Emerging Microbes & Infections 2019, VOL. 8 https://doi.org/10.1080/22221751.2019.1685912 Blocking transmission of Middle East respiratory syndrome coronavirus (MERS-CoV) in llamas by vaccination with a recombinant spike protein Jordi Rodon a*, Nisreen M. A. Okba b *, Nigeer Tea, Brenda van Dierenc, Berend-Jan Boschc, Albert Bensaid a, Joaquim Segalés d,e , Bart L. Haagmans b and Júlia Vergara-Alert a a IRTA, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Bellaterra (Cerdanyola del Vallès), Spain; bDepartment of Viroscience, Erasmus Medical Centre, Rotterdam, The Netherlands; cVirology Division, Department of Infectious Diseases & Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands; dUAB, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Bellaterra (Cerdanyola del Vallès), Spain; eDepartament de Sanitat i Anatomia Animals, Facultat de Veterinària, UAB, Bellaterra (Cerdanyola del Vallès), Spain ABSTRACT The ongoing Middle East respiratory syndrome coronavirus (MERS-CoV) outbreaks pose a worldwide public health threat. Blocking MERS-CoV zoonotic transmission from dromedary camels, the animal reservoir, could potentially reduce the number of primary human cases. Here we report MERS-CoV transmission from experimentally infected llamas to naïve animals. Directly inoculated llamas shed virus for at least 6 days and could infect all in-contact naïve animals 4–5 days after exposure. With the aim to block virus transmission, we examined the efficacy of a recombinant spike S1-protein vaccine. In contrast to naïve animals, in-contact vaccinated llamas did not shed infectious virus upon exposure to directly inoculated llamas, consistent with the induction of strong virus neutralizing antibody responses. Our data provide further evidence that vaccination of the reservoir host may impede MERS-CoV zoonotic transmission to humans. ARTICLE HISTORY Received 21 May 2019; Revised 22 October 2019; Accepted 23 October 2019 KEYWORDS Animal model; llama; Middle East respiratory syndrome coronavirus; MERS-CoV; S1-protein-based vaccine; virus transmission The Middle East respiratory syndrome coronavirus treatments, MERS-CoV has been prioritized for (MERS-CoV) was first identified in September 2012 research and product development in the WHO [1]. This emerging zoonotic pathogen is associated R&D Blueprint for Action to Prevent Epidemics with severe pneumonia, acute respiratory distress syn- [11,12]. The WHO has suggested animal vacci- drome, and multi-organ failure in humans resulting in nation as the best strategy to control MERS-CoV fatal outcomes. As of September of 2019, the World infections, since reduction of virus shedding can Health Organization (WHO) has been notified of potentially prevent both animal-to-animal and zoo- 2,458 laboratory-confirmed cases in humans with at notic transmissions, and might have a faster devel- least 848 deaths [2]. MERS-CoV cases have been opment and licensing pathway compared to human reported in 27 countries, mainly in the Middle East. vaccination [11]. In addition, a major outbreak occurred in South The current MERS-CoV vaccine candidates mainly Korea in 2015 with 186 cases and 39 fatalities [3]. use the entire or sub regions of the spike (S) protein or Therefore, MERS-CoV appears to be a current world- its coding gene. This virus surface structural glycoprotein wide public health threat. binds to the host receptor, dipeptidyl peptidase 4 (DPP4) The dromedary camel is the main reservoir for [13], through its S1 subunit and is therefore the target of MERS-CoV and plays a key role in the infection of pri- choice to raise Nabs [14,15]. The S1 subunit protein is mary human cases [4,5]. In New World immunogenic and can induce both T-cell mediated and camelid species, MERS-CoV infection was evidenced NAb responses mainly directed towards the receptor by the presence of MERS-CoV neutralizing antibodies binding domain (RBD, also named as S1B domain) (NAbs) [6,7]. Furthermore, MERS-CoV experimental [14,16]. Recently, we reported that although most infections in alpacas and llamas confirmed that both NAbs target the S1B domain, antibodies targeting the could serve as potential reservoirs [8–10]. S1 sialic acid binding domain (S1A domain) can also pro- Due to the high human lethality rates and the vide protection against lethal MERS-CoV challenge in a absence of MERS-CoV-licensed vaccines or mouse model [17]. CONTACT Bart L. Haagmans b.haagmans@erasmusmc.nl Department of Viroscience, Erasmus Medical Centre, 3000 CA Rotterdam, The Netherlands; Júlia Vergara-Alert julia.vergara@irta.cat IRTA, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Campus de la UAB, 08193 Bellaterra (Cer- danyola del Vallès), Spain *These authors contributed equally to this work. Supplemental data for this article can be accessed https://doi.org/10.1080/22221751.2019.1685912 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group, on behalf of Shanghai Shangyixun Cultural Communication Co., Ltd This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 1594 J. Rodon et al. Several vaccine prototypes to control MERS-CoV Welfare Committee of IRTA (CEEA-IRTA) and by the have been tested using a wide variety of delivery sys- Ethical Commission of Animal Experimentation of the tems, including DNA vaccines, protein-based vaccines, Autonomous Government of Catalonia (file No. FUE- vector-based vaccines and live attenuated vaccines 2017-00561265). [15,18]. Vector-based-vaccines have been developed using the orthopox modified virus Ankara (MVA) Cell culture and MERS-CoV [19], different host-origin adenovirus (AdV) [20–23], measles virus (MeV) [24], rabies virus (RABV) [25], Vero cells were cultured in Dulbecco’s modified Eagle and Venezuelan equine encephalitis replicons (VRP) medium, DMEM (Lonza) supplemented with 2% [22,26], all expressing different lengths of the S protein. fetal calf serum (FCS; EuroClone), 100 U/ml penicillin These vector-based candidates were tested in human (ThermoFisher Scientific, Life Technologies), 100 μg/ DPP4 (hDPP4) transgenic or transduced mice, except ml streptomycin (ThermoFisher Scientific, Life Tech- the orthopox-based recombinant vaccine, which nologies), and 2 mM glutamine (ThermoFisher Scien- expresses the full-length MERS-CoV spike protein tific, Life Technologies). A passage 2 MERS-CoV and induced efficient protective immunity in dromed- stock (Qatar15/2015 strain) was propagated in Vero aries [19]. Due to reticence in applying live genetically cells at 37°C in a CO2 incubator for 3 days. The infec- modified organisms, protein recombinant subunit or tious virus titre was determined in Vero cells and cal- DNA vaccines mainly based on the S1 protein or culated by determining the dilution that caused gene, respectively, are also under study. A DNA- cytopathic effect (CPE) in 50% of the inoculated cell based vaccine expressing the full-length S protein was cultures (50% tissue culture infectious dose endpoint, shown to induce MERS-CoV specific NAbs and confer TCID50). protection in rhesus macaques [27]. In addition, MERS-CoV protein-based vaccines using the full- Vaccine length or fragments of the S protein were produced in the form of virus-like particles, nanoparticles, pep- Full-length MERS-CoV S1 recombinant protein, tides, or recombinant protein. Partial protection including A and B domains, was produced in house efficacy for some candidates has been demonstrated using baculovirus and HEK 293 T cells production sys- in non-human primates (NHP) [28,29] and hDPP4 tems as previously described [17,38]. In brief, to pro- transgenic mice [30–36.] A more recent study demon- duce soluble MERS-CoV S1 using the baculovirus strated that an S protein subunit vaccine conferred pro- expression system, the gene fragment encoding the tection to MERS-CoV (EMC/2012 strain) in an alpaca MERS-CoV S1 subunit (amino acid 19–748; EMC/ model, although in dromedary camels the vaccine was 2012 isolate; GenBank Accession YP_009047204.1) only able to reduce and delay viral shedding [37]. How- was codon-optimized for insect cell expression and ever, there is no evidence that any of the MERS-CoV cloned in-frame between honeybee melittin (HBM) vaccine candidates developed so far are able to block secretion signal peptide and a triple StrepTag purifi- MERS-CoV transmission in camelids when tested in cation tag in the pFastbac transfer vector. Generation a direct-contact virus transmission setting, mimicking of bacmid DNA and recombinant baculovirus was per- natural transmission in the field. Vaccinating the formed according to protocols from Bac-to-Bac system MERS-CoV animal reservoirs can potentially reduce (Invitrogen), and expression of MERS-CoV S1 was per- transmission to humans and provide a simple and formed by infection of recombinant baculovirus of Sf-9 economical solution to avoid expansion of this threa- cells. Recombinant proteins were harvested from cell tening disease. culture supernatants 3 days post infection and purified In the present study, we show efficient MERS-CoV using StrepTactin sepharose affinity chromatography transmission among llamas. Furthermore, we have suc- (IBA). cessfully used this direct-contact transmission model to Production of recombinant MERS-S1 in HEK 293 T demonstrate the efficacy of a recombinant S1-protein cells was described previously [17,38]. In brief, the vaccine, using a registered adjuvant, to block MERS- MERS-S1 (amino acid 1–747; EMC/2012 isolate; Gen- CoV transmission. Bank Accession YP_009047204.1) encoding sequence was C-terminally fused to a gene fragment encoding the Fc region of human IgG and cloned into the Materials and methods pCAGGS mammalian expression vector, expressed by plasmid transfection in HEK-293 T cells, and affinity Animal welfare and ethics purified from the culture supernatant using Protein- Experiments with MERS-CoV were performed at the A affinity chromatography. The Fc part of S1-Fc fusion Biosafety Level-3 (BSL-3) facilities of the Biocontain- protein was proteolytically removed by thrombin fol- ment Unit of IRTA-CReSA (Barcelona, Spain). The lowing Protein-A affinity purification using the throm- present study was approved by the Ethical and Animal bin cleavage site present at the S1-Fc junction. Emerging Microbes & Infections 1595 Animals, vaccination and experimental design Regarding to the nomenclature used in this study, animals 1–3 and 4–6 corresponded to intranasally Sixteen healthy llamas were purchased and housed at inoculated llamas in boxes 1 and 2, respectively. Llamas IRTA farm facilities at Alcarràs (Catalonia, Spain) 7–11 were naïve contact animals and llamas 12–16 during the immunization period and transferred for were immunized contact animals. challenge at the BSL-3 animal facilities of the Biocon- Animals were monitored daily for clinical signs tainment Unit of IRTA-CReSA, in Barcelona (Spain). (sneezing, coughing, nasal discharge or dyspnea). Rec- Five llamas were prime vaccinated each with 35 µg tal temperatures were recorded with a fast display digi- of a recombinant S1 protein produced in a baculovirus tal thermometer (AccuVet®) until day 13 or 15 post- system, emulsified (1:1 volume) with Montanide™ ISA inoculation (pi) for animals in boxes 1 and 2, respect- 206 VG (Seppic) adjuvant and intramuscularly admi- ively. For llamas housed in box 1, nasal swabs (NS) nistered (2 ml per animal and dose) in the right side were obtained daily until day 14 pi, while in box 2 of the neck. A boosting immunization was conducted NS were collected daily until day 15 pi and two extra 3 weeks later as above (left side of the neck) but with collections were performed on 17 and 19 dpi. Serum 50 µg of recombinant S1 protein produced in HEK samples were obtained before the first and the second 293 T cells, emulsified (1:1 volume) with Montanide™ immunizations, prior to challenge, and weekly after ISA 206 VG (Seppic) adjuvant. The correct structure of the MERS-CoV