《1. Introduction》

1. Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has spread throughout the world, with over 10 000 000 confirmed cases and almost 500 000 deaths worldwide according to the WHO report as of June 29, 2020. The zinc metallopeptidase angiotensinconverting enzyme 2 (ACE2) was first discovered in 2000. The expression level of ACE2 correlates with heart function, hypertension, and diabetes [1,2]. ACE2 is thought to serve as the receptor for both severe acute respiratory syndrome coronavirus (SARSCoV) and SARS-CoV-2 [3,4], and transmembrane protease serine 2 (TMPRSS2) is a serine protease to prime the SARS-CoV-2 spike (S) protein [5]. Thus, ACE2 positive-expressing organs are believed to have a high risk of infection [6]. ACE2 is expressed in lung alveolar epithelial cells, enterocytes of the small intestine [7], a small population of type II alveolar (AT2) cells [8], and respiratory epithelial cells [6]. Furthermore, ACE2 has been reported to highly express in myocardial cells, epithelial cells of the ileum and esophagus, proximal tubule cells of the kidney, and bladder urothelial cells [6].

During gestation, the maternal immune system changes to a general state of immunosuppression to prevent repulsion of the fetal allograft [9], which carries an increasing risk of various virus infections [10]. The placenta serves as the foremost barrier against the maternal–fetal transmission of viruses [11]. However, ACE2 positive-expressing cells have been reported to distribute in syncytiotrophoblasts (STBs), cytotrophoblasts (CTBs) in villi, decidual perivascular cells (dP), decidual stromal cells (dS), and endothelium and vascular smooth muscle cells in the decidua [12,13]. ACE2 regulates angiotensin (Ang) 1–7 to release into the maternal circulation in STBs, leading to maternal vasculature vasodilation [12]. Meanwhile, previous studies have reported that SARS-CoV and SARS-CoV-2 were not detected in newborn babies delivered from SARS-CoV- and SARS-CoV-2-infected pregnant women [14– 16], or in the uteruses of SARS-CoV- and SARS-CoV-2-infected patients [14,17]. Two recent studies claimed that SARS-CoV-2- specific immunoglobulin M (IgM) antibodies were detected in three cases of newborn blood samples [18,19]. Since IgM antibodies cannot generally be transmitted through the placenta to the fetus, and since the production of IgM usually takes 3–7 days after infection, these findings implied that there might be an intrauterine infection, although virus detection of the fetus was negative. Pre-/post-implantation embryos undergo dramatic changes in morphologic and molecular profile [20–22], and embryos are directly exposed to the endometrial cavity in the uterus after zona hatching, which occurs around day 6 (D6) after fertilization [23]. The potential risks for SARS-CoV-2 infection for pre-/postimplantation embryos remains to be elucidated.

To better understand the potential risk of SARS-CoV-2 vertical transmission, we analyzed ACE2 and TMPRSS2 expression patterns in pre-implantation embryos, peri-implantation embryos, and the maternal–fetal interface at the single-cell transcriptome level, with the aim of expounding and providing theoretical bases for the possibility of SARS-CoV-2 vertical transmission.

《2. Materials and methods》

2. Materials and methods

《2.1. Data downloading and processing》

2.1. Data downloading and processing

The pre-implantation embryo data was downloaded from a previously published dataset [24], and the peri-implantation embryo expression data was downloaded from another previously published dataset [22]. The raw data on the pre-implantation embryos was trimmed and mapped to the Homo sapiens genome assembly the Genome Reference Consortium Human Genome Build 37 (GRCh37) reference sequence (RefSeq) with STAR [25]. The fragments per kilobase million (FPKM) was calculated to estimate the expression. The peri-implantation embryo data was handled as described in Ref. [22]. The gene expression matrix and related cell-type annotation file of Smart-seq2 (Smart: switching mechanism at 5' end of the RNA transcript sequencing) single-cell RNA sequencing (scRNA-seq) data of decidual cells and villous cells were respectively downloaded from two previously published datasets [26,27]. The raw count matrix and cell-type annotation file of the droplet scRNA-seq of the human maternal–fetal interface were downloaded from a previously published dataset [27].

