An observational, analytical and prospective study was performed and included 114 women in the third trimester of pregnancy (32 weeks). From these, 62 women were clinically diagnosed with CVD according to CEAP classification (20), with a median age of 33 years [interquartile range (IQR), 22–40 years] and a median gestational age of 40.5 weeks (IQR, 39–41.5 weeks). Simultaneously, 52 controls without a history of CVD [healthy controls (HC)] were also included, with a median age of 34 years (IQR, 27–41 years) and a median gestational age of 41 weeks (IQR, 39–42 weeks). The present study was conducted according to the basic ethical principles of autonomy, beneficence, non-maleficence and distributive justice. The development of the research followed the regulations of Good Clinical Practice, as well as the principles set forth in the last Declaration of Helsinki (2013) and the Oviedo Convention (1997). Patients were informed prior to enrolment, and each participant provided their corresponding written consent. The current study was approved by the Clinical Research Ethics Committee of the Central University Hospital of Defence-University of Alcalá (37/17). During the third trimester consultation, the clinical history was recorded, a general physical examination was performed and lower limb ultrasounds were conducted using an Eco-Doppler (Portable M-Turbo Eco-Doppler; SonoSite, Inc.) at 7.5 MHz.

The inclusion criteria were defined as women over 18 years of age, with clinical evidence of lower limb venous insufficiency (VI) in the third trimester, according to Clinical-Etiology-Anatomy-Pathophysiology classification (≥1) (20). The exclusion criteria included women previously diagnosed with diabetes mellitus, gestational diabetes mellitus or other endocrine diseases; high blood pressure; autoimmune diseases; active infectious diseases; venous malformations; heart, kidney and lung insufficiency; preeclampsia and/or HELLP [an acronym for hemolysis (H), elevated liver enzymes (EL) and a low platelet count (LP)] syndrome; known causes of intrauterine growth restrictions; body mass index ≥25; toxicological habits [tobacco (≥1 cigarette a day), alcohol (≥1 unit a day) or drugs (e.g., cannabis, heroin, cocaine, amphetamines)]; existence of pathological injuries, such as placental infarction, avascular villi, delayed villi maturation or chronic villitis; as well as the appearance of any exclusion criteria in the following months (until delivery); and previous evidence of CVD. There were no significant differences between the groups regarding the number of previous pregnancies: 33 (53.2%) for women with CVD and 19 (36.5%) for women in the HC group (Table SI). There were also no significant differences in the clinical characteristics between the CVD and HC groups (gestational age, c-section delivery, previous pregnancies, previous abortions, regular menstrual cycles and type of profession-sedentary, Table SI).

Placental biopsies were collected once they were expelled after delivery. In all cases, 5 placental fragments were obtained in all cases using a scalpel to ensure that the samples included various cotyledons. These placental pieces were added to two different sterile tubes: One containing Minimum Essential Medium (MEM; Thermo Fisher Scientific, Inc.) with 1% antibiotic/antimycotic (Streptomycin, Amphotericin B and Penicillin) (Thermo Fisher Scientific, Inc.) and another with RNAlater^® (Ambion; Thermo Fisher Scientific, Inc.) solution. The samples were processed in a class II laminar flow hood (Telstar AV 30/70 Müller 220 V 50 MHz; Telstar; Azbil Corporation) in a sterile environment. Preserved samples were stored in 1 ml RNAlater^® at −80°C until they were processed for gene expression analysis. Conserved MEM placentas were used for histological and immunodetection studies.

The samples stored in MEM were washed and rehydrated five times in MEM without antibiotics to remove the blood cells, then they were cut into fragments (2 cm) and fixed in F13 (60% ethanol, 20% methanol, 7% polyethylene glycol and 13% distilled water) following established protocols (20). The samples were then paraffin-embedded in blocks using moulds. After the paraffin had solidified, a HM 350 S rotation microtome (Thermo Fisher Scientific, Inc.) was used to obtain 5-µm thick sections, which were stretched in a hot water bath, then mounted on glass slides, previously treated with 10% polylysine, allowing for improved adhesion of the sections.

RNA was extracted according to the guanidinium thiocyanate-phenol-chloroform method (21,22) and was used to analyze the mRNA expression levels of the genes of interest.

RNA samples at a concentration of 50 ng/µl were used to synthesize complementary DNA (cDNA) by reverse transcription; 4 µl of each sample is mixed with 4 µl of oligo-dT (15) 0.25 µg/µl solution (Thermo Fisher Scientific, Inc.), and incubated at 65°C for 10 min in a dry bath (AccuBlock, Labnet International Inc.), in order to denature the RNA. After this, the samples were placed on ice and 10 µl per sample of a reverse transcription mix containing the following products was added for each sample: 2.8 µl First Strand Buffer 5X (250 mM Tris-HCl and pH 8.3; 375 mM KCl; 15 mM MgCl₂) (Thermo Fisher Scientific, Inc.); 2 µl of 10 mM deoxyribonucleotides triphosphate; 2 µl of 0.1 M dithiothreitol; 1.7 µl of DNase- and RNase-free water; 0.5 µl of RNase inhibitor (RNase Out); 1 µl of reverse transcriptase enzyme (all from Thermo Fisher Scientific, Inc.).

