Intrahepatic fat content correlates with soluble CD163 in relation to weight loss induced by Roux-en-Y gastric bypass
Funding agencies: This research was supported by The Danish Council for Strategic Research (TRAIN:10-092797), the Novo Nordisk Foundation, AP Møller Foundation, and Aarhus University.
Disclosure: The authors have no competing interests besides HJM who has received royalties from IQ-products (Netherlands).
Author contributions: KF designed and performed the clinical study, analyzed the data, searched the literature, generated the figures, wrote the initial manuscript, and interpreted the data together with SBP, HJM, and BR. The RT-PCR, ELISA, immunohistochemistry, MRI, and liver spectroscopy were performed by KF, SPB, BR, HJM, HJS, PR, and AVD. All authors were involved in writing the final paper and had final approval of the submitted and published versions.
Abstract
Objective
Soluble CD163 (sCD163) is a new marker of obesity-related metabolic complications. sCD163 and CD163 mRNA were investigated in relation to the fat distribution at baseline and 12 months after Roux-en-Y gastric bypass (RYGB).
Methods
Thirty-one obese subjects (BMI: 42.3 ± 4.7 kg/m2) were enrolled. Subcutaneous (SAT) and visceral adipose tissue (VAT) volume were determined by MRI, intrahepatic lipid content (IHL) by MR-spectroscopy, and body composition by DXA. Fasting blood samples and adipose tissue samples were obtained, and ELISA and RT-PCR were performed.
Results
RYGB-induced weight loss (36 ± 11 kg) was accompanied by a significant reduction in sCD163 (2.1 ± 0.8 mg/l vs. 1.7 ± 0.7 mg/l), SAT, VAT, and IHL (all, P < 0.001). At baseline, sCD163 was associated with VAT (r = 0.40, P < 0.05) but not with SAT or IHL. Moreover, CD163 mRNA was significantly upregulated in VAT compared with SAT at baseline (P < 0.05) and significantly downregulated in SAT after RYGB (P < 0.001). ΔsCD163 was significantly associated with ΔIHL after RYGB compared with baseline (r = 0.40, P < 0.05).
Conclusions
RYGB-induced weight loss results in a reduction of sCD163 and CD163 mRNA. The association between ΔsCD163 and ΔIHL may reflect a reduction in sCD163-producing Kupffer cells in the liver. Moreover, sCD163 may be a marker of “unhealthy” fat distribution in obese subjects.
Introduction
We and others have recently found that the macrophage-specific serum marker soluble CD163 (sCD163) is a good marker for obesity-related metabolic complications since sCD163 is significantly associated with BMI and with insulin resistance (1-6). Furthermore, sCD163 has been found to be a strong predictor for the development of type 2-diabtes (7). CD163 is a hemoglobin scavenger receptor primary expressed on monocytes and macrophages (8, 9). sCD163 is found to be elevated in pathological conditions with activation of the monocyte-macrophage system (10-12). However, small amounts of sCD163 are constitutively shed to the circulation and the normal range is between 0.7 and 3.9 mg/l in healthy individuals (13). The metalloproteinase tumor necrosis factor α-converting enzyme (ADAM17) mediates the shedding of sCD163 (and TNF-α) from the plasma membrane (14). CD163 is expressed in macrophages resident in several tissues such as the adipose tissue (AT), liver, spleen, and in blood monocytes (15); however, it is unknown which tissues that primarily contribute to the circulating level of sCD163. CD163 is highly expressed in the macrophages in the AT and the expression of CD163 in the subcutaneous AT (SAT) has been found to be closely associated with circulating sCD163 (16, 17).
