on the same well. The assay was performed according to the supplier’s instruction (Bio Rad Laboratories).

In silico bioinformatics analysis

The statistically significant neural network and canonical pathway in which the identified proteins were involved were generated through the use of IPA (Ingenuity System, www.ingenuity.com).

Statistical analysis

Non-parametric Wilcoxon or Mann-Whitney analyses were performed to analyze differences in protein levels between subcutaneous and visceral adipose tissue and between diabetic and non-diabetic patients. Data were expressed as mean ± SEM unless stated. The level of significance was set at p < 0.05. All analyses were conducted with StatView software.

Results

1. Differential protein expression between the subcutaneous and the visceral white adipose tissue

The protein signature of the white adipose tissue was resolved by two-dimensional gel electrophoresis. A total of 72 non-redundant proteins were identified in the differential proteome of SAT and VAT of type 3 obese patients. Since the two fat depots were from the same donor dietary, environmental and genetic factors could be excluded (Supplementary Table 1). Comparative analysis of subcutaneous and visceral fat depots showed significant differential protein expression of twelve proteins, corresponding to different functional groups including coagulation, oxidation, energy homeostasis, chaperones and signaling pathway regulators among others. SAT showed two proteins up-regulated versus VAT (14-3-3 protein beta/alpha and Hemopexin), and ten proteins were down-regulated (Annexin A1, ATP-binding cassette sub-family G member 8 [ABCG8], Ferritin light chain [FTL], Fibrinogen gamma chain [FGG], Glutathione S-transferase P [GSTP1], Kazrin, N(G)-dimethylarginine dimethyl-aminohydrolase 2 [DDAH2], Selenium-binding protein 1 [SBP1], Serum amyloid P [SAP] and Transmembrane and coiled-coil domain-containing protein 7 [TMCO7]) (Table 2 Column I).

Table 2

Significance of changes in differential protein expression of subcutaneous versus visceral adipose tissue in: all patients; obese type 3 non-diabetic patients (OB3 + nDM); and obese type 3 diabetic patients (OB3 + DM).
	Subcutaneous vs Visceral
	I All Patients		II OB3 + nDM	III OB3 + DM
14-3-3 protein b/a		0.05 ↑	0.11	0.17
Actin-related protein 2		0.21	0.75	0.04 ↑
Annexin 1		0.04 ↓	0.60	0.02 ↑
Apolipoprotein A1		0.43	0.34	0.02 ↑
ATP-binding cassette sub-family G member 8		0.01 ↓	0.04 ↓	0.24
Creatine kinase B-type		0.24	0.91	0.04 ↓
Ferritin ligth chain		0.05 ↓	0.028 ↓	0.50
Fibrinogen gamma chain		0.002 ↓	0.02 ↓	0.02 ↓
Glutathione S-transferase P		0.02 ↓	0.02 ↓	0.68
Hemopexin		0.02 ↑	0.24	0.04 ↑
HSP 60		0.08	0.60	0.04 ↓
Inorganic pyrophosphatase		0.18	0.02 ↓	0.91
Kazrin		0.006 ↓	0.02 ↓	0.04 ↓
N(G)-dimethylarginine dimethylaminohydrolase 2		0.004 ↓	0.02 ↓	0.08
Selenium BP1		0.006 ↓	0.02 ↓	0.11
Serum amyloid P		0.02 ↓	0.04 ↓	0.50
Transmembrane and coiled-coil domain-containing protein 7		0.005 ↓	0.06	0.02 ↓
P value < 0.05 is considered significant and marked in bold.
↑ proteins up-regulated in SAT. ↓ proteins down-regulated in SAT.

In order to investigate the influence of the biochemical characteristics on the expression of the different proteins identified by proteomics we analyzed the effect of glucose, tryglicerides, HDL, LDL and Cholesterol levels in protein expression in SAT and VAT. We observed that total cholesterol and LDL levels were the parameters with more influence on protein expression both in SAT and VAT (Table 3).

