key: cord-0972184-j4gv1vep
authors: An, Yaling; Li, Shihua; Jin, Xiyue; Han, Jian-bao; Xu, Kun; Xu, Senyu; Han, Yuxuan; Liu, Chuanyu; Zheng, Tianyi; Liu, Mei; Yang, Mi; Song, Tian-zhang; Huang, Baoying; Zhao, Li; Wang, Wen; Ruhan, A; Cheng, Yingjie; Wu, Changwei; Huang, Enqi; Yang, Shilong; Wong, Gary; Bi, Yuhai; Ke, Changwen; Tan, Wenjie; Yan, Jinghua; Zheng, Yong-tang; Dai, Lianpan; Gao, George F.
title: A tandem-repeat dimeric RBD protein-based COVID-19 vaccine ZF2001 protects mice and nonhuman primates
date: 2021-03-11
journal: bioRxiv
DOI: 10.1101/2021.03.11.434928
sha: 123b7a0fc4b049d8a4cdb3e7a05328cbf1854389
doc_id: 972184
cord_uid: j4gv1vep

A safe, efficacious and deployable vaccine is urgently needed to control COVID-19 pandemic. We report here the preclinical development of a COVID-19 vaccine candidate, ZF2001, which contains tandem-repeat dimeric receptor-binding domain (RBD) protein with alum-based adjuvant. We assessed vaccine immunogenicity and efficacy in both mice and non-human primates (NHPs). ZF2001 induced high levels of RBD-binding and SARS-CoV-2 neutralizing antibody in both mice and NHPs, and also elicited balanced TH1/TH2 cellular responses in NHPs. Two doses of ZF2001 protected Ad-hACE2-transduced mice against SARS-CoV-2 infection, as detected by reduced viral RNA and relieved lung injuries. In NHPs, vaccination of either 25 μg or 50 μg ZF2001 prevented infection with SARS-CoV-2 in lung, trachea and bronchi, with milder lung lesions. No evidence of disease enhancement is observed in both models. ZF2001 is being evaluated in the ongoing international multi-center Phase 3 trials (NCT04646590) and has been approved for emergency use in Uzbekistan.

severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection 67 (Coronaviridae Study Group of the International Committee on Taxonomy of, 2020; 68 Jiang et al., 2020) has led to more than 100 million confirmed cases, with more than 2 69 million deaths (www.who.int). Vaccine is the "final solution" to control the pandemic. In response to the pandemic, we developed a protein subunit vaccine targeting 86

To study the protective efficacy of ZF2001 vaccine against SARS-CoV-2 infection, 133 groups of C57/B6 mice were immunized with two doses of 10 μg ZF2001, 21 days 134 apart ( Figure 2A ). Mice receiving placebo were used as the control. Two doses of 135 ZF2001 vaccine elicited high levels of both RBD-binding IgG (GMT, 81,920) and 136 SARS-CoV-2 NAb (GMT, 959) ( Figure 2B and 2C). These immunized mice were 137 subsequently transduced via intranasal route with adenovirus expressing hACE2 as 138 the SARS-CoV-2-sensitive animal model. Five days later, these transduced mice were 139 intranasally challenged with 5 X 10 5 or 1 X 10 5 50% tissue culture infectious dose 140 (TCID50) of SARS-CoV-2 (HB01 strain) (Wei et al., 2020; Tan et al., 2020). Mice were 141 euthanized and necropsied at 3 or 5 days post infection (DPI) to detect viral loads and 142 exam pulmonary pathology. Encouragingly, in mice challenged with 5 X 10 5 TCID50 143 SARS-CoV-2, the mean titers of viral genomic RNA (gRNA) per gram of lung for 144 placebo and ZF2001 group were 6.5 x 10 8 and 3.0 x 10 6 (218-fold reduction), 145 respectively, at 3 DPI, and 7.0 x 10 8 and 4.8 x 10 5 (1,448-fold reduction), respectively, 146 at 5 DPI ( Figure 2D ). In mice challenged with 1 X 10 5 TCID50 SARS-CoV-2, the mean 147 gRNA titers for placebo and ZF2001 group, were 2.3 x 10 9 and 1.2 x 10 5 , respectively, 148 a reduction of 18,736-fold. The residual viral RNA in lung were probably derived from 149 the high amount of input viruses. Therefore, we measured the magnitudes of 150 subgenomic RNA (sgRNA) to quantify the infectious virus because sgRNA are 151 generated in the infected cells during virus replication but is absent in virions (Kim et 152 sgRNA per gram of lung for placebo recipients were 3.6 x 10 8 and 5.1 x 10 8 at 3 DPI 154 and 5 DPI, respectively. By contrast, this titer was reduced to 1.4 x 10 6 at 3 DPI and 155 undetectable at 5 DPI for ZF2001 vaccine recipients ( Figure 3B ). Accordingly, in mice 156 challenged with 1 X 10 5 TCID50 SARS-CoV-2, these sgRNA titer was 8.1 x 10 8 for 157 placebo recipients, but was undetectable for ZF2001 recipients, at 3 DPI ( Figure 2E ). 158