In obesity, there is an increased amount of activated macrophages in the AT due to adipocyte expansion and possibly local hypoxia (18). The infiltrating macrophages produce inflammatory cytokines that cause systemic inflammation and insulin resistance (19, 20). Furthermore, accumulation of excessive fat in the visceral adipose depot (VAT) and ectopic fat in the liver (IHL), as compared to the subcutaneous adipose depot (SAT), is strongly associated with systemic inflammation and increased risk of developing metabolic complications (21, 22). The obesity-induced systemic inflammation is reversible since weight loss is associated with decreased levels of inflammatory cytokines, reduced SAT and VAT, and improved insulin sensitivity (23, 24). Bariatric surgery is the most effective way to achieve a sustained weight loss and to improve the long-term metabolic profile (25, 26).
The aim of this study was to evaluate the level of sCD163 and the expression of CD163 in relation to the fat distribution (normal and ectopic) and in relation to Roux-en-Y gastric bypass (RYGB)-induced weight loss in obese subjects.
Methods
Forty-six obese subjects were enrolled, who prospectively had been included in a RYGB program at the Department of Endocrinology at Aarhus University Hospital. Thus, meeting the national criteria of RYGB in Denmark (BMI >50 kg/m2 or BMI >35 kg/m2 plus obesity-related comorbidities). Exclusions criteria were glucocorticoid treatment or chronic systemic disease with the exception of type 2 diabetes. 13 subjects failed to lose the required weight (8%) before RYGB and two subjects dropped out of the clinical project. The 31 subjects who completed the study were evaluated at baseline and again at 6 and 12 months after the laparoscopic gastric bypass surgery, which proceeded without any postoperative complications for all subjects. Among the 31 subjects, 11 had been diagnosed with type 2 diabetes. Four of these patients were treated with metformin, one with liraglutide, four with a combination of different treatments, and two did not receive any medical treatment. Two subjects were treated with metformin due to PCOS. Of the 31 subjects, 10 were taking statins at baseline and five of these continued to take the same dose throughout the study. The study took place at the research laboratory at Aarhus University Hospital. All subjects were recruited via our out-patient clinic. The study was approved by the local ethics committee in the county of Aarhus and followed the principles of the Declaration of Helsinki (ClinicalTrialsgov.number:NCT01463449).
AT biopsy
In the beginning of the RYGB surgery, abdominal SAT and VAT biopsies were removed (n = 25). Immediately after the removal, the AT was transported to the laboratory where the samples were rinsed in isotonic NaCl and cut into smaller fragments (between 10 and 20 mg). Visible vessels and connective tissue were carefully removed and the AT samples were snap-frozen in liquid nitrogen and kept at −80°C until RNA extraction. About 12 months after RYGB, a new fat sample was taken (needle aspiration) from the abdominal SAT depot as previously described (27) from the same subjects (n = 25). No biopsy was obtained from VAT after 12 months.
Quantification of mRNA
RNA was isolated as previously described (17). All analyses (n = 25) were performed simultaneously and the mRNA levels of the target genes were expressed relative to the house-keeping gene glyceraldehyde-3-phosphate dehydrogenase (GADPH). The PCR reactions were performed in duplicate using the KAPA SYBR FAST qPCR kit (Kapa Biosystems, Woburn, MA) in a LightCycler 480 (Roche Applied Science) as previously described (17). The relative gene expression was estimated using the default “advanced relative quantification” mode of the software version LCS 480 1.5.0.39 (Roche Applied Science). Following primer pairs were used for amplification of CD163 (5′-3′): CGGCTGCCTCCACCTCTAAGT and ATGAAGATGCTG GCGTGACA, CD68 (5′-3′): GCTACATGGCGG TGGAGTACAA and ATGATGAGAGGCAGCAAGATGG, ADAM17 (5′-3′): ATCTGAACAACGACACCTGCTG and AAGGACTGTTCCTGTCACTGCAC, and those for GADPH (5′-3′) AATGAAGGGGTCATTGATGG and AAGGTGAAGGTCGGAGTCAA. Before analysis of target genes the house-keeping gene was tested for stability and found stable displaying comparable number of CT cycles.
