Hypoxia Disturbs the Metformin‐Evoked Curtailment of Chemokine CX3CL1 (Fractalkine) Signaling in Human Umbilical Vein Endothelial Cells (HUVECs)

Abstract

Metformin (dimethylbiguanid) is an antihyperglycemic agent widely used for treating non‐insulin‐dependent diabetes mellitus (NIDDM). Moreover, it was reported that metformin may prevent nuclear factor kappa‐beta (NF‐κβ) activation in endothelial cells exposed to inflammatory cytokines. These anti‐inflammatory properties resulting from activation of AMP‐activated protein kinase (AMPK). Inflammation and hypoxia activate chemokine CX3CL1 receptor (CX3CR1)‐related signaling pathway. Thus, we hypothesized that lowering of cellular AMP levels during hypoxia may reduce ability of metformin to counteract the CX3CXL1‐CX3CR1‐axis‐dependent mechanism of endothelial dysfunction. In this study we examined comparatively the effect of metformin on lipopolysaccharide (LPS)‐induced CX3CL1 production as well as CX3CR1 expression in HUVECs cultured under normoxic vs. hypoxic conditions. Immediately after childbirth, HUVECs were isolated from a 12‐cm umbilical cord segment via collagenase digestion. The material obtained from primigravidas after term, normal‐course pregnancies was equally divided into two groups: group I (N = 24) – kept under hypoxia (5% O2), and the control normoxic (20% O2) group II (N = 24). After achieving the confluence, HUVECs were exposed to metformin (10 mmol/L; the subgroups IA and IIA) or cultured without metformin (the subgroups IB and IIB, respectively). Following pretreatment with LPS (1 μg/mL), the mean levels of CX3CL1 in the culture media supernatants were determined at predefined time points by ELISA. At the end of the culture period, HUVECs were formalin fixed, paraffin embedded, immunostained for CX3CR1 and cut into 5 μm sections. The expression of CX3CR1 was estimated by quantitative immunohistochemistry. The mean levels of CX3CL1 in group I were 5.9‐fold and 2.7‐fold higher than observed in normoxic group II (IA vs. IIA and IB vs. IIB, respectively). Initially, the mean levels of CX3CL1 were significantly (p < 0.05) lower in IA compared to IB, but these differences gradually disappeared, being non‐significant over the time. The mean expression of CX3CR1 in hypoxia did not differ statistically between IA and IB, but was significantly (p < 0.05) upregulated compared to the both normoxic subgroups IIA and IIB. In normoxia, exposition to metformin produced significant and stable decrease in CX3CL1 production in HUVECs, compared to metformin‐free subgroup IIB. A significant increase in CX3CR1 expression was observed in IIB compared to IIA. In conclusion, upregulation of CX3CR1 in response to hypoxia and/or metformin may serve as a compensatory mechanism to preserve or augment the level of CX3CL1/CX3CR1 signaling. Ineffectiveness of metformin in lowering of CX3CL1 production under hypoxia may suggest insufficient activation of AMPK in HUVECs.Support or Funding InformationWUM grant: 2M2‐W1‐16