The role of proprotein convertase subtillisin/kexin type 9 in placental salvage and lipid metabolism in women with preeclampsia

Abstract

Introduction:

Preeclampsia is associated with decreased maternal low-density lipoprotein cholesterol (LDL-c), which is essential for fetal growth. The underlying mechanisms for decreased LDL-c in preeclampsia remain unknown. Proprotein convertase subtillisin/kexin type 9 (PCSK9) regulates serum LDL-c via LDL receptor (LDL-R) degradation. We describe the possible role of PCSK9 in lipid metabolism in all compartments of the parturient (maternal blood, placental tissue, and fetal blood) in pregnancies with and without preeclampsia.

Methods:

This is an observational study examining PCSK9 levels in maternal sera, umbilical cord blood, and PCSK9 protein content in placental tissue in three different locations (maternal placental interface, fetal placental interface, and umbilical cord) in women with and without preeclampsia at >23 weeks gestation.

Results:

68 parturients with preeclampsia and 55 without preeclampsia were enrolled. Maternal serum LDL-c (116.6 ± 48.9 mg/dL vs 146.1 ± 47.1 mg/dL, p = 0.0045) and PCSK9 (83 [61.8127.6] ng/mL vs 105.3 [83.5142.9] ng/mL, p = 0.011) were also reduced in the preeclamptics versus controls. There were no differences in PCSK9 protein content between preeclamptics and controls at comparative placental interfaces. However, PCSK9 protein content increased between the preeclampsia maternal placental interface (1.87 ± 0.62) and the preeclampsia umbilical cord (2.67 ± 1.08, p = 0.0243).

Discussion:

PCSK9 levels are lower in maternal sera in preeclampsia when compared to controls. Placental PCSK9 protein content in preeclampsia increases from the maternal interface to the umbilical cord; however, this is not seen in controls. This suggests a potential compensatory mechanism for PCSK9 which allows for higher circulating fetal LDL-c levels in preeclampsia.

1. Introduction

Normal placental metabolism requires an intricate balance between energy expenditure, hormone production and fetal development[1]. This includes physiologic hyperlipidemia, which provides the placenta with increased cholesterol, low density lipoprotein cholesterol (LDL-c), and fatty acids essential for hormone production and fetal neurologic development [2–4]. One factor thought to play a central role in regulating this process is proprotein convertase subtillisin/kexin type 9 (PCSK9)[5], which is a serine protease synthesized primarily by the liver and intestine[3,4,6]. In hepatocytes, PCSK9 degrades the LDL-receptor (LDL-R), therefore, increased PCSK9 functionality or gain of function leads to increased LDL-c levels and consequently increases the risk of subsequent atherosclerotic cardiovascular disease (ASCVD)[7]. Conversely, decreased PCSK9 activity leads to lower levels of circulating LDL-c and is associated with reduced ASCVD[8].

The extent to which PCSK9 contributes to lipid metabolism during pregnancy is unknown, especially in women with preeclampsia, a pregnancy specific disease characterized by new-onset or worsening hypertension and end-organ dysfunction in the mother and fetus[9]. Preeclampsia is a result of abnormal placentation in the first trimester which leads to a cascade of inflammation and vascular stiffness in the placental and maternal circulation similar to the development of atherosclerosis [10–14]. In addition, abnormalities in lipid metabolism may play a role as women with metabolic disorders are at increased risk of preeclampsia and abnormal vascular function[15].

Prior studies in women with preeclampsia have reported variable results in lipid abnormalities, with one study revealing elevated triglyceride levels[16], and another reporting higher postpartum LDL-c levels[17]. However, LDL-c levels during pregnancy are perhaps paradoxically lower in women with abnormal pregnancies, e.g. gestational diabetes, compared to those with uncomplicated pregnancies[18]. Lower plasma levels may result from increased attraction and storage of oxidized LDL-c in the placenta for transfer to the rapidly growing fetus; however, the mechanisms underlying the association of preeclampsia with lower LDL-c levels are unclear[19].

PCSK9 activity may have a role in this phenomenon as it plays an important role in pregnancy, placental metabolism, and fetal growth; further, fetal growth restriction is implicated in abnormal PCSK9 activity[20]. Despite its importance, there is little data investigating PCSK9 and its implications on lipid metabolism in the physiologic and anatomic compartments of the parturient: the maternal blood, the placental tissue, and the fetal blood. To address these knowledge gaps, our study measures PCSK9 and lipid levels (including LDL-c) in the maternal and fetal compartments in women with preeclampsia and in those with uncomplicated pregnancies. Further, we examine PCSK9 protein expression in the placenta at the levels of the maternal interface, fetal interface, and umbilical cord.

