Introduction

The current outbreak of the novel severe acute respiratory syndrome (2019-nCov then Covid-19) due to severe acute respiratory syndrome coronavirus 2 (SARS-Cov-2) started in Wuhan, China, in late December 2019 and has spread to many other countries [1,2,3,4]. To date, more than 84,000 cases and more than 4600 deaths have been reported across China due to SARS-Cov2, mostly in the region of Hubei (WHO, [5]). SARS-Cov-2 has disseminated in 188 countries, with currently more than 52 million confirmed cases and 1 million deaths around the world.

Coronaviruses are enveloped, positive single-stranded large RNA viruses that infect also a wide range of animals. The first description of coronavirus was made in 1966 by Tirell and Bynoe, who cultivated the viruses from patients with colds [6]. They were named coronavirus because of their morphology, spherical virions with a core shell and surface resembling to a solar crown, in Latin corona. Coronaviruses are divided into 4 subfamilies alpha, beta, delta and gamma-coronaviruses. The first two originate from mammals, in particular bats, while the other two come from pigs and birds. The genome size of coronaviruses ranges from approximately 27 to 34 kilobases. Severe disease and fatalities are caused essentially by beta-coronaviruses, whereas alpha-coronaviruses cause asymptomatic or mildly symptomatic infections. SARS-CoV and Middle East Respiratory Syndrome coronavirus (MERS-CoV) belong to the beta-coronavirus cluster [7], as well as the SARS-CoV-2 [8].

In this crisis situation, isolation of causative virus is indispensable for developing and evaluating diagnostic tools and therapeutics assays. The first isolation of SARS-CoV-2 was performed on human airway epithelial cells in China [8]. Subsequently, like SARS-CoV and MERS-CoV [9, 10], SARS-CoV-2 was isolated on VERO cells, which are kidney epithelial cells extracted from African green monkey [11,12,13]. In this paper, we investigated the susceptibility of a number of cells lines available in our laboratory collection to SARS-CoV-2. These cells were derived from a variety of species and tissues routinely used for the culture of micro-organisms. After inoculation with SARS-CoV-2, cells were observed for cytopathic effects and quantitative real-time polymerase reaction was used to measure ongoing replication on the cells growing the virus.

Materials and methods

Virus routine propagation

SARS-CoV-2-IHUMI2, IHUMI3, IHUMI669, and IHUMI2123 strains were isolated from human nasopharyngeal swab as previously described [14] and used for all tests. The 4-passage strains were grown in VERO E6 (ATCC® CRL-1586™) before subculture in different cell lines in Minimum Essential Medium culture medium (Ref. 21090022, Thermo Fisher Scientific) with 4% fetal calf serum (Ref. 10270106, Thermo Fisher Scientific) and 1% L-glutamine (Ref. 25030024, Thermo Fisher Scientific), without antibiotics at 37 °C under 5% CO2.

Multiple cell line assays

The cell lines tested are listed in Table 1. These cells are either routinely or occasionally used for microorganism isolation or for various diverse research projects in our laboratory. Cell lines to be tested were inoculated in 96-well flat bottom microplates (Ref. 020035, Dutscher) at 2*105 cells/ml into their specific growth medium (Table 1), without antibiotics and incubated to reach sub-confluence. At this stage, cells were infected with SARS-CoV-2 strains at 10−1 and 10−4 dilutions of VERO E6 supernatant, after filtration of 0.2 μ. Each day, cells were observed for SARS-CoV-2 specific cytopathic effects (CPE) for 7 days using EVOS™ FL Digital Inverted Fluorescence Microscope (Invitrogen™) with × 10 objective. On day 0 and day 7 after infection, supernatants were collected for subsequent quantification using reverse transcription-polymerase chain reaction (RT-PCR) targeting E-gene as previously described [16]. The difference of Ct between day 0 and Day 7 correspond to the Delta Ct (Δ Ct). Each condition was performed in triplicate. For cells for which a CPE effect was observed, cells were incubated with 10-fold dilutions of different viruses and incubated for 7 days at 37 °C. Each condition was performed in quadruplicate. Median tissue culture infectious dose (TCID50) was determined by the Reed and Muench method [17]. All experiments involving SARS-CoV-2 cultures were carried out in a Biosafety level 3 laboratory and conducted under appropriate conditions.

