Abstract

Background

   Effective intrauterine treatments for placental-mediated fetal growth
   restriction (FGR) remain limited, necessitating reliable protein
   biomarkers for early diagnosis and management.

Methods

   In this study, we analyzed differential protein expression in
   peripheral blood plasma samples from 44 placental-mediated FGR patients
   and 44 normal pregnant women using the Olink-Explore-384-Inflammation
   panel. The analysis identified significant differences in protein
   expression levels, followed by enrichment analyses to explore the
   underlying biological mechanisms. Protein-protein interaction (PPI)
   network analysis and Least Absolute Shrinkage and Selection Operator
   (LASSO) modeling were used to identify key proteins as potential
   biomarkers.

Results

   We identified 225 proteins with significantly altered expression
   between FGR patients and normal pregnancies. Proteins such as Placental
   Growth Factor (PGF) and Hepatocyte Growth Factor (HGF) were previously
   found to be strongly associated with FGR. In addition, we discovered
   novel proteins potentially associated with FGR, including ESM1 and
   TIMP3. Enrichment analyses revealed that several pathways, including
   placental dysfunction, inflammatory responses, and oxidative stress,
   may play crucial roles in FGR pathophysiology. PPI network analysis
   further identified key proteins such as ANGPT2, CD40, and HGF, as
   potentially linked to FGR. LASSO modeling validated PGF and ESM1 as
   important biomarkers. Additionally, integrating a multi-protein panel
   with blood flow disruption analysis significantly improved diagnostic
   accuracy.

Conclusion

   Our findings provide valuable insights into the molecular mechanisms of
   FGR, identifying key proteins as potential biomarkers. The
   multi-protein panel model offers a promising tool for early screening
   and diagnosis of FGR.

   Keywords: fetal growth restriction (FGR), Olink proteomics platform,
   proximity extension assay (PEA), biomarkers, targeted proteomics
   analysis

1. Introduction

   Fetal Growth Restriction (FGR) occurs when a fetus fails to achieve its
   genetically determined growth potential within the uterus, manifesting
   as a fetal weight below the 10th percentile for gestational age
   ([44]1). The etiology of FGR is complex, placental-mediated FGR is the
   most prevalent subtype. It holds the greatest promise for improving
   adverse outcomes through clinical prevention and management strategies
   ([45]2). As the mechanism of placental-mediated FGR is unknown and
   there is a lack of effective clinical screening, prevention, diagnosis
   and intervention, related research has become a hotspot of concern at
   home and abroad ([46]3). Placental-mediated FGR is associated with
   several pathological states of pregnancy, with the mother leading to
   the development of preeclampsia and the fetus showing growth
   restriction, poses substantial risks to both maternal and fetal health
   ([47]4). While the global incidence of FGR varies is approximately
   5%-10%, it can exceed 30% among pregnant women with preeclampsia
   ([48]5). The clinical manifestations of FGR are often subtle and
   typically diagnosed through routine prenatal check-ups and ultrasound
   examinations ([49]6). Pregnant women may exhibit symptoms of
   preeclampsia, such as gestational hypertension and proteinuria ([50]7).
   During ultrasound examinations, fetal growth indicators such as
   biparietal diameter, abdominal circumference, and femur length are
   significantly smaller for the corresponding gestational age.
   Additionally, decreased fetal biophysical scores and reduced amniotic
   fluid also signal FGR ([51]8). FGR not only affects fetal growth and
   development but may also lead to a series of adverse pregnancy
   outcomes. Infants born with FGR may experience low birth weight,
   neonatal asphyxia, neonatal death, and in the long term, may face
   intellectual developmental delays, growth retardation, and an increased
   risk of metabolic diseases in adulthood ([52]9). Therefore, early
   diagnosis and intervention for FGR are crucial for improving maternal
   and fetal outcomes.

   Current intervention strategies based on abnormal placental-mediated
   FGR phenotypes are extremely limited, with the underlying reason being
   a lack of understanding of placental-mediated FGR etiological
   mechanisms. It is currently believed that the placenta is the organ of
   origin for placental-mediated FGR pathogenesis, with placental
   development, cellular communication at the maternal-fetal interface,
   and the placenta-fetal gut axis being the key events surrounding
   placental-mediated FGR etiological mechanisms ([53]10). Extensive
   studies have explored potential biomarkers for FGR, including placental
   hormones, inflammatory cytokines, and metabolites ([54]11). However,
   individual biomarkers often lack sufficient sensitivity and specificity
   ([55]12). Noteworthy biomarkers include human placental lactogen (hPL)
   ([56]13) and pregnancy-associated plasma protein A (PAPP-A) ([57]14),
   whose decreased levels correlate with FGR. Elevated Inflammatory
   cytokines like interleukin-6 (IL-6) ([58]15) and tumor necrosis
   factor-α (TNF-α) ([59]16) reflect placental inflammation and oxidative
   stress. Additionally, alterations in metabolites such as placental
   growth factor (PlGF) ([60]17) and soluble fms-like tyrosine kinase-1
   (sFlt-1) ([61]18) have been studied to assess placental function and
   predict FGR risk.

