Plasma Ceramide Concentrations in Full-Term Pregnancies Complicated with Gestational Diabetes Mellitus: A Case-Control Study

Ceramides, a sphingolipid group that acts as a messenger in cellular differentiation, proliferation, apoptosis and senescence, have been associated with cardiovascular disease and type 2 diabetes. The evidence for an association between ceramides and gestational diabetes mellitus (GDM) is scarce. This case-control study aimed to compare women with GDM with healthy, pregnant women in terms of plasma ceramide concentrations at the time of delivery. Ninety-two pregnant women were included in this case-control study, 29 in the GDM group and 63 in the control group. All women were admitted to a tertiary academic hospital for a full-term delivery. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) was applied for the quantification of four molecular ceramides, namely Cer d18:1/16:0 (Cer16:0), Cer d18:1/18:0 (Cer18:0), Cer d18:1/24:0 (Cer24:0) and Cer d18:1/24:1 (Cer24:1) in plasma samples. The raw chromatographic data obtained from the LC-MS/MS analysis were processed using Analyst SCIEX (AB Sciex Pte. Ltd., USA). In a univariate statistical analysis, Cer24:0 concentration was significantly lower in the GDM group compared with the control group (p = 0.01). The present study demonstrated lower Cer24:0 concentrations in pregnancies complicated by GDM. Further prospective studies are required to enhance the results of this study.


Introduction
Gestational diabetes mellitus (GDM) complicates 7.8% (7.2-8.4%) of pregnancies in Europe and can affect the fetus, the mother, and the pregnancy outcome [1]. It is defined as glucose intolerance diagnosed for the first time during pregnancy. There is an increase in the GDM prevalence due to the increased rate of obesity in western countries and the older age of pregnant women [2,3]. Risk factors for GDM include race, age, body mass index (BMI), history of GDM, family history of diabetes mellitus and polycystic ovary syndrome (PCOS) [4]. The main GDM complications are hypertension, preeclampsia, fetal macrosomia, shoulder dystocia and fetal death. Newborns can suffer from hypoglycemia, respiratory distress syndrome and hyperbilirubinemia [5].
Ceramides, a sphingolipid group, act as messengers in cellular differentiation, proliferation, apoptosis and senescence [6]. They are located in the external layer of the Metabolites 2022, 12, 1123 2 of 10 plasma membrane and are messengers of the sphingomyelin transmembrane signaling pathway. Apoptosis and differentiation inducers, damaging agents and inflammatory cytokines can increase the production of ceramides, and thus the plasma concentrations of their free forms [7]. Plasma ceramides are biomarkers of cardiovascular disease, cancer, Alzheimer's disease and type 2 diabetes mellitus (T2DM) [8][9][10]. Elevated plasma ceramide concentrations have been associated with preeclampsia and preterm delivery [11,12].
The evidence on an association between ceramides and GDM is scarce. Placental ceramide concentrations in women with GDM treated with insulin were higher compared with placentas of uncomplicated pregnancies or placentas of women with GDM treated with diet [13]. There are no publications concerning plasma ceramide concentrations in full-term pregnancies complicated by GDM.
Our study hypothesis is that plasma ceramides in pregnant women with GDM should be elevated, because of the similar mechanisms of GDM with T2DM.
To the best of our knowledge, this is the first study that attempts to answer this research question

Study Characteristics
This is a case-control study conducted in the outpatient clinic of a tertiary academic hospital.

Patients
Ninety-two pregnant women participated in the study. All of these women were recruited between August 2020 and August 2021. The patients were enrolled consecutively. Of them, 29 were diagnosed with GDM during their pregnancy (GDM group), and 63 had uncomplicated pregnancies (control group). The latter were matched to the former for gestational age. All women were admitted to our department to deliver as term pregnancies (>37 gestational weeks). The diagnosis of GDM was set according to the International Association of the Diabetes and Pregnancy Study Groups (IADPSG) criteria [14]. Exclusion criteria for both groups were preterm delivery (<37 gestational weeks), type 1 (T1DM) or T2DM, and personal history of chronic diseases and obstetrical complications, such as preeclampsia, IUGR, hypertension or chorioamnionitis. All women were of Caucasian origin.

