key: cord-1026495-9k8zseao
authors: Cinatl Jr., J.; Hoever, G.; Morgenstern, B.; Preiser, W.; Vogel, J.-U.; Hofmann, W.-K.; Bauer, G.; Michaelis, M.; Rabenau, H. F.; Doerr, H. W.
title: Infection of cultured intestinal epithelial cells with severe acute respiratory syndrome coronavirus
date: 2004
journal: Cell Mol Life Sci
DOI: 10.1007/s00018-004-4222-9
sha: 109974d5ee96b2e0c4d14b314d4d415dfb1578bc
doc_id: 1026495
cord_uid: 9k8zseao

To identify a model for the study of intestinal pathogenesis of severe acute respiratory syndrome (SARS) we tested the sensitivity of six human intestinal epithelial cell lines to infection with SARS coronavirus (SARS-CoV). In permissive cell lines, effects of SARS-CoV on cellular gene expression were analysed using high-density oligonucleotide arrays. Caco-2 and CL-14 cell lines were found to be highly permissive to SARS-CoV, due to the presence of angiotensin-converting enzyme 2 as a functional receptor. In both cell lines, SARS-CoV infection deregulated expression of cellular genes which may be important for the intestinal pathogenesis of SARS.

Cells infected with SARS-CoV at multiplicity of infection (MOI) 1 and MOI 10 were collected at different times post infection (p. i.) by trypsinization of adherent cells. Non-adherent cells (the numbers of which increased with time after infection) were collected by centrifugation of culture supernatants. Both adherent and non-adherent cells were fixed on glass slides with 60/40 methanol/acetone for 15 min. Immune peroxidase staining was performed using human immune serum obtained from a SARS patient as described previously [9] .

To investigate whether ACE2 is a functional receptor for SARS-CoV in intestinal epithelial cell cultures, the cells were pre-treated for 60 min at 37°C with goat antibody directed against the human ACE2 ectodomain (R&D Systems; Wiesbaden-Nordenstadt, Germany). After treatment, the cells were washed three times with phosphatebuffered saline (PBS) and infected with one of the SARS-CoV strains at MOI 1. Twenty-four hours p. i. the cells were fixed and stained for viral antigens as described above. Goat anti-ACE1 antibody (R&D Systems) was used as control. Both antibodies were added at a concentration of 50 µg/ml.

To investigate expression of cell surface ACE2, intestinal cell lines were washed twice with PBS and incubated for 30 min with goat anti-ACE2 antibody (R&D Systems). After washing with PBS, the cells were incubated with FITC-conjugated anti-goat IgG (Becton Dickinson, Heidelberg, Germany) for 30 min. As controls, cells were stained with irrelevant primary antibody (goat antimouse IgG; Sigma Biochemicals, Seelze, Germany) or without a primary antibody to determine unspecific and background fluorescence, respectively. Instrument settings of the flow cytometer (FACScan; Becton Dickinson) were adjusted to obtain background mean fluorescence in the histogram mode between 1 and 10 on the logarithmic scale.

Caco-2 cells were infected 2 -3 days after reaching confluence with SARS-CoV at MOI 1. One day p. i., the cells were processed for ultrastructural analysis as described previously [10] . Briefly, cells were pelleted and fixed with 2.5 % glutaraldehyde, post-fixed in 1 % osmium tetroxide, dehydrated in ethanol and embedded in Durupan-Epon. Thin sections were contrasted with uranyl acetate and lead citrate and viewed with a Jeol JEM, 2000 CX electron microscope (Arishima, Japan).

To assess effects of SARS-CoV infection on Caco-2 cell viability, confluent cell layers in 96-well plates were infected at MOI 1 and MOI 10. The viability was measured at different times p. i. using the MTT assay performed as described previously [10] .

Gene array analysis was done according to the principles of Miame [11] We used the Affymetrix HG-U133A chip (Affymetrix, Santa Clara, Calif.). This oligonucleotide microarray targets 22,000 genes. Sample preparation was done by the RNeasy Mini Kit (Qiagen, Hilden, Germany) standard protocol. Generation of biotin-labelled cRNA, hybridization and staining were done according to standard protocols available from Affymetrix. Data analysis was performed using Microarray Analysis Suite (Affymetrix) and GeneSpring software version 4.0 (Silicon Genetics, San Carlos, Calif.) as published previously [12] . In brief, the lowest raw data value was arbitrarily defined as '11' in order not to eliminate genes which are expressed only in one sample. To eliminate false 2102 J. Cinatl Jr. et al.