Table 3

Significance in protein levels corresponding to correlation between SAT/VAT and different biochemical parameters.
Biochemical parameters	Protein	Subcutaneous R² P
Glucose	Alcohol DH [NADP+]	0.039	0.638	0.392	0.05 ↓
	Apolipoprotein E	0.139	0.53	0.635	0.03 ↑
	Protein disulfide isomerase A3	0.389	0.04 ↓	0.101	0.34
Tryglicerides	Annexin A5	0.017	0.7	0.385	0.04 ↑
	Apolipoprotein A1	0.291	0.08	0.331	0.03 ↑
	Breast carcinoma-amplified seq-1	0.558	< 0.01 ↓	0.004	0.82
	Haptoglobin	0.08	0.39	0.371	0.04 ↓
	Hemopexin	0.477	0.01 ↓	0.216	0.15
	Retinal DH	0.412	0.03 ↓	0.113	0.31
	Protein disulfide isomerase A3	0.203	0.61	0.476	0.02 ↓
	Paroxiredoxin 2	0.496	0.02 ↓	0.03	0.61
	Ribonuclease inhibitor	0.552	< 0.01 ↓	0.153	0.23
HDL	Alpha-1 antitrypsin	0.414	0.04 ↓	0.116	0.33
	Annexin A5	0.001	0.82	0.459	0.03 ↑
	D2 Dopamine receptor	0.495	0.02 ↓	0.066	0.47
	L-lactate DH A chain	0.454	0.03 ↓	0.551	0.01 ↓
LDL	Actin. Aortic smooth muscle	0.652	< 0.01 ↓	0.385	0.04 ↓
	Annexin A3	0.41	0. 04 ↓	0.009	0.77
	Apolipoprotein E	0.551	0.15	0.636	0.03 ↓
	Breast carcinoma-amplified seq-1	0.544	0.01 ↓	0.02	0.69
	Creatine kinase B-type	0.844	< 0.01 ↓	0.001	0.91
	Glycerol-3-phosphate DH [NAD+]	0.398	0.03 ↑	0.001	0.92
	Haptoglobin	0.003	0.87	0.373	0.04 ↓
	HSP 60	0.396	0.03 ↓	0.042	0.54
	Intelectin-1	0.399	0.03 ↑	0.234	0.132
	Retinal DH	0.67	< 0.01 ↓	0.029	0.62
	Paroxiredoxin 2	0.588	0.01 ↓	0.082	0.39
	Serine/Threonine protein kinase	0.753	< 0.01 ↓	0.044	0.53
	Tropomyosin beta chain	0.137	0.29	0.424	0.03 ↓
Cholesterol	14-3-3 Protein alpha/beta	0.479	0.01 ↑	0.04	0.55
	14-3-3 Protein gamma	0.57	< 0.01 ↑	0.247	0.24
	14-3-3 Protein zeta/delta	0.514	0.01 ↑	0.338	0.06
	26S protein regulatory subunit 6B	0.179	0.29	0.696	< 0.01 ↑
	ADP/ATP Translocase 3	0.38	0.04 ↓	0.092	0.39
	Alpha-1 antitrypsin	0.582	< 0.01 ↓	0.003	0.87
	Annexin A5	0.001	0.8	0.359	0.05 ↑
	Apolipoprotein A1	0.022	0.66	0.44	0.02 ↑
	Collagen alpha-1 (XIII) chain	0.39	0.05	0.496	0.02 ↑
	Fibrinogen gamma chain	0.117	0.3	0.378	0.04 ↑
	Haptoglobin	0.535	0.01 ↓	0.213	0.15
	Hemopexin	0.196	0.17	0.446	0.02 ↓
	HSP 60	0.002	0.88	0.463	0.03 ↑
	Serine/Threo protein kinase	0.251	0.11	0.433	0.02 ↑
P value ≤ 0.05 is considered significant. (↑ up-regulated and ↓ down-regulated).

2. Influence of diabetes in the proteome of subcutaneous and visceral white adipose tissue of obese patients

Because obesity is associated with metabolic disorders, we studied the effect of type 2 diabetes mellitus (DM) in AT. The influence of DM on the differential proteomic profile between SAT and VAT was striking; changes were affecting different proteins than those observed above in non-diabetic patients. Again the two fat depots were from the same donors and therefore dietary, environmental and genetic factors could be excluded. In obese non-diabetic subjects, expression of Inorganic pyrophosphatase appeared as a new protein significantly reduced in SAT, and these were no proteins showing up-regulation (Table 2 Column II). However, in obese-diabetic patients two new proteins appeared up-regulated (Actin-related protein 2 [ARP2] and Apolipoprotein A1) and two down regulated (Creatine kinase B-type and HSP 60) in SAT versus VAT (Table 2 Column III).

Next, we investigated whether DM may affect AT protein composition in the same fat depot. DM has different effect on AT depending of its localization, thus, not all proteins with a differential profile on SAT showed differences in VAT and vice versa (Fig. 2). 15 proteins in SAT and 18 proteins in VAT were identified with significantly changed due to the presence of diabetes (Table 4), although influence of DM was higher in VAT than SAT. DM patients have many more proteins down-regulated in SAT than those up-regulated in VAT with respect to non-diabetics. However, the affected proteins were different in the two fat depots of the same patient. Only 6 proteins showed the same trend in both tissues: a reduction of Annexin A3 (ANX3), HSP 27 (HSPBI), 14-3-3 protein a/b (YWHAB), and 14-3-3 protein zeta/delta (YWHAZ) and an increased level of haptoglobin (HP), and ADP/ATP translocase 3 SLC24A6) (Fig. 2).

