key: cord-0739017-hsbff9ii
authors: Constantino, F. B.; Cury, S. S.; Nogueira, C. R.; Carvalho, R. F.; Justulin, L. A.
title: Prediction of non-canonical routes for SARS-CoV-2 infection in human placenta cells
date: 2020-08-13
journal: bioRxiv
DOI: 10.1101/2020.06.12.148411
sha: b438f159c5ef3ee64d0256940c8e72159c415e59
doc_id: 739017
cord_uid: hsbff9ii

The SARS-CoV-2 is the causative agent of the COVID-19 pandemic. The data available about COVID-19 during pregnancy have demonstrated placental infection; however, the intrauterine transmission of SARS-CoV-2 is still debated. Intriguingly, while canonical SARS-CoV-2 cell entry mediators are expressed at low levels in placental cells, the receptors for viruses that cause congenital infections such as the cytomegalovirus and Zika virus are highly expressed in these cells. Here we analyzed the transcriptional profile (microarray and single-cell RNA-Seq) of proteins potentially interacting with coronaviruses to identify non-canonical mediators of SARS-CoV-2 infection and replication in the placenta. We show that, despite low levels of the canonical cell entry mediators ACE2 and TMPRSS2, cells of the syncytiotrophoblast, villous cytotrophoblast, and extravillous trophoblast co-express high levels of the potential non-canonical cell-entry mediators DPP4 and CTSL. We also found changes in the expression of DAAM1 and PAICS genes during pregnancy, which are translated into proteins also predicted to interact with coronaviruses proteins. These results provide new insight into the interaction between SARS-CoV-2 and host proteins that may act as non-canonical routes for SARS-CoV-2 infection and replication in the placenta cells.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of the coronavirus disease 2019 (COVID-19) (1) . It was first notified at the end of 2019, in Wuhan, China, and become a worldwide pandemic (2) . At the beginning of July, six months later, COVID-19 infected over 12 million people and is the cause of approximately 557,000 deaths worldwide (https://coronavirus.jhu.edu/).

Older age, laboratory abnormalities, and several comorbidities are associated with the more severe cases of COVID-19 (3) . For specific groups of COVID-19 patients, for example, pregnant women, the potential impacts of SARS-CoV-2 infection remains mostly unknown, and data are limited. However, considering previous works reporting coronaviruses infections (4), pregnant women are at higher risk of SARS-CoV-2 infection due to physiological changes in the immune, cardiorespiratory, and metabolic systems (5) .

Although only a small number of maternal viruses infections are transmitted to the fetus, some may cause life-threatening diseases (6) . These viruses use cellular host entry mediators expressed by placenta cells, as described for the cytomegalovirus and the Zika virus (7, 8) , to infect these cells. The Zika virus (ZIKV) outbreak, associated with fetal brain damage, emphasizes the necessity of further characterization and understanding placental infection or intrauterine (vertical) transmission of SARS-CoV-2, as well as the possible adverse fetal outcomes. The few studies on the subject have provided contradictory findings, with some reports suggesting no evidence of placental infection or vertical transmission of SARS-CoV-2 (9) (10) (11) . Conversely, multiple lines of evidence have shown placental SARS-CoV-2 infection in pregnant women diagnosed with moderate to severe COVID-19. Neonates born from mothers with COVID-19 presented a positive serological test for SARS-CoV-2 immunoglobulin (Ig)M and IgG (12, 13) . While the IgG can be transferred from mother to fetus across the placenta, the detection of IgM in newborns suggests a vertical transmission of the virus, since IgM cannot cross the placental barrier due to its high molecular mass (14) . Accordingly, SARS-CoV-2 RNA transmission was comprehensively confirmed by pathological and virological investigations. (15) Also, it was recently shown SARS-CoV-2 particles in syncytiotrophoblast with generalized inflammation, diffuse perivillous fibrin depositions, and tissue damage in an asymptomatic woman (16) . Remarkably, these placental alterations due to the SARS-CoV-2 infection lead to fetal distress and neonatal multi-organ failure. These results highlight the importance of exploring the expression profile of potential host mediators of the SARS-CoV-2 that may create a permissive microenvironment to placental infection and possible vertical transmission of the virus.