2. Methods

2.1. Study design and participants

This observational cohort study was approved by the Johns Hopkins University School of Medicine Institutional Review Board, and all participants gave written informed consent before enrollment. Female participants were enrolled if they were greater than 23 weeks gestation at index pregnancy from July 2015–May 2021. Preeclampsia was defined per the American College of Obstetricians and Gynecologist Task Force in Pregnancy criteria, which include a systolic blood pressure ≥160 mmHg and/or a diastolic blood pressure ≥100 mmHg measured on two separate occasions, and proteinuria defined by 24-h urine protein ≥300 mg or a protein/creatinine ratio ≥0.3[17]. In the absence of proteinuria, participants could meet the criteria for preeclampsia if they had thrombocytopenia (platelet <100,000/microliter), impaired liver function tests (twice the upper limit of normal), progressive renal dysfunction (new elevation in serum creatinine >1.1 mg/dL), pulmonary edema, or new-onset headaches and/or visual disturbances[17]. For the control group, women were enrolled if they had normotensive pregnancies without a history of essential hypertension or a hypertensive disorder of pregnancy. Exclusion criteria included known or preceding valvular or congenital heart disease, cardiomyopathy, pulmonary hypertension, lupus, prior cardiac surgery, interstitial lung disease, or antiphospholipid antibody syndrome. Maternal demographic and pregnancy characteristics were abstracted from the participants medical records following delivery.

2.2. Maternal blood draw

Following enrollment, blood was collected and then serum separated in a cold centrifuge at 4 ^◦C for 10 min at 2000 RPM and stored at − 80 ^◦C for subsequent analysis.

2.3. Fetal blood draw

Fetal blood draw was performed at the time of placental extraction during delivery. Blood was collected from the umbilical vein and separated into either serum or plasma and stored at − 80 ^◦C.

2.4. Placental samples

After delivery, a 1-inch portion of fresh placental tissue from the maternal and fetal placental interfaces were obtained. Neither the chorionic plates nor basal plates were removed before sampling. Additionally, a segment of umbilical cord was also obtained for analysis. All placental and umbilical cord tissue were stored at − 80 ^◦C.

2.2.1. Lipid profiles

Maternal and fetal lipids were measured by Cholestech LDX Analyzer (Abbott), following the manufacturer’s protocol at the Johns Hopkins Clinical Research Core Laboratory. The intra- and inter-assay coefficients of variance were respectively 2.4% and 3.2% for total cholesterol, 4.1 and 5.4% for HDL-c, 2.6 and 2.8 for triglycerides, and 4.4 and 5.1 for LDL-c.

2.5. C-reactive protein (CRP) assay

CRP was measured via an immunoturbidimetric assay at Quest clinical laboratories.

2.6. PCSK9 measurement

PCSK9 levels were measured on maternal and fetal serum aliquots using a commercial ELISA kit (R & D Systems, MN, USA, Catalog Number DPC900). The minimum detectable dose of human PCSK9 was 0.030 ng/mL and the intra- and inter-assay CVs were 5.4% ± 1.2% and 4.8% ± 1.9%, respectively. All the measurements were performed by utilizing a fully automated ELISA system (DS2, Dynex Technologies, Inc., Chantilly, VA, USA).

2.7. Protein extraction from placental tissue and umbilical cord

For each sample, approximately 5 mg of tissue were cut and placed in 300 μl of ice-cold lysis buffer (Ripa buffer Sigma, cat# R0278 with Pierce protease and Phosphatase inhibitor mini tablet Cat# A32959, Thermo Scientific). Tissue samples were sonicated 3 s/5 times, allowing for sample cool down between sonications. Next, 100 μl of lysis buffer were added and samples incubated on a shaker at 4 ^◦C for 2 h. Finally, the lysate was centrifuged for 30 min at 4 ^◦C and the supernatant transferred to new Eppendorf tubes for downstream applications. Samples were stored in − 80’ C.