Immunohistochemistry (IHC)
SAT (n = 6) and VAT (n = 5) biopsies were obtained in relation to the RYGB surgery and paired SAT biopsies were obtained 12 months after RYGB surgery (n = 6). Immediately after the removal the samples were fixed in 4% formaldehyde (24-48 h at room temperature) and then embedded in paraffin. Sections of 3 μm were deparaffinized, rehydrated and stained with hematoxylin and eosin using routine methods (28). Moreover, 3 μm sections were deparaffinized with EZ prep., and stained for CD163 using anti-CD163 monoclonal mouse antihuman antibody (AbD Serotec, clon edhu-1, MCA1853, dilution 1:450), using standard IHC methods. A single stained whole slide from each biopsy was captured and digitalised at a magnification of 20× using the Hamamatsu Nanozoomer 2.0HT (Hamamamatsu Photonics, Hamamamatsu City, Japan). Virtual slides were loaded into the digital image analysis software Visiopharm Integrator System ver. 4.4.4.0 (Visiopharm, Hørsholm, Denmark). Adipocyte morphology was assessed in HE-slides using a specially designed analysis protocol package (APP). Total tissue area in each slide together with the area and perimeter of all intact adipocytes within the slide were automatically computed and means with standard deviations were calculated for all biopsies, respectively. The numbers of adipocytes were automatically computed and given as number of adipocytes pr. total area adjusted for differences in the total area between biopsies. All digitalized slides were visually screened to ensure correct and satisfactory classification and sufficient intact tissue morphology within each virtual slide after completed automated quantification.
Magnetic resonance (MR) measurements
MR imaging and MR liver spectroscopy were performed using a Signa Excite 1.5 Tesla twin speed scanner (GE Medical Systems, Milwaukee, USA) for all patients and visits (n = 26). Five subjects were not MR-scanned due to claustrophobia. Abdominal SAT and VAT data were determined from axial MRI recordings using a fast spin echo sequence (body coil; echo time: 8.5 ms; repetition time: 600 ms; slice thickness: 8 mm; field of view: 48 cm). A multi-slice technique was used, in which slices were obtained from the upper pole of the femoral heads to the upper pole of the kidney. The protocol of the MR imaging and liver spectroscopy has previously been described (29).
Dual-energy X-ray absorptiometry (DXA)
Body composition was measured by using the same DXA scanner (Hologic Discovery, Hologic, Waltham, MA, USA) for all patients and visits (n = 30). One subject was excluded because the preoperative weight exceeded the limit of the DXA scanner (150 kg). Body composition is presented as total fat mass and total lean body mass measured in kilograms and as percent body fat.
Anthropometrics and blood samples
Anthropometrics and fasting blood samples were collected at baseline, 6 months, and 12 months after RYGB. Venous blood samples were collected after an overnight fast and serum were frozen at −80°C. The homeostasis model assessment insulin resistance index (HOMA-IR) was calculated using the formula: serum fasting insulin (μU/ml)* fasting glucose (mmol/l)/22.5 (30, 31). Soluble CD163 was quantified in serum samples using an in-house enzyme-linked immunosorbent assay, as previously described (32). Adiponectin was measured using a human-specific high sensitive ELISA method (b-Bridge International, USA), high sensitive-CRP by an hs-CRP enzyme immunoassay (DRG, Germany), and TNF-α and IL-6 were measured using a high sensitivity ELISA (R&D Systems, USA).
Statistical analysis
Descriptive statistics for anthropometric, metabolic markers and fat depots are presented as means ± SD.
Baseline associations were determined by a Spearman's rank correlation. Differences in high- and low-VAT groups were analyzed with an unpaired t-test or a Wilcoxon Mann-Whitney rank sum test for those variables, which were not normally distributed. The paired data, VAT versus SAT, was computed with a Wilcoxon signed rank test or a paired t-test where appropriated. Changes in 6 and 12 months after RYGB were determined by one-way repeated measures ANOVA (Holm-sidak corrected). Finally, the relationship between the relative changes before and 1 year after RYGB were analyzed by a Spearman's Rank Correlation. The chosen significance level was a two-tailed P-value of <0.05. The statistical software packet SPSS (SPSS, Chicago, IL) was used for all calculations. Graphs were made in Sigma-Plot.