In fact, like other viruses, SARS-CoV-2 requires diverse host cellular factors for infection and replication. The angiotensin-converting enzyme 2 (ACE2) is the canonical receptor for the SARS-CoV-2 spike protein receptor-binding domain (RBD) for viral attachment (17) . This process is followed by S protein priming by cellular transmembrane serine protease 2 (TMPRSS2) that allows the fusion of the virus with host cellular membranes (17) . Single-cell RNA sequencing (scRNA-Seq) has demonstrated that both ACE2 and TMPRSS2 are co-expressed in multiple tissues affected by COVID-19, including airway epithelial cells, cornea, digestive and urogenital systems (18) . Few cells express ACE2 and TMPRSS2 in the placenta (18, 19) , suggesting that SARS-CoV-2 is unlikely to infect the placenta through the canonical cell entry mediators. Therefore, other host interacting proteins may play a role in the biological cycle of the virus and contribute to the pathogenesis of SARS-CoV-2 in the placenta. In this paper, we demonstrate, through transcriptomic (microarray and scRNA-Seq) analysis and in silico predictions of virus-host protein-protein interactions, that cells of the syncytiotrophoblast, villous cytotrophoblast, and extravillous trophoblast express high levels of potential non-canonical cell-entry mediators dipeptidyl peptidase 4 (DPP4) and cathepsin L (CTSL), despite low-levels of 

We first investigated the gene expression in placental tissues of classical host-virus interacting proteins described in the literature. The canonical entry receptors ACE2 and TMPRSS2 were low expressed throughout gestation. CTSL, which is translated into a lysosomal cysteine proteinase that plays a role in intracellular protein catabolismpresented the highest level of expression in the placenta during the first, second, and third trimester. Similarly, DPP4, which is translated into an intrinsic membrane glycoprotein, was highly expressed throughout gestation ( Figure 1A ).

Next, we analyzed the gene expression profile during gestation in placental tissues. We found 25 differentially expressed genes (DEGs) in the second trimester, and 687 DEGs in the third, when they were independently compared to the first trimester (Supplementary Table 1 ). All DEGs were divided into three clusters by the K-means clustering analysis (Supplementary Figure 1A) . Cluster 1 includes genes that increase expression during pregnancy, and these genes enriched terms related to blood vessels morphogenesis, complement cascade, extracellular matrix organization, and cellular response to nitrogen compounds. Cluster 2 encompasses genes that increase expression specifically in the third trimesters, which are related to female pregnancy, growth hormone signaling pathway, homeostasis, and steroid biosynthetic process. Cluster 3 includes genes that decrease expression during pregnancy. and these genes enriched terms related to cell division, sulfur compounds biosynthetic process, chromosomal segregation, PID MYC active pathway Table 1 ). Next, we selected the human proteins potentially interacting with SARS-CoV using the P-HIPSter database. We found 32 virushost interacting proteins of SARS-CoV (Supplementary Table 4 ), and nine virus-host interacting proteins of the ZIKV (Supplementary Table 5 ). We also found that, from these list of virus-host interacting proteins, 10 DEGs (DAAM1, FRMD3, STX3, HBD, PRKCZ, CTK3, PAICS, EVL, TREM2, and HBE1) are translated into proteins that interact with SARS-CoV proteins, and one with the ZIKV (Supplementary Figures 1F and 1G ).

Among these 10 DEGs, six were up-regulated (DAAM1, FRMD3, STX3, HBD, PRKCZ, and CTK3) and four (PAICS, EVL, TREM2, and HBE1) were down-regulated in the third trimester ( Figure 1B) . The gene DAAM1, which is translated into an intrinsic membrane glycoprotein implicated in cell motility, adhesion, cytokinesis, and cell polarity -showed the highest level of fold change in the third trimester of gestation compared to the first (logFC= 1.43; Figure 1B ). The host-virus protein-protein interactions (PPI) predicted for these 10 DEGs presented a Likelihood Ratio (LR) > 100 (Supplementary Table 6 ), according to P-HIPSTer (21) . Noteworthy, PAICS transcript, which is translated into an enzyme that catalyzes the sixth and seventh steps of the novo purine biosynthesis, were predicted to interact with both SARS-CoV and ZIKV ( Figure 1B and Supplementary Table 7) . We observed a significant interaction of placental proteins predicted to interact with SARS-CoV-2 (based on the DEGs or not), and PAICS was also predicted to interact with other proteins in the PPI network with the highest degree (Figure 1C and   Supplementary Table 8) .

