Circulating levels of Eotaxin-1, -2, and -3 chemokines in morbid obesity and 1 year follow up after bariatric surgery

H. Arefanian1, F. Al-Asfar2, E. Al-Ozairi3, S. Al-Beloushi1, E. Al-Shawaf1, S. Kumari1, S. Sabau4, A. Al-Hubail5, J. Tuomilehto6.

1Dasman Diabetes Institute, Islet Biology Unit- Research Division, Kuwait city, Kuwait.
2Kuwait University, Department of Surgery- Faculty of Medicine, Kuwait City, Kuwait.
3Dasman Diabetes Institute, Medical Division, Kuwait City, Kuwait.
4Dasman Diabetes Institute, Nursing Department- Medical Division, Kuwait City, Kuwait.
5Dasman Diabetes Institute, Clinical Laboratory- Medical Division, Kuwait City, Kuwait.
6Dasman Diabetes Institute, Research Division, Kuwait City, Kuwait.

Background
Obesity is characterized as an accumulation of adipose tissue mass represented by chronic, low-grade inflammation. Chemokines as one of the important mediators in immune cell infiltration have been hypothesized to be involved in macrophage infiltration into adipose tissue in obesity and might therefore play an important role in the development of obesity-related disorders like type 2 diabetes. Eotaxins (eosinophil chemotactic proteins) namely Eotaxin-1 (CCL11), Eotaxin-2 (CCL24), and Eotaxin-3 (CCL26) belong to CC chemokine family, in addition to their important role in asthma, recent reports showed their association to obesity in mice and humans.  However, the role of these chemokines is poorly understood. 

Aim
Our aim was to study the association between circulatory level of Eotaxins (Eotaxin-1, Eotaxin-2, and Eoxtaxin-3), obesity and type 2 diabetes before and after Laparoscopic Sleeve Gastrectomy (LSG), and to determine the occurrence of any possible kinetic change, which is monitored through short- and long-term follow up visits during one year after LSG, with special focus to cases that lost substantial weight and resolved type 2 diabetes. 

Method
38 morbidly obese participants (16 diabetic and 22 non-diabetic) with a BMI ≥ 40 kg/m2 underwent LSG, while 32 normal weight subjects (18 diabetic and 14 non-diabetic) with a BMI < 25 kg/m2 were recruited as controls.  We compared the basal levels of circulatory Eotaxins (pg/mL) in morbid obese participants (i.e. with and without type 2 diabetes) before and after LSG-induced weight loss, to the basal levels in normal weight people (i.e. with or without type 2 diabetes). The correlation between levels of these chemokines and other clinical and biochemical parameters was also studied. 

Results
At baseline, levels of Eotaxin-1 was significantly (p< 0.0001) higher in normal weight diabetic subjects (120.1±8.78) compared to other groups, including morbid obese-diabetics (70.63±6.17), morbid obese non-diabetics (65.47±6.23) and normal weight non-diabetics (66.52±5.17). No significant difference was observed between morbid obese and normal weight non-diabetic groups. Eotaxin-2 levels showed no significant difference between all study groups. However, Eotaxin-3 was significantly higher (p<0.0003) in morbid obese participants (diabetics; 48.42±4.80, and non-diabetics; 48.93±4.06) compared to normal weight participants (diabetics; 27.47±3.94 and non-diabetics; 20.42±1.93). After one year of LSG, the level of Eotaxin3 was significantly (p<0.02) reduced in morbid obese non-diabetic (29.42±5.71), reaching a level that is comparable to normal weight participants. A similar pattern of reduction, but no significance was observed in the level of Eoatxin3 (37.94±10.87) in morbid obese diabetic participants after one year of LSG. Additionally, our results show a significant association between the levels of Eotaxin-1 with body weight (r=-0.64, p<0.0001), BMI (r=-0.56, p<0.0006), Abdominal circumference (r=-0.50, p<0.002), hsCRP (r=-0.50, p<0.002) and Leptin (r=-0.43, p<0.01) in diabetic cases. Eotaxin-3 also showed a significant association to body weight (r=0.56, p<0.0005), BMI (r=0.6156, p<0.0001), Abdominal circumference (r=0.60, p<0.0001), hsCRP (r=0.575, p<0.004), and Leptin (r=0.57, p<0.0004) in diabetic groups. Our results showed that Eoatxin-3 was significantly associated with hsCRP (r=0.345, p<0.039), Leptin (r=0.583, p<0.0002), and adiponectin (r=-0.369, p<0.02) in the non-diabetic groups. While Eoatxin-1 revealed an association with levels of fasting blood glucose (r=0.591, p<0.0004), HbA1c (r=0.631, p<0.0001), and TNF-alpha (r=0.563, p<0.0008) in normal weight participants. No association was observed between level of Eotaxins and waist/hip ratio, fasting Insulin, and fasting c-peptide.

Discussion
We observed an association between the circulatory levels of Eotaxin chemokines (inverse association of Eotaxin-1 and direct association of Eotaxin-3), with the level of obesity and markers of low grade chronic inflammation before and after LSG. Further studies are required to unfold the molecular pathways in adipose tissue and their stromal and vascular components, as a predominant source of these chemokines. 

CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT03038373.