《2.2. Definition of ACE2 and TMPRSS2 gene positive expression》

2.2. Definition of ACE2 and TMPRSS2 gene positive expression

Cells with gene expression (transcripts per kilobase million, TPM) greater than or equal to 1 are defined as "positive expressing cells” in the Smart-seq2 dataset. For droplet scRNA-seq data, cells with a count greater than 0 are defined as "positive expressing cells.”

《2.3. Different expression genes and gene ontology analysis》

2.3. Different expression genes and gene ontology analysis

Differentially expressed genes (DEGs) between cells with different expression levels of ACE2 were identified using the "FindMarkers” function in the "Seurat v3.0” package [28], with the following parameters: "logfc. threshold = log(2), min. pct = 0.4, test.use = ‘roc’, only.pos = F”. Gene ontology (GO) analysis was performed using the "enrichGO” function in "clusterProfiler (3.8.1)” packages [29], with the following parameters: "ont = ‘BP’, pvalueCutoff = 0.05, pAdjustMethod = ‘BH’, qvalueCutoff = 0.1, readable = T”. We used the R packages "VennDiagram (1.6.20)” and "UpSetR (1.3.3)” to show the relationship among different groups of DEGs, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, or GO term lists [30,31]. R (version 3.5.2) was used to carry out all those analyses.

《2.4. Embryo and villous sex determination》

2.4. Embryo and villous sex determination

The sex of each embryo and villous sample was determined by the expression of 15 Y chromosome-specific expressed genes, including: RBMY2FP, RBMY1B, TTTY15, KDM5D, RBMY1J, RBMY1F, RBMY1D, RBMY1E, TSPY4, TSPY2, TSPY8, TSPY1, RPS4Y1, EIF1AY, and DDX3Y. The sex was determined at a whole sample level, so we first calculated the sum expression level of Y chromosomespecific expressed genes at the single-cell level. Embryos with more than 85% cells reached the threshold were defined as male (sum Y chromosome (chrY) gene greater than or equal to 100). Villous sample with more than 65% cells reaching the threshold were defined as male (sum chrY gene greater than or equal to 5 (TPM) in Smart-seq2 cases and sum chrY gene greater than or equal to 1 (scaled count) in droplet cases).

《2.5. Statistical analyses》

2.5. Statistical analyses

For box-dot plots, the Wilcoxon rank-sum test was performed to determine the significance of differences between two groups. Spearman’s correlation coefficient was used to measure the correlation between the expression of ACE2 and TMPRSS2 in D6 periimplantation embryos.