2.8. Western immunoblotting of extracted proteins

All protein lysates were incubated with 1X NuPage LDS sample buffer (ThermoFisher, cat# NP0008) containing 1X NuPage reducing agent (ThermoFisher, cat# NP0004), boiled for 15 min at 95 ^◦C prior, and then proteins were separated on a NuPAGE 4–12% Bis-Tris gel (ThermoFisher, cat# NP0321). Gels were then transferred onto PVDF membranes, and the following primary antibodies were used: PCSK9 (ab28770, 0.3ug/ml dilution, Abcam), and β-actin (ab8226, 1:5000 dilution, Abcam). PVDF membranes were incubated overnight at 4 ^◦C with primary antibodies diluted in 1X TBS Intercept Blocking Buffer (LiCor, cat# 927–60001). The following day, PVDF membranes were washed three times in 1X TBST, then incubated with species-specific IRDye secondary antibodies for 1hr at room temperature (LI-COR, at 1:5000).

2.9. Statistical analysis

Analysis was performed using R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism 8. Continuous variables were examined for normality. Mann-Whitney U tests were used for non-normally distributed continuous variables, and t-tests were used for normally distributed continuous variables. Chi-square tests were used for categorical variables. A p value < 0.05 was considered statistically significant.

To examine PCSK9 levels by race, we stratified participants in both preeclampsia and control groups by Black and non-Black race.

3. Results

3.1. Study design and study population

We recruited 123 pregnant women: 68 with preeclampsia and 55 without preeclampsia. Maternal demographic data are shown in (Table 1). When compared to the normotensive group, the women with preeclampsia had higher median BMI (32.7 [29.0,39.5] vs 30.3 [27.5,32.7] kg/m², p = 0.019), lower gestational age (34.2 [32.1,37.2] vs 39.4 [38.5, 40.2] weeks, p < 0.001), and included more Black women (42 (62%) vs 16 (29%), p = 0.001).

Table 1.

Demographics and biomarkers of study participants with maternal lipid biomarkers.

Variable	Normotensive Control (N = 55)	Preeclampsia (N = 68)	p-value
Maternal and Pregnancy Characteristics
Age (years)	31.0 (6.7)	29.6 (6.6)	0.223
BMI (kg/m2)	30.3 [27.5–32.7]	32.7 [29.0–39.5]	0.019a
Diabetes - n (%)			0.170
Pre-gestational diabetes	1 [2]	6 [9]
Gestational diabetes	1 [2]	3 [4]
Pre-existing chronic hypertension - n (%)	0 (0)	13 [19]	0.002a
Race - n (%)			0.001a
Asian	3 [5]	0 (0)
Black/African American	16 (29)	42 (62)
White/Caucasian	32 (58)	25 (37)
Other	4 [7]	1 [1]
Ethnicity - n (%)
Hispanic	5 [9]	3 [4]	0.500
Non-Hispanic	50 (91)	65 (96)
Fetal Sex – n (%)			0.489
Female	32 (58)	25 (37)
Male Gestational age (weeks) Birthweight (grams)	23 (41) 39.4 [38.5–40.2] 3380 [3105–3721]	43 (64) 34.2 [32.1–37.2] 2020 [1405–3023]	<0.001a <0.001a
Placental weight (grams)b	428 [ 338− 481]	370 [246–494]	0.4065
Maternal Biomarkers PCSK9 ng/mL CRP mg/dL TC mg/dL	105.3 [83.5–142.9] 0.40 [0.20–0.90] 283.5 (80.0)	83 [61.8–127.6] 1.25 [0.56–2.50] 221.3 (68.4)	0.011a <0.001a <0.001a
TRIG mg/dL HDL mg/dL LDL mg/dL	201.5 [150.3–261.0] 77.0 [58.0–96.0] 146.1 (47.1)	185.0 [148.0–255.0] 59.0 [47.8–73.3] 116.6 (48.9)	0.413 <0.001a 0.004a
VLDL mg/dL	40.4 [30.7–53.2]	37.0 [28.6–51.0]	0.32

Data are mean (SD) and median [IQR]. IQR is interquartile range; PCSK9 is Proprotein convertase subtillisin/kexin type 9; CRP is C-reactive protein, TC is total cholesterol; TRIG is triglyceride; HDL is high density lipoprotein cholesterol; LDL is low density lipoprotein cholesterol, VLDL is very low density lipoprotein cholesterol.

^aDenotes statistical significance between the groups.