Methods
All patients provided written informed consent before entering the study. After their admission, a personal, obstetrical, and family history was taken. A blood sample was obtained and centrifuged for 15 min. The plasma used for measuring ceramides was removed and stored at −80 • C. For every woman, the following parameters were recorded: age, gestational age, parity, smoking, body mass index (BMI), gestational weight gain (GWG), oral glucose tolerance test (OGTT) values, delivery type, neonatal birth weight, Apgar score, blood pressure, creatinine, urea, aspartate aminotransferase (SGOT), alanine aminotransferase (SGPT), uric acid, platelets, and presence/absence of protein in urine samples. All parameters were measured at the academic hospital's laboratories, using established techniques.

Sample Preparation
Plasma samples were left to thaw at room temperature before sample preparation. A volume of 100 µL of plasma sample was diluted with 20 µL of IS (mix of isotope-labeled ceramides). The sample was vortex mixed for 2 min, and 1 mL of CH 2 Cl 2 :MeOH, 2:1 v/v was added. The sample was vortex mixed for 5 min and centrifugated at 6700× g for 15 min at 4 • C. A volume of 900 µL of the lower organic phase was evaporated to dryness under vacuum, and the dry residue was reconstituted with 50 µL of IPA:MeOH, 1:1 v/v. The final extract was subjected to LC-MS/MS analysis.

LC-MS/MS Analysis
All samples were analyzed using a previously developed and validated Liquid Chromatography-tandem Mass Spectrometry (LC-MS/MS) method for the simultaneous determination and quantification of the four ceramides, as described by Begou et al. [15]. Briefly, chromatography was performed by an Alliance HT Waters 2790 (Waters Corporation, Milford, MA, USA) system, on a ReproShell ODS-3 (2 mm × 50 mm, 2.7 µm) (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany) column under reversed phase conditions. Intra-day accuracy and precision of the method were between 80-111% and 0.05-10.2% for all ceramides, respectively, while inter-day accuracy and precision ranged from 87.8-106% to 2.2-14%, respectively. The lower limit of quantification (LOQ) was 2.3 ng/mL for Cer16:0 and Cer18:0 and 1.4 ng/mL for Cer24:0 and Cer24:1. To assess the analytical batch precision, a quality control (QC) sample was used, and prepared by mixing equal volumes of all the analyzed samples (pooled plasma sample). The QC sample was analyzed six times at the beginning of the analytical run and then every 10 real samples.

Statistical Analysis
Sample size estimation. No data exists on plasma ceramide concentrations at term in pregnancies complicated by GDM. As the most similar pathology to GDM reporting on ceramide concentrations is preeclampsia, a relevant study was used to define the sample size [11]. In patients with preeclampsia, the Cer24:0 concentration was 2.588 ± 0.574 nmol/mL (0.168 ± 0.037 mg/dL) (mean ± standard deviation) and 4.477 ± 1.986 nmol/mL (0.291 ± 0.129 mg/dL) in uncomplicated pregnancies. With a type 1 error of 5% and a type 2 error of 10%, a sample of 28 patients (14 GDM cases and 14 controls) was calculated as appropriate. Nevertheless, as this was a rough estimation of the expected differences between the groups, it was decided that larger sample sizes were to be recruited in both groups: 29 in the GDM group and 63 in the control group.