《3. Results》

3. Results

《3.1. Potential risk of SARS-CoV-2 infection in peri-implantation embryos》

3.1. Potential risk of SARS-CoV-2 infection in peri-implantation embryos

Our previous studies have profiled the global transcriptional dynamics of human pre-implantation and peri-implantation embryo cells at multiple consecutive stages [22,24]. ACE2 is the cell receptor of SARS-CoV-2. In order to investigate the potential infection risk of pre-/peri-implantation embryos, we further analyzed the ACE2 expression patterns in oocytes and embryos based on our previous studies [22,24]. ACE2 was found to be expressed through the different stages of pre-implantation embryo development. The expression level of ACE2 was highest in zygotes; it then decreased sharply from the 4-cell stage and was exhausted at the morula stage. Next, ACE2 expression was elevated in the blastocyst (Bst) stage (Fig. 1(a)). In peri-implantation embryos, ACE2 was mainly expressed in day 6 trophectoderm (TE) (D6_TE), day 10 primitive endoderm (PE) (D10_PE), and day 12 PE (D12_PE) (Fig. 1(b)). Since the entry of SARS-CoV-2 into cells depends on the expression of ACE2 and TMPRSS2 [5], we also explored TMPRSS2 expression in peri-implantation embryos. TMPRSS2 was highly expressed in the TE lineage and the expression level of TMPRSS2 was highest in D6_TE cells (Figs. 1(c) and (d)). Single-cell sequencing enabled us to determine the cell heterogeneity among different developmental stages. We summarized the positive-expressing (TPM 1) proportion of ACE2 and TMPRSS2 at each stage. The proportion of ACE2 positive-expressing cells was highest in D6_TE cells (38.9%) and the epiblast lineage retained a low proportion of ACE2 positive-expressing cells. In accordance with ACE2 expression levels, the proportion of ACE2 positive-expressing cells was relatively high in D10_PE cells (Fig. 1(e)). Intriguingly, the proportion of TMPRSS2 positive-expressing cells was also highest in D6_TE cells (74.1%), while the epiblast (EPI) and PE lineages had relatively low proportion of TMPRSS2 positive-expressing cells (Fig. 1(f)). Furthermore, we found that the co-expression of ACE2 and TMPRSS2 was highest (33.3%) in D6_TE (Fig. 1(g)). Of 194 ACE2 positive-expressing TE cells, 166 were simultaneously expressing TMPRSS2. The correlation between ACE2 and TMPRSS2 expression was highest in female D6_TE during peri-implantation embryo development (Fig. 1(h)). Of 92 ACE2 positive-expressing female TE cells, 80 were simultaneously TMPRSS2 positive-expressing in female D6_TE (Fig. 1(i)). Taken together, these results indicate that D6 blastocysts are relatively vulnerable to SARS-CoV-2 infection, and a potential risk exists for SARS-CoV-2 infection during periimplantation embryo development.

《Fig. 1》

Fig. 1. ACE2 and TMPRSS2 expression pattern in pre-implantation and peri-implantation embryos. (a) ACE2 expression level in different stages of pre-implantation embryos; (b) ACE2 expression level in different stages of peri-implantation embryos; (c) TMPRSS2 expression level in different stages of pre-implantation embryos; (d) TMPRSS2 expression level in different stages of peri-implantation embryos; (e) proportion of ACE2 positive-and negative-expressing cells in different stages of peri-implantation embryos; (f) proportion of TMPRSS2 positive- and negative-expressing cells in different stages of peri-implantation embryos; (g) proportion of cells with different expression patterns of ACE2 and TMPRSS2 in different stages of peri-implantation embryos (double_pos: cell with both ACE2 and TMPRSS2 positive expression; only_ACE2_pos: cell with ACE2 positive expression but not expressing TMPRSS2; only_TMPRSS2_pos: cell with TMPRSS2 positive expression but not expressing ACE2; double_neg: cell with neither TMPRSS2 expression nor ACE2 expression); (h) correlation between the expression of ACE2 and TMPRSS2 in D6 peri-implantation embryos (R value is the Spearman correlation coefficient); (i) number of cells with both ACE2 and TMPRSS2 positive expression in D6 peri-implantation embryos. Bst: blastocyst; EPI: epiblast; ysTE: yolk sac trophectoderm. Error bars represent means ± standard errors.

