Effects of SARS-COV-2 on molecules involved in vascularization and autophagy in placenta tissues

Clinical data

Study participants were divided into three groups: women negative for SARS-CoV-2 since their samples derived from pregnancy occurred before the virus spread (Controls); pregnant women declared positive at the molecular swab (RT-PCR) for SARS-CoV-2 at the time of delivery (SARS-CoV-2 PCR+); pregnant women who contracted the infection during pregnancy but were RT-PCR negative at the time of delivery (SARS-CoV-2 PCR-). The period of negativity ranged from 199 to 41 days before delivery with a mean of 88.4 ± 71.0 days.

No significant differences were reported between subjects for age, body mass index (BMI) and gestational age at delivery, as shown in the clinical data table (Table 1). The group of SARS-CoV-2 PCR- parturients included one woman with pre-pregnancy obesity. Two women, the first in SARS-CoV-2 PCR- group and the second in the control group, had type 2 diabetes.

Besides the aforementioned conditions, the enrolled women had no comorbidities or preexisting diseases such as hypertension or heart problems, nor did they have pregnancy-related diseases such as preeclampsia.

The hospitalization period for mothers with previous infection was 2.6 ± 0.9 days, shorter than that of the positive mothers at delivery who were hospitalized for 6.4 ± 1.5 days. Regarding COVID-19 complications, one woman of SARS-CoV-2 PCR- group reported fever (2 months before delivery) and respiratory symptoms; in the SARS-CoV-2 PCR + group, a patient with diabetes developed fever and dyspnea that necessitated the use of oxygen therapy.

Biochemical data showed no significant differences in D-Dimer (indicator of clotting risk) and aPTT (prothrombin partial time indicating coagulation rate) values between the groups and these parameters were within the normal range. In contrast, regarding hemoglobin and fibrinogen parameters, the two groups with previous or ongoing SARS-CoV-2 infection showed statistically significant higher levels when compared with the control group.

Table 1 Demographics and clinical data. Controls, samples were collected from pregnancies in pre-COVID-19 era; SARS-CoV-2 PCR+, samples collected from subjects with SARS-CoV-2 infection and PCR positive at delivery; SARS-CoV-2 PCR-, samples collected from subjects with SARS-CoV-2 infection during pregnancy but with negative PCR at delivery. BMI, body Mass Index; aPTT, activated partial Thromboplastin Time. Data are expressed as mean ± standard deviation. Values in bold are those statistically significant with p-value < 0.05Expression of SPIKE, ACE2 and CD147

SPIKE protein expression was detected in decidua and chorionic villi of all samples of the SARS-CoV-2 PCR + women group, whereas it was absent in the placenta of control and previously infected group (Fig. 1).

The SPIKE staining was well outlined between the villi, especially at the level of the syncytiotrophoblast (Fig. 1a). Furthermore, different areas with a more relevant or, at variance, a weaker staining were observable, therefore suggesting a staining that follows viral protein quantity. The decidua was characterized by SPIKE islets of staining, mainly located toward its external surface (Fig. 1b).

Analysis of H-score calculated from IHC images did not report differences in SPIKE staining intensity between villi and decidua in corresponding samples (Fig. 1c).

ACE2 protein was expressed in the chorionic villi, along the surface of syncytiotrophoblast for all the samples, but the expression was increased in the SARS-CoV-2 PCR + as shown in Fig. 1d.

On the contrary in the decidua, ACE2 expression was slightly higher in SARS-CoV-2 PCR + versus control and SARS-CoV-2 PCR- while no differences were detected between control and SARS-CoV-2 PCR- groups (Fig. 1b and d; Table 2).

The transmembrane protein CD147 was present at basal levels in controls and was increased in SARS-CoV-2 PCR + samples, while its expression was lower in SARS-CoV-2 PCR- placentas (Fig. 1a, b and e). In the villi, the labeling is shown tagging along the outer edge of the syncytiotrophoblast laye; while in the decidua, the staining appeared distributed inside the tissue (Fig. 1a, b).

Rho Spearman correlation coefficient was assessed by considering all the three clinical groups and the scores were obtained for each protein (SPIKE and ACE2 or CD147) at the level of the villi and decidua for each woman (Supplementary Fig. 1a, b).

The expression of SPIKE was positively associated to the presence of ACE2 in the placenta tissue (Supplementary Fig. 1a). In contrast, there was a less significant correlation between the expression of SPIKE and CD147 as shown by the Rho- and p-value (Supplementary Fig. 1b).

Fig. 1figure 1

Detection of SPIKE, ACE2, and CD147 in control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- placenta. a: IHC images of the chorionic villi; b: IHC staining of the decidua. Magnifications are 20x. c, d, e: H-score values of SPIKE, ACE2, and CD147, respectively. p-values were obtained by Mann-Whitney test. *0.05 < p-value < 0.01; ** 0.01 < p-value < 0.001

SARS-CoV-2 and the expression of CD34 and VEGF in the placenta

IHC analysis of placenta samples with anti-VEGF and anti-CD34 antibodies was also performed. The purpose was to analyze whether SARS-CoV-2 could influence the vascular organization of the tissue, using CD34 and VEGF as readouts. IHC analysis showed that CD34 staining was localized at the level of the endothelium, nearby the syncytiotrophoblast (Fig. 2a). In the decidua, CD34 was also appreciable on the surface of vascular endothelium, near the vessels of maternal circulation (Fig. 2b). Moreover, the expression of this marker appeared more intense during SARS-CoV-2 infection (SARS-CoV-2 PCR+), even if the increase did not reach the significatively comparing with the other two groups (p = 0.056 versus control and p = 0.095 versus SARS-CoV-2 PCR- in villi and p = 0.095 versus control and p = 0.095 versus SARS-CoV-2 PCR- in decidua) (Fig. 2a, b; Table 2).