^bThere were only n = 18 normotensive and n = 53 placental weights.

Neonatal birthweights were lower in the preeclampsia group (2020 [1405,3023] vs 3380 [3105,3721] grams, p < 0.001) There was no difference in fetal sex and placental weight between groups.

3.2. Maternal LDL-c are reduced in preeclampsia when compared to controls

Maternal serum LDL-c levels were reduced in women with preeclampsia compared to controls (116.6 ± 48.9 vs 146.1 ± 47.1 mg/dL, p = 0.0045). In addition, high density lipoprotein cholesterol (HDL-c) levels were lower in women with preeclampsia when compared to controls (59.0 [47.8, 73.3] vs 77.0 [58.0,96.0] mg/dL, p < 0.001). (Table 1).

3.3. Inflammatory markers are increased in preeclampsia when compared to controls

Maternal serum CRP was greater in women with preeclampsia when compared to controls (1.25 [0.56,2.50] vs 0.40 [0.20,0.90] mg/dL, p < 0.001). (Table 1).

3.4. Maternal PCSK9 levels are reduced in preeclampsia when compared to controls

Maternal serum PCSK9 was lower in women with preeclampsia compared to controls (83 [61.8, 127.6] vs 105.3 [83.5, 142.9] ng/mL, p = 0.011). (Table 1).

When we adjusted PCSK9 levels for chronic hypertension and diabetes status, PCSK9 levels remained significantly lower in preeclampsia versus controls (p = 0.014). However, when we adjusted for chronic hypertension, diabetes, and BMI, PCSK9 was no longer significantly lower, (p = 0.475).

3.5. Maternal biomarkers stratified by race

Non-Black women with preeclampsia were more likely to have a higher BMI and lower gestational age (Supplement Table 1). Black women had no other differences in demographic characteristics when comparing preeclampsia to controls.

In Black women, PCSK9 was lower in women with preeclampsia when compared to controls (72.2 [55.7, 103.2] vs 92.2 [83.1, 288.5] ng/mL (Supplement Table 2). However, PCSK9 levels were not different between non-Black preeclamptic and control women.

3.6. Placental PCSK9 protein content increases from maternal to the fetal compartment in preeclampsia: maternal placental interface, fetal placental interface, and umbilical cord

In a subset of participants, 30 samples of fetal cord blood were obtained, including 12 from women with preeclampsia and 18 without preeclampsia (Table 2). Placental tissue samples were also obtained from the following sites: the maternal placental interface (14 preeclampsia, 7 controls), the fetal placental interface (14 preeclampsia, 7 controls), and the umbilical cord (14 preeclampsia, 4 controls) (Fig. 1). In this sub-cohort, the preeclampsia group had a higher median BMI (35.6 [30.7,40.1] vs 29.1 [26.5, 30.6] kg/m², p = 0.022); higher prevalence of chronic hypertension (n = 6 (50%) vs 0 (0%) p = 0.004), and higher incidence of diabetes (n = 4 (33%) vs 1 (6%), p = 0.03).

Table 2.

Demographics and biomarkers of study participants with placenta and/or cord blood biomarkers.

Variable	Normotensive Controls (N = 18)	Preeclampsia (N = 12)	p-value
Maternal and Pregnancy Characteristics
Age (years)	33.2 (4.7)	32.7 (6.1)	0.817
BMI (km/m2)	29.1 [26.5–30.6]	35.6 [30.7–40.1]	0.023*
Diabetes			0.025*
Pre-gestational diabetes	0 (0%)	4 (33%)
Gestational diabetes	1 (6%)	0 (0%)
Pre-existing chronic hypertension	0 (0%)	6 (50%)	0.003*
Race - n (%)			0.243
Asian	2 [11]	0 (0)
Black/African American	6 (33)	8 (67)
White/Caucasian	1 [6]	4 (33)
Other	9 (50)	0 (0)
Ethnicity - n (%)			1
Hispanic	1 [6]	0 (0)
Non-Hispanic	17 (94)	12 (100)
Gestational age (weeks)	39.2 [39.1–39.4]	36.1 [34.5–36.8]	0.015*
Fetal Biomarkers Cord blood PCSK9 ng/mL	333.2 (102.3)	240.0 (64.6)	0.368
Cord blood CRP mg/dL	0.20 [0.20–0.20]	0.20 [0.20–0.20]	0.589
Cord blood TC mg/dL	100 [100− 100]	100 [100–100.5]	0.452
Cord blood TRIG mg/dL	45.0 [45.0–45.0]	45.0 [45.0–45.0]	0.836
Cord blood HDL mg/dL	22.5 [19.5–31.0]	20 [15.8–35.8]	0.563
Cord blood LDL mg/dL	68.5 [60.5–71.5]	67 [49–76]	0.755

