Molecular evidence suggesting the persistence of residual SARS‐CoV‐2 and immune responses in the placentas of pregnant patients recovered from COVID‐19

Abstract Objectives Recent studies have shown the presence of SARS‐CoV‐2 in the tissues of clinically recovered patients and persistent immune symptoms in discharged patients for up to several months. Pregnant patients were shown to be a high‐risk group for COVID‐19. Based on these findings, we assessed SARS‐CoV‐2 nucleic acid and protein retention in the placentas of pregnant women who had fully recovered from COVID‐19 and cytokine fluctuations in maternal and foetal tissues. Materials and Methods Remnant SARS‐CoV‐2 in the term placenta was detected using nucleic acid amplification and immunohistochemical staining of the SARS‐CoV‐2 protein. The infiltration of CD14+ macrophages into the placental villi was detected by immunostaining. The cytokines in the placenta, maternal plasma, neonatal umbilical cord, cord blood and amniotic fluid specimens at delivery were profiled using the Luminex assay. Results Residual SARS‐CoV‐2 nucleic acid and protein were detected in the term placentas of recovered pregnant women. The infiltration of CD14+ macrophages into the placental villi of the recovered pregnant women was higher than that in the controls. Furthermore, the cytokine levels in the placenta, maternal plasma, neonatal umbilical cord, cord blood and amniotic fluid specimens fluctuated significantly. Conclusions Our study showed that SARS‐CoV‐2 nucleic acid (in one patient) and protein (in five patients) were present in the placentas of clinically recovered pregnant patients for more than 3 months after diagnosis. The immune responses induced by the virus may lead to prolonged and persistent symptoms in the maternal plasma, placenta, umbilical cord, cord blood and amniotic fluid.

The major manifestations of COVID-19 are observed in the respiratory organs. [1][2][3][4] Patients may also show gastrointestinal and neurological symptoms and myocardial dysfunction. [5][6][7][8][9] Most patients develop unilateral or bilateral pneumonia, which is diagnosed by radiological examination. 10 In patients with severe COVID-19, the disease can rapidly progress into acute respiratory distress syndrome (ARDS), severe sepsis with shock or multiple organ failure within 1 week due to the onset of a 'cytokine storm'. 11 SARS-CoV-2 is primarily transmitted through the respiratory tract and infects airway epithelial cells, vascular endothelial cells and macrophages. 12,13 The cellular entry of SARS-CoV-2 is mediated by the spike (S) protein. The binding of the S protein to the cell surface receptor angiotensin-converting enzyme 2 (ACE2) exposes a cleavage site on the S protein. Transmembrane protease serine 2 recognizes this cleavage site and proteolytically cleaves the S protein to initiate fusion and endocytosis. 14,15 Cells in various human tissues, including the small intestine, testes, kidney, heart, thyroid, adipose and placenta, show high ACE2 expression, whereas the lung cells show moderate expression. 16,17 Consistently, autopsies in cases of severe COVID-19 have shown that in addition to the lungs, the virus infects various tissues, including the heart, kidneys and liver, among others. 4 These findings provide evidence of the systemic spread of SARS-CoV-2 in the body during infection.
Following infection, an immune response is typically induced against the pathogen, and patients with severe COVID-19 may suffer from lymphocytopenia and macrophage activation syndrome. 18,19 Moreover, there are reports of increased secretion of a series of cytokines and chemokines in the plasma, including interleukin (IL)-2, IL-7, IL-10, granulocyte colony-stimulating factor, interferon (IFN)-gamma-induced protein 10 (IP-10), monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein 1 alpha and tumour necrosis factor-alpha (TNFα). The plasma cytokine profile of patients was shown to be associated with COVID-19 severity. [20][21][22] Pregnant women are a high-risk population for severe  and exhibit high mortality owing to their unique immune status. 21,23 Maternal inflammatory responses at the maternal-foetal interface, mediated through macrophages and T cells, are induced after SARS-CoV-2 infection, and these responses could persist for as long as 3 months after COVID-19 recovery. 4,[24][25][26] During pregnancy, the placenta acts as a transient endocrine organ that supports foetal growth by extracting nutrients from the maternal blood and serving as a barrier against pathogens or mediators of the maternal immune system. The placental villus is a functional unit of the placenta, composed of a layer of mononucleated cytotrophoblast cells and an outer multinucleated syncytiotrophoblast. 27 Within the placental villi, Hofbauer cells (placental macrophages), placental fibroblasts and foetal endothelial cells are located adjacent to the foetal capillaries. The presence of Hofbauer cells is important for a successful pregnancy, as it regulates placental morphogenesis and the immune system. 28 Although placental SARS-CoV-2 infection may damage the syncytiotrophoblast and disrupt the placental barrier, 29 the presence of SARS-CoV-2 in the placenta post-recovery and the chances of its vertical transmission are debatable, and the impact of SARS-CoV-2 on the placenta and foetus should be investigated further.
In this study, we used the placentas, maternal plasma, neonatal umbilical cords, cord blood and amniotic fluid donated at delivery by pregnant patients who had completely recovered from COVID-19 and confirmed the presence of residual SARS-CoV-2 nucleic acid and protein in the placentas of patients with COVID-19 long after initial diagnosis and complete recovery. The longest interval between diagnosis and sampling was approximately 3 months, and we believe that the immune response induced by SARS-CoV-2 may persist even longer.

