Serum protein profile analysis via label-free quantitation proteomics in patients with early-onset preeclampsia

Abstract Background Preeclampsia (PE) is a serious pregnancy complication, resulting in potentially life-threatening conditions for both mother and foetus. It is worth noting that early-onset PE has become a great challenge for clinicians due to its complex manifestation, rapid progression and serious complications. This study aims to investigate differential serum proteome profiles in patients with early-onset PE. Methods Each serum sample was separated using a nanoliter flow rate Easy-nLC chromatography system. Then the samples were analysed by mass spectrometry. Bioinformatics analyses were conducted to analyse the functional categories or signal transduction pathways for differentially abundant proteins. Key proteins identified by mass spectrometry were verified by ELISA. Results We found 30 and 34 proteins were upregulated and downregulated in early-onset PE patients (n = 3) vs controls (n = 3), respectively. Functional enrichment analysis revealed differentially expressed proteins related to the immune response and regulation of peptidase activity. ELISA confirmed that there were lower CSH1 levels and higher LPA concentrations in the serum samples of early-onset PE patients (n = 22) than in healthy controls (n = 19) (p < 0.05 for CSH1 and p < 0.001 for LPA). Conclusions This study revealed the critical features of serum proteins in early-onset PE patients. LPA and CSH1 may serve as biomarkers for early-onset PE diagnosis and therapy. PLAIN LANGUAGE SUMMARY Early-onset preeclampsia (PE) is still lacking definitive diagnostic or therapeutic strategies. Thus, we tried to identify effective and specific biomarkers for early-onset PE. In this study, we explored the serum protein profiles through the approach of label-free quantitation proteomics between early-onset PE patients and healthy controls. We identified 64 differentially expressed proteins in early-onset PE patients’ serum samples. These differentially expressed proteins are associated with the immune response and regulation of peptidase activity. In addition, our findings suggest that LPA and CSH1 may serve as candidate biomarkers for early-onset PE diagnosis and therapy. These results may help physicians to diagnose early-onset PE clinically. What’s more, our findings provide new insights into the onset and progression of early-onset PE disease.