We next used single-cell RNA sequencing to analyze the expression of the 10 DEGs translated into proteins that potentially interact with SARS-CoV in human placental cells ( Supplementary Figures 1H and 1I) . We selected the genes ACE2, TMPRSS2, CTSL, DPP4, PAICS, and DAAM1 for further investigations using scRNA-seq transcriptome data from cells of the syncytiotrophoblast (n=1144), villous cytotrophoblast (n=8244), and extravillous trophoblast (n=2170) of non-disease human placental tissues, considering the potential relevance of these genes for SARS-CoV infection and replication in the organ ( Figure 1D) . We noticed that the expression of the classical SARS-CoV-2 entry receptor genes (ACE2 and TMPRSS2) was minimally expressed in these cells. In contrast, potential non-canonical cell entry mediator genes DPP4 and CTSL, as well as the genes for the predicted virus-host interaction proteins DAAM1, and PAICS were expressed at higher levels ( Figure 1E and Supplementary Figure 2A) . We also looked for the co-expression of ACE2 and TMPRSS2 with DPP4, CTSL, DAAM1, and PAICS by using scRNA-Seq in these same cells (Figure 2 ). This analysis revealed that DPP4, CTSL, DAAM1, and PAICS are co-expressed at high-levels, while these genes are co-express at low levels with ACE2 and TMPRSS2 (Figure 2) . Remarkably, we found only three cells co-expressing ACE2 and TMPRSS2. For this reason, we consider that DPP4, CTSL, DAAM1, and PAICS may represent candidates of an alternative route for SARS-CoV-2 infection in human placentas.

We used The Human Protein Atlas to predict the subcellular location of proteins encoded by our candidates. DPP4 is predicted as intracellular, membrane, and secreted protein;

CTSL is a protein located in the Golgi apparatus and additionally in vesicles; PAICS is a protein found in the cytosol while DAAM1 is located in the plasma membrane and cytosol ( Figure 1E) . Finally, we analyzed the gene expression profile of ACE2, TMPRSS2, CTSL, DPP4, DAAM1, and PAICS genes on publicly available scRNA-Seq datasets of lung, liver, and thymus fetal tissues. CTSL, DPP4, and DAAM1 were found as highly expressed in fetal cells compared to ACE2 and TMPRSS2 in all tissues analyzed (Supplementary Figure 2) .

Additional investigations may determine the generality and impact of these findings, including the confirmation of vertical transmission.

women diagnosed with moderate to severe COVID-19 (15, 22) with findings supporting the possibility of vertical transmission of SARS-CoV-2 (15, 23) . However, few placenta cells express ACE2 and TMPRSS2 (18, 19) further demonstrated that SARS-CoV-2 can use CTSB and CTSL as well as TMPRSS2 for priming host cells (26, 27) . These data provide evidence that CTSL is a potentially promising treatment for COVID-19 by blocking coronavirus host cell entry and intracellular replication (28) . Secondly, we analyzed whether placental development and growth are associated with transcriptional changes in proteins that potentially interact with SARS-CoV. Among the transcripts translated into proteins predicted as interacting with SARS-CoV, DAAM1 increased with placental development and growth. DAAM1 was previously identified as a regulator of bacteria phagocytosis by regulating filopodia formation and phagocytic uptake in primary human macrophages (29) . We predicted that DAAM1 potentially interact with the viral protein encoded from open reading frame 8 (ORF8) and the hypothetical protein sars7a of SARS-CoV. The SARS-CoV ORF8 shares the lowest homology with all SARS-CoV-2 proteins; however, it was previously demonstrated that SARS-CoV-2 ORF8 expression was able to selectively target MHC-I for lysosomal degradation by an autophagy-dependent mechanism in different cell types (30) .