   However, current research still faces several unresolved problems.
   Firstly, existing biomarkers often lack adequate sensitivity and
   specificity for broad clinical use ([62]19). Secondly, FGR is
   multifactorial, including maternal, fetal, and placental factors,
   complicating comprehensive assessment of FGR through single biomarkers
   ([63]20). Lastly, early diagnosis of FGR remains challenging due to its
   nonspecific nature of early symptoms and the limitations of ultrasound
   examinations ([64]21).

   Until a validated single screening indicator or model is established,
   both domestic and international guidelines do not recommend clinical
   screening for placental-mediated FGR in isolation ([65]22). Currently,
   the majority of clinical studies rely on early-pregnancy preeclampsia
   screening models or first- and second-trimester Down syndrome serum
   screening models to predict FGR. Except for placental growth factor
   (PlGF), which has a sensitivity of 27%, most single biomarker
   screenings exhibit low sensitivity and limited value ([66]14). With the
   widespread adoption of noninvasive prenatal testing, the establishment
   of predictive models for early-onset severe placental-mediated FGR
   during the first trimester, based on fetal free DNA and RNA derived
   from maternal plasma, combined with novel biomarkers such as SPINT1 and
   Ang2, represents a focal area of future clinical research ([67]23,
   [68]24).

   To address these challenges, the objective of our research is to
   identify differentially expressed proteins as potential biomarkers for
   placental-mediated FGR through the application of advanced targeted
   proteomics technology, specifically the OLINK technology. In recent
   years, this technology has gradually been applied in research, offering
   the potential to discover biomarkers through large-scale screening of
   protein changes in maternal plasma ([69]25). However, there are
   currently limited examples of OLINK technology’s application in FGR
   research, which represents an innovative aspect of our study. The FGR
   cases included in this project are all attributed to placental
   perfusion insufficiency, and we have initially established a precise
   diagnostic process for placental-mediated FGR: this involves screening
   for fetal factors (genetic, structural, infectious, etc.) and assessing
   ultrasound and maternal blood biomarkers related to placental function
   ([70]26). Our inclusion criteria are stringent compared to previous
   basic and clinical studies. Ultimately, we aim to identify novel FGR
   biomarkers with higher sensitivity and specificity. This will
   facilitate early diagnosis, risk assessment, and individualized
   treatment of FGR, ultimately improving maternal and fetal outcomes.

2. Methods

2.1. Patient sample collection

   From 2019 to 2024, we conducted a retrospective and prospective study
   at Shanghai First Maternity and Infant Hospital, enrolling 44 patients
   with placental-mediated FGR and 44 normal controls. Inclusion criteria
   for placental-mediated FGR patients included an Abdominal Circumference
   (AC)/Estimated Fetal Weight (EFW) ratio below the 3rd percentile for
   gestational age, abnormal blood flow, negative genetic testing, and a
   gestational age ranging from 19 to 31 weeks (with a mean of 25 weeks).
   Exclusion criteria included multiple gestations and fetal structural
   abnormalities. Control participants were matched for gestational age
   with the FGR group and excluded for any fetal/placental abnormalities
   and abnormal noninvasive prenatal test results. All participants
   provided written informed consent, and the study was approved by the
   Ethics Committee of Shanghai First Maternity and Infant Hospital
   (Approval Number: KS20262). The workflow of this study is illustrated
   in [71]Figure 1 .

Figure 1.

   [72]Figure 1
   [73]Open in a new tab

   Workflow of the study. The analytical workflow and key findings are
   presented in this figure. We analyzed differential protein expression
   in peripheral blood plasma samples from 44 placental-mediated FGR
   patients and 44 normal pregnant women using the
   Olink-Explore-384-Inflammation panel. The analysis identified
   significant differences in protein expression levels, followed by
   enrichment analyses to explore the underlying biological mechanisms.
   PPI network analysis and LASSO modeling were used to identify key
   proteins as potential biomarkers.

   Peripheral blood samples (2 mL) were collected from each participant.
   Plasma samples were obtained by centrifugation at 1200g for 15 minutes,
   processed to remove cellular debris, fats, and other impurities, and
   then stored at -80°C for subsequent analysis.

2.2. Protein abundance measurement

   Plasma samples were analyzed for 368 inflammation-related proteins
   using the Olink-Explore-384-Inflammation panel (Olink™ Proteomics,
   Uppsala, Sweden) with Proximity Extension Assay (PEA) technology on the
   Illumina NGS sequencing platform. The procedure involved three main
   steps: incubation (where plasma samples interacted with 368 antibody
   pairs tagged with unique DNA oligonucleotides), extension
   (amplification and enrichment of DNA fragments), and detection
   (sequencing of enriched DNA fragments for data collection).

   Each detection panel included two quality control (QC) systems. The
   internal QC system monitored amplification using an Immuno control, an
   Extension control, and a Detection control, while external QC system
   consisted of eight references (two sample controls, three negative