Statistical Methods
Quantitative results of ceramides were evaluated using the statistical software Graph-Pad Prism version 7.00 for Windows (GraphPad Software, La Jolla California USA, www. graphpad.com, accessed on 30 January 2022). Data were assessed for normal distribution based on D'Agostino-Pearson (95% de) statistical test. A two-tailed t-test (Mann-Whitney test) with unequal variance and a threshold of p-value < 0.05 was performed. Area under the receiver operating characteristic (AUC-ROC) curves and box plots were constructed for the four ceramides. The correlation of all analytes and patients' personal history, BMI, GWG and other laboratory parameters was assessed using the Spearman correlation.

Results
A total of 92 women were included in the study, 29 in the GDM group and 63 in the control group. The demographic characteristics of the study participants are shown in Table 1. The mean maternal age in the GDM group was 33.0 ± 4.5 and in the control group 30.0 ± 6.6 years (p = 0.029). Cesarian section was more likely in women with GDM than in controls (58.6% vs. 47.6%, respectively, p = 0.327). Results are given as mean ± standard deviation or as number (percentage). BMI: body mass index; p-values marked with bold indicate statistically significant differences between the groups GDM: gestational diabetes mellitus.
The percentage of relative standard deviation (%RSD) of all ceramides in the QC sample was <10%, verifying a satisfactory system precision.