Table 4

Significance of changes in differential protein expression in diabetic *versus* non-diabetic patients in subcutaneous and visceral adipose.
	Diabetes vs non-Diabetes
	Subcutaneous	Visceral
14-3-3 protein b/a	0.02 ↓	0.004 ↓
14-3-3 protein gamma	0.42	0.05 ↓
14-3-3 protein zeta/delta	0.02 ↓	0.03 ↓
26S protease regulatory subunit 6B	0.03 ↓	0.20
Actin. aortic smooth muscle	0.02 ↓	0.34
Actin-related protein 2	0.42	0.01
ADP/ATP Translocase 3	0.05 ↑	0.006 ↑
Alpha-1-antitrypsin	0.26	0.01 ↓
Alpha-1B-glycoprotein	0.47	0.01 ↑
Alpha-2-macroglobulin receptor-associated protein	0.35	0.004 ↑
Annexin A3	0.006 ↓	0.006 ↓
Annexin A8-lije protein 2	0.63	0.004 ↓
Antithrombin III	0.11	0.004 ↓
Apolipoprotein A1	0.01 ↓	0.20
Collagen alpha-1(XIII) chain	0.10	0.04 ↓
Cytochrome c oxidase subunit 7A-related protein. Mitochondrial	0.03 ↑	0.20
Fibrinogen gamma chain	0.11	0.01 ↓
Haptoglobin	0.03 ↑	0.03 ↑
Heat shock cognate 71 kDa protein	0.05 ↑	0.75
Hemopexin	0.2	0.03 ↑
HSP 27	0.05 ↓	0.007 ↓
HSP 60	0.004 ↓	0.75
Protein disulfide isomerase	0.33	0.006 ↑
Selenium-binding protein 1	0.007 ↑	0.52
Serine/threonine-protein phosphatase 2A catalytic subunit alpha	0.05 ↓	0.08
Transhyretin	0.20	0.02 ↑
Ubiquitin carboxyl-terminal hydrolase isozyme 1	0.03 ↑	0.34
P value < 0.05 is considered significant and marked in bold. ↑ proteins up-regulated and ↓ proteins down-regulated in diabetes versus non-diabetes.

To find the protein-interaction networks affected by the presence of diabetes on each tissue we performed an in silico bioinformatics analysis with the Ingenuity Pathway Analysis (IPA) software. For this, we introduce all those proteins with a significant change or a ratio DM versus non DM lower than 0.5 or higher than 1.7. Regarding SAT, the top affected network with a score of 56 and 25 proteins involved was the network “Endocrine System Disorders. Free radical Scavenging” (Supplemental Fig. 1). On the other hand, when we studied VAT the top affected network was associated with “Metabolic disease” with a score of 40 and the contribution of 16 proteins (Supplemental Fig. 2).

3. Identification and detection of Kazrin in adipose tissue

Investigating the proteomic profile of AT, we identified a protein never before described in adipose tissue. This is Kazrin, a protein originally described expressed in stratified squamous epithelia that becomes membrane associated in the suprabasal layers, coincident with up-regulation of periplakin, and is incorporated into the cornified envelope of cultured keratinocytes [16]. It plays a role in desmosome assembly, cell adhesion, cytoskeletal organization, and epidermal differentiation. This protein co-localizes with desmoplakin and the cytolinker protein periplakin. In general, this protein localizes to the nucleus, desmosomes, cell membrane, and cortical actin-based structures [17]. Figure 3 shows Kazrin in the different types of AT studied. Expression is higher in visceral fat than in subcutaneous fat both in non-diabetics and diabetics (Supplemental Fig. 3).

4. Influence of obesity and diabetes on serum proteins.

A sample of blood was extracted for serum isolation at the time of surgery when WATs were collected from the patients. Serum protein analysis was performed with a Bio-Plex kit, which allowed measuring a great amount of proteins at the same time. Forty-nine proteins related to inflammation and diabetes were investigated. A group of non-obese and non-diabetic patients was included as control of plasma values (Supplemental Table 2).

Ten proteins appeared in significantly higher levels in obese patients compared to non-obese controls (independently on the presence of diabetes) (Fig. 4), including pro-inflammatory proteins as C-peptide, Chitinase 3-like 1, Tumor necrosis factor receptor 1 (TNFR-1), gut derived incretin hormones as Ghrelin, Glucose-dependent insulinotropic polypeptide (GIP), Glucagon like-peptide-1 (GLP-1), Glucagon, glucose related adipokines as Insulin, Leptin, Adipsin. Three proteins were minimally affected APRIL (A proliferation-inducing ligand), PAI-1 (Plasminogen activator-inhibitor-1), and TWEAK (TNF-related weak inducer of apoptosis); and three proteins were found in lower levels in obese patients, the obesity-related adipokine adiponectin, the inflammatory proteins matrix metalloproteinase-2 (MMP-2), and Osteocalcin.