The neutralizing titers are negatively correlated with the gRNA and sgRNA, with 159 Spearman correlation r value of -0.7127 and -0.6982, respectively ( Figure 2F ). 160 Consistent with this, immunofluorescence analysis of lung section stained with anti-161 SARS-CoV-2 NP antibody demonstrated the presence of viral protein in mice receiving 162 placebo, but absence in mice receiving ZF2001 vaccine ( Figure 3A ). These results 163 demonstrated that ZF2001 vaccine is protective against pulmonary SARS-CoV-2 164

To further assess the vaccine protection against SARS-CoV-2-induced lung damage, 166 histopathological examination was performed. All lung tissue samples harvested from 167 mice with placebo exhibited apparently moderated to severe viral pneumonia, 168 characterized by thickened alveolar walls, vanishment of alveolar cavities, pulmonary 169 vascular congestion and diffuse inflammatory cell infiltration ( Figure 3B ). By contrast, 170 a markedly relieved histopathological changes were observed in lung tissues of mice 171 vaccinated with ZF2001 vaccine ( Figure 3B ). This result demonstrated that ZF2001 172 protected against SARS-CoV-2-induced lung injury in mice, without observation of 173 vaccine enhanced diseases. 174

To further evaluate ZF2001 vaccine in NHPs, groups of healthy young cynomolgus 177 macaques (n=10) were immunized with four doses of 25 μg or 50 μg vaccine 178 intramuscularly at Week 0, 4, 8 and 10, respectively, to monitor the immunogenicity 179 kinetics ( Figure 4A ). We found after one immunization, ZF2001 vaccine can elicit 180 serological RBD-binding IgG GMTs up to ~100,000 in both the 25 μg or 50 μg groups, 181 and these titers were further enhanced to more than 1000,000 after the second 182 immunization ( Figure 4B ). The third and fourth boosts did not significantly increase the 183 GMTs ( Figure 4B ). For the NAb response, two immunizations of ZF2001 vacine elicited 184 SARS-CoV-2 neutralizing GMTs of 630 in the 25 μg group and 776 in the 50 μg group, 185 respectively. The third immunization did not further enhance NAb titer ( Figure 4C ). 186

To detect the cellular immune responses in NHPs, 6 macaques in each group were 187 euthanized to harvest spleen at 3 days after the last vaccination. Splenocytes were 188 stimulated with RBD protein. ELISPOT assay was perform to detect TH1 (IFN-γ, IL-2) 189 and TH2 (IL-4) cytokine production. We found both 25 μg or 50 μg ZF2001 induced 190 substantial cellular responses, with the significantly enhanced and TH1/TH2 balanced 191 cytokine production ( Figure 4D ). The macaques were challenged with 1 X 10 6 TCID50 SARS-CoV-2 intratracheally at 212

Day 28 post vaccination. Macaques were euthanized at 8 DPI. Tissues from 7 lung 213 lobes (4 sites in each lobe), trachea, left and right bronchus were collected and 214 quantified for viral gRNA. As expected, substantial virus loads can be detected in lung 215 lobes (average 4748 copies/μg gRNA). By contrast, animals receiving either 25 μg or 216 50 μg ZF2001 vaccine exhibited significantly reduced virus loads in lung lobes, with 217 average 1 and 10 copies/μg gRNA for 25 μg and 50 μg group, respectively. In addition, 218 placebo, but was undetectable in all ZF2001-vaccinated macaques ( Figure 6A ). 220

To further assess the protection efficacy in lung, pulmonary histopathology of each 221 macaque was scored based on the thickening of alveolar septa, pulmonary alveolar 222 congestion and inflammatory cell infiltration in alveoli and trachea. Overall, high scores 223 of lung lesions were found in control animals, but were dramatically reduced in 224 ZF2001-vaccinated animals ( Figure 6B ). In particular, all three control animals showed 225 severe pulmonary alveolar congestion. By contrast, macaques receiving either 25 μg 226 or 50 μg vaccine showed almost no alveolar congestion ( Figure 6B and 6C). The 227 typical images of tissue hematoxylin-eosin (H/E) staining were shown in the Figure 6C . 228 We did not observe vaccine enhanced diseases in all these macaques. week-old female BALB/c mice (n=5) vaccinated with 10 μg ZF2001 or placebo. intranasal (i.n.) route. Each groups of mice were further split into three groups (n=6), 308 with the former two groups infected with high-dose (5 x 10 5 TCID50) SARS-CoV-2 and 309 the latter group infected with low-dose (1 x 10 5 TCID50) SARS-CoV-2. Lung tissues 310 95% CI. P values were analyzed with unpaired t test (****, P < 0.0001). The dashed 314 line indicates the limit of detection.

(C) SARS-CoV-2 (HB01 strain) neutralization assay shows the 50% neutralization titer. 316

Data are geometric mean with 95% CI. P values were analyzed with unpaired t test (*, 317 P < 0.05; ****, P < 0.0001). The dashed line indicates the limit of detection. 