《3.2. Potential risk of SARS-COV-2 infection in the maternal–fetal interface》

3.2. Potential risk of SARS-COV-2 infection in the maternal–fetal interface

To evaluate the potential risk of SARS-CoV-2 vertical transmission in post-implantation development, we then examined the ACE2 expression level in the maternal–fetal interface based on scRNA-seq data published previously [26,27]. The researchers had profiled the transcriptional characters of various cell types in the maternal–fetal interface, and we adopted the original definition for each cell type. The expression levels of ACE2 and TMPRSS2 were highly heterogeneous among the different cell types in the maternal–fetal interface (Figs. S1 and S2 in Appendix A). The results from droplet scRNA-seq data revealed that there was no ACE2 positive-expressing cell type in maternal peripheral blood. The top four subgroups with the highest proportion of ACE2 positive-expressing cells were decidual perivascular cells (dP1 and dP2; 1 and 2 mean different cell subtypes), STB and CTB (10%, 5%, 9%, and 5%, respectively; Fig. S3(a) and Table S1). The highest proportions of ACE2 positive-expressing decidual cell types were in Endo (L) (20% of all five cells) and dP (13% of 84 cells), based on Smart-seq2 data (Fig. 2(a) and Table S2). Despite the lower positive-expressing proportion, ACE2 was also expressed in decidual natural killer cells (dNK: 1%) and decidual macrophages (dM: 1%) (Table S2). In villi, STB and CTB had the highest ACE2 positive-expressing cell proportions among all six major types of villous cells in the 8-week group (STB_8W: 41%; CTB_8W: 20%), with an overall positive-expressing proportion of 8.5% (Fig. 2(b) and Table S3). The remarkably increased ACE2 positiveexpressing proportion in STB and CTB might be caused by the detection of low ACE2 expression level in the cells, which was obtained due to the higher sequencing depth. It is worth noting that the positive-expressing cell proportion in extravillous trophoblasts (EVTs) enriched from the maternal surface of a 24- week placenta (EVT_24W) was 63%, which was much higher than that of EVTs from villous samples at 8 weeks (EVT_8W), at 3% (Fig. 2(b) and Table S3). Similar to ACE2, the expression of TMPRSS2 was detected in the villous and decidual cell types mentioned above, and the highest TMPRSS2 positive-expressing proportions were in STB_8W, EVT_24W, and epithelial glandular cells (Epi) (26%, 19%, and 19%, respectively; see Figs. S3(b)–(d) and Tables S1–S3). The co-expression of ACE2 and TMPRSS2 occurred mainly in STB_8W and EVT_24W (13% and 14%, respectively; see Fig. 2(c) and Table S3), while there was no observation of coexpression in any decidual cells (Fig. 2(d) and Table S2), suggesting that both the STB in villi and the EVT in the decidua are susceptible to SARS-CoV-2. Taken together, these results imply that there is a potential risk of SARS-CoV-2 vertical transmission in pregnancy, due to the existence of ACE2 and TMPRSS2 double positiveexpressing cells in the maternal–fetal interface.

《Fig. 2》

Fig. 2. ACE2 and TMPRSS2 expression pattern in the maternal–fetal interface. (a) Proportion of ACE2 positive- and negative-expressing cells in different cell types in decidual cells; (b) proportion of ACE2 positive- and negative-expressing cells in different cell types in villous cells; (c) proportion of cells with different expression patterns of ACE2 and TMPRSS2 in villous cells; (d) proportion of cells with different expression patterns of ACE2 and TMPRSS2 in decidual cells. DC: dendritic cells; dM: decidual macrophages; dNK: decidual natural killer (NK) cells; dP: decidual perivascular cells; dS: decidual stromal cells; Endo: endothelial cells; Epi: epithelial glandular cells; ILC: innate lymphocyte cell; STB: syncytiotrophoblast; CTB: villous cytotrophoblast; EVT: extravillous trophoblast; Marc: fetal macrophages (also called Hofbauer cells); Mes: mesenchymal stromal cells; L: lymphatic; m: maternal; p: proliferative; CD16 is also known as FcγRIII, belonging to the immunoglobulin superfamily (IgSF), and CD16+/– means CD16 positive- or negative-expressing cells in immunofluorescence staining in flow cytometry; T cell: thymus-dependent lymphocyte.

《3.3. Common profile of ACE2 positive-expressing cells among different cell types》