Discussion
This is the first study to quantify plasma ceramide concentrations in women with GDM and women with uncomplicated pregnancies at delivery. The study revealed that Cer24:0 concentrations were significantly decreased in women with GDM, while there were no differences in Cer16:0, Cer18:0, and Cer24:1 between GDM and control groups. These results are not the expected ones based on our initial hypothesis about the similar mechanisms of T2DM and GDM and elevated ceramides as biomarkers in patients with T2DM. On the other hand, our results partially agree with the results of three recent publications about the association of ceramides and GDM in early pregnancy [16,17]. More research is necessary to determine the mechanisms that affect circulating ceramides in pregnant women.
According to Liu et al. who included 486 women, in early pregnancy, high Cer18:0 and Cer18:1 and low Cer24:0 (≤3.60 nmol/mL) have been associated with higher GDM risk [odds ratio (OR) 1.69, 1.72 and 3.59, respectively] [16]. In another large, nested case-control study, including 1008 women, in which the lipidome of first-trimester pregnant women was studied, Cer24:0 and Cer24:1 were inversely associated with GDM [17]. Another recent study, including 135 women, which compared the ceramide concentrations in the first and the second trimester between pregnant women with normal glucose tolerance and women with GDM reported a significantly increased concentration of Cer18:1 in the second trimester in the GDM group [18]. These findings partially agree but are not enough to establish ceramides as clinical biomarkers in early pregnancy.
According to a point of view, GDM not only has similar risk factors but also similar molecular mechanisms with T2DM [19]. Maybe the role of ceramides in T2DM can enlighten the role of ceramides in GDM.
Several mechanisms have been proposed for the role of ceramides in diabetes mellitus (DM). The key to ceramides' biological action may be their double biosynthesis: from hydrolysis of sphingomyelin by sphingomyelinases or de novo. There is a strong link between ceramide accumulation and insulin resistance. In rats, sphingolipids inhibit glucose transport into 3T3-L1 adipose fibroblasts. Ceramides reduce the insulin-induced tyrosine phosphorylation of insulin receptor substrate-1 (IRS-1). Lipotoxicity, defined as increased muscle, liver and plasma fat, plays a key role in the pathogenesis of T2DM. Ceramides, and other bioactive lipids, are related to lipotoxicity and inflammation [20]. Apoptosis and differentiation inducers, damaging agents and inflammatory cytokines increase ceramide production [7]. In addition, ceramides are the second major messengers of the inflammatory response induced by tumor necrosis factor-α (TNFα). Finally, increased Rac activation has been reported, and as a result, reduced glucose transferase-4 (GLUT-4) translocation to the plasma membrane in response to insulin [21].
Ceramide concentrations are elevated in the human adipose tissue of obese men and women compared to lean individuals [22]. A strong positive correlation was found between ceramide and TNFα concentrations in adipose tissue and a negative correlation between ceramide concentrations and adiponectin [23]. The Dallas Heart Study was the first large study (n = 1557) to reveal a positive correlation between shorter-chain saturated fatty acid ceramides (Cer18:0, Cer16:0, Cer20:0) with unfavorable adiposity, and lipid and insulin resistance; nevertheless, plasma ceramides were not independently associated with T2DM after adjustment for clinical factors [9]. A meta-analysis of 2337 participants from the Strong Heart Study and the Strong Heart Family Study showed higher plasma ceramide concentrations (Cer18:0, Cer20:0, Cer22:0) in patients who developed T2DM later. Possible mechanisms include ceramides inhibiting insulin intermediates such as insulin receptor substrate, protein kinase B (Akt) and GLUT-4, promoting β-cell apoptosis and dysfunction [24]. According to the same study, elevated plasma ceramides were correlated with a higher homeostatic model assessment of insulin resistance (HOMA-IR), a proxy of insulin resistance [25]. Based on these findings, ceramides are considered a promising biomarker for T2DM and a possible therapeutic target [26].
Recently, ceramide concentrations have been studied in several pregnancy complications, such as preterm labor, preeclampsia, hemolysis, elevated liver enzymes and low platelets (HELLP) syndrome, intrahepatic cholestasis, and even genetic disorders such as trisomy 21 [27]. Specifically, ceramide concentrations were higher in women with HELLP syndrome and a correlation was found with proteinuria [28]. The role of ceramides in preeclampsia is not yet clarified as there are conflicting results in the literature [11]. Maybe this is because, in the placental blood vessels of pregnancies complicated with preeclampsia, sphingolipid biosynthesis is shifted toward sphingomyelin rather than ceramides [29]. This molecular mechanism and the complex way of ceramide biosynthesis may explain the findings of the current studies. Furthermore, long-chain ceramides, such as Cer24, may have a different, benign, protective role in the vessels, contrary to Cer16 and Cer18 which contain C16 and C18 acyl chains [30]. Maybe the key to using ceramides as biomarkers is in creating ceramide scores, such as Cer18/Cer24 and a more individualized approach taking into consideration age, diet and BMI [31]. Further studies are needed to define the role of plasma ceramides in both uncomplicated pregnancies and pregnancies complicated by GDM.
Other studies provide possible mechanisms that can result in the elevation of ceramide production in preeclampsia [32]. Cer16:0 was found to increase in women with preterm labor [33]. Finally, plasma ceramides were found to increase in pregnant and lactating women compared to non-pregnant women; this finding supports the hypothesis that ceramides contribute to normal insulin resistance during pregnancy [34].
Even though a sample size estimation had been performed, we decided to include a larger number of participants in the study. Considering the small differences in the ceramide concentrations found between the two groups, the larger sample sizes proved to be essential. Although this is one of the first studies to suggest an association between ceramides and GDM, some limitations should be mentioned. The first is that ceramide concentrations were measured only once at delivery; no measurements during the first and second trimesters were available. Thus, the change in concentrations during pregnancy progression cannot be evaluated. Another limitation was the type of study: a cohort study would be more suitable than a case-control study. Finally, the potential molecular pathways by which ceramides are produced in GDM were not studied.
In conclusion, this study demonstrated a lower Cer24:0 plasma concentration during delivery in women with GDM compared to uncomplicated pregnancies. Our findings seem to partially agree with recent research about ceramides as possible biomarkers for GDM in early pregnancy. It has not yet been clarified whether ceramides are involved in metabolic disease pathways, or whether their abnormally increased biosynthesis leads to adverse effects. Further research is needed to establish a role for ceramides in everyday clinical practice.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/metabo12111123/s1, Table S1: Correlation among ceramide concentrations and studied parameters for the GDM group; Table S2: Correlation among ceramide concentrations and studied parameters for the control group. Funding: This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of the Aristotle University of Thessaloniki, Greece (protocol code 207/13-7-20).