The RT process was carried out using a G-Storm GS1 thermal cycler (G-Storm Ltd.). The samples were incubated at 37°C for 1 h and 15 min, to allow cDNA synthesis. The temperature was then increased to 70°C and maintained for 15 min, thus causing the denaturation of the reverse transcriptase enzyme, and the temperature gradually decreased to 4°C.

To verify the absence of genomic DNA contamination in the total RNA samples, a negative reverse transcription was performed in parallel in which the M-MLV RT enzyme is replaced by water free of DNases and RNases. The cDNA produced in RT was diluted 1:20 using water free of DNases and RNases and stored at −20°C until use.

Specific primers for all the genes studied (Table SII) were designed de novo using the Primer-BLAST and AutoDimer online applications (23,24). The constitutively expressed TATA-box binding protein (TBP) gene was used to as a control to normalize the results (25). The gene expression units are expressed as relative quantities of mRNA. RT-qPCR was performed on a StepOnePlus™ System (Applied Biosystems; Thermo Fisher Scientific, Inc.) and the relative standard curve method was used. The total reaction volume was 20 µl and included: 5 µl sample [mixed at 1:20 with 10 µl iQ™ SYBR^® Green Supermix (Bio-Rad Laboratories, Inc.)], 1 µl each forward and reverse primers, and 3 µl DNase and RNase-free water, and added to a MicroAmp^® 96-well plate (Applied Biosystems; Thermo Fisher Scientific, Inc.). The following thermocycling conditions were used: Initial denaturation at 95°C for 10 min, denaturation at 95°C for 15 sec, annealing at variable temperatures depending on the melting temperature of each primer pair for 30 sec, and elongation at 72°C for 1 min, for 40–45 cycles. Followed by a dissociation curve at 95°C for 15 sec, 60°C for 1 min, 95°C for 15 sec, and 60°C for 15 sec. Fluorescence detection was performed at the end of each repeat cycle (amplification) and at each step of the dissociation curve. The data obtained from each gene was added in a standard curve made by serial dilutions of a mixture of the samples, that were included in each plate according to the constitutive expression of TBP (in accordance with the manufacturer's protocols). All tests were performed in duplicate in all samples of placenta tissue.

Immunohistochemical studies were performed on paraffin-embedded placental tissue samples. The antibody retrieval step was described in the protocol specifications (Table SIII). The antigen/antibody reactions were detected using the avidin-biotin complex method, with avidin-peroxidase, as previously described (26). After incubation with the primary antibody (1 h and 30 min; Table SIII), the samples were incubated with 3% BSA Blocker (cat. no. 37525; Thermo Fisher Scientific, Inc.) and PBS overnight at 4°C. Then, the cells were incubated with biotin-conjugated secondary antibody, diluted in PBS, for 90 min at room temperature (RT; Table SIII). The avidin-peroxidase conjugate ExtrAvidin^®-Peroxidase (Sigma-Aldrich; Merck KGaA) was used for 60 min at RT (1:200 dilution with PBS), then the protein expression level was determined using a chromogenic diaminobenzidine (DAB) substrate kit (cat. no. SK-4100; Maravai LifeSciences), which was prepared immediately before exposure (5 ml distilled water, two drops buffer, four drops DAB and two drops hydrogen peroxide). The signal was developed with the peroxidase chromogenic substrate for 15 min at RT; this technique allows for the detection of a brown stain. For the detection of each protein, sections of the same tissue were assigned as negative controls, substituting incubation with the primary antibody for a blocking solution (PBS). In all the tissues, the contrast was performed with Carazzi hematoxylin for 15 min at RT.

For each patient within the defined groups, 5 sections and 10 fields of view were randomly examined. The patients were described as positive when the marked mean area in the analyzed sample was ≥5% of the total, following the immunoreactive score (IRS) from Remmele and Schicketanz (27) and Cristóbal et al (28). Immunostaining in the tissue was assessed by two independent histologists, blinded to the outcome. In each sample, immunohistochemical staining was scored using the following scale: 0–1, minimum staining (≤25%); 2, moderate staining (25–65%); and 3–4, strong staining (≥65-100%). Preparations were viewed using a Zeiss Axiophot optical microscope (Zeiss GmbH).

Once the sections were dried, they were deparaffinized for 30 min in xylol at RT (PanReac AppliChem; Illinois Tool Works, Inc.) and subsequently rehydrated using a descending alcohol series until they were completely hydrated in distilled water. After rehydration, the sections of the samples were: i) Stained with alcoholic orcein for 30 min at RT, ii) washed with distilled water for 30 min, iii) immersed in 96% alcohol for 5 min, iv) immersed in 100% alcohol for 15 min, v) discolored on the bottom with acid alcohol for 2–10 min, vi) washed with water for 10 min, vii) stained with Carazzi hematoxylin for 20 min at RT, viii) washed in running water for 10 min, viiii) dehydrated in 96% alcohol for 5 min, x) dehydrated in 100% alcohol for 5 min, xi) submerged in xylol for 10 min, and xii) mounted using Cytoseal™, which allows for the visualization of the elastic fibers with a brown color using an optical microscope (Zeiss GmbH).