Table 1 Cell lines tested for their susceptibility to SARS-CoV-2
Full size table

Results

Table 1 presents the panel of 34 cell lines present in the laboratory and tested for their susceptibility to the SARS-Cov-2 virus. Among these cell lines, 7 are able to support SARS-CoV-2 multiplication and are presented in Table 2. For these seven cell lines that supported growth of the virus, the Δ Ct between day 0 and day 7 at dilution 10−1 varied between 3.85 and 9.76, as shown in Table 2. Besides VERO E6 in which the virus was isolated and propagated, 4 African green monkey kidney cell lines supported replication of SARS-CoV-2 (VERO 81, VERO SLAM, MA104, and BGM cells) and produced CPE 48 h after SARS-CoV-2 infection. All produced evident CPEs. One human cell line supported virus replication, an epithelial cell line from colorectal adenocarcinoma (Caco-2). Caco-2 showed only discrete modification as compared to control but no real CPE.

Table 2 Tested cell lines permissive to SARS-CoV-2
Full size table

The morphological changes observed in the different cell lines are shown in Fig. 1.

Fig. 1
figure 1

Morphological changes observed in the different cell lines. a Noninfected VERO cells (× 10). b SARS-Cov-2 infected VERO cells at 48 h post-infection (× 10). c Non infected E6 VERO cells (× 10) d SARS-Cov-2 infected E6 VERO cells at 48 h post-infection (× 10). e Noninfected VERO/hSLAM cells (× 10). f SARS-Cov-2 infected VERO/hSLAM cells at 48 h post-infection (× 10). g Noninfected MA104 cells (× 10). h SARS-Cov-2 infected MA104 cells at 48 h post-infection (× 10). i Noninfected BGM cells (× 10). j SARS-Cov-2 infected BGM cells at 48 h post-infection (× 10)

Full size image

LLC-MK2, a rhesus macaque epithelial kidney cell line, did not produce evident CPE.

For these seven cell lines that supported growth of the virus, the Δ Ct between day 0 and day 7 at dilution 10−4 varied between 9.37 and 17.32 as shown in Table 2. Viral multiplication was not associated with the intensity of CPE. Interestingly, according to the strain tested, major differences could be observed between permissivity, level of multiplication, and TCID50. In example, strain IHUMI2 as nearly always the highest TCID50 in spite its level of multiplication tested by RT-PCR can be lower. Twenty-seven other cell lines, derived from various species like insect, human, rodent, bovine, dog, sheep and bat cell lines, did not present any morphological changes or CPE and no difference of Δ Ct was observed.

Discussion

In the context of the SARS-CoV-2 epidemic, it was first important to develop rapid methods to isolate the virus. This was done easily using the common VERO E6 cell line, a highly virus permissive interferon deficient cell line [18]. In order to produce the virus in large quantities for vaccine research, to identify potential antiviral compounds, to understand intracellular trafficking and to develop innovative therapeutic approach, it is important to have other cell line, especially from human origin. In this paper, we explored the species specificity and tissue tropism of SARS-CoV-2 in vitro on a panel of cells available in our microbiology laboratory and identified human and animal cell lines susceptible to support SARS-CoV-2 multiplication.

Previous published reports showed that several monkey kidney cell lines are susceptible to SARS-CoV-2, specifically classical VERO cells, VERO E6 cells, VERO h/SLAM cells [8, 11,12,13, 19,20,21,22]. In this paper, we showed that all kidney cells derived from two species of monkey (African green monkey and rhesus macaque) support the growth of SARS-CoV-2, and all these cells, except for LLC-MK2 cell lines, presented CPE at 48 h post-infection. Unsurprisingly, MA104, BGM and LLC-MK2 already tested for SARS-CoV with very early CPE [23] and not previously tested with SARS-CoV-2, supported its growth.

In our first tests, HEP-2, an endothelial cell line suspected to be derived from laryngeal epidermoid carcinoma but in fact a clone derived from HELA cells [24], was first identified as susceptible to SARS-CoV-2 infection. SARS-CoV2 infection on our HEP-2 cells induced CPE after 120 h of infection with high virus multiplication. This result was unexpected, as previous studies on SARS-CoV showed that this virus did not infect HEP-2 cell lines, with no observable CPE or virus multiplication [23]. Based on PCR amplification of HPV18, a virus that chronically infects HEP-2 and HELA cells, we could detect that our HEP-2 cell line was misidentified. A new batch of cells ordered to the ATCC allowed to confirm that HEP-2 cells do not support growth of SARS-CoV-2.