We then compared the differential serum protein levels in obese and diabetic patients (Fig. 5) and found seven proteins in higher levels in the serum of diabetic patients than in non-diabetics, related with glucose metabolism, satiety and inflammation: Ghrelin, GLP-1, Glucagon, B-cell activating factor (BAFF), Chitinase 3-like, TNFR-1 and TNFR-2; one minimally modified Matrix metalloproteinase-3 (MMP-3), and one protein in lower levels in diabetic patients, the cytokine interleukin-8 (IL-8).

Results in the three patient groups; 1) non-obese & non-diabetic; 2) Obese & non-diabetic; 3) Obese & diabetic are shown in Fig. 6 and Table 5. C-Peptide, GLP-1, glucagon, insulin, adipsin, chitinase 3-like 1 and TNF-R1 levels increased with increasing cardiovascular risk, suggesting an alteration in both the inflammatory and metabolic systems. In contrast, Adiponectin level was reduced. Levels of gut derived incretins, ghrelin, GIP and leptin were significantly lower in group 1 (non-obese and non-diabetic) whereas no changes were observed between both groups of obese patients. On the contrary, MMP-2 and Osteocalcin serum levels were significantly higher in group1 (non-obese and non-diabetic) compared with the other two obese patient groups. Additionally, Osteopontin, MMP-3, TNF-R2 and BAFF levels were significantly lower in the obese non-diabetic group compared with the obese diabetic group, suggesting a reduced inflammatory state. Finally, TWEAK level of the obese non-diabetic group was reduced compared with the control group and IL-6RA was also reduced compared to both the control and the obese diabetic group.

Table 5

Differential serum-circulating protein concentration due to obesity and diabetes
	OB-nDM vs nOB-nDM	OB-DM vs nOB-nDM	OB-DM vs OB-nDM
Leptin (ng/ml)	< 0.01 ↑	< 0.01 ↑	0.713
Insulin (pg/ml)	0.01 ↑	< 0.01 ↑	0.52
Chitinase 3-like 1 (ng/ml)	0.66	< 0.01 ↑	0.01 ↑
C-peptide (ng/ml)	0.03 ↑	< 0.01 ↑	0.86
TNF-R1 (ng/ml)	0.03 ↑	< 0.01 ↑	< 0.01 ↑
Ghrelin (pg/ml)	< 0.01 ↑	< 0.01 ↑	0.82
GLP-1 (pg/ml)	0.01 ↑	< 0.01 ↑	0.18
Glucagon (pg/ml)	0.03 ↑	< 0.01 ↑	0.08
IL-6Ra (ng/ml)	0.04 ↓	0.84	0.04 ↑
TWEAK/TNFSF12 (pg/ml)	0.03 ↓	0.16	0.33
Adiponectin (µg/ml)	0.01 ↓	< 0.01 ↓	0.75
Osteocalcin (ng/ml)	< 0.01 ↓	0.01 ↓	0.33
MMP-2 (ng/ml)	< 0.01 ↓	< 0.01 ↓	0.59
PAI-1 (ng/ml)	0.05 ↓	0.16	0.96
GIP (pg/ml)	0.06	0.02 ↑	0.71
Adipsin (µg/ml)	0.07	0.01 ↑	< 0.01 ↑
TNF-R2 (pg/ml)	0.41	0.10	0.03 ↑
BAFF/TNFSF13B (ng/ml)	0.55	0.06	0.05 ↑
Osteopontin (OPN) (ng/ml)	0.16	0.58	0.05 ↑
MMP-3 (ng/ml)	0.24	0.26	0.03 ↑
Statistical significance was accepted at P ≤ 0.05; ↓ Down-regulated; ↑ up-regulate

Discussion

In the present study we have used a proteomic approach to compare, for the first time, subcutaneous and visceral fat from the same patient with type 3 obesity and with or without T2DM as well as their serum proteome.

Obesity is an important risk factor for cardiovascular diseases (CVD). However, a protective role of the adipose tissue in certain clinical conditions has been described, giving rise to the “obesity paradox”, defined as the duality behavior of the adipose tissue [18]. Until now, most of the studies carried out on AT have been performed by using non-invasive imaging technologies such as MRI, X-ray or ultrasound [19–21] as well as in different animal models. However, some limitations appear when animal findings are translated to human studies [22, 23]. Concerning human studies, there are a few studies that have compared the proteomic profile of both SAT and VAT of the same individual, but results have been quite different. Pérez-Pérez et. al. [24] found an increased expression of metabolic related proteins in VAT compared with SAT, but Insener et. al. [21] did not reproduce similar findings. Additionally, obesity is associated with the development of insulin resistance and type-2 diabetes mellitus (T2DM), although not all obese patients become diabetic or insulin resistant. Therefore, serum and adipose tissue composition, or adipose tissue location might differ in obese patients with or without any other metabolic complication [20, 25]. The proteomic signature of VAT from diabetic and non-diabetic patients was previously described but a little number of changes was reproduced among the different groups [26]. However SAT proteomic signature in diabetic and non-diabetic has not been investigated.

