DISCUSSION
Early prediction of PE remains a challenge in current clinical practice.
Known traditional risk factors inadequately identify women who will
develop PE early in gestation . To aim the discovery of novel
serological markers with better predictive power for PE at early
gestations, we applied a multi-omics approach, integrating
differentially expressed gene (DEGs) from placental mRNA expression
multiplex analysis and lipidomic database mining of existing literature,
to identify novel PE biomarkers, which are Lep and Cer. By
characterizing the maternal serological profiles of Lep and Cer using
commercial ELISA and reported liquid chromatography-tandem mass
spectrometry (LC/MS/MS) assays, we validated the up-regulated Lep and
down-regulated Cer(d18:1/25:0) in a case-control testing cohort with
maternal sera collected at confirmative diagnosis of PE. With a cohort
of longitudinally collected maternal sera from both PE and non-PE women,
we further assessed the PE-predictive power of the Lep/Cer (18:1/25:0)
ratio early in gestation. Our results validated this ratio as a better
serological predictor of impending PE than the established sFlt-1/PlGF
ratio. Our findings demonstrated that the use of the Lep/Cer
(d18:1/25:0) ratio can identify women at high risk of developing PE at a
substantially earlier time window during pregnancy (at 5 to 25 wks) than
the sFlt-1/PlGF ratio (after 25 wks). Moreover, our results showed that,
compared with the sFlt-1/PlGF ratio, the Lep/Cer (d18:1/25:0) ratio has
better sensitivity (40% vs 85%), specificity (45% vs 90%), PPV (42%
vs 89%), and NPV (43% vs 86%). Therefore, early in gestation, the
Lep/Cer (d18:1/25:0) ratio outperforms the established sFlt-1/PlGF ratio
and is a predictor of impending PE. The fact that the Lep/Cer
(d18:1/25:0) ratio increases early in gestation in pregnant women who
later develop PE offers an opportunity for predicting PE prior to the
onset of clinical signs and symptoms. Integration of the ratio into a
high-risk screening tool might allow patient identification at a
pre-symptomatic stage. In addition, the concept of integrating a
transcriptomic approach in placenta tissue with a lipidomic approach in
serum is novel, as it combines the merits of studies in tissue whose
focuses are more towards the pathogenesis and pathophysiology with those
study in serum whose focuses are more towards the clinical translation.
Taking the candidates obtained from the discovery phase to the
validation phase makes the findings of this study translatable into
clinical practice.
Previous studies have suggested that placental trophoblast cells are a
leading source of circulating Lep in pregnancy , where Lep increases
progressively in the first and second trimesters, peaks in the third,
and returns to pre-pregnancy levels prior to parturition . In early
gestation, Lep may play a critical role in modulating essential
biological activities such as proliferation, protein synthesis,
invasion, and apoptosis of trophoblast cells . Failure of trophoblastic
invasion might result in incomplete remodeling of the maternal spiral
arteries and inadequate placental perfusion to the embryo , leading to
various disorders of reproduction and gestation such as intrauterine
growth restriction , PE , gestational diabetes mellitus , and recurrent
miscarriage . Other recent studies documented significant elevations of
Lep expression in preeclamptic placentas . In the current study, we
found a significant upregulation of Lep in maternal sera of PE women,
which is consistent with previous reports .
The gestational dysregulation of Cer metabolism is believed to induce
the aberrant de novo synthesis and lysosomal breakdown of Cer, which
leads to trophoblast cell autophagy, dysfunctional development of
placenta, and eventually pregnancy complications like PE . Cer
(d18:1/25:0) is an unusual odd-chain species of the Cer family that is
generated by de novo synthesis based on 25:0 fatty acid. Such odd-chain
fatty acids are mainly from dairy products and meat from ruminant
animals . Cer (d18:1/25:0) was previously described as a potential
urinary marker of inflammation-induced alcoholic liver disease .
Recently, an elevation of the Cer (d18:1/25:0) was also identified as a
serum prognostic marker to predict various acute diseases, including
cardiovascular death, myocardial infarction, and stroke in patients with
acute myocardial infarction within an ensuing 12-month period . Our
results suggested the pathological implications of Cer (d18:1/25:0) in
the development of pregnancy complicated by PE, which might provide
additional insights into the mechanistic roles of Cer (d18:1/25:0) and
other odd-chain Cer species in PE pathophysiology.
An association between Lep and Cer has been reported in several studies.
Lep was shown to exert its anorexigenic action by promoting
mitochondrial lipid oxidation in both adipose and non-adipose tissues to
alleviate ectopic accumulation of lipotoxic Cer via the activation of
AMP-activated protein kinase (AMPKK/AMPK) . The de novo synthesis of Cer
was found to play prominent roles in modulating downstream signaling of
central Lep’s activity via mediation of malonyl-CoA, carnitine palmitoyl
transferase-1c, and serine palmitoyl transferase . Persistent elevation
of circulating Lep also appears to induce resistance at the level of the
Lep receptor, which accounts for attenuated potency of Lep to alleviate
the accumulated cytotoxic Cer . Our data suggest a crosstalk between
Lep’s activity and de novo Cer synthesis. Lep functions well as a
predictor of PE early in gestation, while Cer (d18:1/25:0) performs
better at mid-gestation. The Lep/Cer (d18:1/25:0) ratio has a better
predictive performance than Lep or Cer (d18:1/25:0) levels alone. Our
findings revealed a correlation between the biological patterns of the
two markers during PE progression, which might add value to existing
knowledge about the Lep-Cer relationships.
Our study has several limitations. First, the sample sizes of the
cohorts were small and lacked racial heterogeneity–thus, the
generalizability of the results awaits larger and more racially diverse
study populations. Second, longitudinal collections of blood samples
were not evenly distributed over gestation. Finally, we did not
investigate the exact tissue of origin, where Lep is overexpressed in PE
women. By using a conditional knock-in placental Lep transgenic mouse
model, it may be possible to elucidate the mechanistic role of placental
Lep in the pathogenesis of PE early in pregnancy.