For the statistical analysis, the GraphPad Prism^® v6.0 (GraphPad, Inc.) program was used. The Mann-Whitney U test was used to compare the 2 groups, and the data was expressed as the median and the IQR. For the categorical variables, Pearson's χ² or Fisher's exact test was used. P<0.05 was used to indicate a statistically significant difference.

There was a significant decrease in the EGFL7 gene expression level in the placental villi of women with CVD during pregnancy (P=0.0012; Fig. 1A). The results from the protein expression levels showed that the IRS score was significantly lower in the placental villi from the CVD group [CVD, 0.00 (IQR, 0.00–1.25); HC, 1.00 (IQR, 0.00–2.00) P=0.0077; Fig. 1B and D]. Notably, there was a decrease in the percentage of decidual cells with EGFL7 protein expression [CVD, 15.50% (IQR, 7.00–41.00%); HC, 31.00% (IQR, 12.00–82.00%); P=0.0059; Fig. 1C and E]. The arrows show the positive expression in the tissue.

An increase in TE gene expression was observed in the placental villi of women with CVD during pregnancy compared with that in women from the HC group (P=0.0355; Fig. 1F). The analysis of TE protein expression level using immunohistochemistry showed a significant increase in the IRS in patients with CVD, with intense staining throughout the ECM [CVD, 2.50 (IQR, 0.50–3.00); HC, 1.00 (IQR, 0.00–2.500); P=0.0003; Fig. 1G and I]. The percentage of decidual cells with TE protein expression level was significantly higher in the placenta of women with CVD during pregnancy [CVD, 52.00 (IQR, 16.00–96.00%); HC, 22.00 (IQR, 10.00–54.00%); P=0.0005; Fig. 1H and J]. The arrows show the positive expression in the tissue.

The FBLN-4 gene was significantly increased in the placental villi of women with CVD compared with that in women in the HC group (P=0.0456; Fig. 2A). An increase in FBN-1 gene expression was also observed in the CVD group [CVD, 76.26 (IQR, 35.39–259.70); HC, 55.39 (IQR, 8.81–156.02); P=0.0185; Fig. 2D]. The arrows show the positive expression in the tissue.

The protein expression level of FBLN-4 did not differ significantly in the placental villi between the 2 groups [CVD, 0.87 (0.00–2.00); HC, 0.50 (IQR, 0.00–2.00); P=0.40], using the IRS score; however, FBLN-4 protein expression was observed around the large vessels in the placenta of women with CVD during pregnancy (Fig. 2B and C). By contrast, there was a significant increase in the IRS for FBN-1 in the placental villi of women with CVD during pregnancy [CVD, 1.25 (IQR, 0.50–3.00); HC, 1.00 (IQR, 0.00–3.00); P=0.0188; Fig. 2E and G]. The protein expression level of FBN-1 was particularly found around the large vessels in the placenta of patients with CVD (indicated by the arrow; Fig. 2E and F). No protein expression of FBLN-4 or FBN-1 was observed in decidual cells. The arrows show the positive expression in the tissue.

A significant increase in LOX gene expression level was observed in the placental villi of women with CVD during pregnancy [P=0.0344; Fig. 3A]. Similarly, the gene expression level of LOXL-1 was significantly higher in the placental villi from women with CVD [CVD, 96.63 (IQR, 41.99–321.38); HC, 68.26 (IQR, 27.30–247.54); P=0.0390; Fig. 3F).

The results from protein expression showed an increase in the IRS in the placental villi of women with CVD during pregnancy for LOX [CVD, 2.00 (IQR, 0.50–3.00); HC, 1.00 (IQR, 0.00–2.50); P=0.0036] (Fig. 3B and D) and LOXL-1 [CVD, 2.50 (IQR, 2.00–3.00); HC, 1.00 (IQR, 0.00–2.50); P<0.0001] (Fig. 3.G and I). In addition, an increase in the protein expression level of LOX [CVD, 50.50% (IQR, 21.00–85.00%); HC, 21.00 (IQR, 9.00–41.00%); P<0.0001] (Fig. 3C and E) and LOXL-1 [CVD, 18.00% (IQR, 7.00–45.00%); HC, 9.00% (6.00–21.00%), P=0.0021] (Fig. 3H and J) was observed in the placental decidual cells (Fig. 3.C and H) of women with CVD during pregnancy. The arrows show the positive expression in the tissue.

Orcein staining revealed the presence of a significant increase in the elastic fibers in the placenta from women with CVD during pregnancy (P=0.0110; Fig. 4A), particularly in the placental villi (Fig. 4B) and around the decidual cells (Fig. 4C). In the HC group, the staining was not as intense in either the placental villi (Fig. 4D) or in the decidual cells (Fig. 4E).