challenge. Animals were euthanized the S1 antigens was previously confirmed by reactivity 3-weeks after challenge, with an overdose of pentobar- of conformational antibodies, DPP4 solid phase and bital. An extra sampling of NS was performed prior to sialic acid binding assays [17]. Two weeks later, necropsy procedures. MERS-CoV challenge was performed. The experiments on virus transmission and vaccine efficacy were con- ducted in two separate boxes. In box 1, a group of lla- mas (n = 3) were intranasally inoculated with a 107 Environmental samples TCID50 dose of MERS-CoV Qatar15/2015 strain (Gen- Three different types of environmental samples (ES) Bank Accesion MK280984) in 3 ml saline solution were collected to determine viral loads in the boxes (1.5 ml in each nostril) using a nebulization device throughout the study (see Supplementary Fig. S1), as (LMA® MADgic®, Teleflex Inc.). At 2 days post-inocu- previously described [39]. An air filtering device (Sartor- lation (dpi) naïve llamas (n = 5) were put in contact ius MD8, Sartorius Stedim) was used for testing one with infected llamas (Figure 1a, Supplementary Fig. thousand litres of air during 20 min (50 L/min air S1). In box 2, the same protocol as in box 1 was fol- volume) through a gelatine membrane filter (ES1). lowed but using vaccinated llamas (n = 5) as a contact One wall was scrubbed with two swabs (ES2 and ES3) group (Figure 1b). Each box was set up as in a previous and a water sample from the drinking point (ES4) was transmission study performed in pigs [39]. also obtained. ES were collected daily until 10 dpi. Figure 1. Schematic diagram of the llama transmission (a) and vaccination (b) experiments. (a) Three llamas (black, LL1-3) were intranasally inoculated with MERS-CoV (Qatar15/2015) and two days later were brought in contact with five naïve llamas (grey, LL7- 11). (b) Vaccination, challenge and sampling scheme showing vaccinated llamas (red, n = 5, LL12-16) and directly inoculated llamas (black, n = 3; LL4-6) used as a transmission challenge model for MERS-CoV. Dpi, days post-inoculation. 1596 J. Rodon et al. Viral RNA detection by RT-qPCR amount of infectious virus in swabs was calculated by determining the TCID50. Viral RNA in collected samples was detected by RT- qPCR as previously described [10,39]. Briefly, NS and ES, except water samples, were transferred into cryo- MERS-CoV S1-ELISA tubes containing either 500 µL DMEM (Lonza) or Specific S1-antibodies in serum samples from all col- PBS (Lonza) supplemented with 100 U/ml penicillin lected time-points and from all animals were deter- (ThermoFisher Scientific, Life Technologies) and 100 mined by a MERS-CoV S1-ELISA as previously μg/ml streptomycin (ThermoFisher Scientific, Life described [10,19]. Briefly, 96-well high-binding plates Technologies), vortexed and stored at −80°C until (Sigma-Aldrich) were coated with 100 µl of S1 protein use. Water samples were directly frozen at −80°C [42] at 1 µg/ml in PBS o/n at 4°C. After blocking with instead. Viral RNA from NS and ES was extracted 1% bovine serum albumin/PBS/0.5% Tween20 for 1 h with a NucleoSpin® RNA virus kit (Macherey-Nagel) at 37°C, serum samples were tested at a 1:100 dilution, following the manufacturer’s instructions. The RNA followed by 1 h incubation at 37°C. Plates were washed extracts were tested by using the UpE PCR [40]. RT- 4 times with PBS, and wells were incubated with a goat qPCR was carried out using AgPath-IDTM One-Step anti-llama biotin conjugate (Abcore, 1:1,000 diluted in RT–PCR Reagents (Applied Biosystems, Life Technol- blocking buffer), followed by incubation with streptavi- ogies), and amplification was done by using a 7500 Fast din peroxidase (Sigma-Aldrich). After 1 h of incu- Real-Time PCR System (Applied Biosystems, Life bation at 37°C, wells were washed 4 times with PBS, Technologies) programmed as follows: 10 min at 50° and a TMB substrate solution (Sigma-Aldrich) was C, 10 sec at 95°C, and 45 cycles of 15 s at 95°C and added and allowed to develop for 8–10 min at room 30 sec at 58°C. Samples with a quantification cycle temperature, protected from light. Optical density (Cq) value ≤40 were considered positive for MERS- was measured at 450 nm. CoV RNA. To test for viral replication, viral RNA extracted form NS was tested for the presence of M mRNA according to the previously published protocol MERS-CoV N-LIPS by Coleman et al. [41] We tested llama sera for MERS-CoV nucleocapsid (N) specific antibody responses using a luciferase immuno- precipitation (LIPS) assay [43]. The N protein was Viral RNA sequencing expressed as an N-terminal Renilla luciferase (Ruc)- Viral RNA was extracted from llama NS using the tagged protein (Ruc-N) using pREN2 expression vec- QIAamp viral RNA mini kit (Qiagen) according to tor. The cells were lysed, and the luminescence units the manufacturer’s instructions. cDNA was produced (LU)/μl was measured in cell lysates. LIPS assay was from RNA using Superscript III first strand synthesis done according to a previous protocol with minor system (Invitrogen Corp) using random hexamers. modifications [44]. Briefly, serum samples were diluted The cDNA was then used as a template to PCR amplify 1:100 and mixed with 1× 107 LU of Ruc-N in a total the MERS-CoV spike S1 encoding region (nucleotides volume of 100 μl in buffer A (20 mM Tris, pH 7.5, positions 21,304–25,660, GenBank Accession 150 mM NaCl, 5 mM MgCl2, 1% Triton X-100). The JX869059) using the PfuUltra II Fusion HS DNA poly- mixture was incubated on a rotary shaker for 1 h at merase (Aligent Technologies). The PCR was carried room temperature. Then, the mixture was transferred out as follows: 95°C for 5 min, 39 cycles of 20 sec at into MultiScreenHTS BV Filter Plate (Merk Millipore) 95°C, 20 sec at 48°C, and 45 sec at 72°C, and a final containing 5 μl of a 30% suspension of UltraLink extension at 72°C for 1 min. The amplicons were protein A/G beads and further incubated for one sequenced