Effective SARS-CoV-2 replication of monolayers of intestinal epithelial cells differentiated from human induced pluripotent stem cells

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes severe acute respiratory symptoms in humans. Controlling the coronavirus disease pandemic is a worldwide priority. The number of SARS-CoV-2 studies has dramatically increased, and the requirement for analytical tools is higher than ever. Here, we propose monolayered-intestinal epithelial cells (IECs) derived from human induced pluripotent stem cells (iPSCs) instead of three-dimensional cultured intestinal organoids as a suitable tool to study SARS-CoV-2 infection. Differentiated IEC monolayers express high levels of angiotensin-converting enzyme 2 and transmembrane protease serine 2 (TMPRSS2), host factors essential for SARS-CoV-2 infection. SARS-CoV-2 efficiently grows in IEC monolayers. Using this propagation system, we confirm that TMPRSS2 inhibition blocked SARS-CoV-2 infection in IECs. Hence, our iPSC-derived IEC monolayers are suitable for SARS-CoV-2 research under physiologically relevant conditions.

Next, we checked the expression levels of the host factor genes involved in SARS-CoV-2 infection, such as ACE2, FURIN, CTSL, TMPRSS2, and TMPRSS4, in IEC#17 monolayers, using qRT-PCR. During IEC differentiation, the mRNA expression of each gene gradually increased (Fig. 1B); particularly, ACE2 transcription drastically increased. Moreover, the increase in TMPRSS2 transcription was remarkable. FURIN and CTSL transcriptions were also increased, but TMPRSS4 transcription did not increase. The protein expressions of ACE2 and TMPRSS2 in IEC monolayers on days 2, 4, and 6 post-differentiation were confirmed by immunofluorescence staining in the horizontal or vertical plane ( Fig. 1C−E). Although ACE2 protein expression was consistent with the mRNA expression, TMPRSS2 protein expression seemed to be lower than the mRNA expression (Fig. 1B,D). ACE2 was mainly located on the cell surface at the luminal side in polarized IEC#17 monolayers (Fig. 1E). Taken together, these data suggested that fully differentiated IEC#17 monolayers (6 days post-differentiation) expressed SARS-CoV-2-related genes, indicating their potential for effective SARS-CoV-2 infection.

SARS-CoV-2 growth in human iPSC-derived IECs.
Although it has been reported that enterocytes in human small intestinal organoids are susceptible to SARS-CoV-2 infection 12,13 , the susceptibility of 2D-cultured IEC monolayers to SARS-CoV-2 infection has not been evaluated. To investigate the infection of SARS-CoV-2 in IEC#17 monolayers, IEC#17 and Vero cells were seeded in 96-well and 24-well plates, respectively. The cells were infected with SARS-CoV-2 at a multiplicity of infection (MOI) of 0.1. At each time point, supernatants were collected, and the viral titers were measured by determining the median tissue culture infectious dose (TCID 50 ). SARS-CoV-2 grew well in both cell types, and the viral titers peaked at 48 h post-infection (hpi) ( Fig. 2A). Although the peak viral growth rate was higher in Vero cells (ΔTCID 50 /mL: 10 9 vs. 10 6 ), IEC#17 showed comparable levels of viral growth to Vero cells at 24 hpi (ΔTCID 50 /mL: ≈10 5 ). In addition, the replication rate of the virus in IEC#17 was remarkably higher than that previously reported in human cell lines or organoids [12][13][14]25 .
SARS-CoV-2 particles were visualized using transmission electron microscopy (TEM). IEC#17 cells were seeded onto Matrigel-coated Transwell membranes or plastic coverslips (Cell Desk) and infected with SARS-CoV-2 at an MOI of 0.1. The infected cells in Transwells and Cell Desks were observed vertically and horizontally, respectively (Fig. 2B). SARS-CoV-2 particles were visible in IEC#17 monolayers from both angles (Fig. 2C,D). Budding virus particles were mainly observed at the apical side of the IEC#17 cells (Fig. 2C, red square). Viral replication organelles (ROs) were also observed (Fig. 2C, blue square), indicating an active viral cycle. Although some secretions were observed to be released from the pore in the Transwell membrane, budding virus particles were not observed at the basal side of the IEC#17 cells within our field of view (Fig. 2C, cyan square). Horizontally, several SARS-CoV-2 particles that replicated inside the cells were clearly observed per cell (Fig. 2D). Budding virus particles were also observed. Although clear visible ROs composed of a membrane were not observed, we found assembled viruses (Fig. 2D). These observations indicated that SARS-CoV-2 replication and budding were active in IEC#17. Further, immunostaining revealed the SARS-CoV-2 nucleoprotein (NP) and SP www.nature.com/scientificreports/ in post-infected IEC#17 (Fig. 2E). Interestingly, SP-positive IECs appeared to have downregulated expression of ACE2 (Fig. 2F). These data supported that SARS-CoV-2 could infect and replicate in IEC#17 monolayers. Therefore, we propose that IEC#17 is suitable for SARS-CoV-2 research.
