Genetically Engineered Live-Attenuated Middle East Respiratory Syndrome Coronavirus Viruses Confer Full Protection against Lethal Infection

The emergence of the new highly pathogenic human coronavirus SARS-CoV-2 that has already infected more than 80 million persons, killing nearly two million of them, clearly indicates the need to design efficient and safe vaccines protecting from these coronaviruses. Modern vaccines can be derived from virus-host interaction research directed to the identification of signaling pathways essential for virus replication and for virus-induced pathogenesis, in order to learn how to attenuate these viruses and design vaccines.

The study of the molecular mechanisms of virus pathogenesis allows the identification of viral factors involved in virulence and virus-host interaction (67). Using reverse genetics systems, several virulence factors can be eliminated to obtain live-attenuated viruses in which the composition and genetic stability are precisely known (68)(69)(70)(71)(72), thus minimizing the risk of reversion to a virulent phenotype (73). Furthermore, these vaccines can be easily designed and prepared in a relatively short period of time as a quick response to outbreaks or epidemics.
Our group has described that the Envelope (E) protein of SARS-CoV is implicated in virulence through the interaction with several cellular proteins (69,74,75). These interactions have been studied in the context of SARS-CoV infection, and it is unknown whether they are reproduced by E proteins from other CoVs. The study of these interactions in CoVs such as MERS-CoV could help to identify potential targets for the development and application of effective treatments against infections caused by these viruses.
In this study, we describe the construction of a collection of mutants carrying partial deletions within the C-terminal domain of the MERS-CoV E protein. Among them, one mutant was attenuated and conferred protection with a single dose in a lethal MERS-CoV challenge. The attenuation of this mutant seemed related to the absence of one predicted linear motif: a ligand of forkhead-associated (FHA) domain.

RESULTS
Viability and selection of MERS-CoV-E* mutants. The C-terminal domain of the envelope protein (E) was analyzed with the Eukaryotic Linear Motif resource (http:// elm.eu.org/) to identify potential functional motifs. In addition, a multiple sequence alignment of 10 different CoV E proteins was performed to localize several conserved residues (https://www.ebi.ac.uk/Tools/msa/clustalo/). Combining the results of the two analyses, five MERS-CoV mutants (MERS-CoV-E*) were designed to contain deletions of 9 to 11 amino acids spanning the C-terminal domain exposed to the cytosol (Fig. 1). Each of the five mutations deleted different predicted linear motifs and conserved positions ( Fig. 1C and D; see also Table S1 in the supplemental material). One of these functional motifs was the PDZ binding motif (PBM), previously identified in our laboratory as a virulence factor in SARS-CoV infection (69). It has been demonstrated that these PBMs are highly relevant for the virulence of different viruses, like influenza virus (76) and rabies virus (77), among others (78). In silico, the engineered deletions or the substitutions did not generate new motifs in the resultant MERS-CoV-E* mutants (Table S1).
Three independent infectious clones (bacterial artificial chromosomes [BACs]) were rescued for each mutant: MERS-CoV-E*D1, E*D2, E*D3, E*D4, and E*D5. These viruses were passaged four times in Huh-7 cells every 72 h ( Fig. 2A). The original sequence was compared with the sequence of the virus from passage 4. Virus titers and viability were analyzed during these passages. Among the five mutants, only four were rescued: MERS-CoV-E*D1, E*D2, E*D4, and E*D5. Despite the E*D1 mutant being viable, its titers were 1 logarithmic unit lower than MERS-CoV-WT (wild type). E*D1 lysis plaques were smaller than MERS-CoV-WT (data not shown). Although the E*D4 mutant was rescued, it was lost through the passages, indicating that it was not viable. In the case of the E*D5 mutant, no lysis plaques were detected in any of the passages, although genomic and subgenomic RNAs were present (data not shown). To confirm that the E*D5 mutant was, indeed, rescued and viable, the titer from passages 1 to 4 was analyzed by a focus-forming immunofluorescence assay. At 17 h postinfection (hpi), there were no differences in the size and shape of E*D5 foci compared to MERS-CoV-WT (Fig. 3). Nonetheless, the titer was 100-fold lower ( Fig. 2A).
