and decreases neutralization sensitivity to individual ......2020/06/20 · 15 Ai-Long Huang, Ni Tang, Kai Wang, Key Laboratory of Molecular Biology for 16 Infectious Diseases (Ministry

The D614G mutation of SARS-CoV-2 spike protein enhances viral infectivity 1

and decreases neutralization sensitivity to individual convalescent sera 2

Running Title： D614G mutant spike increases SARS-CoV-2 infectivity 3

Jie Hu1, #, Chang-Long He1, #, Qing-Zhu Gao1, #,Gui-Ji Zhang1 #, Xiao-Xia Cao1, 4

Quan-Xin Long1, Hai-Jun Deng1, Lu-Yi Huang1, Juan Chen1, Kai Wang1,*, Ni Tang1,*, 5

Ai-Long Huang1,* 6

7

1Key Laboratory of Molecular Biology for Infectious Diseases (Ministry of 8

Education), Institute for Viral Hepatitis, Department of Infectious Diseases, The 9

Second Affiliated Hospital, Chongqing Medical University, Chongqing, 400010, 10

China 11

#These authors contributed equally to this work. 12

13

*Corresponding authors: 14

Ai-Long Huang, Ni Tang, Kai Wang, Key Laboratory of Molecular Biology for 15

Infectious Diseases (Ministry of Education), Institute for Viral Hepatitis, Department 16

of Infectious Diseases, The Second Affiliated Hospital, Chongqing Medical 17

University, Chongqing, China. Phone: 86-23-68486780, Fax: 86-23-68486780, 18

E-mail: [email protected] (N.T.), [email protected] (A.L.H.), 19

[email protected] (K.W.) 20

21

.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is

The copyright holder for this preprintthis version posted June 20, 2020. ; https://doi.org/10.1101/2020.06.20.161323doi: bioRxiv preprint

https://doi.org/10.1101/2020.06.20.161323

http://creativecommons.org/licenses/by-nc-nd/4.0/

Abstract 22

Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory 23

syndrome coronavirus 2 (SARS-CoV-2). The spike protein that mediates 24

SARS-CoV-2 entry into host cells, is one of the major targets for vaccines and 25

therapeutics. Thus, insights into the sequence variations of S protein are key to 26

understanding the infection and antigenicity of SARS-CoV-2. Here, we observed a 27

dominant mutational variant at the 614 position of S protein (aspartate to glycine, 28

D614G mutation). Using pseudovirus-based assay, we found that S-D614 and 29

S-G614 protein pseudotyped viruses share a common receptor, human 30

angiotensin-converting enzyme 2 (ACE2), which could be blocked by recombinant 31

ACE2 with the fused Fc region of human IgG1. However, S-D614 and S-G614 32

protein demonstrated functional differences. First, S-G614 protein could be cleaved 33

by serine protease elastase-2 more efficiently. Second, S-G614 pseudovirus 34

infected 293T-ACE2 cells significantly more efficiently than the S-D614 pseudovirus, 35

Moreover, 93% (38/41) sera from convalescent COVID-19 patients could neutralize 36

both S-D614 and S-G614 pseudotyped viruses with comparable efficiencies, but 37

about 7% (3/41) convalescent sera showed decreased neutralizing activity against 38

S-G614 pseudovirus. These findings have important implications for SARS-CoV-2 39

transmission and immune interventions. 40

Keywords: antiviral therapeutics, coronavirus, D614G mutation, neutralizing 41

antibodies, pseudovirus, SARS-CoV-2, spike protein 42

43



https://doi.org/10.1101/2020.06.20.161323


Introduction 44

SARS-CoV-2 is a novel coronavirus reported in 2019 that caused the recent 45

outbreak of coronavirus disease-2019 (COVID-19)1. By June 17, 2020, the World 46

Health Organization (WHO) reported that 8.06 million people worldwide had been 47

infected with SARS-CoV-2, and 440,290 individuals died of COVID-19. This 48

pandemic had a significant adverse impact on international social and economic 49