3.3. Common profile of ACE2 positive-expressing cells among different cell types

As previously mentioned, D6 female TE cells, EVT_24W cells, and STB_8W cells were the top three cell types with the highest risk of SARS-CoV-2 infection. To obtain further understanding of those cells, GO and KEGG pathway enrichment analyses were performed on DEGs between ACE2 positive-expressing and negativeexpressing subtypes in each of these three cell groups (for details, see Tables S4 and S5). Common DEGs, GO terms, and KEGG pathways were found among different lists in the subsequent integration analysis (Figs. 3(a), S4(a), and S4(b)). The processes of the "viral life cycle,” "process utilizing autophagic mechanism,” and "membrane fusion” were enriched in all three up-regulated DEG groups; the former two were even enriched in all five analyzed groups (Figs. 3(b), S4(c), and S5). Meanwhile, "sphingolipid metabolism” was the common KEGG pathway in three up-regulated groups and "lysosome” was enriched in all five analyzed groups (Figs. 3(c) and S6). Many studies have also reported that the mechanisms involved in the sphingolipid metabolism and lysosomal degradation pathway took part in the processes of virus infections and related cell response [32–35]. These findings suggest that the ACE2 positive-expressing and ACE2 negativeexpressing cells are probably in a different state when responding to viral infection, aside from having a different sensitivity to SARSCoV-2. This hypothesis was further supported by our observations of immune and virus-related processes in GO enrichment in all three cell groups (Fig. S7(a)). The DEGs in these terms contained many interferon-stimulated genes (ISGs), which have already been reported, and genes that are involved in inflammation and defense against viral infection (Figs. S7(b)–(g)). Furthermore, ACE2 positive-expressing and ACE2 negative-expressing EVT_24W or STB_8W cells might have different metabolisms, as revealed by the metabolic processes/pathways, "regulation of cholesterol metabolic process” in GO and "steroid biosynthesis” in KEGG, for example (Figs. S4(c) and S6(a)). There were 12 common upregulated expression DEGs in all three ACE2 positive subtypes (Fig. 3(d)), such as MDK, UGCG, and RAB25, which are implicated in the processes of inflammatory response, cell proliferation, cell growth, and cell migration. Overall, these results indicate that the ACE2 positive-expressing cells in D6 female TE cells and the EVT_24W and STB_8W cell groups show certain common characteristics in comparison with the ACE2 negative-expressing cells, which suggest a general profile of sensitivity to SARS-CoV-2 in early and later TE lineage cells.

《Fig. 3》

Fig. 3. Differentially expressed genes and enriched biological processes/pathways between ACE2 positive-expressing and ACE2 negative-expressing expression cells in D6 female TE and in EVT_24W and STB_8W cell groups. (a) Number of cell-type specific or inter-cell-type common biological processes in GO enrichment among D6 female TE, and EVT_24W and STB_8W cell groups; (b) common enriched GO terms of all three up-regulated DEGs among D6 female TE, and EVT_24W and STB_8W cell groups; (c) common enriched KEGG pathway of all three up-regulated DEGs among D6 female TE, and EVT_24W and STB_8W cell groups; (d) number of common DEGs in all three upregulated DEGs among D6 female TE, and EVT_24W and STB_8W cell groups.

《3.4. Gender bias in ACE2 expression of peri-implantation embryos and villi》

3.4. Gender bias in ACE2 expression of peri-implantation embryos and villi

As the ACE2 gene is located in the X chromosome and escapes from chromosome inactivation [36], we speculated that different expression levels of ACE2 might exist, as the X chromosome copy number differs in male and female embryos. After comparing embryos of the different genders, we found that the ACE2 expression level was high in D6 female TE and PE cells, but not in the latter development stages of post-implantation (Fig. 4(a)). The TMPRSS2 expression showed no significant difference between male and female embryos, except for D12_TE cells during periimplantation development (Fig. 4(b)). We further analyzed the ACE2 and TMPRSS2 expression levels in subsequent villous cells of different genders. Three male villous samples were identified in droplet data. The ACE2 expression level showed a significant gender difference in ACE2 positive-expressing STB and CTB, while the TMPRSS2 expression level showed no gender bias in TMPRSS2 positive-expressing cells (Figs. 4(c), 4(d), and S8(a)). Only one male villous sample was identified in Smart-seq2 data, and there was no significant gender bias in both STB_8W and EVT_8W, in accordance with the post-implantation TE cells (Figs. S8(b)–(d)). In conclusion, ACE2 showed no global gender bias in post-implantation embryos, but was highly expressed in D6 female TE and PE cells, along with a gender difference in ACE2 positive-expressing STB and CTB. These findings indicate that there may be different SARS-CoV-2 infection susceptibilities in D6 embryos of different genders, as well as during gestation.