Comparison of the propagation efficiency of SARS-CoV-2 among IEC#17 cells, other iPSC-derived IECs, and tissue-derived primary IECs. It has been reported that SARS-CoV-2 can replicate in tissue-derived 3D-cultured IECs 13 . We directly compared the growth of SARS-CoV-2 in primary IECs derived from jejunum, ileum, and colon tissues with that in IEC#17 cells, which were in a monolayered state. The viral titers in all primary-derived IECs were significantly lower than those in IEC#17 (Fig. 3A). This suggests that IECs derived from human iPSCs are more susceptible to SARS-CoV-2 infection than tissue-derived IECs.   www.nature.com/scientificreports/ To confirm the effective replication of SARS-CoV-2 caused by the culture system of iPSC-derived IEC monolayers, we next investigated the possibility of in vitro propagation of SARS-CoV-2 in other IECs differentiated from other human iPSC lines. We established IEC#20, IEC#25, and IEC#29 from the human iPSC lines 1231A3, 1383D6, and TkPP7, respectively. Since the viral titers reached the peak at 48 hpi in IEC#17 ( Fig. 2A), we comparatively analyzed the viral titers at 48 hpi in these IECs. Although the viral titers in IEC#29 were as high as those in IEC#17, they were lower in IEC#20 and IEC#25 cells (Fig. 3B). To determine the characteristics of these IECs, qRT-PCR and immunostaining were conducted. The results showed that there were no marked differences in IEC #20 and IEC #25 compared with those in IEC #17, except for a slight increase in CHGA expression ( Supplementary Fig. 1). IEC#29 also showed gene expression changes similar to those of IEC#17, in addition to exhibiting a slight increase in MUC2 expression. The protein expression of ACE2 and TMPRSS2 was increased at 6 days post-differentiation, similar to the observation in IEC#17; however, the expression of ACE2 in IEC#20 was lower than that in the other IECs ( Supplementary Fig. 2). These results suggest that not only IEC#17 but Influence of the inhibition of TMPRSS2 in IEC#17. The most important advantage of using IEC#17 monolayers for SARS-CoV-2 research is their ability to replicate the in vivo condition and perform quantitative cellular assays. To investigate the interferon (IFN) response, mRNA transcripts in IEC#17 infected with SARS-CoV-2 were quantified by qRT-PCR. As the IFN response against SARS-CoV-2 in IEC#17 was normal (Supplementary Fig. 3), IEC#17 could be expected to be available for several analyses. To evaluate whether IEC#17 monolayers are suitable for drug screening, cells were pretreated with the TMPRSS2 inhibitors camostat and nafamostat before infection with SARS-CoV-2. These inhibitors prevent the cleavage of the SP, which is required  Each value is representative of at least three independent experiments and is shown as the mean ± SD from three wells of supernatants of each culture group. The significant differences were determined using one-way ANOVA. The significant differences were determined using one-way ANOVA. (B) IEC#17 monolayers were untreated or treated with DMSO, 10 µM camostat and nafamostat for 1 h. After incubation, the pretreated cells were infected with 2 × 10 6 genome equivalents of norovirus GII.4. The supernatants of infected IEC#17 cells were collected at 3 and 72 hpi. The viral RNA copies of norovirus GII.4 in IEC#17 were measured by qRT-PCR. Each value is representative of at least three independent experiments and is shown as the mean ± SD from between four and six wells of supernatants of each culture group. The significant differences were determined using one-way ANOVA. *0.01 < P < 0.05. www.nature.com/scientificreports/ for the membrane fusion of SARS-CoV-2 with the host cell, thereby inhibiting viral replication. As expected, both inhibitors completely inhibited SARS-CoV-2 infection at a concentration of 10 µM (Fig. 4A). Nafamostat can also inhibit replication at a concentration of 0.1 µM. To confirm the conservation of cell properties, the inhibitory assay was conducted using human norovirus (NoV) as a negative control. Propagation of human NoV GII.4, which infects IECs in a TMPRSS2-independent manner, was not affected by the two inhibitors (Fig. 4B). Altogether, these findings suggest that monolayered IEC#17 cells are a suitable alternative to alveolar epithelial type II cells and that IECs more closely resemble native cells than Vero cells.