bidirectionally using the BigDye Terminator hour. The wells were then washed and luminescence v3.1 cycle sequencing kit on an ABI PRISM 3130XL was measured for each well after adding 100 μl of 0.1 Genetic analyzer (Applied Biosystems). μM coelenterazine (Nanolight Technology) in assay buffer (50 mM potassium phosphate, pH 7.4, 500 mM NaCl, 1 mM EDTA). The sera were tested in duplicates in at least two independent assays and the Virus titration data was averaged to determine the LU value for each NS and ES collected at different times pi were evaluated sample. for the presence of infectious virus by titration in Vero cells, as previously reported [10,19]. Ten-fold dilutions Hemagglutination inhibition (HI) assay were done, starting with a dilution of 1:10, and dilutions were transferred to Vero cells. Plates were To test llama sera from the vaccine efficacy study for daily monitored under the light microscope and wells functional antibodies against the sialic acid binding were evaluated for the presence of CPE at 5 dpi. The S1 N-terminal domain (S1A), a nanoparticle-based HI Emerging Microbes & Infections 1597 assay was used. S1A lumazine synthase (LS) nanoparti- vaccinated llamas throughout the study. Despite higher cles were produced as described previously [17,45]. basal body temperatures, no animals housed in box 1 Two-fold diluted sera were mixed with 4 HA units of (inoculated and non-vaccinated in-contact llamas) S1A-LS and incubated for 30 min at 37°C. Following showed a significant increase in body temperatures incubation, 0.5% washed turkey RBCs were added above 40°C upon MERS-CoV challenge. In box 2 and further incubated for 1 h at 4°C. HI titres were (inoculated and vaccinated in-contact llamas), body determined as the reciprocal of highest serum dilution temperatures in llamas remained constant all along showing inhibition of hemagglutination. the experiment and never exceeded 39.5°C. Receptor binding inhibition (RBI) assay MERS-CoV RNA and infectious virus We tested llama sera from the vaccine efficacy study for All MERS-CoV inoculated llamas shed viral RNA in antibodies able to block MERS-CoV binding to its the nasal cavity during a 2-week period (Figure 2a, receptor (DPP4) using a competitive ELISA. ELISA b). The amount of viral RNA was still high (Cq values plates were coated with 2 μg/ml recombinant soluble < 25) in all inoculated llamas at 6-7 dpi, but a decrease DPP4 protein [13] overnight at 4°C. The plates were in RNA load was observed from 8 dpi onwards. In- washed with PBS and blocked with 1% BSA in PBS/ contact naïve llamas from box 1 revealed evidence of 0.1% Tween-20 at 37°C for 1 h. Serum samples were infection (detectable viral RNA) 4–5 days after contact, tested at a 1:20 dilution. Recombinant MERS-CoV with viral RNA loads and duration of shedding similar S1-mFc was mixed with diluted sera, incubated for 1 to those of the inoculated animals (Figure 2a). In box 2, hr at 37°C, added to the plate and further incubated only one out of the five vaccinated llamas (No. 15) had for 1 h. The plates were then washed and HRP-labelled viral RNA in the nasal cavity to levels comparable to rabbit anti-mouse Igs was added to detect S1 bound to non-vaccinated in-contact animals, while the other DPP4. Following 1 h of incubation, the plates were four animals had very low levels of viral RNA (Figure washed and the signal was detected using TMB as 2b). Additionally, the viral RNA from this llama was described above. Optical density was measured at 450 sequenced at days 9–12 pi and used for comparative nm. analysis of the S1 protein (see Supplementary Fig. S3). A substitution of serine for phenylalanine was found at the amino acid position 465 (S465F) in com- Plaque reduction neutralization assay parison with the inoculum isolate S1 protein (see Sup- Serum samples and nasal swabs were further tested for plementary Fig. S3a). This mutation was also found in neutralizing antibodies against MERS-CoV (Qatar15/ another vaccinated llama (No. 13) at 10 dpi. Interest- 2015 and EMC/2012 isolates) using a plaque reduction ingly, we identified the S465F mutation arising at 5- neutralization (PRNT) assay. PRNT assay was carried 6 dpi in three directly inoculated llamas (No.1, 4, 5). out using according to the previously published proto- Furthermore, the naïve contact animals were also col [19] with some modification. Briefly, samples were investigated and the same mutation was found in first inactivated at 56°C for 30 min. Then, 50 μl of 2- llama No. 9 at 10 dpi (see Supplementary Fig. S3b). fold serial dilutions of heat-inactivated serum were To ensure that this mutant is not a neutralization mixed 1:1 with virus (400 PFU) prior to over-layering escape mutant, the mutant virus was plaque-purified onto Huh7 cells. After 8 h of infection, the cells were form the nasal swab of llama No. 4 at 6 dpi. The fixed and stained using mouse anti-MERS-CoV virus was sequenced (Llama-passaged-Qatar15; Gen- nucleocapsid protein (SinoBiological) and HRP-conju- Bank Accession MN507638) to ensure no other gated goat anti-mouse IgG1 (SouthernBiotech). The mutations were present in the spike protein and then number of infected cells were detected using a precipi- used to carry out neutralization assays. The virus was tate-forming TMB substrate (True Blue, KPL) and neutralized by serum of all five vaccinated animals counted using an ImmunoSpot® Image analyser (CTL (Supplementary Fig. S4a). Europe GmbH). The PRNT titre was calculated based RT-qPCR positive nasal swab samples were tested on a 50% or greater reduction in infected cells counts. for the presence of infectious virus. All intranasally inoculated llamas excreted infectious MERS-CoV at some point until 8 dpi (Figure 2c, d). The duration of Results infectious virus shedding varied among individual ani- mals ranging from 1 up to 6 consecutive days. In each Clinical signs box, at least one inoculated llama (animals No. 2 and 5) Three out of