《Fig. 4》

Fig. 4. ACE2 expression level in peri-implantation embryos and villous cells with different genders. (a) ACE2 expression in all cells in different stages of peri-implantation embryos of different genders; (b) TMPRSS2 expression in all cells in different stages of peri-implantation embryos in different genders; (c) ACE2 expression in ACE2 positiveexpressing villous cells of different genders in droplet data analysis; (d) TMPRSS2 expression in TMPRSS2 positive-expressing villous cells of different genders in droplet data analysis. Nor_count: normalized counts; HB, Hofbauer cells; fFB1, fetal fibroblasts 1. The p-value between two gender groups was determined by the Wilcoxon rank-sum test.

《4. Discussion》

4. Discussion

The COVID-19 pandemic has posed huge challenges to global public health. Pregnant patients are subject to special attention due to concerns about compounding pregnancy complications and potential negative influences on fetal development in utero. Whether SARS-CoV-2 can infect the embryo or placenta and disturb the establishment or maintenance of pregnancy, and the possible existence of vertical transmission, are hot topics. In this study, we comprehensively investigated the single-cell transcriptome data of pre-/post-implantation embryos at consecutive stages and the maternal–fetal interface to clarify the expression profiles of ACE2 and TMPRSS2, as well as other molecular signatures involved in SARS-CoV-2 infection. We found that ACE2 and TMPRSS2 were strongly co-expressed in D6_TE, STB_8W in villi, and EVT_24W cells in the decidua. Common DEGs, GO terms, and KEGG pathways were observed among the top three cell types with the highest potential risk of SARS-CoV-2 infection, which indicated that ACE2 positive-expressing cells shared conserved changes involved in the autophagy-lysosome system, and immune and virus-related processes when compared with ACE2 negativeexpressing cells. Furthermore, we revealed that ACE2 was highly expressed in D6 female TE and PE cells, along with a significant gender difference in ACE2 positive-expressing STB. Collectively, this research presents an ACE2 expression landscape in periimplantation embryos and the maternal–fetal interface for the first time at the single-cell transcriptome level, and suggests a potential risk of SARS-CoV-2 vertical transmission during the periimplantation period and pregnancy.

 ACE2 was found to be expressed in peri-implantation embryos. The ACE2 positive-expressing cell proportion was highest in the D6_TE cells of peri-implantation embryos, although other lineages showed low expression. This was consistent with the TMPRSS2 positive-expressing cell proportion and the co-positiveexpressing cell proportion in D6_TE cells, indicating a relatively high susceptibility to SARS-CoV-2 infection in D6 embryos. Nevertheless, since ACE2 displayed a relatively low expression level, the possibility of infection may still exist at other developmental stages. These findings also suggest a potential infection risk during embryo transplantation in clinic in vitro fertilization (IVF).

In the maternal–fetal interface, the data showed that CTB, STB, EVT, and dP cells had a high ACE2 positive-expressing proportion; the highest co-positive-expressing proportion of ACE2 and TMPRSS2 occurred in STB_8W and EVT_24W. It is known that CTB derives from TE and then differentiates into STB and EVT, with the former being infiltrated by maternal blood and the latter invading uterine spiral arteries [23,37]. In a previously published study, SARS-CoV-2 was detected in three of 307 blood samples collected from 205 patients [38], although the detection proportion was low. STB were directly exposed to the mother’s blood, so the virus in the blood would mainly infect STB. EVT also showed a high infection possibility in utero for part of EVT serving as a kind of blood vessel wall cell and located in the decidua. Although the ACE2 positiveexpressing cell proportion was much higher in EVT_24W than in EVT_8W, the various risks of vertical transmission at the different stages of pregnancy have yet to be assessed. Taken together, these results indicate a potential risk of SARS-CoV-2 vertical transmission during gestation.