VAT associates with higher metabolic complications than SAT [27] and accordingly, it is not surprising to see powerful changes on visceral adipose tissue compared with subcutaneous tissue, as a result of diabetes. Likewise, changes on SAT are related to the endocrine system, whereas changes in VAT are linked to metabolic disease, supporting the involvement of VAT in metabolic disorders in comparison with that of SAT. Thus, an altered metabolic and inflammatory state was observed both in adipose tissue and in serum suggesting a coordinated effect probably caused by the altered protein secretion from adipocites to the bloodstream. Among all the identified proteins in AT, only Hemopexin (HPX), actin related protein-2 (Arp2) and 14-3-3-protein beta/alpha (YWHAB) were reduced in VAT compared with the SAT. HPX and Arp2 are both only modified in the diabetic group, and both associated with adipocyte differentiation and development [28, 29] suggesting adipocyte dysfunction in VAT of diabetic patients. On the other hand, a relative increased expression in VAT compared with SAT, was observed in many other proteins, highlighting the presence of two proteins, kazrin and N(G)-dimethylarginine dimethylaminohydrolase 2 (DDAH2).

Kazrin was initially identified bounded to the N-terminal domain of the cytolinker periplakin [17]. Multiple functions have been reported, including regulation of desmosome assembly [30], epidermal differentiation by regulation of keratinocytes adhesion and differentiation [31], but its association with visceral adiposity was unknown, in fact, kazrin had never before been detected in any adipose tissue. Desmosomes are intercellular junctions of epithelia and cardiac muscle. Disruption of desmosomal adhesion implicates human health problems, mutations in genes encoding desmosomal proteins have been related with arrhythmogenic right ventricular dysplasia, a heart muscle disease characterized by life-threatening arrhythmias and increased risk of sudden heart failure [32]. The presence of Kazrin in AT could indicate the degree of AT dystrophy.

By contrast, DDAH2, protein narrowly related with insulin secretion [33], was previously detected in VAT, but not in SAT, and therefore, no one has previously compare the expression of this protein in different fat depots. The presence of these two proteins in AT underwrites the need to further investigate adipose cells function and regulation that is highly dependent on fat location

Besides its undoubted link to metabolic disorders, obesity has also a direct effect in oxidative stress that is present in countless diseases. This might be one potential connector of obesity with related disorders such as T2DM and cardiovascular diseases [34, 35]. Indeed, diets rich on carbohydrates induce oxidative stress and inflammation state in obese individuals [34]. Also, Murri et al have previously described the effect of DM in the visceral oxidative proteome [26]. They described an antioxidant state in the visceral fat of diabetic patients that could act as a defense against the metabolic stress induced by insulin resistance, diabetes and hyperglycaemia. In line with these, we aimed to compare the differential oxidative stress among the different fat depots, and we observed an up-regulation of different oxidation related proteins in VAT compared with SAT, including antioxidant proteins as Paroxiredoxin-6 [36], gluthatione-S Transferase (GSTP1) [37], and the glutathione related protein disulfide isomerase-3 (PDIA3) [38], supporting the statement that visceral adipose tissue from diabetic patients could try to compensate the characteristic oxidative stress that comes up with the impaired glucose handling. However, not only oxidation-related proteins were up-regulated in VAT but also proteins related with coagulation, inflammation and cholesterol efflux indicating that VAT is a much more active fat depot than SAT. Besides the differential adipose tissue protein composition, we also observe changes at serum levels, related both to obesity and DM. Obesity is known to be associated with a low-grade pro-inflammatory state and many inflammatory adipokines are altered in the secretome of obese patients compared with lean subjects [39, 40]. This also happens with the proteins secreted by the adipose tissue, adipokines such Leptin, Adiponectin and Visfatin among others, are dysregulated as a response of fat accumulation [41]. Thus, as it could be expected, the pro-inflamatory adipokines Leptin, Adipsin, Insulin, C-Peptide, Chitinase 3-like 1, Tumor necrosis factor receptor 1 (TNF-R1), and TNF-ligand superfamily member 13 (TNFSF13 or APRIL) were found in higher levels in the serum profile of the obese group compared with lean subjects and osteocalcin, which improves insulin resistance by decreasing inflammation [42], was found in lower levels in the obese patients group. A narrow relation between insulin resistance, DM and inflammation is also generally accepted [43]. Therefore, a coordinated pro-inflammatory state is observed not only in the adipose tissue, but also in the serum proteins as a response to obesity and diabetes. However, serum protein composition is more complex than just adipose tissue secreted proteins, being also affected by the secretion of other cytokines produced by other organs such as the pancreas, the liver, and the hypothalamus [44] and cells from the cardiovascular system.