The E*D2 mutant was particularly interesting, since viable virus was rescued from just one of the three independent cDNA clones: E*D2 clone 2 (her called E*D2in) ( Fig. 2A). The sequence of the virus at passage 4 was analyzed, and an insertion of 15 nucleotides (39-TCTCAGATGGTAAAA-59) was identified, encoding five amino acids (SQMVK) following the engineered original mutation (Fig. 4A). The E*D2 deletion eliminates three predicted functional motifs: (i) a GSK3 (glycogen synthase kinase 3) phosphorylation site; (ii) an Atg8 (autophagy-related protein) protein family ligand, also named LIG_LIR_GEN_1; and (iii) an FHA (forkhead-associated) phosphopeptide ligand (Table S1). The insertion of five amino acids identified in the MERS-CoV-E*D2in mutant after four passages restored a GSK3 phosphorylation site, but not the FHA phosphopeptide ligand or the Atg8 protein family ligand (Table S1).
In order to identify the origin of the introduced sequence, a Basic Local Alignment was performed with BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) using the 15-nucleotide sequence insertion (39-TCTCAGATGGTAAAA-59) against all the annotated MERS-CoV sequences. There was a 100% identity with a sequence of 15 nucleotides within the S gene, although the frame sequence was different from the E*D2in gene, rendering a different sequence of amino acids (Fig. 4B). Most likely, the nucleotide sequence incorporated into the E gene was derived from the S gene, restoring the GSK3 phosphorylation site and E protein activity and leading to a viable E*D2in mutant. E*D4 and E*D5 mutants were discarded, due to the lack of viability or the low titer, respectively. On the other hand, plaques were isolated from passage 4 of E*D1 and E*D2in mutants for further characterizations. The growth kinetics of these mutants were analyzed in Huh-7 cells infected with a multiplicity of infection (MOI) of 0.001 (Fig. 2B). E*D1 and E*D2in mutants reached titers similar to MERS-CoV-WT.
Evaluation in vivo of the attenuation of E*D1 and E*D2in mutants. The attenuation of the selected MERS-CoV-E* mutants was evaluated in transgenic K18-hDPP4 mice (79) (Fig. 5). Each animal was intranasally inoculated with 5 Â 10 3 PFU of each virus. As expected, the mice infected with MERS-CoV-WT lost weight rapidly, and all of them died between 6 and 9 days postinfection (dpi). Interestingly, the mice infected with the E*D1 mutant lost weight and died too, indicating that the E*D1 mutant was not attenuated, and therefore, it was discarded. In contrast, mice infected with the E*D2in mutant lost weight slightly, but all of them survived and recovered from the infection, indicating that this mutant was attenuated and a promising vaccine candidate.
The attenuation of the E*D2in mutant is not associated with deficiencies in growth, replication, or transcription in the lungs. The E*D2in mutant was selected as a vaccine candidate due to its capacity to reach titers similar to MERS-CoV-WT in cell cultures (Fig. 2B). For a better description of E*D2in mutant infection in the lung, mice were intranasally infected with 5 Â 10 3 PFU to study lung pathology, viral growth, and viral replication and transcription at 3 and 6 dpi. No significant differences in lung viral titer between the E*D2in mutant and MERS-CoV-WT were observed (Fig. 6A), similar to what was observed in cell cultures. Furthermore, no significant differences were found in either genomic or subgenomic viral RNAs levels between the E*D2in mutant and MERS-CoV-WT (Fig. 6B).
During the histological examination, the appearance of lungs from mice infected with the E*D2in mutant at 3 dpi was similar to the lungs of the noninfected mice (Fig. 7). In contrast, the lungs of the mice infected with MERS-CoV-WT showed incipient  infiltrates in several perivascular and peribronchiolar areas. At 6 dpi, the lungs of the mice infected with MERS-CoV-WT presented generalized infiltration and parenchyma consolidation, as well as edema in the airspaces, whereas the lungs of the mice infected with the E*D2in mutant were similar to those from noninfected mice. In general, the histopathology associated with E*D2in infection was significantly lower than that associated with MERS-CoV-WT infection, although some little and scattered infiltrates could be observed surrounding bronchioles and blood vessels at 6 dpi (Fig. 7). These results demonstrated that the survival of the mice infected with the E*D2in mutant was correlated with a mild or absent pulmonary infiltration, and its attenuation was not due to a decrease in viral titers, or a difference in viral replication and transcription in the lungs compared to MERS-CoV-WT.