the six directly-inoculated and one out of shed infectious virus continuously from days 1–6 pi the five contact naïve llamas showed moderate nasal (Figure 2c, d). Three out of the five direct contact mucus secretion at 8–15 dpi (see Supplementary Fig. naïve llamas from box 1 shed infectious virus at 8, 9 S2). No clinical signs were noticed in any of the five and 10 dpi (Figure 2c). These non-vaccinated in- 1598 J. Rodon et al. Figure 2. Viral shedding in llamas after experimental inoculation or contact with MERS-CoV-infected llamas. Viral RNA detected in nasal swab samples collected from naïve (a) and S1 vaccinated (b) llamas at different time points after contact with directly inocu- lated animals. Panels c) and d) display infectious MERS-CoV in nasal swab samples collected from naïve (c) and S1 vaccinated ani- mals (d) at different time points after inoculation. Each line/bar represents an individual animal. Orange lines/bars indicate experimentally inoculated llamas. Blue and green lines/bars indicate in-contact naïve animals, while purple lines/bars indicate vac- cinated llamas. Dashed lines depict the detection limit of the assays. Cq, quantification cycle; MERS-CoV, Middle East respiratory syndrome coronavirus; TCID50, 50% tissue culture infective dose. contact animals (No. 7, 9 and 10) exhibited virus titres until 5 dpi, suggesting that a contamination occurred at least equal to those observed in inoculated llamas during the collection or the processing of this sample. (Figure 2c, d). The peaks of viral RNA coincided with Additionally, no infectious virus was detected in this the highest levels of infectious virus shed. Although animal at 1 dpi (Figure 2c). llama No. 15 had MERS-CoV mRNA indicative of Relatively low levels of viral RNA were detected in replication in the nasal cavity to levels comparable to all types of environmental samples that were taken in non-vaccinated in-contact animals (Supplementary the boxes during the experiment (≥30 Cq) (Table 1). Fig. S5), as assessed by the specific RT-qPCR described The highest MERS-CoV RNA levels were found in by Coleman and collaborators [41], none of the drinking water samples. However, titration of infec- vaccinated animals from box 2 (including llama No. tious virus was not successful. 15) shed infectious virus at any point in the study (Figure 2d), Humoral immune response Llama No. 7 showed low levels of MERS-CoV RNA at 1 dpi before in-contact challenge (Figure 2a). How- We evaluated the MERS-CoV specific antibody ever, this animal remained negative to RT-qPCR responses induced in llamas following infection and Table 1. MERS-CoV RNA detection in environmental samples expressed in Cq values at different times after inoculation. Swab 1 and 2 correspond to ES2 and ES3 of the Suppl. Fig. S1, respectively. Cq, quantification cycle; MERS-CoV, Middle East respiratory syndrome coronavirus; nc, non-collected samples. Days post-inoculation 0 1 2 3 4 5 6 7 8 9 10 Box 1 – transmission study Sartorius − − 35,04 − 36,22 39,43 38,52 nc 38,21 38,72 31,91 Swab 1 − − − − 36,57 − 39,53 nc 32,23 39,64 38,01 Swab 2 − 39,90 − 38,31 35,85 35,35 37,00 nc 34,30 38,12 36,58 Water − 36,31 − − − − 36,01 nc 38,70 33,42 33,24 Box 2 – vaccine trial Sartorius − − − − − − − 36,79 38,54 − − Swab 1 − − − − 36,42 − 39,09 37,88 37,34 30,93 34,01 Swab 2 − − − 36,76 − 37,20 − 37,63 − 31,40 35,99 Water − − − 37,89 35,20 30,92 31,91 33,94 34,62 38,77 37,52 Emerging Microbes & Infections 1599 Figure 3. Serum antibodies elicited against MERS-CoV in inoculated and in-contact naïve llamas. (a) MERS-CoV spike S1, (b) MERS- CoV neutralizing (Qatar15/2015 strain), (c) hemagglutination inhibition (HI; anti-S1A N-terminal domain), and (d) receptor binding inhibition (RBI; anti-S1 receptor binding domain) antibodies. The horizontal dotted lines indicate the cutoff of each assay. HI, hemagglutination inhibition; LL, llama; PRNT, plaque reduction neutralization assay; RBI, receptor binding inhibition; W, week vaccination. Regarding the transmission study, all antibodies in the nasal swabs of three out of the five directly inoculated and in-contact naïve llamas sero- vaccinated animals (Figure 4e, f). converted to MERS-CoV as detected by MERS-CoV S1 ELISA (Figure 3a) and virus neutralization (Figure 3b). In contrast, only three of those, two directly inocu- Discussion lated and one in-contact, also developed anti-N anti- In this study, experimental MERS-CoV transmission body responses (see Supplementary Fig. S6a). from infected llamas to naïve in-contact llamas has Antibodies against the S1A sialic acid binding domain been demonstrated for the first time. Consistent with were detected in one of the directly inoculated and previous studies [10], all MERS-CoV inoculated llamas four in-contact naïve animals using a HI assay (Figure got infected, shed infectious virus and were able to 3c). Receptor-binding blocking (mainly RBD-directed) transmit the virus to all naive contact animals as antibodies were detected in the sera of all directly assessed by MERS-CoV RNA and viral titration of inoculated animals and in four out of the five in-con- the nasal swabs. We confirmed that 3 infected llamas tact naïve llama sera using a competitive RBI ELISA were able to transmit MERS-CoV to at least 5 naïve (Figure 3d). animals; nonetheless, further studies are needed to Following MERS-CoV S1 vaccination, all vaccinated determine the basic reproduction ratio of this virus animals (Figure 4a-d, red) developed high titres of transmission in camelids. Interestingly, the three con- serum S1-reactive antibodies (Figure 4a) and virus neu- tact llamas shedding infectious MERS-CoV showed tralizing antibodies against both clade B Qatar15/2015 the highest viral RNA loads, while the remaining two and a clade A EMC/2102 isolates as detected by PRNT had higher Cq values and no infectious virus was iso- (Figure 4b, Supplementary Fig. S4b). In particular, the lated. Altogether, taking into account