Results
Baseline characteristics of the subjects are given in Table 1. Mean BMI was 42.3 ± 4.7 kg/m2 (range: 35-51 kg/m2). Age was between 25 and 57 years with a mean of 43.9 ± 7.7 years, and 61% of the subjects were females. There was a relative large individual variation in the fat depots at baseline with a range of SAT at 7941-19422 cm3 and a range of VAT at 1936-7621 cm3.
| Baseline | 6 Months | 12 Months | |
|---|---|---|---|
| Number | 31 | 31 | 31 |
| Age | 43.9 ± 7.7 | — | — |
| Sex (female/male), n | 19/12 | — | — |
| BMI (kg/m2) | 42.3 ± 4.7 | 31.8 ± 3.8b | 29.9 ± 3.7b |
| Waist (cm) | 118.9 ± 11.5 | 94.5 ± 9.7b | 91.8 ± 12.3b |
| ALAT (U/l) | 36.3 ± 13.6 | 30.1 ± 18.3a | 27.5 ± 12.4a |
| Glucose (mmol/l) | 7.2 ± 1.7 | 6.1 ± 1.2a | 5.9 ± 0.9a |
| Insulin (pmol/l) | 102.0 ± 49.4 | 45.7 ± 34.2b | 37.7 ± 27.3b |
| HOMA-IR | 5.5 ± 3.0 | 2.2 ± 2.1b | 1.7 ± 1.3b |
| sCD163 (mg/l) | 2.1 ± 0.8 | 1.8 ± 0.6b | 1.7 ± 0.7b |
| Adiponectin (μg/ml) | 5.8 ± 3.2 | 8.5 ± 3.7b | 11.2 ± 5.5bc |
| IL-6 (pg/ml) | 2.7 ± 1.4 | 1.9 ± 0.8a | 1.8 ± 0.8a |
| TNF-α (pg/ml) | 1.8 ± 0.4 | 2.1 ± 1.4 | 1.7 ± 0.6 |
| hs-CRP (mg/l) | 5.3 ± 3.4 | 2.0 ± 1.9b | 1.2 ± 1.1b |
| Lean mass (kg) | 67.1 ± 12.8 | — | 57.2 ± 11.6b |
| Fat mass (kg) | 51.0 ± 9.3 | — | 25.7 ± 7.5b |
| Fat percent (%) | 42.4 ± 6.7 | — | 30.1 ± 8.4b |
| SAT (cm3) | 13944.6 ± 3573.7 | — | 7654.4 ± 2833.2b |
| VAT (cm3) | 4820.9 ± 1550.7 | — | 1933.8 ± 1112.9b |
| VAT/SAT-ratio (%) | 38.4 ± 20.1 | — | 33.7 ± 25.8b |
| IHL, arbitrary fraction | 0.23 ± 0.23 | — | 0.021 ± 0.018b |
- Data are given for the obese subjects at baseline, 6, and 12 months after RYGB surgery. Data are means ± SD. The difference between baseline, 6, and 12 months after RYGB is measured by one-way repeated measures ANOVA.
- Baseline vs. 6 months and baseline vs. 12 months,
- a P-value < 0.05,
- b P-value < 0.001.
- 6 months vs. 12 months after RYGB,
- c P < 0.001.
Effect of RYGB-induced weight loss on sCD163
The mean weight loss was 36 ± 11 kg (29%), and the fat mass was reduced from 51.0 ± 9.3 kg to 25.7 ± 7.5 kg after RYGB (Table 1). The weight loss was primary achieved within the first 6 months after RYGB (31 ± 8 kg). The fat depots were significantly decreased (all, P < 0.001)—the SAT by 45 ± 16%, the VAT by 57 ± 28%, and the IHL by 86 ± 13%. The relative change in IHL was significantly higher than the relative change in VAT and SAT (both, P < 0.05). Furthermore, the change in VAT was significantly higher compared with SAT (P < 0.05).