A statistically significant difference was observed for the distribution of this marker between decidua and villi with high levels in the last one (Supplementary Table 1).

Regarding VEGF, its expression was lower in control samples and SARS-CoV-2 PCR- groups when compared with SARS-CoV-2 PCR + samples, both in the villi and decidua (Fig. 2d; Table 2). In the villi, VEGF was expressed at the level of syncytiotrophoblast and in the vascular endothelium (Fig. 2a) while in the decidua the expression was most detectable in the vascular endothelium lining the blood vessels (Fig. 2b).

Rho Spearman test revealed a significant positive correlation between CD34 and VEGF (Supplementary Fig. 2a) and between SPIKE and VEGF (Supplementary Fig. 2b).

Analysis of H-score calculated from IHC images confirmed that CD34, detectable also in control samples, was more appreciable during infection, while its expression was lower in SARS-CoV-2 PCR- samples. Similarly, VEGF expression was higher during infection, while it was less appreciable in control samples and past infection (Fig. 2c, d).

Fig. 2figure 2

Expression of CD34 and VEGF in control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- placenta. a: IHC images of the chorionic villi; b: IHC staining of the decidua. Magnifications are 20x. c, d: H-score values of CD34 and VEGF respectively. p-values were obtained by Mann-Whitney test. ** 0.01 < p-value < 0.001. The arrows indicate the staining localization of CD34 and VEGF nearby the syncytiotrophoblast in the villi and on the surface of vascular endothelium in the decidua

Table 2 Analysis of the expression of the markers in the villi or decidua between control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- groups. p-value was calculated using Kruskal-Wallis or Mann-Whitney test to compare marker expression, estimated as H-score, between all the three or the two specified groups, respectively. vs.: versus. Values in bold are those statistically significant with p-value < 0.05SARS-CoV-2 induced autophagy

With the aim of analyzing the possible modulation of autophagy in our samples, the expression of LC3B protein was assessed. IHC analysis showed that in the normal placenta after delivery, LC3B was expressed either in the villi and in the decidua (Fig. 3a, b). In SARS-CoV-2 PCR + villi and decidua, an increase in both the spread and staining intensity was observed when compared with control and SARS-CoV-2 PCR- samples (Fig. 3; Table 2). Indeed, the H-score analysis showed a relevant increase of the PCR + samples (H-score 17–18) in comparison with the levels of control and PCR- ones. Surprisingly, LC3B expression was higher in the control than in the SARS-CoV-2 PCR- group in villi (Fig. 3; Table 2). The activation of autophagy during ongoing infection was also confirmed by the positive correlation between SPIKE and LC3B expression with the Rho Spearman correlation test (Supplementary Fig. 3).

Fig. 3figure 3

Expression of LC3B in control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- placenta. a: IHC images of chorionic villi; b: IHC staining of decidua. Magnifications are 20x. c: H-score values of LC3B. p-values were obtained by Mann-Whitney test. The arrows indicate the staining localization of LC3B in the villi, more widespread in the decidua

Analysis of tissue vascular and autophagy readouts

To first analyze the co-distribution of SPIKE, CD34 and VEGF, MICSSS was performed in placenta samples at the different conditions (Fig. 4).

Similar to previous results, CD34 characterized the areas of vascular endothelium in all the three conditions, with greater expression in villi than in decidua. In particular, CD34 demarcated the capillaries in the inner part, while the more intense and thicker marking precisely indicated the surface of the vascular endothelial cells that constitute the chorionic villi. Notably, in control placenta, CD34 represented essentially the only visible staining, as there was no detection of SARS-CoV-2 and only a slight expression of VEGF. Likewise, SPIKE was not detected in the SARS-CoV-2 PCR- condition, thus indicating the absence of the virus.

The appreciable aspect of multiple staining is the co-expression, and in this case, it was most strongly expressed in the positive condition. Indeed, the strong presence of SPIKE was distributed in villi and decidua areas characterized by a high expression of VEGF and CD34 (Fig. 4).

To further investigate the correlation of SARS-CoV-2 with the autophagic process, MICSSS with SPIKE protein and LC3B staining was also performed to detect whether the presence of the SPIKE protein was associated with increased autophagy (Fig. 5).

In both villi and decidua areas of SARS-CoV-2 PCR + placenta samples, the infection appeared very extensive as SPIKE staining was distributed throughout the tissue and LC3B is distributed in the proximity of the viral protein staining, both at the level of the endothelial cells that characterize the villi and in the extravillous trophoblast in the decidua (Fig. 5).

Fig. 4figure 4

Analysis of co-expression of SPIKE, CD34 and VEGF in control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- groups in villi and decidua. The expression of the three markers SPIKE, CD34 and VEGF was co-localized. In detail, SPIKE protein was assigned the color cyan, VEGF the color red, and CD34 the color magenta. H-E: Hematoxilin Eosin. Magnifications are 20x. The arrows indicate points of co-localization of the three markers in the villi

Fig. 5figure 5

Analysis of co-expression of SPIKE and LC3B. Comparison of MICSSS of control, SARS-CoV-2 PCR + and SARS-CoV-2 PCR- conditions showing the co-localization of SPIKE and LC3B in the villi and decidua sections. With overlapping images, SPIKE protein was assigned the color cyan and LC3B the color magenta. Magnifications are 20x. The arrows indicate points of co-localization of the two markers in the villi and in the decidua

Comments (0)

No login
gif