activities. The RNA genome of SARS-CoV-2 was rapidly sequenced to facilitate 50

diagnostic testing, molecular epidemiologic source tracking, and development of 51

vaccines and therapeutic strategies2. Coronaviruses are enveloped, 52

positive-stranded RNA viruses that contain the largest known RNA genomes to 53

date. The mutation rate for RNA viruses is extremely high, which may contribute to 54

its transmission and virulence. The only significant variation in the SARS-CoV-2 55

spike (S) protein is a non-synonymous D614G (Aspartate (D) to Glycine (G)) 56

mutation3. Primary data showed that S-G614 is a more pathogenic strain of 57

SARS-CoV-2 with high transmission efficiency3, however whether D614G 58

conversion in S protein affect the viral entry and infectivity in cell model is still 59

unclear. 60

61

The S protein of coronavirus, the major determinant of host and tissue tropism, is a 62

major target for vaccines, neutralizing antibodies, and viral entry inhibitors4,5. 63

Similar to SARS-CoV, the cellular receptor of SARS-CoV-2 is 64

angiotensin-converting enzyme 2 (ACE2); however, the SARS-CoV-2 S protein has 65



https://doi.org/10.1101/2020.06.20.161323


a 10- to 20-fold higher affinity for ACE2 than the corresponding S protein of 66

SARS-CoV6,7. Coronaviruses use two distinct pathways for cell entry: 67

protease-mediated cell-surface pathway and endosomal pathway8. The S proteins 68

of several coronaviruses are cleaved by host proteases into S1 subunit for receptor 69

binding and S2 subunit for membrane fusion at the entry step of infection. Several 70

cellular proteases, including furin, transmembrane protease serine 2 (TMPRSS2) 71

and cathepsin (Cat)B/L, are critical for priming the SARS-CoV-2 S protein to 72

enhance ACE2-mediated viral entry4. Recently, Chandrika Bhattacharyya et al 73

reported that a novel serine protease (elastase-2) cleavage site was introduced in 74

S-G614 protein of SARS-CoV-2 variant9. However, it is unknown whether S-G614 75

protein can be processed and activated by elastase-2 in cell model. The S protein 76

plays a key role in the evolution of coronavirus to evade the host immune system. It 77

is still uncertain whether D614G mutation affect the antigenic properties of the S 78

protein. Whether elastase-2 inhibitors and convalescent serum samples of 79

COVID-19 can block the infection of SARS-CoV-2 D614G variant remains unclear. 80

81

In this study, we analyzed the S gene sequences of SARS-CoV-2 submitted to the 82

Global Initiative on Sharing All Influenza Data (GISAID) database. We showed the 83

expression and cleavage of S-D614 and S-G614 protein in cell lines. Using a 84

luciferase (Luc)-expressing lentiviral pseudotype system, we established a 85

quantitative pseudovirus-based assay for the evaluation of SARS-CoV-2 cell entry 86

mediated by the viral S protein variants. We also compared the neutralizing 87



https://doi.org/10.1101/2020.06.20.161323


sensitivity of the S-D614 and S-G614 protein pseudovirus to convalescent sera 88

from COVID-19 patients. Our study provide further insight into the transmission and 89

immune interventions of this newly emerged virus. 90

91

Methods 92

Plasmids. The codon-optimized gene encoding SARS-CoV-2 S protein (GenBank: 93

QHD43416) with C terminal 19 amino acids deletion was synthesized by Sino 94

Biological Inc. (Beijing, China), and cloned into the Kpn I and Xba I restriction sites 95

of pCMV3 vector (pCMV3-SARS-CoV-2-S-C19del, denoted as pS-D614). The 96

D614G mutant S expressing plasmid (denoted as pS-G614) was constructed by 97

site-directed mutagenesis, with pS-D614 plasmid as a template. The HIV-1 NL4-3 98

ΔEnv Vpr luciferase reporter vector (pNL4-3.Luc.R-E-) constructed by N. Landau10, 99

was provided by Prof. Cheguo Cai from Wuhan University (Wuhan, China). The 100

vesicular stomatitis virus G (VSV-G)-expressing plasmid pMD2.G was provided by 101