A unique human cell line, Caco-2, epithelial cells from colorectal adenocarcinoma, was susceptible to SARS-CoV-2 with medium virus multiplication, but no specific CPE. Instead of CPE, we observed that the cell layer appears to be mottled more rapidly than in the control. This effect is rather seen in aging uninfected Caco-2. Previous studies showed that SARS-CoV and SARS-CoV-2 can infect Caco-2 cell lines [25, 26]. For SARS-CoV infections, CPE appeared on Caco-2 cell line 48 h post-infection [26], whereas, as observed, no obvious cell damage was found for SARS-CoV-2 infections [25]. This capability of SARS-CoV-2 to infect Caco-2 cells, could explain why patients infected with the virus present commonly gastrointestinal symptoms [27]. Moreover, SARS-CoV-2 RNA was detected in stools of patients infected with the virus, raising the question of viral gastrointestinal infection and fecal-oral transmission routes [28, 29]. However, to our knowledge, the virus could not be isolated from stools of infected patients.

We showed that 8 other human cells lines were not susceptible to SARS-CoV-2 (HEP-2, HT-29, HELA, HCT-8, ECV-304, HL-60, MRC5, and THP1 cell lines). In a recent paper of Chu et al. [25], SARS-CoV-2 was inoculated on 9 human cell lines. They showed that SARS-CoV-2 replicates also on Calu3 (lung adenocarcinoma), Huh7 (hepatocellular carcinoma), U251 (glioblastoma), and 293T (embryonic kidney) cell lines, whereas no growth was observed on A549 (lung adenocarcinoma), HFL (embryonic lung fibroblasts), and RD (rhabdomyosarcoma) cell lines. These data are consistent with the results observed in our study.

In this latter study, they evaluated the cell tropism profile of SARS-CoV-2 in non-human and non-primate cells originating from different animal species and showed that SARS-CoV-2 replicate in cat (feline kidney CRFK cells), rabbit (RK-13 rabbit kidney cells), and pig cells (PK-15 porcine kidney cells). In our study, we evaluated the susceptibility of SARS-CoV-2 in 19 animal cell lines. SARS-CoV-2 did not infect insect cells (Aa23, C6/36, S2, ISE6, and IPL-LD-65Y cells), rodent cells (BHK-21, McCoy, L929, P388 D1, and RAW 264.7 cells), bovine cells (BA886), bat cells (R05T, R06E, TB1 Lu cells), frog cells (XTC-2), dog cells (DH-2, MDCK cells), and sheep cells (OA3.Ts, MDOK cells).

Cellular entry of coronaviruses depends on the binding of the spike (S) protein to a specific cellular receptor and subsequent S protein priming by cellular proteases. Similarly to SARS-CoV [30, 31], SARS-CoV-2 seems to employ angiotensin-converting enzyme 2 (ACE2) as a receptor for cellular entry, and priming to be performed by the Transmembrane Serine Protease 2(TMPRSS2) [20, 32, 33]. This likely explains the specific permissivity of animal and kidney cell lines to the virus. ACE2 is expressed in various human tissues, such as heart, kidney, and testes, in addition to the lungs [34], indicating that SARS-CoV-2 may infect other tissues aside from the lungs. Moreover, Zhou et al. demonstrated that overexpressing ACE2 from different species in HeLa cells with human ACE2, pig ACE2, civet ACE2 (but not mouse ACE2) allowed SARS-CoV-2 infection and replication [20]. Hoffmann et al. reported similar findings for human and bat ACE-2 [35]. Additionally, Hoffmann et al. showed that treating VERO E6 cells, a monkey kidney cell line known to permit SARS-CoV replication, with an Anti-ACE-2 Antibody blocked the entry of vesicular stomatitis virus pseudotypes expressing the SARS-CoV-2 S protein [35]. A recent study conducted by Wang et al. reported that the existence of the novel SARS-CoV-2 (CD147-SP) route in host cells [36]. All these data suggest that SARS-CoV-2 is able to infect different tissues in human, but is also able to infect animals, and these information are concomitant with the variety of cell line that SARS-CoV-2 is able to infect. Finally, using 4 different isolates sampled at different time during the epidemic, we can observe that result of growth measured by RT-PCR and TCID50 can vary greatly and independently. This is of paramount importance because most studies testing drugs, survival on surfaces and efficiency of disinfectants are usually based on test performed on a unique strain. Our work could be an explanation of some discrepant results obtained i.e. about survival of SARS-CoV-2 on different supports [37,38,39]. Our results suggest all phenotypic tests should rather include a panel of strains.