Data are mean (SD) and median [IQR]. IQR is interquartile range; PCSK9 is Proprotein convertase subtillisin/kexin type 9; CRP is C-reactive protein, TC is total cholesterol; TRIG is triglyceride; HDL is high density lipoprotein cholesterol; LDL is low density lipoprotein cholesterol, VLDL is very low density lipoprotein cholesterol.

^*denotes statistical significance between the groups.

Fig. 1.

Proprotein convertase subtilisin/kexin type 9 (PCSK9) protein content in placental tissue (maternal and fetal) and umbilical cord samples from pre-eclamptic patients and healthy control pregnancies Panel (A) illustrates a significant gradient in PCKS9 protein content from the maternal placenta to the umbilical cord in preeclampsia but not in healthy controls. Thus, PCSK9 protein content is higher in preeclamptic umbilical cord (PEC UC) compared with PEC MP, p = 0.0243. In Panel (B) representative immunoblot with PCSK9 and beta actin specific antibodies. CNTL FP = control fetal placenta; PEC FP = preeclamptic fetal placenta; CNTL UC = control umbilical cord. N = 14 for PEC MP, PEC FP, PEC UC; N = 7 for CNTL MP and CNTL FP; N = 4 for CNTL UC.

PCSK9 protein content among preeclamptic participants increased with progression towards the fetal compartment, and there was a significant difference in PCSK9 protein content from the preeclampsia maternal placental interface (1.87 ± 0.62 densitometry units normalized to beta-actin) to the preeclampsia umbilical cord (2.67 ± 1.08 densitometry units normalized to beta-actin), p = 0.0243 (Fig. 1).

In contrast, there were no differences in PCSK9 protein content in the maternal placental interface, fetal placental interface, and umbilical cord among control patients. Fig. 1. Based on a prior study assessing the stability of reference proteins in human placental tissue, we also performed general protein stains using amido black as a reference protein marker and normalization approach[21].The results in our cohort using beta actin as a reference protein for normalization and amido black as a general protein stain were comparable.

3.7. Fetal LDL-c and PCSK9 serum levels

PCSK9 levels in the circulating fetal blood were twice that of the maternal sera in the preeclampsia and control subgroups: maternal sera PCSK9 of preeclampsia (83 [61.8–127.6] ng/mL) and controls (105 [83.5–142.9] ng/mL); fetal sera PCSK9 of preeclampsia (240 + 64.6 ng/mL) and controls (332 + 102.3 ng/mL). In addition, LDL-c was less than half that of the maternal sera: maternal LDL-c sera of preeclampsia (116 + 48.9) mg/dL and controls (146.1 + 47.1); fetal LDL-c sera of preeclampsia (67 [49–76] mg/dL) and controls (68.5 [60.5–71.5] mg/dL). Notably, there were no differences in fetal sera CRP samples between preeclamptic and control participants.

4. Discussion

This is the first study to describe a possible link between PCSK9 and other lipid markers in women with preeclampsia and in those with healthy pregnancies. In addition, this study is the first to report markedly higher levels of PCSK9 expression in the fetal circulation compared to maternal circulation, irrespective of the presence of preeclampsia. This provides further evidence of its role in fetal lipid metabolism and development. Further, our study highlights differences in lipid biomarkers, particularly for women with preeclampsia, who have markedly lower PCSK9 and LDL-c levels compared to women with normal pregnancies. Finally, women with preeclampsia are noted to have a lower amount of PCSK9 protein at the maternal placental interface which incrementally increases as it reaches the fetal compartment (umbilical cord) (Fig. 2). Difference in PCSK9 levels remained significant when adjusted for hypertension and diabetes; however, when BMI was added, significance between groups no longer existed.

Fig. 2.