The E*D2in mutation might be associated with a reduced brain tropism. Since K18-hDPP4 mice developed brain disease after infection with MERS-CoV (79), the presence of the virus was analyzed in this organ to test whether the attenuation of E*D2in  was related to a reduction of the viral replication (Fig. S1). No virus was detected at 3 dpi in the brain of mice infected with MERS-CoV-WT or E*D2in, an observation consistent with previous data in which it was described that virus titers in the brain of K18-hDPP4 mice were low or null at short times after infection (79). Although at 6 dpi the wild-type virus showed relatively high titers (around 10 5 PFU/g of tissue), the levels of virus in the deletion mutant E*D2in were reduced by 1,000-fold, illustrating the significant attenuation of this deletion mutant virus. The presence of both MERS-CoV-WT and E*D2in in the brain of the transgenic mice was, most likely, due to the artificial expression of large amounts of hDPP4 in this tissue (79). In addition, it must be considered that the potential growth of the virus in the brain of humans should not be promoted by the lower levels of hDPP4 in human brains (80), and by the administration of the vaccine candidate intramuscularly instead of intranasally, as done in this study. Nevertheless, these observations also indicate the need to introduce complementary safety measures in the vaccine candidate before it could be administered to humans.
Protection elicited by E*D2in mutant in a lethal challenge with MERS-CoV-WT. Mice immunized with 5 Â 10 3 PFU of E*D2in were challenged with 5 Â 10 3 PFU of MERS-CoV-WT at 21 days postimmunization (dpim) (Fig. 8). Nonimmunized mice lost weight and died at 6 and 7 days postchallenge (dpc). However, the mice immunized with E*D2in vaccine candidate survived, and no weight loss was observed. These results demonstrated that a single immunization with 5 Â 10 3 PFU of E*D2in elicited full protection in a challenge with a lethal dose of MERS-CoV-WT.
The E*D2in mutant elicited sterilizing immunity in a challenge with a high dose of MERS-CoV-WT. Mice immunized with 5 Â 10 3 PFU of E*D2in mutant were challenged at 21 dpim with a high dose of MERS-CoV-WT (1 Â 10 5 PFU). Lung samples were taken at 2, 4, and 8 dpc to determine lung histopathology, viral titer, and levels of viral replication and transcription.
No weight loss was observed in the immunized mice group, and all of them sur-   Engineered MERS-CoV Live-Attenuated Vaccines ® vived to challenge, demonstrating that immunization with the E*D2in mutant conferred protection even against a higher dose of MERS-CoV-WT (Fig. 9A). Compared to nonimmunized mice, the replication and transcription levels of MERS-CoV-WT were reduced throughout the course of the experiment in the lungs of the immunized mice, being undetectable at 8 dpc (Fig. 9B). No virus was detected in the lungs of the immunized mice at any time of sampling, in contrast to nonimmunized mice (Fig. 9C). Moreover, while the lungs of nonimmunized mice showed large infiltrates at 2 and 4 dpc, with appearance of edema at 8 dpc, the lungs of the immunized mice presented a healthy and functional aspect during the whole experiment (Fig. 10). All together, these results indicated that the protection elicited by the E*D2in mutant promoted a sterilizing immunity, as no histopathological damage or viable virus was observed in the lungs of the immunized mice. Stability of the attenuation of the E*D2in mutant. Biosafety is one of the main concerns associated with live-attenuated vaccines. To determine the stability and the possibility of reversion to virulence, the E*D2in mutant and MERS-CoV-WT were passaged 10 times and 6 times in Huh-7 and MRC-5 cells, respectively. Then, the sequence of the final passage was compared with that of passage 1 virus (Fig. 11). It was observed that the mutation introduced in the E gene of the E*D2in mutant was stable after 10 and 6 passages in Huh-7 and MRC-5 cells, respectively.