The level of sCD163 was significantly reduced at 12 months compared with baseline (1.7 ± 0.7 mg/l vs. 2.1 ± 0.8 mg/l) (P < 0.001). A significant reduction was also seen from baseline to 6 months (P < 0.001); however, no significant change was found in sCD163 from 6 to 12 months after RYGB (P > 0.05) (Figure 1). The insulin resistance measured by HOMA-IR decreased significantly with a reduction at 69% 12 months after RYGB (P < 0.001). Same result was found when excluding those with type 2 diabetes at baseline (n = 20) (P < 0.001). Diabetes remission was seen in eight out of the 11 subjects with type 2 diabetes (data not shown). The level of IL-6 and hs-CRP were significantly reduced (P < 0.05; P < 0.001) and adiponectin was significantly increased by 93% after 12 months (P < 0.05). The level of TNF-α was not affected by the RYGB-induced weight loss.
Changes in the level of sCD163 at 6 and 12 months post after RYGB. The Level of sCD163 in obese subjects (n = 31) at baseline (mean 2.1 ± 0.8 mg/l), at 6 months after RYGB (mean 1.8 ± 0.6 mg/l), and at 12 months after RYGB (1.7 ± 0.7 mg/l). One-way repeated measures ANOVA, **P-value < 0.001. The lines represent mean values.
Gene expression of CD163 in VAT versus SAT
The gene expression of CD163, ADAM17, and CD68 were examined in VAT and SAT at baseline. The expression of CD163 and the general macrophage marker CD68 was significantly higher in VAT compared with SAT (both, P < 0.05) (Figure 2). The metalloproteinase ADAM17, which is known to regulate the shedding of sCD163, was equally expressed in VAT and SAT (P > 0.05) (Figure 2).
Gene expression levels of CD163 and other macrophage markers in VAT and SAT at baseline. SAT and VAT samples from obese subjects at baseline (n = 25). Gene expression levels of macrophage markers measured by RT-PCR relative to the house-keeping gene GADPH. (A) Expression of CD163 in SAT vs. VAT, (B) expression of the general macrophage marker CD68 in SAT vs. VAT, and (C) expression of ADAM17 in SAT vs. SAT. *P-value <0.05. Mean value with a solid line and outliners represent the 5th/95th percentile.
The morphology of the adipocytes in VAT and SAT were determined in a small subgroup (n = 5). The number of adipocytes pr. total area was significantly higher in VAT compared with SAT (P < 0.05). The mean area and the mean perimeter of the adipocytes were similar in VAT compared with SAT (both, P > 0.05).
Changes in CD163 mRNA in SAT after RYGB
Abdominal SAT biopsies were taken in association with the RYGB surgery and 12 months after the surgery. The expression of CD163 was significantly reduced after RYGB by 46% (P < 0.001) (Figure 3) and also the expression of CD68 was significantly decreased (P < 0.05) (Figure 3). No change was found in the expression of ADAM17 (P > 0.05) (Figure 3). In a subgroup (n = 6), the adipocyte morphology was determined by IHC. The mean perimeter, diameter and area of the adipocytes were significantly decreased after RYGB compared with baseline (both, P < 0.001) (Figure 4). Moreover, the number of adipocytes pr. total area was significantly higher after RYGB compared with baseline (P < 0.05) (Figure 4). Crown-like-structures (CLS) with CD163-positive macrophages were observed in five out of six samples at baseline; however, only one sample with CLS was observed after RYGB (Figure 5).
Gene expression levels in SAT 12 months post after RYGB compared with baseline. Gene expression levels relative to the house-keeping gene GADPH were determined by RT-PCR in SAT samples from 25 obese subjects at baseline and 12 months after RYGB. (A) Expression of CD163, (B) expression of CD68, and (C) expression of ADAM17. Differences measured by paired t-test or Wilcoxon Mann-Whitney rank sum test where appropriate. *P-value <0.05, **P-value <0.001. Mean value with a solid line and outliners represent the 5th/95th percentile.