that (i) viral vaccination induced antibodies against the two func- genomic replication was observed in all in-contact tional domains of S1, the S1A binding N-terminal naïve llamas for an extended period, (ii) 3 out of 5 domain as detected by HI assay (Figure 4c) and the in-contact animals shed detectable infectious virus RBD as detected by a competitive RBI ELISA (Figure and (iii) one of them exhibited nasal discharges, this 4d). Additionally, only one directly inoculated but in-contact model of virus transmission is valuable to none of the vaccinated animals developed antibodies test vaccine efficacy. However, before stating that lla- against the N protein (Supplementary Fig. S6b). Aim- mas can be surrogates of dromedaries for vaccine test- ing to assess mucosal immunity elicited upon vacci- ing in an in-contact model, it would be important to nation, we evaluated the presence of antibodies in the assess whether infectious viral pressure elicited by the nasal cavity. Remarkably, we detected low levels of experimental challenge are similar between these two both MERS-CoV S1-directed and neutralizing animal species. In that respect, in a previous report, 1600 J. Rodon et al. Figure 4. Antibody responses to MERS-CoV elicited in directly inoculated (LL4-6; black) and in-contact MERS-CoV S1 vaccinated (LL12-16; red) llamas in sera (a-d) and nasal swabs (e,f). (a,e) MERS-CoV S1-reactive antibodies, (b,f) MERS-CoV neutralizing anti- bodies (Qatar15/2015 strain), (c) hemagglutination inhibition (HI; anti-S1A N terminal domain) antibodies, and (d) receptor binding inhibition (RBI; anti-S1 receptor binding domain) antibodies. The horizontal dotted lines indicate the cutoff of each assay. HI, hemagglutination inhibition; LL, llama; PRNT, plaque reduction neutralization assay; RBI, receptor binding inhibition; W, week two dromedaries inoculated with the MERS-CoV from cells harvested in the nasal swabs; nonetheless, EMC/2012 strain shed viral RNA and infectious virus we could not detect any infectious virus. Neutralization for 13 and 6 days, respectively [19], similar to what of the virus by antibodies at mucosal level may have we found in the present study in llamas infected intra- inhibited infectious viral particle production. The nasally with the MERS-CoV Qatar15/2015 strain. lack of detectable infectious virus in the vaccinated lla- Based on the in vivo protective capacity of mono- mas despite being infected, renders these animals unli- clonal antibodies directed against different domains kely to transmit the virus further to other animals and of the spike protein [17], a broader protective immune thus blocking the transmission chain. In addition, our response can be achieved using multi-domain vaccines studies revealed a mutation (S465F) in the spike (S1A and S1B domains) compared to RBD-focused vac- protein encoded by this viral RNA, which may suggest cines. Thus, the efficacy of an S1 recombinant protein a potential escape variant being produced. However, emulsified with the adjuvant Montanide™ ISA 206 the emergence of the same mutation in another vacci- VG was evaluated as a potential vaccine candidate. nated llama, in one naïve in-contact animal and in We showed that immunized llamas were efficiently other three directly inoculated llamas was revealed. In protected against MERS-CoV infection; no infectious addition, the capacity of vaccinated animals to induce virus was detected in the nose of any of the vaccinated NAbs against this variant when isolated, indicate that animals and viral RNA shedding remained low (Cq ≥ it is unlikely an escape variant induced under antibody 34), with the exception of one llama (No. 15). Viral pressure. Mutation at this site (S465F) is not directly mRNA was also detected in the nasal cavity of this involved in receptor binding but has been previously llama, which might be from intracellular viral mRNA reported to occur as a result of virus adaptation to its Emerging Microbes & Infections 1601 host receptor [46]. Overall, this indicates a probable protein, in combination with a commercial adjuvant, adaptive mutation rather than a vaccine escape efficiently limits infectious viral shedding from vacci- mutation. nated llamas upon exposure to directly inoculated Immunization with the S1 protein induced anti- ones. bodies against the RBD as confirmed by the RBI and virus neutralization assays as well as antibodies to the S1A domain as confirmed by HI assay. These latter Acknowledgements antibodies may be important in blocking virus attach- The authors want to thank Dr. Joan Pujols and Dr. Xavier ment to sialic acid present in camelids, as it has been Abad for scientific discussions. The pREN2 expression vec- demonstrated in the dromedary camel upper respirat- tor was kindly provided by Dr. Peter D Burbelo. ory tract [45]. Importantly, serum NAbs were gener- This study was performed as part of the Zoonotic Antici- pation and Preparedness Initiative (ZAPI project) [Innova- ated in all vaccinated animals after the boosting tive Medicines initiative (IMI) grant 115760], with immunization and were maintained during challenge. assistance and financial support from IMI and the European Therefore, a correlation of NAb levels in serum upon Commission and contributions from EFPIA partners. J.R. vaccination and protection occurred, as previously was supported by the VetBioNet project (EU Grant Agree- described in another vaccination study in camelids ment INFRA-2016-1 N°731014). IRTA is supported by CERCA Programme / Generalitat [37]. Notably, we detected mucosal NAb in the nasal de Catalunya. cavity of 3 out of 5 vaccinated llamas, as also reported B.-J. B., A.B., J.S., B.L.H., J.V.-A. conceived and designed in dromedary camels immunized with an MVA-based the experiment. J.R., N.M.A.O., N.T., B.D., B.-J.B., A.B., J.S., candidate [19]. In addition, we demonstrate that vacci- B.L.H., J.V.