At circulating level, we have observed significant higher levels of the gut derived incretin hormones, specifically the obese patients showed higher levels of Ghrelin, GIP, GLP-1 and Glucagon. Incretin hormones have an important role in glucose homeostasis [45]. In fact, T2DM patients show a decreased bioavailability of these proteins. In previous studies, incretins have showed an anti-inflammatory, antioxidant and anti-apoptotic effect and they are major players on appetite and satiety, being also widely related with obesity [46]. Interestingly, serum levels of PAI-1, an inhibitor of fibrinolysis usually up-regulated in obesity [47] were not much affected in our patients. This could be caused by the hypocaloric diet prior bariatric surgery received by the patients that could have affected their metabolic profile. However, as expected, the weight loss of those patients (around the 5 to 8% of the total body mass) was not enough to achieve a total improvement on the serum profile of type 3 obese subjects, and the down-regulation of Adiponectin, protein negatively correlated with body mass [39] and the up-regulation of adipsin and leptin, both proteins usually increased with body mass gain [41], were observed in the serum proteome of this type 3 obese individuals.

Not only obesity, but also diabetes affects the serum proteome. However, unlike in adipose tissue composition where metabolic, oxidation, coagulation and inflammatory proteins were altered, mainly inflammatory cytokines as IL-8, TNFSF13 B, TNFR1 and TNFR2 [41] and the metalloproteinase MMP-3 were affected in the serum of diabetic patients, independently on the presence of obesity. This pro-inflammatory state is however accompanied with the increase of insulin in DM patients and adipsin levels in obese diabetic patients. Adipsin, which is a member of the complement system, is also associated with the protection of β-cells and therefore positively related to insulin levels [48].