'fold-change' calls, genes that were classified as 'up-regulated' had to be flagged as 'present' in the infected samples, while genes that were classified as 'down-regulated' had to be flagged as 'present' in the mock-infected samples. Within those parameters, genes were selected if they were either up-or down-regulated at least threefold in duplicate. Following microarray analysis, genes related to apoptosis, cytokines, chemokines or interferons were confirmed by RT-PCR, according to standard protocols [10] . PCR primer and amplification conditions were determined by the software Primer3 (Whitehead Institute for Biomedical Research, Cambridge, Mass.) [13] .

Previously, we demonstrated that Caco-2 cells are highly permissive to infection with SARS-CoV strain FFM-1 [8] . 

Since ACE2 was identified as a functional SARS-CoV receptor in different cell types [3] we measured whether its expression may correlate with the sensitivity of intestinal cell lines to SARS-CoV infection. Caco-2 and CL-14 expressed ACE2 mRNA and protein which were not detectable in the other cell lines tested ( fig. 3 A 

The relative abundance of specific mRNA in SARS-CoVinfected cells was compared to mock-infected confluent Caco-2 cell cultures (same passage and identical culture conditions) 24 h p. i. when cell viabilities were similar ( fig. 2 B) . All gene expression experiments were done in duplicate and only genes which were up-or down-regulated in both samples underwent further evaluation. After applying strong restrictions as described in Materials and Methods, resulting genes were grouped according to their function (table1). We focussed on genes related to apoptosis, chemokines, interferon-induced genes and transcription factors, since these gene groups may play an important role in the pathogenesis of SARS. Expression of the selected genes was confirmed by RT-PCR ( fig. 5 ). In the infected cells, we found an up-regulation of some anti-apoptotic genes including Bcl-2 (only in Caco-2 but not in CL-14 cells) and A20, while several pro-apoptotic genes including Bid, Bad, caspase-2 and caspase-6 were down-regulated.

On the other hand, the anti-apoptotic programmed cell death 4 gene (PDCD4) was down-regulated in infected cells. Increased levels of mRNA of members of the AP-1 family of cellular transcription factors including c-jun . Some viral particles attach onto the microvilli, whereas some detach from the cell surface (E). c, cytoplasm; n, nucleus. and c-fos were observed in the infected cells. Concerning cytokine/chemokine-related genes, our results showed an up-regulation of several CXC chemokines; among the down-regulated genes we found interleukin (IL)-18 and macrophage migration inhibitory factor (MIF). Several interferon-induced genes were up-regulated, including the human 2¢-5¢ oligoadenylate synthetase 2 gene (OAS2) and human myxovirus resistance-1 gene (MXA). SARS-CoV strains FFM-1 and 6109 influenced similarly the expression of the selected genes ( fig. 5 ). Neither UV-inactivated virus nor virus-free filtered cell culture supernatants caused any changes in gene expression pattern compared to mock-infected cells (data not shown).

To provide an experimental model for the study of SARS gastrointestinal pathology, we tested the sensitivity of six intestinal cell lines to SARS-CoV infection. In addition to Caco-2 cells which were previously shown to be per-missive to SARS-CoV [8] , only CL-14 cells promoted SARS-CoV replication. CL-14 cells show features of well-differentiated enterocytes [14] while Caco-2 cells show an undifferentiated phenotype with the ability to undergo spontaneous enterocytic differentiation after reaching confluence [15] . We infected Caco-2 cells 2 -3 days after confluence, i. e. when electron microscopy identified mostly poorly differentiated enterocytes and only few well-differentiated (villus) enterocytes. Both poorly and well-differentiated enterocytes supported SARS-CoV replication suggesting that the sensitivity of intestinal epithelial cells does not depend on a particular stage of cellular differentiation. ACE2 has recently been shown to be a functional receptor for SARS-CoV [3] and surface ACE2 is abundantly present on enterocytes of the small intestine [16] . while it down-regulated pro-apoptotic genes such as Bid, Bad, caspase-2 and caspase-6. In a murine model, Bcl-2 overexpression in gut epithelial cells, decreased the apoptosis [17] and protected against intestinal injury [18] . Although Bcl-2 was detectable only in Caco-2 cells, other regulators of the apoptotic mitochondrial pathway including pro-apoptotic Bcl-2 homologues Bid and Bad were down-regulated by SARS-CoV in both Caco-2 and CL-14 cells. Since Bid and Bad are involved in the regulation of intestinal epithelial cell survival [19] , their role in SARS-CoV intestinal infection should be studied further. In both cell lines, SARS-CoV up-regulated A20 which may protect different cell types against tumour necrosis factor (TNF)-mediated programmed cell death and is critical for limiting inflammation by terminating TNF-induced nuclear factor (NF)-kB responses in the intestine and other organs [20] . In addition, down-regulation of caspase-2 and caspase-6 in infected intestinal cell lines may be of interest as these caspases were shown to be important mediators of apoptosis in gastrointestinal epithelium [21, 22] . The results show that SARS-CoV-infected epithelial cells develop an anti-apoptotic response which may be important to inhibit or delay destruction of infected enterocytes. These findings are consistent with clinical observations demonstrating a relatively normal endoscopic and microscopic appearance of the intestine in patients with SARS [5] .

