Oxidative stress in preeclampsia and the role of free fetal hemoglobin

Abstract

Preeclampsia is a leading cause of pregnancy complications and affects 3-7% of pregnant women. This review summarizes the current knowledge of a new potential etiology of the disease, with a special focus on hemoglobin-induced oxidative stress. Furthermore, we also suggest hemoglobin as a potential target for therapy. Gene and protein profiling studies have shown increased expression and accumulation of free fetal hemoglobin in the preeclamptic placenta. Predominantly due to oxidative damage to the placental barrier, fetal hemoglobin leaks over to the maternal circulation. Free hemoglobin and its metabolites are toxic in several ways; (a) ferrous hemoglobin (Fe(2+)) binds strongly to the vasodilator nitric oxide (NO) and reduces the availability of free NO, which results in vasoconstriction, (b) hemoglobin (Fe(2+)) with bound oxygen spontaneously generates free oxygen radicals, and (c) the heme groups create an inflammatory response by inducing activation of neutrophils and cytokine production. The endogenous protein α1-microglobulin, with radical and heme binding properties, has shown both ex vivo and in vivo to have the ability to counteract free hemoglobin-induced placental and kidney damage. Oxidative stress in general, and more specifically fetal hemoglobin-induced oxidative stress, could play a key role in the pathology of preeclampsia seen both in the placenta and ultimately in the maternal endothelium.

Keywords: alpha-1-microglobulin; fetal hemoglobin; hemolysis; oxidative stress; placenta.

Figure 1

Increased placental expression of HbF in preeclampsia. Representative images from in situ hybridizations of human placenta, displaying the villous section of the placenta. HbF mRNA expression (α-chain) in a control placenta shown as light field image (A) and as dark field image (B), and in a preeclamptic placenta shown as light field image (D) and as dark field image (E). HbF expression was especially seen in and around blood vessels, with several scattered cells detected in the villous section of the preeclamptic placenta. HbF protein expression (γ-chain) is shown with a red fluorescent marker in control placenta (C) and in preeclamptic placenta (F). In the preeclamptic placenta there is a strong expression of HbF in the vascular lumen (lu), but HbF is also expressed in the vascular endothelium (arrow) as well as in the extravascular section of the villous stroma. The placenta from control shows no expression of HbF in the vascular lumen (lu), but expression is detected in the vascular endothelium (arrow). Scale bars for (A,B,D,E) = 100 μm and for (C,F) = 25 μm. Modified from Centlow et al. (2008).

Figure 2

Free Hb causes placental damage ex vivo which can be ameliorated by A1M. Transmission electron microscopy analysis of ex vivo perfused human placenta. (A) Non-perfused human placenta. (B) Free Hb was added to the fetal circulation and caused severe damage to the extracellular matrix with an almost complete elimination of the collagen fibers. (C) A1M was added to the maternal circulation at the same time as Hb was added to the fetal circulation, and prevented the damaging effects of Hb on the extracellular matrix, displaying normal collagen fibers. Scale bars: 200 μm. Adapted from May et al. (2011).

Figure 3

Free HbF causes structural and cellular damage to placenta tissue in vivo. Transmission electron microscopy analysis of placenta tissue from pregnant rabbits infused with free HbF. (A,B) Normal placental tissue from control animals showing extracellular matrix with dense bundles of collagen fibers. (C,D) HbF causes severe damage to the extracellular matrix with loss of collagen fibers (indicated by arrows), extracellular apoptotic bodies, cell debris and a lot of empty extracellular space (indicated by stars). Scale bars: (A,C) = 500 nm; (B,D) = 1 μm.

Figure 4

Free Hb causes kidney damage in pregnant ewes which can be ameliorated by A1M. The renal tissue from starved pregnant ewes was studied by light microscopy on cortical specimens stained with hematoxylin and eosin (A–C) and by transmission electron microscopy (D–F). (A) Normal cortical tissue morphology in control ewes. The renal tubules show minor signs of postmortem changes but the height of the epithelium is normal, and only discrete cytoplasmic vacuolization's are seen. The glomeruli demonstrate open capillary loops and no signs of segmentation. (B) Clear signs of glomerular endothelial swelling, seen as closure of capillary loops, and non-isometric vacuolization of the tubular epithelium in the starved ewes. (C) In the A1M-treated starved ewes, only small tubular changes compatible with acute tubular necrosis are present and to a milder degree than seen in starved ewes. Also, no signs of glomerular endothelial swelling can be seen. (D) Normal ultrastructure of the glomerular area in control ewes. (E) Disturbed morphology of the podocytes in the starved animals. The arrows point at abnormal regions on the basement membrane with fenestrations underneath. The asterisks mark the basal membrane and P indicates the podocyte foot extensions. (F) Normalized ultrastructure was observed after A1M treatment. Magnification: (A,C) 100×; (B) 200×. Scale bars: (D–F) 1 μm. Adapted from Wester-Rosenlof et al. (2014).