Introduction
Preeclampsia (PE) is a multisystem disease unique to pregnancy, with an incidence of approximately 3−5%, resulting in 60,000 maternal deaths and more than half a million preterm births annually (Mol et al. 2016).It is also a leading cause of maternal morbidity and mortality (Lo et al. 2013).PE is typically characterised by maternal hypertension and proteinuria, as well as other systemic manifestations such as edoema, renal failure, and CNS disturbances (Filipek and Jurewicz 2018).The effects of PE in women are long-term, with an increased risk of cardiovascular disease, cerebrovascular disease, and premature death (Tranquilli et al. 2012).PE increases the risk of foetal growth restriction and placental abruption and is associated with adverse neonatal outcomes, including respiratory distress syndrome, bronchopulmonary dysplasia, retinopathy of prematurity, necrotising enterocolitis, the need for neonatal intensive care, neurodevelopmental delay, and foetal or neonatal death (Carter et al. 2017, Hung et al. 2018).
According to varied gestational ages at onset, PE has been recognised as two different subtypes: early-onset PE (< 34 weeks of gestation) and late -onset PE (� 34 weeks of gestation) (Durst et al. 2016).As a severe form of PE, early-onset PE contributes to higher rates of perinatal mortality and severe neonatal morbidity than late-onset PE (van Esch et al. 2017).Many independent risk factors are related with the occurrence of early-onset PE, such as abnormal gene expressions (Kang et al. 2021), impaired villous development (Travaglino et al. 2019), immune maladaptation (Quinn et al. 2011) and defective immunoregulatory (Quinn et al. 2011).Strong heterogeneity and high subjective assessment affect the diagnostic accuracy rate of early-onset PE.Currently, there are no effective clinical methods to predict early-onset PE.With the rapid development of proteomics technologies and novel instruments, new tools for biomarker research in earlyonset PE have emerged.It is worth mentioning that maternal serum biomarkers have been extensively studied and have enabled the rapid development of non-invasive prenatal diagnostic techniques.sFlt-1/PlGF ratio and soluble endoglin (sEng) have been identified as candidate biomarker in the assessment of early-onset PE (Perales et al. 2017, Hirashima et al. 2008).Erbil , a high molecular weight glycoprotein, was elevated in the severe PE group than that in the mild PE and control groups.Besides, maternal serum CA-125 levels were positively correlated with systolic and diastolic blood pressure as well as proteinuria, and negatively correlated with birth weight and gestational age at birth.This indicated that the increased levels of CA-125 might be a useful marker for the severity of PE (Karaman et al. 2014).Despite such advances, no single reliable marker has performed sufficiently well owing to the heterogeneity of this disorder.However, there are no well -established primary prevention measures.Therefore, robust biomarkers are urgently needed to predict the occurrence and development of preeclampsia.
Proteomics, which refers to a large-scale characterisation of proteins coded by the genome, aims to systematically investigate detailed information of protein abundance, localisation, modifications, interactions, activities and functions.With the development of proteomics technologies, proteomics has proven to be a powerful tool in screening specific biomarkers for PE from maternal body fluids or foetal appendages.Recent evidence has demonstrated that the differentially expressed proteins identified in cerebrospinal fluid collected from women who have normal or preeclamptic pregnancies related to four key regulators of signal transduction, including transforming growth factor-b (TGF-b), vascular endothelial growth factor A (VEGFA), angiotensinogen, and interleukin-6 (IL-6) (Ciampa et al. 2018).Furthermore, several potential PE biomarkers are identified by urine peptidome/proteome, such as prostaglandin-H2 D-isomerase (PTGDS), lipocalin-type prostaglandin D synthase (L-PGDS), fibrinogen alpha chain, collagen alpha chain, and uromodulin fragments (Guo et al. 2019, Kononikhin et al. 2016, Carty et al. 2011).Impaired uterine spiral artery remodelling, which is caused by incomplete infiltration of nourishing cells, might play a fundamental role in the development of PE.Consequently, placental proteomics analysis is of great value towards PE diagnose and treatment.Yu et al.' study have revealed critical features of placental proteins between PE and control group through proteome and acetyl proteome analysis (Shangguan et al. 2021).In summary, proteomics provides a powerful approach for potential protein biomarkers discovery, which would be meaningful to conduct early clinical diagnosis and provide effective management and treatment for PE patients.
In this study, a label-free quantitative proteomics method was used to detect and screen the proteomes of patients with early-onset PE and healthy pregnant women.Bioinformatics analysis of differential proteome profiles was performed using gene ontology (GO) annotations and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways between the two groups.Enzyme-linked immunosorbent assay (ELISA) was used to detect the expression of candidate differentially expressed proteins to verify the reliability of the label-free results.Our study provides insight into the proteomic landscape of earlyonset PE.

Participants
The Second Hospital of Shandong University Ethics Committee approved our research protocol.Written informed consent was obtained from all study participants before their participation.All the participants were singleton pregnancies.The details of early-onset PE patients and controls recruitment have been previously published (Xin et al. 2022).Three early-onset PE patients and three healthy subjects matched for age and BMI were enrolled for the serum proteome profiles.An additional 19 healthy controls and 22 early-onset PE patients were recruited for ELISA validation of candidate protein biomarkers.Demographic details of the study participants are presented in Table 1 and Supplemental Table 1.Serum samples were collected from all the participants from May 2020 to December 2020 at the Department of Obstetrics of the Second Hospital of Shandong University.The isolated serum samples were stored at −80 � C until use.