On the other hand, SARS-CoV suppressed expression of the anti-apoptotic gene PDCD4 which is constitutively expressed in most normal tissues including lung and intestine [23] . Apart from its effects in the regulation of apoptosis, PDCD4 was shown to play a role in inhibition of translation by direct interaction with eukaryotic translation initiation factor 4A (eIF4A) [23, 24] . This finding is of interest since activity of eIF4E (together with eIF4A and eIF4G forming the translation initiation factor complex eIF4F) was shown to be important for replication of murine coronavirus [25] . Moreover, PDCD4 has the ability to suppress transactivation of activator protein (AP)-1 [23] . Since the SARS-CoV nucleocapsid was shown to activate the AP-1 transduction pathway [26] , it will be of interest to show whether there may be a mechanistic link between PDCD4 suppression and enhancement of AP-1 tranactivation in SARS-CoV-infected cells. The present observations demonstrate that SARS-CoV infection elevated mRNA levels of AP-1 subunits c-Fos and c-Jun in intestinal cells which could also increase AP-1 transactivation. Infection with most viruses up-regulates different interferon (IFN)-induced genes which may establish an anti-viral state within cells. Its major effectors and indicators include double-stranded RNA-dependent protein kinase (PKR), OAS and MX proteins [27] . SARS-CoV infection of Caco-2 cells up-regulated OAS2 and MXA but not PKR genes. The discrepancy between transcriptional activation of IFN-induced genes and the ability of SARS-CoV to replicate in Caco-2 cells could be explained by the existence of a specific viral mechanism for escaping IFNinduced anti-viral effects common to most viruses [28] . The enteropathogenic potential of HCoV-OC43 (strain Paris) has been suggested to be due to its inability to induce IFN-a [29] . Recently, we showed that IFN-a and IFN-b (type I IFN) inhibited SARS-CoV replication in Caco-2 cells while IFN-g (type II IFN) was not effective [8] . Moreover, IFN-b was 50-90 times more potent than IFN-a against different SARS-CoV strains. The differences in anti-viral activity of different types of IFN could result from their ability to differentially influence expression of cellular genes important for anti-viral activity. For example, treatment of the human fibrosarcoma cell line HT1080 (expressing both type I and type II IFN receptors) with IFN-b stimulated PKR which was not stimulated by IFN-a or IFN-g [30] . Since PKR was not up-regulated in the infected Caco-2 cells and virus replication progressed despite up-regulation of OAS2 and MXA, a role for PKR in SARS-CoV replication must be elucidated. Although enteric pathogens such as viruses, protozoans, multicellular helminths and enteroinvasive bacteria vary in their mode of infection, enterocytes display a common chemokine/cytokine profile in response to infection. Several ELR+CXC chemokines (containing a conserved glutamate-leucine-arginine sequence) including CXCL1 (groa), CXCL2 (grob), CXCL3 (grog) and CXCL8 (IL-8) were up-regulated in SARS-CoV infected Caco-2 cells. These chemokines mainly regulate neutrophil trafficking [31] . In addition, SARS-CoV induced in Caco-2 cells non-ELR CXC chemokines CXCL10/IFN-g inducible protein-10 (IP-10) and CXCL11/IFN-inducible T cell alpha chemoattractant (I-TAC) which are potent CD4+ T cell chemoattractants [32, 33] . Rotavirus infection was shown to induce CXCL1, CXCL8 and CXCL10 in intestinal cell lines [34] . Mucosal inflammation associated with rotavirus infection is predominantly mononuclear, i. e. consists of monocytes and T lymphocytes [35] , although neutrophil infiltration is found in some cases [34] . In contrast, biopsy specimens taken from the colon and terminal ileum of patients with SARS failed to demonstrate any inflammatory infiltrates [5] . Neutrophil infiltration in the intestine of SARS patients may be limited despite neutrophilia due to changes of cytokine/ chemokine levels in the intestinal environment. We observed that SARS-CoV infection of Caco-2 cells inhibited expression of IL-18 which is constitutively expressed in intestinal epithelial cells [36] . Suppression of IL-18 levels reduces neutrophil accumulation in liver and lungs [37] . The absence of T lymphocyte infiltration of the intestine in SARS may be a consequence of the profound decline of both CD4+ and CD8+ lymphocytes in the blood [38] , possibly resulting from lymphocyte apoptosis [39] . Although macrophage counts were increased in lungs [40] , macrophage infiltration was absent from the gut of SARS patients [5] . In Caco-2 cells, SARS-CoV down-regulated MIF. Recently, MIF was identified as a major factor produced by intestinal cells in response to microbial infection regulating macrophage emigration, inflammation and cell metabolism [41] . Some of the chemokines we found up-or down-regulated in vitro were also changed in serum samples from SARS patients. For example, serum levels of CXCL10 and IL-8 were increased whereas IL-18 was decreased [42, 43] . This justifies the use of intestinal cell lines as a model to study the direct effects of SARS-CoV infection on gene expression in permissive human cells. Given the intestinal tropism of SARS-CoV, the results presented here provide several important hints at possible mechanisms of intestinal pathogenesis and potential novel therapeutic targets in SARS.

