Participant characteristics
Fifteen participants were invited, and thirteen participants were included in the study. We evaluated 13 obese, non-diabetic patients (mean age 37.7 ± 8.2 years; 100% women; Height 1.64 ± 0.05 m; BMI 42.2 ± 4.2 kg/m2), that underwent Roux-en-Y gastric bypass (RYGB). The clinical and biochemical characteristics of the evaluated group are summarized in Table 1.
Table 1
Clinical characteristics of the study participants. Data are mean ± SD. Was applied one-way ANOVA with Tukey post-hoc test. * Baseline vs 3 months. # Baseline vs 6 months. $ 6 months vs 3 months
|
Before surgery
|
After 3 months
|
After 6 months
|
Characteristic
|
(n = 13)
|
(n = 13)
|
(n = 13)
|
Age (year)
|
37.7 ± 8.2
|
-
|
-
|
Female (%)
|
100
|
-
|
-
|
Height (m)
|
1.64 ± 0.05
|
-
|
-
|
BMI (kg/m2)
|
42.2 ± 4.2
|
35.9 ± 4.7*
|
33.0 ± 2.9###
|
Body weight (Kg)
|
113.4 ± 14.7
|
94.2 ± 13.9*
|
90.1 ± 12.4##
|
Waist circunference (cm)
|
123.2 ± 12.0
|
107.8 ± 7.7
|
107.2 ± 13.7#
|
Fat Mass (kg)
|
52.8 ± 9.3
|
37.8 ± 9.1*
|
34.4 ± 7.1###
|
Lean Mass (Kg)
|
61.0 ± 5.9
|
56.4 ± 5.2
|
55.7 ± 6.1
|
Fasting glucose (mg/dL)
|
97.5 ± 25.1
|
87.0 ± 19.2
|
82.5 ± 11.6
|
A1C (%)
|
5.4 ± 0.7
|
5.2 ± 0.5
|
5.1 ± 0.4
|
Cholesterol (mg/dL)
|
165 ± 20.5
|
155.8 ± 27.1
|
160.9 ± 24.1
|
HDL cholesterol (mg/dL)
|
39.0 ± 9.5
|
39.0 ± 10.6
|
42.0 ± 10.2
|
LDL cholesterol (mg/dL)
|
105.7 ± 20.3
|
105.3 ± 20.1
|
100.4 ± 16.5
|
Triglycerides (mg/dL)
|
121.5 ± 67.0
|
86.5 ± 36.2
|
91.9 ± 40.1
|
Insulin (mU/L)
|
17.1 ± 7.7
|
9.0 ± 5.0
|
7.5 ± 4.4#
|
HOMA-IR
|
4.0 ± 1.7
|
1.9 ± 1.0*
|
1.4 ± 0.8##
|
Systolic BP (mmHg)
|
136 ± 9.6
|
113 ± 5.4***
|
114 ± 7.5###
|
Diastolic BP (mmHg)
|
102 ± 6.8
|
76 ± 5.4***
|
78.5 ± 6.4###
|
% of weight loss
|
-
|
-17%
|
-20.50%
|
% blood glucose reduction
|
-
|
-12.90%
|
-17.50%
|
% Cholesterol reduction
|
-
|
-5.50%
|
-2.50%
|
% Triglycerides reduction
|
-
|
-30.90%
|
-26.60%
|
% LDL reduction
|
-
|
-0.40%
|
-5.00%
|
% HDL change
|
-
|
-
|
7.40%
|
Abbreviations: BMI: Body mass index; A1C: Hemoglobin A1C; HDL: High Density Lipoprotein; LDL: Low Density Lipoprotein; HOMA-IR: Homeostatic model assessment- Insulin resistance. |
Effects of RYGB on body composition, biochemicals parameters and blood pressure
After RYGB, we observed a significant reduction in body weight and BMI after 3 months (3M) (-16.9%, P<0.05; -14.9%, P<0.05 respectively) and 6 months (6M) (-20.5%, P<0.01; -21.8%, P<0.001respectively) Table 1. Waist circumference had not changed at 3M, but after 6M decreased by -12.9% (P<0.05). In addition, fat mass was decreased by -28.4% (P<0.05) at 3M and − 34.8% (P<0.001) at 6M. However, there was no difference in lean mass Fig. 1.
The fasting glucose levels and A1C, despite having a downward trend, did not show significant differences after RYGB, Table 1. All participants in this study were not using any medication that could interfere in the glycemic control. The total cholesterol, HDL cholesterol, LDL cholesterol and triglycerides did not show significant differences after RYGB, Table 1. The fasting insulin levels did not show significant change at 3M; however, in 6 months, there was a reduction of -56.1% (P<0.05), Table 1.