The E*D2in mutant and MERS-CoV-WT were then passaged six additional times in Huh-7 cells. The potential attenuation of the E*D2in virus obtained after 16 passages (E*D2in-P16) was evaluated in vivo (Fig. 12). The MERS-CoV-WT and the MERS-CoV-WT passaged 16 times (WT-P16) were used as controls. Both MERS-CoV-WTand MERS-CoV-WT-P16-infected mice rapidly lost weight and died at 10 dpi. In  Engineered MERS-CoV Live-Attenuated Vaccines ® contrast, all the mice inoculated with the E*D2in-P16 deletion mutant survived. Indeed, the sequences of the inoculated E*D2in-P16 deletion mutant and the E*D2in-P16 deletion mutant isolated from the lung of infected mice at 6 dpi were compared, and it was observed that both viruses retained the attenuated E gene sequence of the E*D2in mutant.

DISCUSSION
Modern LAV are based on the study of the molecular mechanisms of pathology in viral infections, allowing the identification of viral genes specifically implicated in virulence. Through a reverse genetics system, these virulence factors can be deleted to obtain an attenuated virus of which we know its exact composition, pathogenicity, and stability, minimizing the risk of reversion to virulence (67). Currently, there are no approved vaccines against MERS-CoV infection for human use, while LAV can be designed and produced with a reverse genetics system like the one developed in our laboratory (81). In this work, a collection of MERS-CoV mutants have been generated by introducing small deletions within the C-terminal domain of the E protein (MERS-CoV-E* mutants). These deletions removed predicted functional sites. As a result, an attenuated vaccine candidate has been obtained that induced protection in a challenge with a lethal dose of MERS-CoV-WT. The characterization of the mutant led to the identification of a functional motif in the E protein, potentially related to MERS-CoV virulence.
From the five engineered deletion mutants of MERS-CoV E protein (MERS-CoV-E*), the E*D2 mutant was selected for further characterization. Apparently, three predicted linear motifs were deleted in the MERS-CoV-E*D1 mutant: the GSK3 motif, an Atg8 protein binding motif (Atg8-BM) and an FHA phosphopeptide ligand binding motif (FHA-BM) (82). Nevertheless, in the E protein sequence of the MERS-CoV-E*D1 mutant, the FHA-BM and the GSK3 motifs were recovered, but not the Atg8-BM one. Both the MERS-CoV-E*D1 mutant and MERS-CoV-WT were viable and virulent, and the unique difference between them was the absence of Atg8-BM, suggesting that this motif was not implicated in MERS-CoV virulence. All the mice infected with the MERS-CoV-E*D1 mutant lost weight and died just 1 day earlier than MERS-CoV-WT-infected mice (Fig. 5). This small difference may indicate the possible presence of an attenuating motif in E protein that, when removed, slightly increased virus virulence. In the MERS-CoV-E*D1 mutant, the only motif deleted was the Atg8-BM, whereas the attenuated MERS-CoV-E*D2in mutant lacks two known motifs: FHA-BM and Atg8-BM (Table 1 and also Table S1 in the supplemental material). It might be interesting to analyze which of these motifs is responsible for virus pathogenicity, an issue that will be addressed in future work.
In the MERS-CoV-E*D2 mutant, a GSK3 phosphorylation site (GSK3-PS), an FHA-BM motif, and an Atg8-BM motif were deleted. The MERS-CoV-E*D2 mutant was not viable in cell culture. However, the spontaneous insertion of five amino acids (SQMVK), introduced into the protein E gene, gave rise to a viable virus (MERS-CoV-E*D2in) in which the GSK3-PS, but not the FHA-BM or the Atg8-BM, was restored. These data suggest that the presence of the GSK3-PS motif could be necessary for either replication or release of the virus. In this work, it has been shown that the attenuation of the mutant was not due to a decrease in the levels of replication or growth in the lung, since the MERS-CoV-E*D2in mutant grew and replicated at the same level as the MERS-CoV-WT. Likewise, it was observed that the MERS-CoV-E*D2in mutant did not produce significant pulmonary pathology compared to that produced by the MERS-CoV-WT. When comparing the E protein sequences of the MERS-CoV-E*D1 and MERS-CoV-E*D2in mutants, it was observed that they differed by the presence in MERS-CoV-E*D1 (virulent) or absence in MERS-CoV-E*D2in (attenuated) of the FHA-BM motif (Table 1 and  Table S1). Taken together, these results suggest that FHA-BM may be involved in the virulence of MERS-CoV and brain tropism, which is relevant in the K18-hDDP4 mouse model, as it develops brain disease upon MERS-CoV infection (79). FHA domains are the only signaling domains that specifically recognize phosphothreonine residues (pThr) (83)(84)(85)(86). These protein-phosphoprotein interaction modules are present in a wide variety of prokaryotic and eukaryotic proteins (87) and participate in various cellular functions such as signal transduction, vesicular trafficking, and cell cycle control (84,(88)(89)(90)(91)(92)(93)(94)(95)(96).