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

Critically ill patients with severe acute respiratory syndrome

Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus

Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study

Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection

Clinical features and short-term outcomes of 144 patients with SARS in the greater Toronto area

Viral replication in the nasopharynx is associated with diarrhea in patients with severe acute respiratory syndrome

Treatment of SARS with human interferons

Glycyrrhizin, an active component of liquorice roots, and replication of SARS-associated coronavirus

Bovine seminal ribonuclease selectively kills human multidrug-resistant neuroblastoma cells via induction of apoptosis

Minimum information about a microarray experiment (MIAME) toward standards for microarray data

Relation between resistance of Philadelphia-chromosome-positive acute lymphoblastic leukaemia to the tyrosine kinase inhibitor STI571 and geneexpression profiles: a gene-expression study

Primer3 on the WWW for general users and for biologist programmers

Growing colorectal tumors: minimizing microbial and stromal competition and assessing in vitro selection pressures

Enterocyte-like differentiation and polarisation of the human colon carcinoma cell line Caco-2 in culture

Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus: a first step in understanding SARS pathogenesis

Inhibition of intestinal epithelial apoptosis and survival in a murine model of pneumonia-induced sepsis

Total parenteral nutrition-induced apoptosis in mouse intestinal epithelium: modulation by keratinocyte growth factor

Human intestinal epithelial cell survival: differentiation state-specific control mechanisms

Failure to regulate TNF-induced NF-kappaB and cell death responses in A20-deficient mice

Apoptosis-programmed cell death and its relevance to gastrointestinal epithelium: survival signal from the matrix

Hierarchical cleavage of focal adhesion kinase by caspases alters signal transduction during apoptosis of intestinal epithelial cells

Programmed cell death protein 4 (pdcd4): a novel target for antineoplastic therapy?

A novel function of the MA-3 domains in transformation and translation suppressor Pdcd4 is essential for its binding to eukaryotic translation initiation factor 4A

Murine coronavirus replication-induced p38 mitogen-activated protein kinase activation promotes interleukin-6 production and virus replication in cultured cells

Activation of AP-1 signal transduction pathway by SARS coronavirus nucleocapsid protein

Viruses and interferons

Is flavivirus resistance interferon type I-independent?

Comparison of the replication of distinct strains of human coronavirus OC43 in organotypic human colon cells (Caco-2) and mouse intestine

Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays

Immunobiology of epithelial chemokines in the intestinal mucosa

Chemokine receptor specific for IP10 and mig: structure, function, and expression in activated T-lymphocytes

Interferon-inducible T cell alpha chemoattractant (I-TAC): a novel non-ELR CXC chemokine with potent activity on activated T cells through selective high affinity binding to CXCR3

Rotavirus infection of cultured intestinal epithelial cells induces secretion of CXC and CC chemokines

Structural and functional abnormalities of the small intestine in infants and young children with rotavirus enteritis

IL-18, a novel immunoregulatory cytokine, is up-regulated in Crohn's disease: 2112 J. Cinatl Jr. et al. SARS-CoV infection of intestinal cells expression and localization in intestinal mucosal cells

Interleukin-18 and host defense against infection

Expression of lymphocytes and lymphocyte subsets in patients with severe acute respiratory syndrome

SARS: understanding the coronavirus. Apoptosis may explain lymphopenia of SARS

Lung pathology of fatal severe acute respiratory syndrome

Ubiquitous production of macrophage migration inhibitory factor by human gastric and intestinal epithelium

Dynamic changes in blood cytokine levels as clinical indicators in severe acute respiratory syndrome

Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome

Acknowledgements. The authors wish to acknowledge valuable technical support by G. Meincke, L. Stegmann and C. Sippel. The authors are grateful to H. Kabickova for transmission electron microscopy experiments.