Discussion
Recently, the number of studies investigating SARS-CoV-2 in vitro and in vivo has drastically increased. However, most in vitro studies use Vero cells or human cells with forced expression of ACE2 (e.g., HEK293T/ACE2) as SARS-CoV-2 grows efficiently in these cells. SARS-CoV-2 also grows in several human cell lines, such as Calu-3 and Caco-2. These cell lines highly express ACE2, the receptor of SARS-CoV-2. Pseudotyped vesicular stomatitis virus containing the SARS-CoV-2 SP grows to comparable titers in these cell lines 26 . 3D-cultured IECs and lung organoids and monolayered alveolar type II-like cells are also susceptible to SARS-CoV-2 infection [10][11][12][13][14] . However, SARS-CoV-2 grew much less effectively in these human primary cells than in Vero cells. Our established IEC#17 monolayers enabled SARS-CoV-2 to grow effectively; however, the peak viral titer in IEC#17 was still lower than that in Vero cells. Although the well-differentiated IECs are unable to grow further, some of the non-infected or low-infected Vero cells grow and can support virus replication. Thus, the viral titer of SARS-CoV-2 in Vero cells was higher than that in IEC#17. In addition, it should be noted that Vero cells are type I IFN-deficient 27 . The circular polymerase extension reaction, a great tool for producing recombinant SARS-CoV-2, was established using HEK293-3P6C33 cells, in which IFN alpha and beta receptor subunit 1 (IFNAR1) has been knocked out 28 .
These findings indicate that SARS-CoV-2 grows better in cells with an impaired IFN signaling is required to grow SARS-CoV-2 to high titers. Interestingly and importantly, IEC#17 monolayers induced type I IFNs, especially IFN alpha (IFNA), in response to SARS-CoV-2, indicating that IEC#17 has an intact IFN response. These observations could explain why SARS-CoV-2 grew better in Vero cells than in IEC#17. Therefore, knockdown or knockout of the IFNAR1 gene might further enhance viral replication in IEC#17. Furthermore, IEC#17 appears to be a suitable human cell model to analyze type I IFN responses induced by SARS-CoV-2.
Although previous studies have reported SARS-CoV-2 infection in 3D-cultured IEC organoids 12,13 , the monolayers from iPSC-derived IECs showed a more efficient SARS-CoV-2 growth. The difference between previous reports and our findings might be because we used iPSC-derived IEC monolayers since pluripotent stem cellderived IECs have been reported to be similar to fetal, but not adult, IECs 29 . It was suggested that differentiated fetal IECs are more susceptible to SARS-CoV-2 infection than differentiated adult tissue-derived IECs. Notably, fetal-derived and/or pluripotent stem cell-derived IECs are more susceptible to SARS-CoV-2 infection, but in general, COVID-19 is more likely to cause severe symptoms in the elderly [30][31][32] .
3D-cultured organoids serve as an excellent model to reflect in vivo conditions. However, as with bronchial epithelial cells 33 , SARS-CoV-2 is expected to easily invade and bud from the apical side of IECs. Therefore, 3D-cultured organoids will be slightly difficult to analyze for apical infection, and 2D-cultured monolayer cells seem to be more suitable than organoids. We previously reported that the expression levels of several intestinal markers and functional genes of IECs do not differ between 2D-and 3D-cultured conditions 18 . Although the reasons why the growth of SARS-CoV-2 was better in the monolayer IECs than 3D-cultured ones are still unclear, one possibility is that the luminal side of 3D-cultured IECs is enclosed, and replicating viruses and cellular metabolites accumulate in a limited space, leading to a rapid increase in their concentration. SARS-CoV-2 is an enveloped virus, which is sensitive to unfavorable environmental conditions, such as unsuitable pH 1 . Therefore, the luminal side of 3D-cultured IECs might die faster after viral infection than that of 2D-cultured IECs. Thus, the 2D-cultured monolayer IECs will contribute to a more easy and effective analysis of the virus.