Prof. Ding Xue from the Tsinghua University (Beijing, China). The expression 102

plasmid for human ACE2 and ELANE (elastase-2) were obtained from 103

Genecopoeia (Guangzhou, China). 104

105

Cell lines. HEK293T cells were purchased from the American Type Culture 106

Collection (ATCC, Manassas, VA, USA). Cells were maintained in Dulbecco’s 107

Modified Eagle Medium (DMEM; Hyclone, Waltham, MA, USA) supplemented with 108

10% fetal bovine serum (FBS; Gibco, Rockville, MD, USA), 100 mg/mL of 109



https://doi.org/10.1101/2020.06.20.161323


streptomycin, and 100 unit/mL of penicillin at 37 °C in 5% CO2. HEK293T cells 110

transfected with human ACE2 (293T-ACE2) were cultured under the same 111

conditions with the addition of G418 (0.5 mg/mL) to the medium. 112

113

Antibodies and inhibitors. The anti-RBD (receptor-binding domain) monoclonal 114

antibody against the SARS-CoV-2 S protein was kindly provided by Prof. Aishun Jin 115

from Chongqing Medical University. Recombinant human ACE2 linked to the Fc 116

domain of human IgG1 (ACE2-Ig, Sino Biological Inc.). Camostat mesylate (Tokyo 117

Chemical Industry, Tokyo, Japan) and aloxistatin (E-64d; MedChemExpress, 118

Monmouth Junction, NJ, USA) were dissolved in dimethyl sulfoxide (DMSO) at a 119

stock concentration of 50 mM. 120

121

Serum samples. A total of 41 convalescent COVID-19 patient sera (at 2-4 weeks 122

after symptom onset) were collected from three designated hospitals in Chongqing 123

in the period from February 5 to February 10, 2020 (Supplementary Table 1). All 124

sera were tested positive using MCLIA kits supplied by Bioscience Co. (Tianjin, 125

China)11. Patient sera were incubated at 56 °C for 30 min to inactivate the 126

complement prior to experiments. 127

128

SARS-CoV-2 genome analysis 129

The online Nextstrain analysis tool (https://nextstrain.org/ncov) was used to track 130

D614G mutation in SARS-CoV-2 genomes. Global of 2834 genomes sampled 131



https://doi.org/10.1101/2020.06.20.161323


between Dec 20, 2019 and Jun 12, 2020 were visualized using the ‘rectangular’ 132

layout. The mutations were labeled in branch. 133

All complete SARS-CoV-2 S gene sequences were downloaded from NCBI website 134

(https://www.ncbi.nlm.nih.gov/sars-cov-2/) on Jun 1, 2020. We obtained 4701 S 135

coding sequences from NCBI, after excluded partial and frameshift sequences, 136

4649 completed S sequences were used to further analysis. All S nucleotide 137

sequences were translated to amino acid sequences. The nucleotide and amino 138

acid sequences of S were aligned with multiple sequence alignment software 139

MUSCLE separately. The ‘Wuhan-Hu-1’ strain (NC_045512) was used to be the 140

reference sequence, and the mutations were extracted using private PERL scripts. 141

142

Western blot analysis of SARS-CoV-2 S protein expression. To analyze S 143

protein expression in cells, S-D614 and S-G614 expressing plasmids were 144

transfected into HEK293T cells respectively. Total protein was extracted from cells 145

using RIPA Lysis Buffer (CWbiotech, Beijing, China) containing 1 mM phenyl methyl 146

sulfonyl fluoride (PMSF; Beyotime, Shanghai, China). Equal amounts of protein 147

samples were electrophoretically separated by 10% sodium dodecyl sulfate 148

polyacrylamide gel electrophoresis (SDS-PAGE), and then transferred to 149

polyvinylidene difluoride (PVDF) membrane (Millipore, Billerica, MA, USA). The 150

immunoblots were probed with the indicated antibodies. Protein bands were 151

visualized using SuperSignal™ West Pico Chemiluminescent Substrate kits 152

(Bio-Rad, Hercules, CA, USA) and quantified by densitometry using ImageJ 153



https://www.ncbi.nlm.nih.gov/sars-cov-2/

https://doi.org/10.1101/2020.06.20.161323


software (National Center for Biotechnology Information [NCBI], Bethesda, MD, 154