CoV E protein appears to be involved in many processes of viral morphogenesis, such as assembly, induction of envelope curvature, excision, or release of virions (97). In the case of MERS-CoV, the absence of the E gene led to a virus that is competent in replication but deficient in propagation (81). The requirement of protein E for virion maturation and release is different for each CoV. For example, partial deletions introduced in the SARS-CoV E gene attenuate the virus, but it is still viable (70), maybe because accessory protein 3a can partially compensate for deficiencies of protein E (71). In fact, the complete deletion of the E gene in SARS-CoV produces a viable virus, too, although the virus titer is reduced between 20 and 200 times (98). In the prediction of functional motifs of the MERS-CoV protein E, different potential domains were identified and deleted in each of the constructed MERS-CoV-E* mutants, leading to viable or nonviable viruses. MERS-CoV-E*D3 and MERS-CoV-E*D4 mutants were not viable, and MERS-CoV-E*D5 showed very low titers. The phenotype of these three mutants was like that observed when the E gene of the murine hepatitis virus (MHV) was mutated. In MHV, the replacement of several positively charged amino acids throughout the C-terminal domain by alanine residues leads to the appearance of temperature-sensitive mutants and to deficiencies in morphogenesis and virion release (99). Since protein E is key to the spread of MERS-CoV, it is likely that the deletions introduced into MERS-CoV-E*D3, MERS-CoV-E*D4, and MERS-CoV-E*D5 mutants were altering the structure of the E protein, rendering nonfunctional viruses, or a virus with reduced tissue tropism in the case of MERS-CoV-E*D2in.  The engineered MERS-CoV-E*D2in mutant seems to be the basis of a promising vaccine candidate, since it is an attenuated virus that conferred full protection against a lethal challenge with virulent MERS-CoV. Its high levels of replication and growth in cell cultures make it ideal from the point of view of large-scale vaccine production. Likewise, the E protein of the E*D2in mutant was stable after 10 passages in Huh-7 cells and 6 passages in MRC-5 cells; furthermore, the resulting virus after 16 passages in Huh-7 cells was attenuated in vivo. However, since the MERS-CoV-E*D2in mutant is a live-attenuated vaccine candidate, it will be mandatory to address what would be the potential spread of this mutant to contacts, and how stable it is in vivo. Besides, it is still necessary to introduce additional safety measures in order to administer this vaccine to humans. The introduction of partial deletions into the SARS-CoV nsp1 nonstructural protein has been shown in our laboratory to attenuate the virus in vivo (70). Mutations in the nsp16 protein of MERS-CoV also lead to virus attenuation (100). Additionally, accessory genes of MERS-CoV, involved in the antagonism of the interferon-mediated antiviral response, could be deleted (101)(102)(103)(104)(105)(106)(107)(108) to introduce an additional safeguard. Incorporation of these deletions into the MERS-CoV-E*D2in mutant would increase its safety by considerably reducing the probability of reversion to a virulent phenotype in a recombination event. Plasmids and bacterial strains. Bacterial artificial chromosome (BAC) pBeloBAC11 (109), kindly provided by H. Shizuya (California Institute of Technology, Pasadena, CA), was used to assemble recombinant MERS-CoV infectious cDNA clones. This plasmid is a low-copy-number plasmid (one to two copies per cell) based on the Escherichia coli F factor (110) that allows the stable maintenance of large DNA fragments in bacteria. E. coli DH10B (Gibco/BRL) cells were transformed by electroporation using a MicroPulser unit (Bio-Rad) according to the manufacturer's instructions. BAC plasmid and recombinant BACs were isolated and purified using a large-construct kit (Qiagen), following the manufacturer's specifications.