| Patients
The biological specimens used in this study were collected from 11 pregnant women who had recovered from COVID- 19  Results: Residual SARS-CoV-2 nucleic acid and protein were detected in the term placentas of recovered pregnant women. The infiltration of CD14+ macrophages into the placental villi of the recovered pregnant women was higher than that in the controls. Furthermore, the cytokine levels in the placenta, maternal plasma, neonatal umbilical cord, cord blood and amniotic fluid specimens fluctuated significantly.

Conclusions:
Our study showed that SARS-CoV-2 nucleic acid (in one patient) and protein (in five patients) were present in the placentas of clinically recovered pregnant patients for more than 3 months after diagnosis. The immune responses induced by the virus may lead to prolonged and persistent symptoms in the maternal plasma, placenta, umbilical cord, cord blood and amniotic fluid. by a designated diagnostic laboratory in accordance with the novel coronavirus pneumonia diagnosis and treatment program (5th edition).
Among the women, four (#30, #32, #47 and #49) had an asymptomatic infections. At the time of delivery, the patients met the requirements for clinical discharge or release from quarantine, with approval provided by clinicians. However, owing to the limited information about the pathogen and the lack of sensitive testing methods at the onset of suspected symptoms, the time point of initial infection could not be determined in some patients. All the patients were treated in isolation.
The diagnosis was based on SARS-CoV-2 nucleic acid detection from throat swabs, serum antibody testing and chest CT imaging before delivery (Table S2)

| Ethical considerations
The study was approved by the Research Ethics Committee (refer-

| Real-time fluorescent quantitative polymerase chain reaction (qRT-PCR)
RNA extracted from maternal throat swabs, placental villi (10 mg) and neonatal throat swabs were used to test for SARS-CoV-2 S protein using qRT-PCR. The RNA titre was measured using a fluorescent probe targeting the S protein RNA. The sequences of the probes were as follows: CoV-F (5'-TCCTGGTGATTCTTCTTCAGGT-3'), CoV-R

| Immunohistochemical staining
Immunohistochemical staining was performed as described by Fu et al 32

| Cytokine and chemokine measurement
To characterize the cytokine profiles of maternal plasma, placenta, neonatal umbilical cord, cord blood and amniotic fluid specimens, The mean cytokine level in uninfected controls was considered the baseline and the cytokine levels in the specimens were expressed relative to the baseline level and termed "relative secretion levels".
For negative controls, we used maternal plasma, placenta, umbilical cord, cord blood and amniotic fluid donated by three, four, two, two and three uninfected pregnant women at delivery respectively.

| Statistical analysis
Results are expressed in terms of means ± standard error of the mean. Statistical analysis was performed using a paired-sample t-test with Statistical Package for Social Science (SPSS; SPSS Inc., Chicago, IL, USA). Significance was defined as follows: *, P <.05; **, P <.01; ***, P <.001.

| Clinical manifestations
The clinical details of the pregnant patients with COVID-19 and their fetuses have been summarized in Tables 1 and 2. The pregnant women had been discharged from the hospital before delivery, and routine prenatal examinations were performed after the completion of the quarantine period (Tables S1 and S2). After recovery, the women delivered 11 infants (one set of monochorionic diamniotic twins). Five patients underwent caesarean deliveries, five patients had natural childbirth and one patient experienced foetal malformation-induced labour. The twins were delivered preterm at 36 +4 W, whereas the other infants were full-term newborns (Table   S3). Subsequent assessments were based on the biological samples obtained.