PCSK9 metabolism: From mother to baby

Fig. 2 shows the relationship of PCSK9 to LDL-R and LDL-c in the (A) maternal sera, (B) the maternal placental interfaces, and (C) fetal sera in preeclampsia. The figure shows an increase in PCSK9 and PCSK9 protein upregulation from A to C. It further shows the increase in PCSK9 upregulation within the placenta at the [1] maternal placental interface [2], the fetal placental interface [3], and the umbilical cord in preeclampsia. The upregulation in PSCK9 may allow for shunting of LDL-c from the mother to fetus for uptake at the level of the fetal adrenals and/or neurologic system. Normal pregnancy does not show an upregulated gradient in PCSK9 or PCSK9 protein expression, thus indicating a possible mechanism of fetal salvage via LDL-c and PCSK9 in preeclampsia. (D) is the anatomy of the umbilical cord.

The lower LDL-c levels observed in our preeclampsia cohort are consistent with prior studies evaluating women with hypertensive disorders of pregnancy[5,22]. Reasons for lower LDL-c in preeclampsia may be due to hydrolysis and exclusion from maternal and fetal circulation in abnormal pregnancies involving fetal growth restriction or hypertensive disorders of pregnancy[22,23]. Further, LDL-c transport may be insufficient from the placenta to maternal circulation in preeclampsia[24,25]. Our study suggests that lower maternal PCSK9 levels may be implicated in this lipid regulating process, since women with preeclampsia expressed significantly lower maternal PCSK9 protein expression in their placentas when compared to healthy controls.

In normal healthy pregnancies LDL-c levels increase up to 50%[17], which is needed for placental hormone production and fetal fatty acid synthesis[3,4,16]. Thus, healthy pregnancies may require higher levels of PCSK9 (leading to increased LDL-R degradation and resulting in higher serum LDL-c levels). In fact, one study reports greater levels in the first trimester with notable decline in levels entering the second trimester[26]. In our study, preeclamptic participants had significant incremental increases in the expression of PCSK9 from maternal placental interface to the umbilical cord. Control participants did not have this increase making umbilical cord samples similar between the two groups (Fig. 1). Therefore, maternal PCSK9 levels may be a biomarker associated with states of distress (i.e., preeclampsia) and indicate a greater need for placental and/or fetal salvage of lipids. These potential compensatory mechanisms in preeclamptic placentas and fetal circulation may allow for comparable PCSK9 levels to fetuses of healthy normal pregnancies. This in turn could lead to similar fetal lipid profiles between the groups.

Our study does have several limitations. First, there are demographic differences between the control and preeclampsia groups, which is likely due to the nature of higher risk pregnancies in the US and in our population. When we stratified the maternal serum by race, we only saw PCSK9 differences in Black women, which were significantly lower in preeclampsia and tended to be lower when compared to non-Black women. Interestingly, non-pregnant Black Americans tend to have higher PCSK9 levels when compared to their white counterparts [24]; however, Black women have higher rates of preeclampsia. Therefore, lower PCSK9 levels in Black women may be a dramatic marker of an abnormal placental metabolic process, and it is important to further explore adverse social and environmental exposures that have a disproportionate effect on Black women.

When adjusting for chronic hypertension and diabetes PCSK9 levels were significantly lower in preeclamptic women. However, with the addition of BMI to the model, we saw loss of significance in PCSK9 levels between controls and preeclamptic women with preeclampsia having higher BMI. This is contrary to other published data in PCSK9 and BMI as people with higher BMIs typically have higher PCSK9 levels[27]. This is likely secondary to the use of adjustment models in cohort with a smaller number of participants. We also did not perform mechanistic experiments to prove the role of PCSK9. We also note differences in gestational age from iatrogenic prematurity, and we were only able to obtain placental and fetal samples from a subset of participants. However, even with these relatively small numbers, meaningful comparisons for future research can be made between the women with preeclampsia and normotensive pregnancies.

5. Conclusion

In conclusion, we observed lower maternal PCSK9 and LDL-c levels in women with preeclampsia when compared to controls, especially in Black women. We further showed that placental PCSK9 expression increased from the maternal to the fetal interface with equal levels of PCSK9 and lipoproteins at the fetal level. These findings may provide mechanistic insights into placental metabolic abnormalities in preeclampsia and may explain why fetal lipid and PCSK9 levels remain similar. Future studies should replicate these findings in larger cohorts in order to determine further mechanisms underlying maternal, placental, and fetal lipid metabolism in healthy and complicated pregnancies.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.placenta.2022.12.008.