-A. performed the experiments and analyzed the nation of llamas with a spike protein from a clade A data. The manuscript was written by J.R., N.M.A.O., B.L.H., MERS-CoV (EMC/2012 isolate) provides protection J.V.-A. All the authors discussed the results and revised the manuscript. against a challenge with a clade B virus (Qatar15/ 2015 islolate). Since evidence of MERS-CoV reinfec- tion has been reported in camels in the field [47], Disclosure statement further studies to determine whether intramuscular administration of the subunit vaccine can boost muco- No potential conflict of interest was reported by the authors. sal immunity in the upper respiratory tract of animals that have been previously exposed to MERS-CoV are Funding needed. A critical component of a vaccine that influences the This work was supported by European Commission: [Grant duration and the quality of immune responses is the Number INFRA-2016-1 N°731014, Innovative Medicines initiative (IMI) grant 115760]. adjuvant. Here we used the Montanide™ ISA 206 VG adjuvant, which was shown to induce long-term protective immunity in large animal species by stimu- ORCID lating both cell-mediated and humoral immune responses [48]. Further studies should be conducted Jordi Rodon http://orcid.org/0000-0002-1032-9091 Nisreen M. A. Okba http://orcid.org/0000-0002-2394-1079 in target species in order to determine the optimal anti- Albert Bensaid https://orcid.org/0000-0001-6493-3756 gen dose and the persistence of NAb following S1 Joaquim Segalés https://orcid.org/0000-0002-1539-7261 recombinant vaccination. In fact, here, two doses of Bart L. Haagmans http://orcid.org/0000-0001-6221-2015 35 and 50 µg were enough to induce protection, as Júlia Vergara-Alert https://orcid.org/0000-0001-7484-444X opposed to a recent study which used 3 doses of 400 µg of the S1 antigen with a combination of adju- vants [37]. Unlike vector-based vaccines, protein- References based vaccines do not require safety testing in high [1] Zaki A, van Boheemen S, Bestebroer T, et al. Isolation containment facilities and field studies could be directly of a novel coronavirus from a man with pneumonia in conducted; thus, reducing the cost of the proposed vac- Saudi Arabia. New Engl J Med. 2012;367:1814–1820. cine. The registered adjuvant used in this study, Mon- [2] World Health Organization (WHO). Middle East res- piratory syndrome coronavirus (MERS-CoV) – tanide™ ISA 206 VG, offers economical and practical [updated 2019 March; cited 2019 Sep 26]. Available use for field applications. Therefore, the S1 recombi- from: http://www.who.int/emergencies/mers-cov/en/ nant vaccine tested in this study appears as a good can- [3] World Health Organisation (WHO). Middle East res- didate to prevent animal-to-animal and, eventually, piratory syndrome coronavirus (MERS-CoV): WHO animal-to-human transmission. MERS-CoV Global Summary and Assessment of Risk; 2018. Available from: https://www.who.int/csr/disea Overall, this work revealed that the llama model can se/coronavirus_infections/risk-assessment-august-201 be a surrogate for dromedary camel in MERS-CoV 8.pdf transmission and vaccination studies. Moreover, [4] Sabir JSM, Lam TT-Y, Ahmed MMM, et al. Co-circu- immunization with the MERS-CoV S1 recombinant lation of three camel coronavirus species and 1602 J. Rodon et al. recombination of MERS-CoVs in Saudi Arabia. syndrome coronavirus vaccine in BALB/c mice. Science. 2016;351:81–84. Vaccine. 2014;32:5975–5982. [5] Mohd HA, Al-Tawfiq JA, Memish ZA. Middle East [22] Cockrell AS, Yount BL, Scobey T, et al. A mouse model respiratory syndrome coronavirus (MERS-CoV) origin for MERS coronavirus-induced acute respiratory dis- and animal reservoir. Virol J. 2016;13:87. tress syndrome. Nat Microbiol. 2017;2:16226. [6] Reusken CB, Schilp C, Raj VS, et al. MERS-CoV infec- [23] Hashem AM, et al. A highly immunogenic, protective tion of alpaca in a region where MERS-CoV is ende- and safe adenovirus-based vaccine expressing MERS- mic. Emerg Infect Dis. 2016;22:1129–1131. CoV S1-CD40L fusion protein in transgenic human [7] David D, Rotenberg D, Khinich E, et al. Middle East DPP4 mouse model. J Infect Dis. 2019;pii:jiz137. respiratory syndrome coronavirus specific antibodies [24] Malczyk AH, Algaissi A, Agrawal A, et al. A Highly in naturally exposed Israeli llamas, alpacas and camels. immunogenic and protective Middle East respiratory One Health. 2018;5:65–68. syndrome coronavirus vaccine based on a recombinant [8] Crameri G, Durr PA, Klein R, et al. Experimental infec- measles virus vaccine platform. J Virol. 2015;89:11654– tion and response to rechallenge of alpacas with 11667. Middle East respiratory syndrome coronavirus. [25] Wirblich C, Coleman CM, Kurup D, et al. One-Health: Emerg Infect Dis. 2016;22:1071–1074. a safe, efficient, dual-use vaccine for humans and ani- [9] Adney DR, Bielefeldt-Ohmann H, Hartwig AE, et al. mals against Middle East respiratory syndrome coro- Infection, replication, and transmission of Middle navirus and rabies virus. J Virol. 2017;91:1–15. East respiratory syndrome coronavirus in alpacas. [26] Zhao J, Li K, Wohlford-Lenane C, et al. Rapid gener- Emerg Infect Dis. 2016;22:1031–1037. ation of a mouse model for Middle East respiratory [10] Vergara-Alert J, van den Brand JMA, Widagdo W, syndrome. Proc Natl Acad Sci USA. 2014;111:4970– et al. Livestock susceptibility to infection with Middle 4975. East respiratory syndrome coronavirus. Emerg Infect [27] Muthumani K, Falzarano D, Reuschel EL, et al. A syn- Dis. 2017;23:232–240. thetic consensus anti-spike protein DNA vaccine [11] World Health Organization (WHO). WHO Target induces protective immunity against Middle East res- Product Profiles for MERS-CoV Vaccines, 2017. piratory syndrome coronavirus in nonhuman pri- Available from: https://www.who.int/blueprint/what/ mates. Sci Transl Med. 2015;7:301ra132. research-development/MERS_CoV_TPP_15052017. [28] Wang C, Zheng X, Gai W, et al. MERS-CoV virus-like pdf particles produced in insect cells induce specific [12] Aguanno R, ElIdrissi A, Elkholy AA, et al. MERS: humoural and cellular imminity in rhesus macaques. Progress on the global response, remaining challenges Oncotarget. 