In a subgroup (n = 6), the adipocyte morphology was determined at baseline and 12 months after RYGB. SAT samples were formalin-fixed and paraffin-embedded, sectioned (3 μm sections), and stained with hematoxylin and eosin using routine methods. Histopathological evaluation of the adipocytes was performed using Visiopharm integrator system ver. 4.4.4.0. (A) Adipocyte number, (B) mean adipocyte area, (C) mean adipocyte diameter, and (D) mean adipocyte perimeter. Paired t-test or Wilcoxon Mann-Whitney rank sum test where appropriate. *P-value<0.05, **P-value <0.001. The lines represent mean values.
Characterization of CD163-positive macrophages in SAT before and 12 months after RYGB. Human AT was stained with antibodies to CD163 (brown). (A) Baseline SAT sample. Adipocytes surrounded by CD163-stained macrophages placed in a crown-like-structure (20×). (B) SAT sample 12 months after RYGB. No crown-like-structure observed (20×). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Differences in sCD163 in relation to the amount of VAT
To determine the possible effects of excessive fat in VAT on sCD163 the obese subjects were divided into two groups, one with high amount of VAT and one with low VAT with a cutoff at 4820 cm3 corresponding to the mean VAT in the obese subjects at baseline (Table 2). Age, BMI, and waist circumference were similar in both groups (P > 0.05). In the high-VAT group, there were more men and 64% of the subjects had type 2 diabetes compared with only 13% in the low VAT group. sCD163 was significantly higher in the high-VAT group (2.4 ± 0.9 mg/l vs. 1.8 ± 0.6 mg/l) (P < 0.05). Furthermore, IHL was significantly higher (P < 0.001), HOMA-IR tended to be higher (P = 0.05), SAT was significantly lower, and the level of adiponectin was significantly lower (both, P < 0.05) in the high-VAT group compared with the low VAT group. There was no significant difference in the level of TNF-α, hs-CRP, or IL-6 between the high- and low-VAT groups (Table 2).
| Low VAT (<4820 cm3) | High VAT (>4820 cm3) | |
|---|---|---|
| Number, n | 15 | 11 |
| Sex (female/male), n | 12/3 | 5/6 |
| DM2, n | 2 | 7 |
| Age (years) | 41.7 ± 7.3 | 46.5 ± 8.2 |
| BMI (kg/m2) | 43.1 ± 4.6 | 41.0 ± 4.5 |
| Waist (cm) | 116.2 ± 9.9 | 121.3 ± 11.9 |
| Fat mass (g) | 54726.2 ± 8993.0 | 46913.6 ± 9467.9a |
| Fat percent (%) | 45.3 ± 5.9 | 39.1 ± 7.0a |
| SAT (cm3) | 15370.9 ± 3118.2 | 11999.5 ± 3332.2a |
| VAT (cm3) | 3775.4 ± 981.6 | 6246.5 ± 887.6b |
| VAT/SAT ratio (%) | 25.4 ± 9.0 | 56.1 ± 17.2b |
| IHL, arbitrary fraction | 0.12 ± 0.08 | 0.39 ± 0.26b |
| HOMA-IR | 4.1 ± 2.7 | 6.2 ± 2.2 |
| sCD163 (mg/l) | 1.8 ± 0.6 | 2.4 ± 0.9a |
| Adiponectin (μg/ml) | 7.0 ± 3.9 | 4.8 ± 1.4a |
| IL-6 (pg/ml) | 2.3 ± 0.8 | 3.2 ± 1.9 |
| TNF-α (pg/ml) | 1.8 ± 0.4 | 1.8 ± 0.5 |
| Hs-CRP (mg/l) | 5.7 ± 3.6 | 5.0 ± 3.6 |
- Data are means ± SD. Data are given for the obese subjects, who were MR-scanned at baseline (n = 26). The subjects are divided into two groups, one with high VAT and one with low VAT. The cutoff level is 4820 cm3 VAT corresponding to the mean VAT in the obese subjects at baseline. Comparison of the high- and low-VAT group by unpaired t-test or Wilcoxon Mann-Whitney rank sum test where appropriate.