SARS-CoV-2 growth in tissue-derived IECs was lower even under 2D culture conditions than that in IEC#17 monolayer. Regarding the comparison with primary cultured cells, we have recently reported that the characteristics of IECs in monolayer culture are indistinguishable between those derived from iPSCs (TkDN4-M) and those derived from small intestinal tissue 19 . In addition, SARS-CoV-2 growth varied in IEC monolayers differentiated from different iPSCs. Although the expression of SARS-CoV-2 infection-related genes was analyzed, critical differences were not observed among various IECs. Histo-blood group antigens are highly expressed in the intestine, and some enteric viruses (e.g., NoV and rotavirus) use these glycans for infection 34,35 . The H antigen expression in IECs requires fucosyltransferase 2 (FUT2) activity. IEC#17 and IEC#29 were derived from human iPSCs whose FUT2 activity is intact. In contrast, the FUT2 gene in IEC#20 and #25 is mutated, rendering it inactive (Sato et al., unpublished data). Thus, the fucosylation on the apical side of IECs by FUT2 might influence the susceptibility to SARS-CoV-2. In addition, iPSCs may differ in their properties as pluripotent stem cells depending on the origin of the original somatic cells and the method of reprogramming, which may affect the properties of the differentiated IECs. However, further studies are required to elucidate the differences in responsiveness of different IECs to SARS-CoV-2 infection and propagation.
In conclusion, differentiated human iPSC-derived IECs might reflect the in vivo condition better than Vero cells, which are type I IFN-deficient monkey cells. SARS-CoV-2 grew better in 2D-cultured IEC#17 cells than in 3D-cultured IECs and 2D-cultured primary IECs. In vitro assays using IEC#17 monolayers are the easiest way to characterize severe COVID-19. Furthermore, IEC#17 monolayers are useful for screening drugs and neutralization antibodies and investigating the type I IFN response during SARS-CoV-2 infection.
Viruses. The ancestral SARS-CoV-2/Hu/DP/Kng/19-020 strain (GenBank accession no. LC528232; lineage B) was propagated in VeroE6/TMPRSS2 cells. The supernatant of infected cells at an MOI of 0.1 was collected at 48 hpi and centrifuged at 1600×g at 4 ℃ for 5 min to remove cell debris. The supernatant was stocked at − 80℃ until used. The viral titer was determined using the TCID 50 method, as described below. To inoculate the virus at the appropriated MOI, the TCID 50 was converted to PFU using a previously reported algorithm 39 . SARS-CoV-2 was handled in a biosafety level 3 facility. NoV GII.4-positive stool was dissolved in phosphate-buffered saline (PBS) at 10% (w/v) by vigorous vortexing. After centrifugation and filtration 38 , the supernatants were prepared for infection. Growth kinetics. Vero cells were seeded in a 24-well plate at 2 × 10 5 cells per well. The differentiated IEC#17 cells were seeded in a 96-well plate at approximately 1 × 10 5 cells per well. The virus solution was diluted appropriately with DMEM containing 2% FBS. The cells were washed with PBS and infected with SARS-CoV-2 at an MOI of 0.1 for 1 h at 37 °C. After inoculation, the cells were washed with PBS twice, and 100 µL of DMEM containing 5% FBS or differentiation medium was added for Vero cells or IEC#17, respectively. Supernatants were collected at the indicated time points, and the TCID 50 was determined. Inhibition assay. IEC#17 cells were seeded at a concentration of 1 × 10 5 per well in a 96-well plate and pretreated with the inhibitors, camostat (Cayman Chemical Company, MI, USA) and nafamostat (Cayman Chemical Company) for 1 h before virus infection. Dimethyl sulfoxide (DMSO; Sigma-Aldrich, MO, USA) was used as a control. After removal of the inhibitors, the cells were washed three times and infected with SARS-CoV-2 at an MOI of 0.1 or 2 × 10 6 genome equivalents of NoV GII.4 for 1 h at 37 °C. At 48 hpi, the supernatant was collected, and the viral titers of SARS-CoV-2 and NoV GII.4 were measured by the TCID 50 method and qRT-PCR, respectively. Each value is representative of at least three independent experiments and is presented as the mean ± SD from three wells of supernatants of each culture group.

qRT-PCR. Total
Statistical analysis. Two-way ANOVA was used to determine the significant differences in the growth kinetics in IECs. For the other experiments, one-way ANOVA was used. The statistical analysis was conducted using GraphPad Prism 9 software (GraphPad Software).

Ethics declarations.
The experiments using primary human organoids and NoV-positive human stools were performed in accordance with the relevant guidelines and were approved by the human ethical committee of Osaka University (approval #27-5 and #28-3), Nihon University (approval #2022-05), and Wakayama Medical University (approval #3565, #3566, and #3716). We received written informed consent of all participants. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.