MATERIALS AND METHODS
Generation of recombinant MERS-CoV infectious clones. The cDNA from the MERS-CoV strain EMC/2012 (GenBank accession number JX869059), assembled into the pBAC (pBAC-SA-FL) (81), was used as the basis to generate a collection of five mutants with partial deletions within the C-terminal domain of the E protein. Each mutant has a deletion of 9 to 11 amino acids. Deletion fragments were generated by overlapping mutagenic PCR. Primers used for mutant engineering are described in Table S2 in the supplemental material. First, a PCR was performed with the oligonucleotide SA27502VS and the corresponding reverse-sense (RS) oligonucleotide of each mutant (PCR 1), and another PCR with the oligonucleotide SA28319RS and the corresponding virus-sense (VS) oligonucleotide of each mutant (PCR 2). As a template for these two PCRs, a pBAC was used that included nucleotides 25841 to 30162 of the MERS-CoV EMC/2012 genome (pBAC-SA-F6) and that contains the ORF4a, ORF4b, ORF5, E, M, and N viral genes. Between 50 and 150 ng of each product from PCRs 1 and 2 was taken as the template for a third PCR. The products of PCRs 1 and 2 were amplified with the oligonucleotides SA27502VS and SA28319RS, generating an overlapping PCR product (PCR 3) that included the partial deletion of the E gene flanked by the KflI and Pfl23II restriction sites.
Viruses. Parental wild-type and recombinant viruses, rescued from infectious cDNA clones generated in a BAC, were grown and titrated on Huh-7 cells using closed flasks or sealed plastic bags, respectively. All the work was performed at the CNB-CSIC biosafety level 3 facility (Madrid, Spain) following the security guidelines and standard procedures.
Recovery of recombinant rMERS-CoV-E* mutants from the cDNA clones. BHK cells were grown to 95% confluence in 12.5-cm 2 flasks and transfected with 6 mg of infectious cDNA clone and 18 ml of Lipofectamine 2000 (Invitrogen), according to the manufacturer's specifications. Three independent cDNA clones of each mutant were transfected. At 6 h posttransfection (hpt), cells were trypsinized, added to confluent Huh-7 cell monolayers grown in 12.5-cm 2 flasks, and incubated at 37°C for 72 h. Cell supernatants were harvested (p0) and passaged four times on fresh cells. At passage 4 (p4), viability, titer, and sequence of the mutants were analyzed. Selected mutants were cloned by three rounds of plaque purification following standard procedures (81).
Focus-forming immunofluorescence assay. Huh-7 cells at 5 Â 10 4 per well were seeded in 96well plates in 100 ml of medium 1 day prior to the immunofluorescence assay. Next day, cells were infected with 20 ml of undiluted or serially 10-fold-diluted virus. At 16 h postinfection, cells were fixed with 4% (wt/vol) paraformaldehyde for 40 min, washed, and permeabilized with chilled methanol at room temperature (R/T) for 20 min. Unspecific binding sites were blocked with 10% FBS in phosphate-buffered saline (PBS) for 1 h at R/T. Then, cells were incubated for 90 min at R/T with polyclonal antibody rabbit anti-N-MERS (BioGenes, Germany). Secondary monoclonal antibody goat anti-rabbit conjugated with Alexa 488 (Invitrogen) was incubated for 45 min to detect and count MERS-CoV infectious foci.
Growth kinetics. Subconfluent monolayers (.90% confluence) of Huh-7 cells in 12.5-cm 2 flasks were infected at a multiplicity of infection (MOI) of 0.001 with the indicated viruses. Culture supernatants were collected at 0, 24, 48, and 72 h postinfection (hpi), and virus titers were determined as described above.
Mice. MERS-CoV-susceptible transgenic mice (K18-hDPP4) were kindly provided by Paul McCray (University of Iowa, USA) (79). A colony was established in the CNB-CSIC animal care facility (Madrid, Spain). For infection experiments, female K18-hDPP4 mice at 16 to 24 weeks of age were anesthetized with isoflurane and intranasally inoculated. All work with infected animals was performed in a biosafety level 31 facility by workers wearing personal protection equipment (3M).