- a P-value < 0.05.
- b P-value < 0.001.
Baseline correlations
In a Spearman correlation, no association was found between sCD163 and respective sex or age at baseline (r = 0.08; r = 0.21, P > 0.05). A positive association was found between sCD163 and type 2 diabetes (r = 0.35, P = 0.06). Furthermore, sCD163 was significantly associated with VAT (r = 0.40, P < 0.05); however, when adjusting for sex and type 2 diabetes, the association between sCD163 and VAT was no longer significant (P = 0.12) (data not shown). No association was found between sCD163 and BMI, total fat mass, SAT, or IHL (all, P > 0.05). A strong association was found between sCD163 and IL-6 (r = 0.54, P < 0.05), whereas, no association was found between sCD163 and hs-CRP, adiponectin, and TNF-α (P > 0.05).
Relationship between sCD163, CD163 mRNA, and fat depots after RYGB
The relative change in the level of sCD163 was significantly associated with change in IHL (r = 0.40, P < 0.05) but not with changes in BMI, VAT or SAT (P > 0.05) (Table 3). Furthermore, the decrease in CD163 mRNA in SAT was significantly associated with a reduction in the fat percent (r = 0.41, P < 0.05) (Table 3). No significant association was found between the change in CD163 mRNA in SAT and the change in circulating sCD163 (r = 0.31, P > 0.05) (data not shown).
| ΔsCD163 | ΔCD163 mRNA | |
|---|---|---|
| ΔBMI | 0.01 | 0.11 |
| ΔWaist | 0.11 | 0.01 |
| ΔFat mass | −0.03 | 0.35 |
| ΔFat percent | −0.10 | 0.41a |
| ΔSAT | −0.10 | 0.40 |
| ΔVAT | 0.27 | 0.12 |
| ΔIHL | 0.40a | −0.04 |
- Relative changes in circulating sCD163 and CD163 mRNA expressed in SAT and the association with relative changes in the fat depots in obese subjects (n = 31) at baseline compared with 12 months after RYGB. Statistical tests: Spearman's correlation, r = correlations coefficient.
- a P-value < 0.05.
Discussion
Circulating sCD163 is significantly associated with BMI and waist circumference (1-6); however, the level of sCD163 has until now not been examined in relation to the fat distribution in obese subjects. Thus, in this study, we examined the level of sCD163 in relation to the fat distribution at baseline and 12 months after RYGB surgery-induced weight loss.
RYGB induced an impressive weight loss at 29% (mean 36 ± 11 kg) and the weight loss was accompanied by a significant reduction in the level of sCD163 (2.1 ± 0.8 mg/l vs. 1.7 ± 0.7 mg/l, P < 0.001) which is in line with previous studies showing a reduction in sCD163 in relation to diet-induced weight loss (1, 16). Furthermore, the weight loss was accompanied by a significant reduction in the total fat mass, SAT, VAT, and IHL (all, P < 0.001) with a preferential reduction of IHL and VAT compared with SAT (both, P < 0.05) which has been described previously (33, 34). Massive weight loss in relation with RYGB is associated with improvement of the metabolic profile (24, 26) and likewise, we found a reduction in the circulating inflammatory cytokines IL-6 and hs-CRP, decreased level of HOMA-IR, diabetes remission, and increased level of adiponectin after RYGB.
Recently, we have demonstrated that the gene expression of CD163 and the general macrophage marker CD68 are elevated in SAT in obese subjects compared with lean subjects (17). In this study, we found that CD163 mRNA was significantly upregulated in VAT compared with SAT at baseline. Furthermore, the expression of CD68 mRNA, which is a close correlate to the total number of macrophages in the AT (35), was also significantly upregulated in VAT compared with SAT at baseline. The expression of CD163 mRNA has previously been found equally upregulated in the two depots; however, in a study by Bauer et al., there was a tendency towards a higher level of CD163 mRNA in VAT compared with SAT (36). Upregulated expression of CD68 mRNA in VAT has also been described previously (35), but the results are conflicting since no depot difference has been described in other studies (6, 37). However, other general macrophage markers have been found upregulated in VAT compared with SAT in obese subjects (38) which is in accordance with the general assumption that the VAT is more infiltrated by macrophages than SAT in obese subjects.