We found that PAICS transcript, which is also translated into a protein predicted as interacting with the Nsp3 of SARS-CoV, decreased its levels with placental development and growth. PAICS is an enzyme of de novo purine biosynthesis pathway, which was previously predicted as having a putative interaction with the human influenza A virus (IAV) nucleoprotein and was up-regulated during IAV infection (31) . The putative interaction of Nsp3 with PAICS is relevant. Nsp3 is one of the 16 non-structural proteins in the replicase ORF1ab gene of SARS-CoV and SARS-CoV-2 (1, 32) , and binds to virus RNA and proteins, including the nucleocapsid protein (33) . Considering the low-levels of ACE2 and TMPRSS2 that we found in villous trophoblast cells, our results provide evidence of alternatives cell-entry mediators in the placenta. Moreover, the description of the potential interaction between host and SARS-CoV-2 may provide insights into the mechanisms of placental infection in women with COVID-19.

transmission of viruses (34, 35) . Thus, we finally used single-cell analysis of the syncytiotrophoblast, villous cytotrophoblast, and extravillous trophoblasts -the fetal component of the placenta -that confirmed the low-levels of expression of the canonical entry receptor ACE2 and TMPRSS2 genes (18, 19) . Conversely, we found that these cells express high-levels of DPP4 and CTSL, two potential mediators of SARS-Cov-2 host cell entry (24, 27) that may contribute to the SARS-CoV-2 infection (15, 22, 23) . It is essential to highlight that our scRNA-Seq analysis showed that the non-canonical DPP4 and CTSL entry genes (24, 27) are highly co-expressed in syncytiotrophoblast, villous cytotrophoblast, and extravillous trophoblasts cells. The transcripts DAAM1 and PAICS, which are translated into proteins predicted as potentially interacting with SARS-CoV-2, were also highly co-expressed with DPP4 and CTSL. These results demonstrate that, although ACE2 and TMPRSS2 are poorly expressed in the placenta, other mediators that potentially interact with the virus are highly co-expressed in villous trophoblast cells and, therefore, may represent a valuable alternative route for infection and viral replication.

Although our in-silico analyses are a starting point, they have limitations. We reanalyzed published placenta datasets with a limited number of samples (microarray = 12 individuals, and scRNA-seq = 5 individuals) and, consequently, validations in a large cohort of samples must be performed. The predicted interactions presented here should also be experimentally validated in infected cells or placenta to circumvent these limitations.

Moreover, the analyses of single-cell data should be conducted for the entire period of gestation, considering that we only evaluated single-cell data from the first trimester of pregnancy. The detection of SARS-CoV-2 in the placenta needs to be performed in a large cohort of pregnant women with COVID-19, including asymptomatics. Considering that vertical transmission of SARS-CoV-2 is still debated, the follow-up of newborns from mothers with COVID-19 during pregnancy should be necessary since, if it occurs even in the non-asymptomatic, the long-term consequences are mostly unknown.

In conclusion, despite low-levels of ACE2 and TMPRSS2, our analyses demonstrate that villous trophoblast cells express high levels of potential non-canonical cell-entry mediators DDP4 and CTSL. We also found changes in the expression of the DAAM1 and PAICS genes coding for proteins predicted to interact with SARS-CoV proteins during pregnancy. These results provide new insight into the interaction between SARS-CoV-2 and the host proteins and indicate that coronaviruses may use multiple mediators for virus infection and replication.

We reanalyzed microarray data from villous trophoblast tissues of first (45-59 

We used the Metascape tool (https://metascape.org) (37) to perform functional enrichment analysis of the genes lists generated in the clustering and differential expression analyses.

SARS-CoV-2 cell entry mediators were selected for first screening using the human proteins already described in the COVID-19 Cell Atlas (https://www.covid19cellatlas.org/) and the literature (18, 24, 27, 38) . Next, we used gene expression data to generated a list of PPIs that potentially interact with human coronaviruses and the ZIKV from the Pathogen-Host Interactome Prediction using Structure Similarity (P-HIPSTer, http://phipster.org/) database (23) (Supplementary Table 4 and 5, respectively). SARS-CoV was included in the analysis considering the evolutionary relationship between the novel SARS-CoV-2 (38) , and ZIKV considering its impact on placental infection and fetal microcephaly (7, 34) .

The HumanBase webtool (https://hb.flatironinstitute.org) (39) was used to generate the tissue-specific gene network of placental genes coding for proteins that potentially interact with SARS-CoV-2 based on P-HIPSTER or that we identified as either associated with COVID-19 or that we hypothesized may be associated with the disease, based on the literature (Supplementary Biorender was used to design cell images (https://app.biorender.com/).

All data is available in the manuscript. The images were generated using COVID-19 Cell Atlas (https://www.covid19cellatlas.org)

functionalities.

