Divergent Relationships Among Soluble Tumor Necrosis Factor-α Receptors 1 and 2, Insulin Resistance, and Endothelial Function

Inflammation has the capacity to impair flow-mediated vasodilatation, which is regarded as a causal factor in the development of atherosclerosis (1). Tumor necrosis factor-α (TNF-α) is a proinflammatory cytokine that is also implicated in the pathogenesis of insulin resistance and endothelium dysfunction linked to this event (2). Contradictory effects of TNF-α on endothelial function have been described in different studies (3). Acute intrabrachial TNF-α infusion impairs endothelium-dependent vasodilatation, but TNF-α also enhances protective mechanism (3).

After binding of TNF-α to TNF-α receptors (TNFR1 and TNFR2), a proteolytic cleavage of the extracellular parts of these receptor elicits the soluble forms, named sTNFR1 and sTNFR2 (4).

We aimed to evaluate brachial artery vascular reactivity (high-resolution external ultrasound) and insulin sensitivity (minimal model analysis [5]) in relation with plasma sTNRF1 and sTNFR2 levels (commercially available solid-phase enzyme-amplified sensitivity immunoassays [EASIA]; Medgenix, Biosource Europe, Fleunes, Belgium) in 100 consecutive, apparently healthy, Caucasian men, 70 with normal glucose tolerance (NGT) and 30 with impaired glucose tolerance (IGT), enrolled in a prospective study of insulin sensitivity in Northern Spain.

In multiple regression analysis, serum sTNFR1 independently contributed to endothelium-dependent vasodilatation (EDVD) in subjects with NGT, after adjusting for age, BMI, smoking status, systolic and diastolic blood pressure, and insulin sensitivity (β = 0.414, P = 0.002). In fact, we observed a positive correlation between sTNFR1 levels and endothelium-dependent vasodilatation (r = 0.291, P = 0.02) (Table 1).

In all subjects as a whole, circulating sTNFR2 was negatively associated with insulin sensitivity (r = −0.20, P = 0.04) and a trend was observed with EDVD (r = −0.190, P = 0.058). In IGT subjects, serum sTNFR2 levels correlated negatively with EDVD (r = −0.366, P = 0.047) (Table 1). The relationship, however, was not significant after adjusting for confounding variables. No association was found between endothelium-independent vasodilatation and circulating sTNFR1 or sTNFR2 levels (Table 1).

This study shows divergent relationships between circulating sTNFRs levels and endothelial function. While sTNFR1 was positively associated with EDVD, opposite relationships regarding sTNFR2 were observed, mainly in subjects with IGT.

Shedding of TNFR1 leads to increased sTNFR1, which antagonizes TNF-α (6). Increased sTNFR1 expression reduced TNF-α bioactivity and protected the myocardium from infarction following ischemia and reperfusion in animal models (7). sTNFR1 might have other protective roles through the stimulation of endothelial cell growth. These antiatherosclerotic mechanisms induced by sTNFR1 are in line with our findings. On the other hand, sTNFR2 levels have been linked to coronary artery disease (8), insulin resistance, and hypertension (5) in concordance with the inverse association between sTNFR2 levels and endothelium vasodilatation in subjects with IGT described here.

Sustained upregulation of human TNFR2 in transgenic mice leads to a chronic accumulation of cell surface and plasma receptor (9), providing them the capacity to be hyperresponders to circulating TNF-α. It is tempting to speculate that similar findings in subjects with IGT may contribute to both insulin resistance and endothelial dysfunction induced by TNF-α.

In summary, we found divergent associations between both sTNFR and endothelium vasodilatation. The knowledge of how these interactions occur may have therapeutic implications.