Virus infection and growth in mice. K18-hDPP4 transgenic mice (79) were intranasally inoculated with 50 ml of DMEM mixed with virus. The indicated virus at 5 Â 10 3 PFU per mouse was inoculated for attenuation experiments, and 5 Â 10 3 PFU and 1 Â 10 5 PFU of MERS-CoV-WT (recombinant MERS-CoV reproducing the EMC/2012 strain engineered in the CNB-CSIC laboratory) (81) were used for the lowand high-dose challenge, respectively. Weight loss and mortality were evaluated daily. To determine viral titers, lungs and brains were homogenized in 2 ml of PBS containing 100 IU/ml penicillin, 0.1 mg/ml streptomycin, 50 mg/ml gentamicin, and 0.5 mg/ml amphotericin B (Fungizone), using a gentleMACS dissociator (Miltenyi Biotec, Inc.). Virus titrations were performed in Huh-7 cells as described above. Viral titers were expressed as PFU counts per gram of tissue.
Extraction and analysis of viral RNA. RNA from infected cells or homogenized mouse lungs was collected and purified using the RNeasy kit (Qiagen). Total cDNA was synthesized using a highcapacity cDNA reverse transcription kit (ThermoFisher Scientific) with random hexamers and 150 ng of purified RNA in a final volume of 30 ml. cDNA products were subsequently subjected to PCR for sequencing using Vent polymerase (New England Biolabs). Only cDNA products from mouse lungs were analyzed by real-time qPCR for viral RNA synthesis quantification. MERS-CoV genomic RNA (gRNA) (forward primer 59-GCACATCTGTGGTTCTCCTCTCT-39, reverse primer 59-AAGCCCAGGCCCTA CTATTAGC-39, and minor groove binder (MGB) probe 59-TGCTCCAACAGTTACAC-39) and MERS-CoV subgenomic RNA (sgmRNA) N (forward primer 59-CTTCCCCTCGTTCTCTTGCA-39, reverse primer 59-TCA TTGTTATCGGCAAAGGAAA-39, and MGB probe 59-CTTTGATTTTAACGAATCTC-39) custom assays were designed for this analysis; forward and reverse primers were purchased from Sigma-Aldrich, and MGB probes from Eurofins Genomics. Data were acquired with a 7500 real-time PCR system (Applied Biosystems) and analyzed with ABI Prism 7500 software, version 2.0.5. The relative quantifications were performed using the cycle threshold (2 2DDCT ) method (111). To normalize differences in RNA sampling, the expression of mouse 18S rRNA was analyzed using a specific TaqMan gene expression assay (Mm03928990_g1; ThermoFisher Scientific).
Histopathology. Mice were sacrificed at the indicated day postinfection or postchallenge. The left lung of infected mice was fixed in 10% zinc formalin for 24 h at 4°C and paraffin embedded. Serial longitudinal 5-mm sections were stained with hematoxylin and eosin (H&E) by the Histology Service at CNB-CSIC (Madrid, Spain) and subjected to histopathological examination with a Zeiss Axiophot fluorescence microscope. Samples were obtained using a systematic uniform random procedure, consisting in serial parallel slices made at a constant thickness interval of 50 mm. Histopathology analysis was conducted in a blind manner by acquiring images of 50 random microscopy fields from around 40 nonadjacent sections for each of the three independent mice analyzed per treatment group.
Stability of mutants. Stability of the selected MERS-CoV-E*D2in mutant was analyzed in Huh-7 and MRC-5 cells. Cells were seeded in 12.5-cm 2 flasks and infected with the virus. Every 24 h, a third of the supernatant was passaged to a new 12.5-cm 2 flask. After each passage, the remaining supernatant was conserved at 280°C, and the cells were lysed to extract the RNA as described above. The 39-end third of the MERS-CoV genome (S, ORF3, ORF4a, ORF4b, ORF5, E, M, and N) was amplified by PCR, and analyzed by agarose gel electrophoresis and sequencing.
Statistical analysis. Two-tailed, unpaired Student's t tests were used to analyze the differences in mean values between groups. All results were expressed as means 6 standard deviations, except weight losses, which were expressed as means 6 standard errors of the means; P values ,0.05 were considered significant.