Protein extraction, quantification, and digestion
Serum proteins were extracted using a ProteoPrep Blue Albumin and IgG Depletion Kit (Sigma-Aldrich, USA).Protein concentration was determined using the Bradford method.Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed to separate proteins and evaluate sample quality.Protein samples were digested with sequencegrade modified trypsin (Promega, USA).

Liquid chromatography (LC)-MS/MS analysis
For total proteome analysis, the digests of each sample were separated using an EASY-nLC 1000 UPLC system (Thermo Fisher Scientific, USA).Subsequently, eluting peptides were analysed using Q-Exactive TM Plus mass spectrometers (Thermo Fisher Scientific, USA) based on the following parameters: PRM mode; MS scan range, 350-1800 m/z; automatic gain control target set to 3e6 for full MS; maximum injection time, 50 ms for full MS and 45 ms for MS/MS; and isolation window of MS/MS, 2 m/z.

Database search
We used MaxQuant software (version 1.5.5.1) for the database search against the protein database Uniprot_HomoSapiens_ 161584_20180123 for MS/MS data, and the Label-Free Quantitation (LFQ) algorithm was applied for quantitative analysis.

GO and KEGG signalling pathway analyses
Our GO annotations for human proteins were performed for three different aspects: biological process (BP), molecular function (MF), and cellular component (CC).KEGG pathways were used to determine the pathways enriched for the identified proteins.The HiPlot online tool (https://hiplot.com.cn/) was used to perform GO annotations and KEGG pathways.

ELISA
ELISA was performed to validate the candidate protein biomarkers identified in the maternal serum by LC-MS/MS.Standard curves were established using the standards supplied by the manufacturers for reference sample concentrations.The experiments were performed according to the manufacturer's protocol.All measurements were done in duplicate, and the inter-and intra-assay coefficients of variation ranged from 3% to 8%.Human chorionic somatomammotropin hormone 1 (CSH1) and lysophosphatidic acid (LPA) ELISA kits were purchased from Jiangsu Meimian Industrial Company.

Statistical analysis
The differences between the two groups were calculated using fold change as a criterion.As for Table 1, median and range were used for the comparison of clinical characteristics between controls and patients.We used mean and standard deviation to analyse the demographic data in Supplemental Table 1.A protein species was considered differentially expressed if its fold-change (FC) was >1.5 or <0.67 and the p-value was <0.05.GraphPad Prism version 8 software (GraphPad Software, USA) was used to calculate statistical significance.P-values of <0.05 were considered significant.

Proteomics analysis showed 64 differentially expressed serum proteins in PE patients compared with healthy controls
The demographic and clinical characteristics of the participants are shown in Table 1.There was no difference in maternal age or BMI between early-onset PE patients (n ¼ 3) and gestational age-matched normotensive controls (n ¼ 3).Systolic and diastolic blood pressure in the early-onset PE group were remarkably higher than that in the control group (p < 0.01).The median values of 24 h proteinuria and gestational age at onset in the case group were 50.1 mg/24 h and 28.6 weeks, respectively.The ALT and AST liver enzymes were significantly higher in the PE group.Gestational age at delivery was lower among PE group.No differences were observed in factors of foetal birthweight, neonatal sex, caesarean section, live-born infants, uneventful pregnancy and gestational age at blood sampling.Uterine and umbilical Doppler indices were elevated in patients with PE.Label -free analysis of the early-onset PE samples is summarised in a flowchart (Figure 1).According to the label-free report, 511 proteins were detected, and 64 species of serum proteins were differentially expressed in the serum of early-onset PE patients compared with that of normal controls, including 30 proteins that were upregulated and 34 that were downregulated (P < 0.05, ratio of change > 1.5 or < 0.67).Hierarchical clustering analysis of the top ten significantly up/downregulated proteins is shown in Figure 1(B), indicating a distinction between healthy controls and early-onset PE patients.