The reduction in insulin levels resulted in a significant increase in insulin sensitivity at 3M (52.5%, P<0.05) and 6M (65%, P<0.01), as measured by the HOMA-IR index, Table 1. Blood pressure was reduced at 3M (-16.9% systolic BP; -25.4% diastolic BP, P<0.001 for all) and 6M (-16.1% systolic BP; -23% diastolic BP, P<0.001 for all), Table 1.
Time course effects on expression of genes in subcutaneous adipose tissue
To identify possible changes in the gene expression of adipose tissue in non-diabetic obese individuals before and after RYGB, we performed a screening of several genes on the fat biopsies. We investigated interleukin genes, genes involved in energy homeostasis, adipogenesis, mitochondrial biogenesis, ERS and amino acid metabolism, summarized in Supplementary Table 1. Below we report the difference in the gene expression fold change between each time point evaluated.
Three months (3M) after RYGB was long enough to cause a decrease in the expression levels of IL6 (Interleukin 6) (-0.62 ± 0.2, P = 0.020) and MCP1 (monocyte chemoattractant protein 1) (-0.72 ± 0.2, P = 0.003) while a decrease in the expression of TNF-ɑ (Tumor necrosis factor ɑ) was observed only after 6M (-0.82 ± 0.2, P = 0.010) in adipose tissue, Fig. 2a.
Additionally, we observed an increase in ADIPOQ (adiponectin) expression after 3M (2.37 ± 0.5, P<0.001) and a more pronounced increase after 6M (4.15 ± 0.4, P<0.001). Similarly, levels of PGC1ɑ (PPARγ coactivator-1) was increased after 3M (1.43 ± 0.3, P<0.001) and 6M (0.83 ± 0.2, P = 0.004) of RYGB, Fig. 2b. In contrast, mRNA levels of PPARϒ (Peroxisome proliferator-activated receptor gamma) were reduced after 3M (-0.56 ± 0.1, P = 0.002) and 6M (-0.35 ± 0.1, P = 0.022), Fig. 2b.
Next, we examined the subcutaneous adipose tissue expression of genes involved in the ERS. EIF2AK3 (Eukaryotic Translation Initiation Factor 2 Alpha Kinase 3) was decreased at 3M (-0.85 ± 0.2; P = 0.006) and 6M (-0.84 ± 0.2; P = 0.001). Also, the mRNA levels of ATF4 (Activating Transcription Factor 4) was increased after 3M (1.66 ± 0.3; P<0.001) and remained increased after 6M (1.54 ± 0.5; P = 0.014). Notably, we found that CARL (Calreticulin) expression was significant decreased at 3M (-0.95 ± 0.2; P = 0.006) and 6M (-0.84 ± 0.2; P = 0.003) after RYGB, Fig. 2c. The mRNA expression of other ERS-related genes did not show significant differences.
We also tested several genes related to the cellular response to oxidative stress and sensor of cellular energy homeostasis and amino acid concentration. In general, at 3M after RYGB, increased expression of SIRT1 (2.40 ± 0.3; P<0.001) and SIRT3 (1.05± 0.4; P = 0.021). At 6M, the mRNA expression of SIRT1 and SIRT3 remained increased compared to the baseline (1.84 ± 0.3, P<0.001; 0.75 ± 0.3, P = 0.048, respectively). Moreover, compared with baseline, the expression of AMPK, was higher at 3M (3.36 ± 0.2; P<0.001) and 6M (1.84 ± 0.3; P<0.001) after RYGB. Finally, we verify that GCN2 mRNA levels was increased after 3M (2.28 ± 0.2; P<0.001) and 6M (1.58 ± 0.1; P<0.001), Fig. 2d.
We sought to determine whether RYGB could induce changes in the expression of genes related to oxidative stress directly in adipose tissue. Notably, we found that NRF2 expression was increased at 3M (3.74 ± 0.4; P<0.001) and 6M (1.85 ± 0.2; P<0.001) compared to baseline. In addition, SOD2 expression was increased in adipose tissue after 3M (0.56 ± 0.2; P = 0.027). However, we found no differences in expression at SOD1 and SOD3 in adipose tissue after RYGB, Fig. 2e. These data demonstrate the genetic dynamism of adipose tissue post-RYGB.
BMI is positively correlated with PGC1ɑ, SIRT1, AMPK and Adiponectin expression in human adipose tissues
To address possible associations between the genes evaluated in this study and the biochemical and anthropometric changes that occurred after RYGB, we performed correlation analysis between gene expression in adipose tissue samples and the biochemical and anthropometric variables.