| Laboratory detection of residual SARS-CoV-2 nucleic acid and protein in the placenta
To explore the existence of SARS-CoV-2 in the placenta of clinically recovered patients, we first performed qRT-PCR on isolated total RNAs of the placentas from six patients (#26, #30, #42, #46, #47 and #49). As shown in Figure 1A, the placenta from patient #46

| Immune response in the placenta and maternal plasma
After SARS-CoV-2 infection, the immune response in the patient is activated, and the infiltration of macrophages and lymphocytes into the infected locus is a common mechanism for virus clearance. 18,35 To clarify the nature of the immune response that might be acti- To further explore the immune fluctuation in vivo, we profiled the cytokines in maternal plasma samples collected at delivery from nine recovered pregnant women (#14, #20, #26, #27, #32, #35, #42, #47 and #49) and three normal controls using the Luminex assay ( Figure 4A). The relative secretion levels of MCP-3, IL-8, LIF, G-CSF, IL-4, IFNγ, IL-1β and IL-13 were higher, and those of IL-12 (p70) were lower than those in normal controls ( Figure 4B). Among these, MCP-3, G-CSF and IFNγ have been reported as being pro-inflammatory mediators in ARDS caused by a 'cytokine storm' . 41,42 The blockade of IL-12, which is known to be a propathogenic cytokine, has been used for the treatment of autoimmune and autoinflammatory diseases. 43 The reduction in IL-12 levels in the plasma of recovered pregnant women may indicate the activation of an anti-inflammatory response. Interestingly, the level of IL-1 receptor antagonist (IL-1ra), which can act as an antagonist to the pro-inflammatory cytokine IL-1 by binding to its receptor, was elevated in the maternal plasma samples collected from patients #26, #42, #47 and #49, who and #35, whose chest CT images indicated convalescence ( Figure 4C and S3). Thus, the elevated IL-1ra level may indicate an advanced inflammatory response, which was consistent with the CT imaging results.
Of the nine pregnant women who donated maternal plasma, four also donated their placentas (#26, #42, #47 and #49). Among them, the plasma IL-2 and IL-15 levels in patients #26, #42 and #49, who showed the presence of viral protein in the placenta, were elevated compared with those in normal controls ( Figure 4D). Plasma IL-2 signalling can stimulate the expansion of regulatory T cells, 44 and IL-15 is a known T-cell growth factor, 44 both IL-2 and IL-15 show overlapping activities in pathogen elimination. 44 Collectively, we observed specific patterns of cytokine secretion in the plasma of pregnant women who had recovered from COVID-19.

F I G U R E 3
Inflammatory response in term placentas of pregnant women recovered from COVID-19. A, Immunostaining for CD14+ macrophages in the term placenta of patient #46. Two placentas from uninfected pregnant women served as the negative controls (Ctrl 1 and Ctrl 2). Nuclei were stained blue with haematoxylin. Scale bars: 100 μm. B, Immunostaining for CD14+ macrophages in the placentas of five COVID-19-recovered pregnant patients (#42, #26, #30, #47 and #49). The placenta of an uninfected pregnant woman was used as the negative control sample (Ctrl 3). Scale bars: 50 μm. C, Ratio between CD14+ staining area and the area occupied by all cells in the placental villi of six recovered pregnant women (#26, #30, #42, #46, #47 and #49) and three uninfected pregnant women, indicating the extent of CD14+ macrophage infiltration in the placental villi. D, Heatmap showing the relative cytokine profiles of the indicated placentas evaluated using the Luminex assay compared with those of the placentas collected from four uninfected pregnant women. E, Differentially expressed cytokines in the placentas of COVID-19-recovered pregnant women compared with those of normal controls. F, The relative IP-10 and MIG secretion levels increased in the placenta of patient #46 (red circles) compared with that in patients #26, #30, #42, #47 and #49 (green squares). G, The relative secretion level of MIF in the placenta of patient #47 was higher than that in the indicated placentas based on comparison with the level in normal controls. The blue dotted lines in E, F and G represent the baseline 1 or −1

| Clinical outcomes of pregnant women and newborns
To determine whether placental SARS-CoV-2 infection affected the pregnancy outcome and led to the intrauterine infection of the foetus, we further investigated the foetal and accessory abnormalities at delivery. Of all newborns, the monochorionic diamniotic twins weighed 2500 g and 2100 g, whereas the other full-term infants had body weights ranging from 2810 g to 3900 g. All infants received 1-minute and 5-minute Apgar scores of 8 -10 (Table S3). Maternal infection-associated abnormalities were not observed in the foetuses and appendages of any of the 11 pregnant women (Table S4).
To further explore the possibility of intrauterine infection of the foetus, we performed a COVID-19 test on the newborns, and no viral nucleic acid was detected in the nasopharyngeal swab samples.

CO N FLI C T O F I NTE R E S T
The authors declare there are no competing interests and all authors consent to publish the data.