2017;8:12686–12694. and the way forward. Antiviral Res. 2018;159:35–44. [29] Lan J, Yao Y, Deng Y, et al. Recombinant receptor [13] Raj VS, Mou H, Smits SL, et al. Dipeptidyl peptidase 4 binding domain protein induces partial protective is a functional receptor for the emerging human coro- immunity in rhesus macaques against Middle East res- navirus-EMC. Nature. 2013;495:251–254. piratory syndrome coronavirus challenge. [14] Okba NM, Raj VS, Haagmans BL. Middle East respir- EBioMedicine. 2015;2:1438–1446. atory syndrome coronavirus vaccines: current status [30] Zhang N, Channappanavar R, Ma C, et al. and novel approaches. Curr Opin Virol. 2017;23:49– Identification of an ideal adjuvant for receptor-binding 58. domain-based subunit vaccines against Middle East [15] Cho H, Excler J-L, Kim JH, et al. Development of respiratory syndrome coronavirus. Cell Mol Middle East respiratory syndrome coronavirus vac- Immunol. 2016;13:180–190. cines – advances and challenges. Hum Vaccin [31] Lan J, Deng Y, Chen H, et al. Tailoring subunit vaccine Immunother. 2018;14:304–313. immunity with adjuvant combinations and delivery [16] Stalin Raj V, Okba NMA, Gutierrez-Alvarez J, et al. routes using the Middle East respiratory coronavirus Chimeric camel/human heavy-chain antibodies pro- (MERS-CoV) receptor-binding domain as an antigen. tect against MERS-CoV infection. Sci Adv. 2018;4: PLoS One. 2014;9:e112602. eaas9667. [32] Tang J, Zhang N, Tao X, et al. Optimization of antigen [17] Widjaja I, Wang C, van Haperen R, et al. Towards a dose for a receptor-binding domain-based subunit vac- solution to MERS: protective human monoclonal anti- cine against MERS coronavirus. Hum Vaccin bodies targeting different domains and functions of the Immunother. 2015;11:1244–1250. MERS-coronavirus spike glycoprotein. Emerg [33] Tai W, Wang Y, Fett CA, et al. Recombinant receptor- Microbes Infect. 2019;8:516–530. binding domains of multiple Middle East respiratory [18] Xu J, Jia W, Wang P, et al. Antibodies and vaccines syndrome coronaviruses (MERS-CoVs) induce cross- against Middle East respiratory syndrome coronavirus. neutralizing antibodies against divergent human and Emerg Microbes Infect. 2019;8:841–856. camel MERS-CoVs and antibody escape mutants. J [19] Haagmans BL, van den Brand JMA, Raj VS, et al. An Virol. 2017;91:e01651–16. orthopoxvirus-based vaccine reduces virus excretion [34] Tao X, Garron T, Agrawal AS. Characterization and after MERS-CoV infection in dromedary camels. demonstration of the value of a lethal mouse model Science. 2016;351:77–81. of Middle East respiratory syndrome coronavirus [20] Guo X, Deng Y, Chen H, et al. Systemic and mucosal infection and disease. J Virol. 2016;90:57–67. immunity in mice elicited by a single immunization [35] Tai W, Zhao G, Sun S, et al. A recombinant receptor- with human adenovirus type 5 or 41 vector-based binding domain of MERS-CoV in trimeric form pro- vaccines carrying the spike protein of Middle East res- tects human dipeptidyl peptidase 4 (hDPP4) trans- piratory syndrome coronavirus. Immunology. genic mice from MERS-CoV infection. Virology. 2015;145:476–484. 2016;499:375–382. [21] Kim E, Okada K, Kenniston T, et al. Immunogenicity [36] Jiaming L, Yanfeng Y, Yao D, et al. The recombinant of an adenoviral-based Middle East respiratory N-terminal domain of spike proteins is a potential Emerging Microbes & Infections 1603 vaccine against Middle East respiratory syndrome cor- Mild Middle East respiratory syndrome coronavirus onavirus (MERS-CoV) infection. Vaccine. 2017;35:10– infections. Emerg Infect Dis. 2019;25:1868–1877. 18. [43] Burbelo PD, Goldman R, Mattson TL. A simplified [37] Adney RD, Wang L, van Doremalen N, et al. Efficacy of immunoprecipitation method for quantitatively an adjuvanted Middle East respiratory syndrome coro- measuring antibody responses in clinical sera samples navirus spike protein vaccine in dromedary camels and by using mammalian-produced Renilla luciferase-anti- alpacas. Viruses. 2019;11:212. gen fusion proteins. BMC Biotechnol. 2005;5:22. [38] Mou H, Raj VS, van Kuppeveld FJM, et al. The receptor [44] Burbelo PD, Ching KH, Klimavicz CM, et al. Antibody binding domain of the new Middle East respiratory profiling by luciferase immunoprecipitation systems syndrome coronavirus maps to a 231-residue region (LIPS). J Vis Exp. 2009;32:1549. in the spike protein that efficiently elicits neutralizing [45] Li W, Hulswit RJG, Widjaja I, et al. Identification of antibodies. J Virol. 2013;87:9379–9383. sialic acid-binding function for the Middle East respir- [39] Vergara-Alert J, Raj VS, Muñoz M, et al. Middle East atory syndrome coronavirus spike glycoprotein. Proc respiratory syndrome coronavirus experimental trans- Natl Acad Sci. 2017;114:E8508–E8517. mission using a pig model. Transbound Emerg Dis. [46] Letko M, Miazgowicz K, McMinn R, et al. Adaptive 2017;64:1342–1345. Evolution of MERS-CoV to species Variation in [40] Corman VM, Eckerle I, Bleicker T, et al. Detection of a DPP4. Cell Rep. 2018;24:1730–1737. novel human coronavirus by real-time reverse-tran- [47] Hemida MG, Alnaeem A, Chu DK, et al. Longitudinal scription polymerase chain reaction. Euro Surveill. study of Middle East respiratory syndrome coronavirus 2012;17:20285. infection in dromedary camel herds in Saudi Arabia, [41] Coleman CM, Frieman MB. Growth and quantifi- 2014–2015. Emerg Microbes Infect. 2017;6:1–7. cation of MERS-CoV infection. Curr Protoc [48] Ibrahim EE-S, Gamal WM, Hassan AI, et al. Comparative Microbiol. 2015;37(1):15E.2.1–15E.2.9. study on the immunopotentiator effect of ISA 201, ISA 61, [42] Okba NMA, Raj VS, Widjaja I, et al. Sensitive and ISA 50, ISA 206 used in trivalent foot and mouth disease specific detection of Low-level antibody responses in vaccine. Vet World. 2015;8:1189–1198.