Bioinformatics analysis and identification of differentially accumulated protein species
To further clarify the functions of the differentially expressed proteins between patients with early -onset PE and control subjects, GO annotation and KEGG pathway enrichment analyses were performed.According to biological process analysis, the most differentially expressed proteins were involved in humoral immune response (Figure 2A).Cellular component enrichment analysis revealed that the largest number of differentially expressed proteins belonged to blood microparticles (Figure 2B).Molecular function analysis revealed that peptidase regulator activity accounted for the most differentially expressed proteins (Figure 2C).Furthermore, KEGG pathway analysis revealed that the differentially expressed proteins were enriched in the complement and coagulation cascades, cholesterol metabolism, and Staphylococcus aureus infection pathways (Figure 2D).

Verification of protein expression in serum
It is noteworthy that CSH1 is produced by the syncytiotrophoblast layer of the placenta during early pregnancy (Baines and Renaud 2017).LPA signalling regulates cytotrophoblast differentiation and syncytiotrophoblast function (Fujii et al. 2019).This suggests that CSH1 and LPA are closely related to the pathogenesis of early-onset PE.Thus, levels of CSH1 and LPA were detected for serum samples from 19 healthy controls and 22 early-onset PE patients by ELISA.The data showed that the expression of CSH1 decreased in the earlyonset PE group (Figure 3A).Serum LPA concentrations were higher in the early-onset PE group than in the control group (Figure 3B).Our ELISA results were consistent with the mass spectrometry results.