References

Ray I, Bhattacharya A, De RK. OCDD: an obesity and co-morbid disease database. BioData mining. 2017;10:33.
Fuster JJ, Ouchi N, Gokce N, Walsh K. Obesity-Induced Changes in Adipose Tissue Microenvironment and Their Impact on Cardiovascular Disease. Circulation research. 2016;118(11):1786–807.
Badimon L, Bugiardini R, Cenko E, Cubedo J, Dorobantu M, Duncker DJ, Estruch R, Milicic D, Tousoulis D, Vasiljevic Z, et al. Position paper of the European Society of Cardiology-working group of coronary pathophysiology and microcirculation: obesity and heart disease. European heart journal. 2017;38(25):1951–8.
Wozniak SE, Gee LL, Wachtel MS, Frezza EE. Adipose tissue: the new endocrine organ? A review article. Digestive diseases sciences. 2009;54(9):1847–56.
Shibata R, Ouchi N, Ohashi K, Murohara T. The role of adipokines in cardiovascular disease. Journal of cardiology. 2017;70(4):329–34.
Antonopoulos AS, Tousoulis D. The molecular mechanisms of obesity paradox. Cardiovascular research. 2017;113(9):1074–86.
Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev. 2000;21(6):697–738.
Badimon L, Cubedo J. Adipose tissue depots and inflammation: effects on plasticity and resident mesenchymal stem cell function. Cardiovascular research. 2017;113(9):1064–73.
Peinado JR, Jimenez-Gomez Y, Pulido MR, Ortega-Bellido M, Diaz-Lopez C, Padillo FJ, Lopez-Miranda J, Vazquez-Martinez R, Malagon MM. The stromal-vascular fraction of adipose tissue contributes to major differences between subcutaneous and visceral fat depots. Proteomics. 2010;10(18):3356–66.
Onate B, Vilahur G, Camino-Lopez S, Diez-Caballero A, Ballesta-Lopez C, Ybarra J, Moscatiello F, Herrero J, Badimon L. Stem cells isolated from adipose tissue of obese patients show changes in their transcriptomic profile that indicate loss in stemcellness and increased commitment to an adipocyte-like phenotype. BMC Genomics. 2013;14:625.
Onate B, Vilahur G, Ferrer-Lorente R, Ybarra J, Diez-Caballero A, Ballesta-Lopez C, Moscatiello F, Herrero J, Badimon L. The subcutaneous adipose tissue reservoir of functionally active stem cells is reduced in obese patients. FASEB journal: official publication of the Federation of American Societies for Experimental Biology. 2012;26(10):4327–36.
Bejar MT, Ferrer-Lorente R, Pena E, Badimon L. Inhibition of Notch rescues the angiogenic potential impaired by cardiovascular risk factors in epicardial adipose stem cells. FASEB journal: official publication of the Federation of American Societies for Experimental Biology. 2016;30(8):2849–59.
Ferrer-Lorente R, Bejar MT, Badimon L. Notch signaling pathway activation in normal and hyperglycemic rats differs in the stem cells of visceral and subcutaneous adipose tissue. Stem Cells Dev. 2014;23(24):3034–48.
Tous M, Ferrer-Lorente R, Badimon L. Selective inhibition of sphingosine kinase-1 protects adipose tissue against LPS-induced inflammatory response in Zucker diabetic fatty rats. American journal of physiology Endocrinology metabolism. 2014;307(5):E437–46.
Cubedo J, Padro T, Garcia-Moll X, Pinto X, Cinca J, Badimon L. Proteomic signature of Apolipoprotein J in the early phase of new-onset myocardial infarction. J Proteome Res. 2011;10(1):211–20.
DiColandrea T, Karashima T, Maatta A, Watt FM. Subcellular distribution of envoplakin and periplakin: insights into their role as precursors of the epidermal cornified envelope. J Cell Biol. 2000;151(3):573–86.
Groot KR, Sevilla LM, Nishi K, DiColandrea T, Watt FM. Kazrin, a novel periplakin-interacting protein associated with desmosomes and the keratinocyte plasma membrane. J Cell Biol. 2004;166(5):653–9.
Amundson DE, Djurkovic S, Matwiyoff GN. The obesity paradox. Crit Care Clin. 2010;26(4):583–96.
Abbott RD, Borowsky FE, Quinn KP, Bernstein DL, Georgakoudi I, Kaplan DL. Non-invasive Assessments of Adipose Tissue Metabolism In Vitro. Ann Biomed Eng. 2016;44(3):725–32.
Despres JP, Lemieux I. Abdominal obesity and metabolic syndrome. Nature. 2006;444(7121):881–7.
Insenser M, Montes-Nieto R, Vilarrasa N, Lecube A, Simo R, Vendrell J, Escobar-Morreale HF. A nontargeted proteomic approach to the study of visceral and subcutaneous adipose tissue in human obesity. Molecular cellular endocrinology. 2012;363(1–2):10–9.
Oikonomou EK, Antoniades C. The role of adipose tissue in cardiovascular health and disease. Nature reviews Cardiology. 2019;16(2):83–99.
Peinado JR, Pardo M, de la Rosa O, Malagon MM. Proteomic characterization of adipose tissue constituents, a necessary step for understanding adipose tissue complexity. Proteomics. 2012;12(4–5):607–20.
Perez-Perez R, Ortega-Delgado FJ, Garcia-Santos E, Lopez JA, Camafeita E, Ricart W, Fernandez-Real JM, Peral B. Differential proteomics of omental and subcutaneous adipose tissue reflects their unalike biochemical and metabolic properties. J Proteome Res. 2009;8(4):1682–93.
Kusminski CM, Bickel PE, Scherer PE. Targeting adipose tissue in the treatment of obesity-associated diabetes. Nature reviews Drug discovery. 2016;15(9):639–60.
Murri M, Insenser M, Bernal-Lopez MR, Perez-Martinez P, Escobar-Morreale HF, Tinahones FJ. Proteomic analysis of visceral adipose tissue in pre-obese patients with type 2 diabetes. Molecular cellular endocrinology. 2013;376(1–2):99–106.