Cynomolgus macaques were divided into three groups (n=10), with five females 520 weighing 2.97 to 3.62 kg and five males weighing 2.95 to 3.67 kg in each group. 

Monomeric RBD protein of SARS-CoV-2 used in ELISA assay was expressed and 538 purified as previously described (Dai et al., 2020) . Briefly, the coding sequence for 539 SARS-CoV-2 RBD (S protein 319-541, GISAID accession No. EPI_ISL_402119) was 540 codon-optimized for mammalian cell expression and synthesized. For this construct, 541 signal peptide sequence of MERS-CoV S protein (S protein residues 1-17) was added 542 to the protein N terminus for protein secretion, and a hexa-His tag was added to the C 543 terminus to facilitate further purification processes. The construct, which synthesized 544 by GENEWIZ, China, was cloned into the pCAGGS vector and transiently transfected 545 into HEK293T cells. After 3 days, the supernatant was collected and soluble protein 546 was purified by Ni affinity chromatography using a HisTrap TM HP 5 mL column (GE 547 Healthcare). The sample was further purified via gel filtration chromatography with 548 ELISA 551

For mice and rhesus macaque, ELISA plates (3590; Corning, USA) were coated 552 over-night with 3 μg/ml of RBD protein in 0.05 M carbonate-bicarbonate buffer, pH 9.6, 553 and blocked in 5% skim milk in PBS. Serum samples were serially diluted and added 554 to each well. Plates were incubated with goat anti-mouse IgG-HRP antibody or goat 555 anti-monkey IgG-HRP antibody and subsequently developed with 3,3',5,5'-556 tetramethylbenzidine (TMB) substrate. Reactions were stopped with 2 M hydrochloric 557 acid, and the absorbance was measured at 450 nm using a microplate reader 558 (PerkinElmer, USA). The endpoint titers were defined as the highest reciprocal dilution 559 of serum to give an absorbance greater than 2.5-fold of the background values. 560

Antibody titer below the limit of detection was determined as half the limit of detection. 561

For cynomolgus macaques, ELISA plates were coated over-night with 1 μg/ml of 562 RBD-dimer protein and blocked in 3% skim milk in PBST. Serum samples were serially 563 diluted and added to each well. 10ul serum from cynomolgus macaques of blank 564 control group was mixed as one blank control serum and diluted as the same with initial 565 fold of vaccine group. The subsequent operations were as described above. Cut off 566 value was defined 2.1-fold of OD450 values of the blank control serum. When OD450 > 567 Cut off value, the serum was determined positive. The endpoint titer was defined as 568 the highest reciprocal dilution of serum with an absorbance greater than 2.1-fold of the 569 blank control serum value. When OD450 value of the serum to test ≤ Cut off value, the 570

The neutralizing activity of serum was assessed using a previously described SARS-573

CoV-2 neutralization assay (Nie et al., 2020) . Briefly, mouse serum samples were 4-574 fold serially diluted and mixed with the same volume of 100 TCID50 SARS-CoV-2, then 575 incubated at 37°C for 1 hour. Then, 100 μL virus-serum mixture was transferred to pre- Viral loads were expressed on a log10 scale as viral copies/gram after calculation 622 with a standard curve. Viral copy numbers below the limit of detection were set as 623 the half of the limit of detection. 624

The total RNA of tissues from rhesus macaques were extracted with TRIzol reagent China), and the copy number of each sample was calculated. 633

Mice lung tissues were fixed in 4% paraformaldehyde, dehydrated, embedded in 635 paraffin, and then sectioned. Tissue sections (4 μm) were deparaffinized in xylene and 636

A universal design of betacoronavirus vaccines against COVID-19, MERS, and SARS

A SARS-CoV-2 infection model in mice 689 demonstrates protection by neutralizing antibodies

Establishment and validation of a pseudovirus neutralization assay for

Emerging microbes & infections 9

Delayed severe cytokine storm and immune cell infiltration in SARS-695

CoV-2-infected aged Chinese rhesus macaques

Lower respiratory tract samples are reliable for severe acute respiratory 698 syndrome coronavirus 2 nucleic acid diagnosis and animal model study

Virological assessment of hospitalized patients 702 with COVID-2019

Recombinant chimpanzee adenovirus vaccine AdC7-M/E protects against zika 705 virus infection and testis damage

Translocation of microbes and changes of immunocytes in the gut of rapid-708 and slow-progressor Chinese rhesus macaques infected with SIVmac239

protein of SARS-CoV-2 (for cynomolgus macaques) was added to the wells for 596 stimulation. Phytohemagglutinin (PHA) was added as a positive control. Cells without 597 stimulation were employed as a negative control. After 12-36 hours of incubation, the 598 cells were removed, and the plates were processed with biotinylated IFN-γ, IL-2 or IL-599 4 detection antibody, streptavidin-HRP conjugate, and substrate. When the colored 600 spots were intense enough to be observed, the development was stopped by 601 thoroughly rinsing samples with deionized water. The numbers of the spots were 602 determined using an automatic ELISPOT reader and image analysis software.