Discussion
In the present study, we explored the differentially expressed proteins in the serum of patients with early-onset PE and healthy controls using label-free quantitative mass spectrometry.Compared with serum samples from healthy pregnant women, those from women with early-onset PE displayed 64 proteins that were significantly differentially expressed.
Pathway analysis revealed that PE was associated with several key signalling pathways, such as complement and coagulation cascades, as well as cholesterol metabolism pathways.These findings could potentially be used for diagnostic and therapeutic purposes in early-onset PE.
Researchers have made great efforts to elucidate the molecular mechanisms underlying PE pathogenesis using proteomics-level analysis of umbilical cord tissues, maternal urine, placental tissues, and blood serum or plasma samples (Conrad et al. 2022, He et al. 2022, Navajas et al. 2022, Shangguan et al. 2021).Consistent with previous studies, our study showed an abnormal expression of pregnancy-associated plasma protein A (PAPPA) and sex hormone-binding globulin (SHBG).PAPPA encodes a secreted metalloproteinase that cleaves the insulin-like growth factor-binding protein (IGFBPs) (Lawrence et al. 1999).PAPPA is associated with placenta disease and has been assessed as a potential biomarker in the prediction or detection of PE (Yliniemi et al. 2015, Sifakis et al. 2018).Most studies have shown that decreased PAPPA levels during the first trimester of pregnancy are associated with an increased risk of PE (Kirkegaard et al. 2010).However, PAPPA is not PE-specific and needs to be combined with Doppler ultrasonography and other biochemical markers to obtain high sensitivity for PE screening (Anderson et al. 2012).Human SHBG is a circulating glycoprotein produced by the liver that binds to circulating oestradiol and testosterone (Knochenhauer et al. 1998).SHBG participates in the regulation of steroid responses and mediates the balance between inactive bound sex hormones and biologically active free sex hormones.Wolf et al. found that first trimester SHBG levels may be a candidate biomarker for PE, especially in lean women who would otherwise be perceived to be at low risk (Wolf et al. 2002).It is meaningful to investigate protein profile of the placenta cells, which could reflect the physiological or pathologic status of placenta development.Rolfo A and college have characterised cytokine expression profiles of placentaderived mesenchymal stromal cells (PDMSCs) isolated from normal and PE chorionic villous tissue through cytokine array (Rolfo et al. 2013).This meaningful study depicts chorionic mesenchymal stromal cells as central players in placental physiopathology.However, no overlap in differentially expression proteins was observed between our current results and their findings.Their results differ from our data most likely because we analysed protein isolated from serum instead of PDMSCs.We think our study is helpful for the disease diagnosis in the non-invasive way.
The protein encoded by CSH1 is a member of the somatotropin/prolactin family of hormones that plays a key role in growth control.CSH1 is expressed mainly in the placenta, and mutations in this gene result in placental lactogen deficiency and Silver-Russell syndrome, suggesting a special role in placental development (Hill et al. 1988, Eggermann et al. 1998).In addition, M€ annik et al. revealed significantly reduced mRNA expression of the CSH1 gene cluster in PE placentas (Mannik et al. 2012).With borderline hypertension before conception, the BPH/5 mouse strain develops PE during pregnancy and has been identified as a PE mouse.The mRNA expression levels of CSH1 were lower in BPH/5 mouse placentas than in control mice during the gestation period (Dokras et al. 2006).Our study showed that patients with early-onset PE had lower serum CSH1 concentrations than healthy controls did.Current evidence indicates that CSH1 may be a potential biomarker, in combination with other well-known biomarkers, for early-onset PE assessment.
LPA is a serine proteinase that inhibits the activity of tissue-type plasminogen activator I. Signalling associated with LPA has been reported to stimulate the growth of endothelial cells, fibroblasts, vascular smooth muscle cells, as well as keratinocytes (Sturm et al. 1999).It has been suggested that LPA signalling may play a pathological role during pregnancy.As a receptor of LPA, LPAR3 levels were increased in the placentas of patients with gestational hypertension and PE at both the mRNA and protein levels (Fujii et al. 2019).In addition, LPA can potentiate endothelin-1-induced vasoconstriction (Yakubu andLeffler 2000).Our results showed that serum levels of LPA were higher in patients with early-onset PE than in healthy pregnant control subjects.Our data and previous evidence suggest that LPA might contribute to the development of early-onset PE.
An increasing number of studies have shown that abnormal expression in the complement system is associated with pregnancy-related disorders.HELLP (Hemolysis, elevated liver enzymes and low platelet count) syndrome, a rare pregnancy complication, is thought to be a variant of PE.Bazzan M et al.' study has confirmed a marked activation of complement in patients with HELLP.In addition, the complementrelated gene variants for HELLP cohort showed significantly higher frequencies compared to controls (Bazzan et al. 2020).It would be valuable to investigate possible phenotypegenotype correlations in control and early-onset group for our screened differentially expression proteins-coding genes in our future studies.The results of our study are comparable with those of previous studies.Moreover, our results are in agreement with those of Yu et al., who found that acetylated proteins could affect complement and coagulation cascade activation pathways, which are closely linked to early-onset severe PE (Youssef et al. 2021, Shangguan et al. 2021).It is well established that PE is associated with disturbed maternal cholesterol metabolism (Lee et al. 2019).Cholesterol metabolism was an interesting finding in our study.Lee et al.'s research suggested increased cholesterol biosynthesis and accumulation of cholesterol in the maternal blood of PE patients.
Nonetheless, there were certain limitations to consider in our current study.Our study is limited by its small sample size.Although the results seem persuasive, our findings need to be confirmed in larger prospective studies.In addition, the detailed mechanisms should be explored in future studies to enrich our understanding of early-onset PE occurrence.

Conclusion
In the present study, label-free quantitative proteomics and systematic analysis revealed the critical features of serum proteins between women with early-PE patients and healthy controls.GO and KEGG analyses demonstrated that differentially expressed proteins connected with signalling pathways of immune response and regulation of peptidase activity.In addition, we have identified LPA and CSH1 may act as potential biomarkers in diagnosis and therapeutic of early-PE disease.

Figure 1 .
Figure 1.Proteome profiling in serum from healthy controls and PE patients.

Figure 2 .
Figure 2. Function enrichment analysis of between two groups.

Figure 3 .
Figure 3. Protein level of CSH1 and LPA in serum.

Table 1 .
Characteristics of the PE patients and healthy controls.