Ibrahim MM. Subcutaneous and visceral adipose tissue: structural and functional differences. Obesity reviews: an official journal of the International Association for the Study of Obesity. 2010;11(1):11–8.
Lawson HA, Zayed M, Wayhart JP, Fabbrini E, Love-Gregory L, Klein S, Semenkovich CF. Physiologic and genetic evidence links hemopexin to triglycerides in mice and humans. Int J Obes. 2017;41(4):631–8.
Yang W, Thein S, Lim CY, Ericksen RE, Sugii S, Xu F, Robinson RC, Kim JB, Han W. Arp2/3 complex regulates adipogenesis by controlling cortical actin remodelling. Biochem J. 2014;464(2):179–92.
Karashima T, Watt FM. Interaction of periplakin and envoplakin with intermediate filaments. Journal of cell science. 2002;115(Pt 24):5027–37.
Liu C, Wang J, Hu Y, Xie H, Liu M, Tang H. Upregulation of kazrin F by miR-186 suppresses apoptosis but promotes epithelial-mesenchymal transition to contribute to malignancy in human cervical cancer cells. Chinese journal of cancer research = Chung-kuo yen cheng yen chiu. 2017;29(1):45–56.
Ohno S. The genetic background of arrhythmogenic right ventricular cardiomyopathy. Journal of arrhythmia. 2016;32(5):398–403.
Hasegawa K, Wakino S, Kimoto M, Minakuchi H, Fujimura K, Hosoya K, Komatsu M, Kaneko Y, Kanda T, Tokuyama H, et al. The hydrolase DDAH2 enhances pancreatic insulin secretion by transcriptional regulation of secretagogin through a Sirt1-dependent mechanism in mice. FASEB journal: official publication of the Federation of American Societies for Experimental Biology. 2013;27(6):2301–15.
Fernandez-Sanchez A, Madrigal-Santillan E, Bautista M, Esquivel-Soto J, Morales-Gonzalez A, Esquivel-Chirino C, Durante-Montiel I, Sanchez-Rivera G, Valadez-Vega C, Morales-Gonzalez JA. Inflammation, oxidative stress, and obesity. Int J Mol Sci. 2011;12(5):3117–32.
Matsuda M, Shimomura I. Increased oxidative stress in obesity: implications for metabolic syndrome, diabetes, hypertension, dyslipidemia, atherosclerosis, and cancer. Obesity research clinical practice. 2013;7(5):e330–41.
Fisher AB. Peroxiredoxin 6 in the repair of peroxidized cell membranes and cell signaling. Arch Biochem Biophys. 2017;617:68–83.
Doukali H, Ben Salah G, Hamdaoui L, Hajjaji M, Tabebi M, Ammar-Keskes L, Masmoudi ME, Kamoun H. Oxidative stress and glutathione S-transferase genetic polymorphisms in medical staff professionally exposed to ionizing radiation. Int J Radiat Biol. 2017;93(7):697–704.
Okada K, Fukui M, Zhu BT. Protein disulfide isomerase mediates glutathione depletion-induced cytotoxicity. Biochem Biophys Res Commun. 2016;477(3):495–502.
Trujillo ME, Scherer PE. Adipose tissue-derived factors: impact on health and disease. Endocr Rev. 2006;27(7):762–78.
Vasilenko MA, Kirienkova EV, Skuratovskaia DA, Zatolokin PA, Mironyuk NI, Litvinova LS. The role of production of adipsin and leptin in the development of insulin resistance in patients with abdominal obesity. Doklady Biochemistry biophysics. 2017;475(1):271–6.
Fantuzzi G. Adipose tissue, adipokines, and inflammation. J Allergy Clin Immunol. 2005;115(5):911–9. quiz 920.
Mizokami A, Kawakubo-Yasukochi T, Hirata M. Osteocalcin and its endocrine functions. Biochem Pharmacol. 2017;132:1–8.
Patel P, Abate N. Body fat distribution and insulin resistance. Nutrients. 2013;5(6):2019–27.
Lontchi-Yimagou E, Sobngwi E, Matsha TE, Kengne AP. Diabetes mellitus and inflammation. Curr Diabetes Rep. 2013;13(3):435–44.
Nauck MA, Meier JJ. Incretin hormones: Their role in health and disease. Diabetes Obes Metab. 2018;20(Suppl 1):5–21.
Joao AL, Reis F, Fernandes R. The incretin system ABCs in obesity and diabetes - novel therapeutic strategies for weight loss and beyond. Obesity reviews: an official journal of the International Association for the Study of Obesity. 2016;17(7):553–72.
Supriya R, Tam BT, Yu AP, Lee PH, Lai CW, Cheng KK, Yau SY, Chan LW, Yung BY, Sheridan S, et al. Adipokines demonstrate the interacting influence of central obesity with other cardiometabolic risk factors of metabolic syndrome in Hong Kong Chinese adults. PloS one. 2018;13(8):e0201585.
Lo JC, Ljubicic S, Leibiger B, Kern M, Leibiger IB, Moede T, Kelly ME, Chatterjee Bhowmick D, Murano I, Cohen P, et al. Adipsin is an adipokine that improves beta cell function in diabetes. Cell. 2014;158(1):41–53.

Untargeted and Targeted Proteomic Analysis of the Impact of Diabetes Mellitus on Two Adipose Tissue Depots and Blood of Obese Patients

Abstract

Background

Methods

Study population and design

Sample collection and preparation

Biochemical analysis

Two-Dimensional Gel Electrophoresis (2-DE)

Mass spectrometry analysis

Quantification of serum protein levels

In silico bioinformatics analysis

Statistical analysis

Results

1. Differential protein expression between the subcutaneous and the visceral white adipose tissue

2. Influence of diabetes in the proteome of subcutaneous and visceral white adipose tissue of obese patients

3. Identification and detection of Kazrin in adipose tissue

4. Influence of obesity and diabetes on serum proteins.

Discussion

Conclusions

List of Abbreviations

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and materials

Supplementary information

Competing interests

Funding

Authors' contributions

Acknowledgements

References