Although there is no direct evidence for SARS-CoV-2 vertical transmission, two studies published in The Journal of the American Medical Association (JAMA) reported three cases of newborn babies with positive SARS-CoV-2 IgM antibodies, suggesting that vertical transmission may exist. Moreover, our study found that ACE2 and TMPRSS2 co-positive-expressing cells were present in the maternal–fetal interface, providing a theoretical possibility of SARS-CoV-2 vertical transmission during gestation. If pregnant women are infected with SARS-CoV-2 in the first trimester, the health monitoring of the mother and baby should be strengthened. Intrauterine SARS-CoV-2 transmission might directly affect the clinical therapeutic schedules for infected pregnant women. With this research, we hope to provide theoretical bases for the clinical diagnosis and treatment of pregnant women. As for the existence of vertical transmission, further accumulation of clinical cases and more confirmation from basic research are necessary.

We also analyzed the differentially expressed genes between ACE2 positive- and negative-expressing cells in three cell groups with a high risk of SARS-CoV-2: D6 female TE, EVT_24W, and STB_8W cell groups. We found that some common virus-related and cell response processes and pathways were enriched in these three cell types. The GO term "process utilizing an autophagic mechanism” and the KEGG pathway "lysosome” were enriched in all five analyzed groups. As a lysosome-dependent process, autophagy has been thought to play an important role in maintaining cellular homeostasis and defending against viral infection [39,40]. Many studies suggest that viruses may hijack the conserved autophagic machinery during infection for their advantage [41,42]. Our results revealed that the ACE2 positive-expressing cells in TE lineage cells at different development stages shared some common changes involved in the autophagy process, when compared with ACE2 negative-expressing cells. This finding supports the view that it is reasonable to set the autophagy process as a target in therapeutic strategies for SARS-CoV-2 infection [43]. We also revealed that the DEGs in these three cell groups showed enrichment in the biological processes of the immune and defense responses to the virus. Many ISGs were involved in the DEGs in those GO terms. Ziegler et al. [44] indicated that ACE2 was actually a neglected ISG. It partly interprets the coordinated changes of ACE2 and ISGs observed in three TE lineage cells.

X chromosome inactivation (XCI) is a crucial epigenetic mechanism to balance X chromosome expression dosage between XY males and XX females. As an escape gene, ACE2 is reported to show a heterogeneous sex bias [36]. In this study, we found that the ACE2 expression level in D6 female TE cells was significantly higher than in male, albeit not in the later development stages. ACE2 positiveexpressing STB and CTB also showed significant differences between female and male samples. These findings suggest a gender bias in the risk of SARS-CoV-2 vertical transmission during the peri-implantation period and gestation. These results also imply that ACE2 might not escape X-inactivation in human postimplantation embryos, which enhances our knowledge about XCI during early embryo development.

《5. Conclusions》

5. Conclusions

Although IgM antibodies have been detected in blood samples in neonates, it is still controversial whether SARS-CoV-2 can be transmitted in utero. More definitive evidence is still urgently needed to validate the vertical transmission of SARS-CoV-2. In this study, we provided theoretical bases for the vertical transmission possibility of SARS-CoV-2 infection during peri-implantation embryo development and gestation, and provided recommendations for SARS-CoV-2-infected women during pregnancy or who are preparing for pregnancy.

《Acknowledgements》

Acknowledgements

We show our admiration and gratitude to those people who are fighting on the front line of defense of SARS-CoV-2. This work was supported by the Peking University Health Science Center (BMU2020HKYZX001), the Chinese Academy of Engineering (2020-KYGG-01-06), and the National Key Research and Development Program of China (2018YFC1004000 and 2019YFA0801400).

《Authors’ contribution》

Authors’ contribution

Wei Chen, Peng Yuan, and Ming Yang wrote the manuscript. Wei Chen and Peng Yuan performed bioinformatic analysis. Zhiqiang Yan, Siming Kong, Jie Yan, Xixi Liu, and Yidong Chen were helpful for the discussion of this study. Jie Qiao and Liying Yan developed the study conception and designs. All the authors have read and approved the final manuscript.

《Compliance with ethics guidelines》

Compliance with ethics guidelines

Wei Chen, Peng Yuan, Ming Yang, Zhiqiang Yan, Siming Kong, Jie Yan, Xixi Liu, Yidong Chen, Jie Qiao, and Liying Yan declare that they have no conflict of interest or financial conflicts to disclose.

《Appendix A. Supplementary data》

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2020.07.013.