USA). 155

156

Production and titration of SARS-CoV-2 S pseudoviruses. SARS-CoV-2 157

pseudotyped viruses were produced as previously described with some 158

modifications12. Briefly, 5×106 HEK293T cells were co-transfected with 6 μg 159

pNL4-3.Luc.R-E- and 6 μg recombinant SARS-CoV-2 S plasmids (pS-D614, or 160

pS-G614) using the Lipofectamine 3000 transfection reagent (Invitrogen) according 161

to the manufacturer’s instructions. The cells were transferred to fresh DMEM 12 h 162

later. The S-D614 and S-G614 protein pseudotyped viruses in supernatants were 163

harvested 48 h after transfection, centrifuged, filtered through a 0.45 μm filter and 164

stored at −80°C. The pMD2.G was co-transfected with the pNL4-3.Luc.R-E- plasmid 165

to package the VSV-G pseudovirus. 166

The titers of the pseudoviruses were calculated by determining the number of viral 167

RNA genomes per milliliter of viral stock solution using real-time (RT)-qPCR with 168

primers and a probe that target LTR13. Sense primer: 169

5'-TGTGTGCCCGTCTGTTGTGT-3' 3’, anti-sense primer: 170

5'-GAGTCCTGCGTCGAGAGAGC-3', probe: 171

5'-FAM-CAGTGGCGCCCGAACAGGGA- BHQ1-3'. Briefly, viral RNAs were 172

extracted with Trizol (Invitrogen, Rockville, MD) and treated with RNase-free DNase 173

(Promega, Madison, WI, USA) and re-purified using mini columns. Then RNA was 174

amplified using the TaqMan One-Step RT-PCR master mix reagents (Applied 175



https://doi.org/10.1101/2020.06.20.161323


Biosystems, ThermoFisher). The known quantity of pNL4-3.Luc.R-E- vector was 176

used to generate standard curves. The S-D614 and S-G614 protein pseudotyped 177

viruses were adjusted to the same titer (copies/ml) for the following experiments. 178

179

SARS-CoV-2 spike-mediated pseudovirus entry assay. To detect S variants 180

mediated viral entry, 293T-ACE2 cells (2×104) grown in 96-well plates were 181

respectively infected with same amount of S-D614 or S-G614 pseudovirus (3.8x104 182

copies in 50 μL). Cells were transferred to fresh DMEM medium 8 h post-infection, 183

and relative luminescence units (RLU) were measured 24-72 h post-infection by 184

Luciferase Assay Reagent (Promega, Madison, WI, USA) according to the 185

manufacturer's protocol14. 186

187

Neutralization and inhibition assays. The 293T-ACE2 cells (2×104 cells/well) 188

were seeded in 96-well plates. For the neutralization assay, 50 μL pseudoviruses, 189

which is equivalent to 3.8x104 vector genomes, were incubated with serial dilutions 190

of sera sample from patients and human normal serum as a negative control for 1 h 191

at 37 °C, then added to the 96-well 293T-ACE2 cells with 3 replicates. For the 192

inhibition assay, the cells were pretreated with the protease inhibitors (camostat 193

mesylate or E-64d) 2 h before infection. After 12 h incubation, the medium was 194

replaced with fresh cell culture medium. Luciferase activity was measured 72 h 195

after infection and the percent neutralization was calculated using GraphPad Prism 196

6.0 software (GraphPad Software, San Diego, CA, USA). Percentages of RLU 197



https://doi.org/10.1101/2020.06.20.161323


reduction (inhibition rate) were calculated as: 1- (RLU of sample sera - control 198

wells)/( RLU from mock control sera - control wells)) X100%. The titers of 199

neutralizing antibodies were calculated as 50% inhibitory dose (ID50). 200

201

Statistical analyses. Statistical analyses of the data were performed by using 202