RYGB-induced weight loss resulted in a significant reduction of the expression of both CD163 and CD68 mRNA the latter indicates that the total number of infiltrating macrophages was reduced in relation to the weight loss. A study by Cancello et al. also found at reduced number of macrophages in SAT 3 months after RYGB compared with baseline (39); however, increased level of CD163-positive macrophages has been described 3 months after RYGB (40). Macrophages expressing CD163 are considered anti-inflammatory since they are elevated in the down-regulatory phase of an inflammatory response. Anti-inflammatory macrophages have a function in tissue repair and remodeling (9); thus, it could be speculated that there is a temporary increased level of CD163-positive macrophages due to a need of tissue repair the first 6 months after RYGB along with the main weight loss. We found indeed that the morphology of the adipocytes had changed with a significant reduction in the adipocyte area, diameter and perimeter after RYGB.
Accumulation of fat in VAT is known to represent an unhealthy obese phenotype compared with accumulation of fat in SAT (22). We found that sCD163 was significantly associated with VAT but not with SAT or IHL. We divided the obese subjects into low- and high-VAT groups and found that the high-VAT group was characterized by a high sCD163 level, high IHL, high HOMA-IR, low SAT, and low adiponectin level, despite similar BMI in the two groups. Moreover, we found that there were more male subjects and a preponderance of type 2 diabetics in the high-VAT group. Thus, we confirm that sCD163 is a marker of obese subjects with an “unhealthy” metabolic phenotype; however, this association may in part be explained by the strong association between sCD163 and type 2 diabetes (7).
Interestingly, the relative reduction in sCD163 in relation to RYGB was not associated with the change in VAT (r = 0.27, P > 0.05) but significantly associated with the change in IHL (r = 0.40, P < 0.05). We found a dramatic reduction in IHL after RYBG compared with baseline which coherently may have induced a reduction in the amount of sCD163-producing Kupffer cells (macrophages) in the liver. Thus, activated macrophages in the liver and in the VAT may be responsible for the increased level of sCD163 in obese subjects. We found no difference in the expression of ADAM17 in VAT and SAT at baseline which could indicate that sCD163 is released equally from the two depots; however, the gene expression level may not reflect the enzyme activity. A study by Bauer et al. showed comparable levels of sCD163 in portal venous, hepatic venous, and systemic venous blood of liver healthy controls indicating that the VAT, liver, and SAT equally contribute to the circulating sCD163. However, only nine subjects were enrolled in the study, all of them were lean, and six of the subjects had malignant abdominal tumors, which could affect the level of sCD163 (36). Further investigations are needed to sort out which tissue that primarily contributes to the increased production of sCD163 in obese subjects.
Our study had some limitations. First, the SAT samples were not obtained by the same procedure before and after RYGB which could influence the results. Furthermore, it would have been interesting with SAT samples after 3-6 months and VAT samples after 12 months. Secondly, we were not able to differentiate between weight loss-induced changes or changes in relation to RYGB which may induce other hormonal changes that may affect the inflammatory status. Finally, HOMA-IR may not be the best measurement of insulin sensitivity in this study.
In conclusion, we found that the BMI, SAT, VAT, and IHL were significantly reduced after RYGB. A concomitant reduction was seen in the level of sCD163 and in the expression of CD163 mRNA in SAT. The change in sCD163 was significantly associated with the change in IHL after RYGB which may reflect a reduction in sCD163-producing Kupffer cells in the liver. Moreover, we found that sCD163 may be a marker of an “unhealthy” fat distribution in obese subjects at baseline.
Acknowledgments
We thank Lenette Pedersen and Pia Hornbek for their very skillful technical assistance.
