*3.2. Homocysteine and Pregnancy*

The homocysteine plasma levels fall in normal pregnancy [89,90]. An increase in plasma volume and associated haemodilution, glomerular hyperfiltration and postulated raised foetal need for methionine are mechanisms considered to contribute to this effect [91]. The importance of homocysteine to early foetal metabolism is demonstrated in a number of studies [92].

The reference values for HHcys in pregnancy that are proposed in one study were established as: higher than 7.7 mmol/L in the second trimester, and 10.5 mmol/L in the third trimester [93], although different authors usually defined their own cut-off values.

In the pathology of pregnancy, the disturbance of maternal homocysteine metabolism has been linked with recurrent pregnancy loss, deep venous thrombosis, foetal neural tube defects, and various conditions characterized by placental vasculopathy, such as preeclampsia, foetal growth restriction, and abruption [94–98].

#### **4. NO, ADMA, and Homocysteine in Preeclampsia**

### *4.1. NO Pathway Dysfunction in PE*

In preeclamptic patients, like in normotensive pregnancies, the measurements of total NO concentration have shown variable results, ranging from decreased [81,99,100], unchanged [101], and increased [102,103] levels of circulating NO metabolites. The dietary intake of these substances could also influence the disparity, although a study that was conducted on women with PE, subjected to a reduced nitrate/nitrite diet, did not show decreased endogenous NO production [78]. In fact, NO measurements may be difficult to interpret, since they reflect the total activity of all three isoforms of NOS, not just endothelial. While the whole body NO may not change in PE, in view of the evidence for reduced endothelial NO signalling and decrease in vascular relaxation in PE, tissue-specific differences in NOS expression and NO bioavailability could be expected. For instance, in late pregnan<sup>t</sup> rats, renal eNOS decreases by 39%, while iNOS and nNOS increase by 31% and 25%, respectively [104].

PE is associated with abnormalities eNOS-NO pathway that probably exists at different stages of signal transduction process. There is not one particular defect, but multiple changes in key regulatory aspects in NO signaling. In studies where serum from PE women was placed on isolated vessels, nitric oxide-mediated vasorelaxation appeared to be absent [105].

Some studies have indicated that measuring plasma nitrite levels may reflect endogenous NO formation because NO is rapidly oxidized to nitrite. This is because 70% of plasma nitrites derive from NO synthase activity in the endothelium and its inhibition was associated with corresponding decreases in plasma nitrite concentrations [106,107]. By only using nitrite levels (which may be a better measure than total nitrite + nitrate), a reduction from 40 to 60% of total whole blood or plasma nitrite concentration was reported in PE women [108–110].

There is clinical evidence for the link between impaired NO formation and antiangiogenic factors overexpression in preeclampsia. Significant negative correlation between two antiangiogenic factors: sEng and sFlt-1, and nitrite concentrations was described [110], which suggested a possible inhibitory effect caused by these substances on the production of NO in patients with preeclampsia. Experimental studies have shown that NO increases proangiogenic VEGF and PIGF and it decreases sFlt-1 in hypoxic human trophoblast cells [111].

Using the nitrate reductase assay to measure NO, also a correlation between reductions in plasma NO with disease severity was identified, such that the levels were about 30% lower in severe PE vs. healthy pregnan<sup>t</sup> controls [112].

Attempts to assess eNOS activity in PE led to the conclusion that it is still unknown whether eNOS deficiency plays a causal role there. In the murine model, chronic NOS inhibition reversed systemic vasodilation and glomerular hyperfiltration in pregnancy, which suggested its role for endothelial damage and decreased NO in the pathogenesis of preeclampsia [113]. However, different study with the use of eNOS knockout mice showed reduced uterine artery diameter, spiral artery length, and, as a consequence, diminished uteroplacental blood flow, resulting in elevated markers of placental hypoxia in the junctional zone. Even so, interestingly, sFlt-1concentration was not elevated in the eNOS knockout mice [43].

Data from PE women is quite limited and without consensus on eNOS expression, as higher, lower, and unchanged levels of mRNA or enzyme have been reported. Several human studies failed to detect any significant differences in the circulating levels of eNOS [114,115]. One of the earliest studies on eNOS activity found an increase in eNOS expression in syncytiotrophoblast, foetal terminal villous capillary, and stem villous vessel endothelium, whereas the lack of eNOS expression in vascular terminal villi and weak expression in endothelial cells of villous vessels in placenta from normal pregnancy was noted [116]. These results are supported with a recent finding that caveolar eNOS expression is increased in PE placentas [117]. However, this stays in contrast to the observed similar placental levels of eNOS activity in PE patients [118], and to a more recent study in which placental syncytiotrophoblast eNOS expression was even decreased [119]. Apart from these confusing findings, altered placental eNOS levels may not directly relate to peripheral vascular endothelial function.

The reduced availability of eNOS substrate (L-Arginine) or the competitive inhibition by ADMA constitute other factors that may contribute to dysfunctional endothelial NO signalling in PE. One of proposed animal PE models involves administering the aforementioned L-NAME, a competitive inhibitor of arginine, which leads to maternal hypertension and proteinuria, and reduced foetal weight in a dose-dependent way [120]. L-NAME-induced hypertension and high circulating levels of sFlt-1 could be attenuated by the administration of exogenous sodium nitrite, and in this way restoring NO bioavailability [121].

In pregnancies that are complicated by PE, the ADMA levels are significantly higher than in both normotensive gestational age-matched and the nonpregnan<sup>t</sup> control group [24,115,122–124]. Even higher concentration of ADMA in patients with early-onset PE may sugges<sup>t</sup> a relationship between disease severity and determining the time of PE clinical manifestation [125,126]. Moreover, ADMA may have a predictive value in PE, since its modestly ( +26%) elevated concentrations were observed as early as in the first trimester [127] and significantly elevated during second trimester in pregnancies that developed PE in more advanced gestational age [128]. Another hypotesis is that increased ADMA concentration may contribute to development of PE in early pregnancy, leading to impaired placentation and its consequences [126]. The association of abnormal uterine artery Doppler waveforms with elevated ADMA levels [129,130] supports the role of endogenous NOS inhibitors adversely a ffecting maternal vasodilation and blood pressure.

It has been postulated that hyperhomocysteinemia (HHcy) may contribute to the development of PE, as it leads to endothelial dysfunction and accumulation af ADMA [29,30].