A new coronavirus associated with human respiratory disease in China

An interactive web-based dashboard to track COVID-19 in real time

OpenSAFELY: factors associated with COVID-19-related hospital death in the linked electronic health records of 17 million adult NHS patients

Potential Maternal and Infant Outcomes from Coronavirus 2019-nCoV (SARS-CoV-2) Infecting Pregnant Women: Lessons from SARS, MERS, and Other Human Coronavirus Infections

Pregnancy affected by SARS-CoV-2 infection: a flash report from Michigan

Congenital Viral Infection: Traversing the Uterine-Placental Interface

The emergence of Zika virus and its new clinical syndromes

Temporal dynamics of protein complex formation and dissociation during human cytomegalovirus infection

Clinical characteristics and intrauterine vertical transmission potential of COVID-19 infection in nine pregnant women: a retrospective review of medical records. The Lancet

Factors preventing materno-fetal transmission of SARS-CoV-2

Possible Vertical Transmission of SARS-CoV-2 From an Infected Mother to Her Newborn

Antibodies in Infants Born to Mothers With COVID-19 Pneumonia

Can SARS-CoV-2 Infection Be Acquired In Utero?: More Definitive Evidence Is Needed

Vertical transmission of COVID-19: SARS-CoV-2 RNA on the fetal side of the placenta in pregnancies with COVID-19 positive mothers and neonates at birth

SARS-CoV-2 placental infection and inflammation leading to fetal distress and neonatal multi-organ failure in an asymptomatic woman

Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses

SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes

Does the human placenta express the canonical cell entry mediators for SARS-CoV-2? eLife

Placental Pathology in COVID-19

A Structure-Informed Atlas of Human-Virus Interactions

Detection of severe acute respiratory syndrome coronavirus 2 in placental and fetal membrane samples

SARS-CoV-2 Infection of the Placenta

The MERS-CoV Receptor DPP4 as a Candidate Binding Target of the SARS-CoV-2 Spike. iScience

Inhibitors of cathepsin L prevent severe acute respiratory syndrome coronavirus entry

Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and its immune cross-reactivity with SARS-CoV

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Cathepsin L-selective inhibitors: A potentially promising treatment for COVID-19 patients

Daam1 is a regulator of filopodia formation and phagocytic uptake of Borrelia burgdorferi by primary human macrophages

The ORF8 Protein of SARS-CoV

Mediates Immune Evasion through Potently Downregulating MHC-I

Identification of putative interactions between swine and human influenza A virus nucleoprotein and human host proteins

Coronavirus replication and reverse genetics

Nsp3 of coronaviruses: Structures and functions of a large multi-domain protein

Congenital Zika syndrome is associated with maternal protein malnutrition

The Mechanisms of Placental Viral Infection

Profiling Gene Expression in Human Placentae of Different Gestational Ages: An OPRU* Network and UW SCOR Study

Metascape provides a biologist-oriented resource for the analysis of systems-level datasets

Understanding multicellular function and disease with human tissue-specific networks

Single-cell reconstruction of the early maternal-fetal interface in humans

Funding This research was supported by the National Council for Scientific and Technological Development, CNPq (Process 311530/2019-2 to RFC, and Process 310663/2018-0 to LAJ, and scholarship #870415/1997-2 to SSC

Process 2012/13961-6 to RFC

Supplementary Figures: Supplementary Figure 1. Gene expression profile of gestational trimesters and its potential to interact with SARS-CoV-2 proteins. A. Heatmap illustrates the expression pink bar, and cluster 3 by the blue bar. B. Functional enrichment analysis generated in the Metascape tool

Circus plot showing overlapped functional enriched terms from the gene lists of the clusters (1, 2 and 3). E. Circus plot showing the overlapping between the functional enriched terms generated from DEGs of first versus second trimester and DEGs from first trimester versus term placentas. F. Venn Diagram showing the DEGs shared by the placenta (up and down), human protein of expected interaction with SARS-CoV, and human protein of expected interaction with ZIKA virus

Dimensionality reduction demonstrates different cell populations identified in five human placental samples in a Uniform Manifold Approximation and Projection (UMAP) plot. SCT, syncytiotrophoblast; VCT, villous cytotrophoblast; EVT, extravillous trophoblast

) functionalities. RMA: Robust Multichip Average

The author declares no competing interests.

Not applicable as we used publicly available data.