Interestingly, we noted that after 3M the expression of PGC1ɑ, SIRT1, and AMPK was positively correlated with the BMI changes in the same period (Pearson’s correlation r = 0.961, P = 0.009; r = 0.958, P = 0.010; r = 0.947, P = 0.014 respectively), Table 2. Similar patterns of significant correlations were found between the gene expression of ADIPOQ and SIRT1 after 6M of RYGB and changes in BMI at 6M (r = 0.950, P = 0.001; r = 0.844, P = 0.016 respectively), Table 2. Collectively, these data suggest that the relationship between the expression of genes related to the control of energy homeostasis and mitochondrial biogenesis may be dependent on post-RYGB time dynamics in adipose tissue.
Table 2
Correlations between BMI changes and delta of gene expression in adipose tissue at 3 and 6 months after RYGB.
|
Correlation statistics
|
|
|
Correlation statistics
|
3 Months
|
Pearson’s r
|
P value
|
|
6 Months
|
Pearson’s r
|
P value
|
IL6
|
0.362
|
0.548
|
|
IL6
|
-0.305
|
0.5046
|
TNFα
|
0.344
|
0.570
|
|
TNFα
|
-0.555
|
0.1954
|
MCP1
|
0.497
|
0.394
|
|
MCP1
|
0.183
|
0.6942
|
ADIPOQ
|
-0.066
|
0.915
|
|
ADIPOQ
|
0.950
|
0.0010
|
PGC1α
|
0.961
|
0.009
|
|
PGC1α
|
0.386
|
0.3915
|
PPARγ
|
0.384
|
0.523
|
|
PPARγ
|
0.591
|
0.1623
|
EIF2AK3
|
0.511
|
0.379
|
|
EIF2AK3
|
0.165
|
0.7233
|
ATF4
|
0.648
|
0.237
|
|
ATF4
|
-0.373
|
0.4097
|
ATF6
|
0.410
|
0.492
|
|
ATF6
|
0.677
|
0.0945
|
CHOP
|
0.254
|
0.680
|
|
CHOP
|
-0.286
|
0.5332
|
GRP78
|
0.261
|
0.671
|
|
GRP78
|
0.46
|
0.2987
|
GADD34
|
-0.554
|
0.332
|
|
GADD34
|
-0.05
|
0.9141
|
XBP1
|
-0.609
|
0.275
|
|
XBP1
|
0.232
|
0.6164
|
CALR
|
0.004
|
0.994
|
|
CALR
|
0.416
|
0.3523
|
CANX
|
0.490
|
0.401
|
|
CANX
|
0.743
|
0.0554
|
CCT4
|
0.637
|
0.247
|
|
CCT4
|
0.272
|
0.5539
|
SIRT1
|
0.958
|
0.010
|
|
SIRT1
|
0.844
|
0.0169
|
SIRT3
|
0.407
|
0.496
|
|
SIRT3
|
0.489
|
0.2646
|
AMPK
|
0.947
|
0.015
|
|
AMPK
|
0.692
|
0.0846
|
GCN2
|
-0.511
|
0.379
|
|
GCN2
|
0.278
|
0.5459
|
NRF2
|
0.127
|
0.839
|
|
NRF2
|
-0.014
|
0.9757
|
SOD1
|
-0.342
|
0.573
|
|
SOD1
|
-0.527
|
0.2232
|
SOD2
|
-0.710
|
0.178
|
|
SOD2
|
-0.112
|
0.8102
|
SOD3
|
-0.466
|
0.428
|
|
SOD3
|
0.105
|
0.8227
|
Abbreviations: IL6: Interleukin 6; TNF-ɑ: Tumor necrosis factor-α; MCP-1: Monocyte chemoattractant protein-1; PGC1ɑ: Peroxisome proliferator-activated receptor gamma coactivator 1-alpha; PPARϒ: Peroxisome proliferator activated receptor gamma; EIF2AK3: Eukaryotic translation initiation factor 2 alpha kinase 3; ATF4: Activating Transcription Factor 4; ATF6: Activating Transcription Factor 6; CHOP: C/EBP homologous protein; GRP78: Heat shock protein family A (Hsp70) member 5; GADD34: growth arrest and DNA damage-inducible protein; XBP1: X-box binding protein 1; CARL: Calreticulin; CANX: Calnexin; CCT4: Chaperonin Containing TCP1 Subunit 4; SIRT1: Sirtuin 1; SIRT3: Sirtuin 3; AMPK: AMP-activated protein kinase; GCN2: General control nonderepressible 2; NRF2: Nuclear factor erythroid 2-related factor 2; SOD1: Superoxide dismutase 1; SOD2: Superoxide dismutase 2, SOD3: Superoxide dismutase 3.