GraphPad Prism version 6.0 software. Statistical significance was determined 203

using ANOVA for multiple comparisons. Student’s t-tests were applied to compare 204

the two groups. Differences with P-values < 0.05 were deemed statistically 205

significant. 206

207

Ethical approval. The study was approved by the Ethics Commission of 208

Chongqing Medical University (ref. no. 2020003). Written informed consent was 209

waived by the Ethics Commission of the designated hospital for emerging infectious 210

diseases. 211

212

Results 213

D614G mutation of SARS-CoV-2 spike protein was globally distributed 214

The spike (S) protein of SARS-CoV-2, containing 1273 amino acids, forms a 215

trimeric spike on the virion surface that plays essential roles in viral entry. We 216

analyzed the SARS-CoV-2 S protein amino acid sequence from viral genomic 217

sequences in GISAID database. In line with prior reports, we found a globally 218

distributed S protein mutation, D614 to G614, which represented in 64.6% of all the 219



https://doi.org/10.1101/2020.06.20.161323


analyzed sequences (Fig.1 and Table 1). Among the top 10 most abundant 220

non-synonymous mutations observed in S protein, the relative abundance of the 221

D614G mutant (the clade G) was the highest around the world, indicating that G614 222

strain may be selectively advantageous. Since the S protein is critical to 223

coronavirus infection, we sought to explore the impact of the most prevalent D614G 224

mutation on the S protein expression and function. 225

226

D614G mutation enhanced the cleavage of S protein variant by proteases 227

We predicted potential cleavage sites of proteases in S protein variants by 228

PROSPER15, and found a novel serine protease (elastase-2) cleavage site at 229

615-616 residues on S1-S2 junction region of S-G614 protein (Fig. 2a, 230

Supplementary Table 2). To evaluate the expression and cleavage of SARS-CoV-2 231

S protein in the human cell line, the codon-optimized S expressing plasmids 232

(pS-D614 and pS-G614) were transfected into HEK293T cells, respectively. 233

Immunoblot analysis of whole cell lysates revealed that both S-D614 and S-G614 234

proteins showed two major protein bands (unprocessed S and cleaved S1 subunit), 235

when reacted with the monoclonal antibody targeting the RBD on the SARS-CoV-2 236

spike (Fig. 2b). However, the pS-G614-transfected cells showed stronger S1 signal 237

than pS-D614-transfected cells, indicating that D614G mutation altered cleavability 238

of S protein by the cellular protease. Moreover, the elastase-2 inhibitor sivelestat 239

sodium significantly decreased the S1 signal of S-D614 protein (Fig. 2b). These 240

data indicated that the D614G mutation of SARS-CoV-2 S facilitates its cleavage by 241



https://doi.org/10.1101/2020.06.20.161323


host serine protease elastase-2. The coronavirus S protein must be cleaved by host 242

proteases to enable membrane fusion, which is critical for viral entry. Next, we 243

sought to explore the impact of D614G mutation on virus entry. 244

245

Evaluating viral entry efficacies between S-D614 and S-G614 pseudotyped 246

lentiviral particles 247

Lentiviral vectors can be pseudotyped with various heterologous viral glycoproteins 248

that modulate cellular tropism and cell-entry properties16. Due to the highly 249

pathogenic nature of SARS-CoV-2, infectious SARS-CoV-2 must be handled in a 250

biosafety level 3 (BSL-3) facility. We generated a pseudotyped SARS-CoV-2 based 251

on the viral S protein using a lentiviral system, which introduced a Luc reporter 252

gene for quantification of coronavirus S-mediated entry. pNL4-3.Luc.R-E- was 253

co-transfected with pS-D614, pS-G614, respectively to package the SARS-CoV-2 254

S pseudotyped single-round Luc virus in HEK293T cells. 255

The titers of S-D614 and S-G614 protein pseudotyped viruses were determined by 256

RT-qPCR expressed as the number of viral RNA genomes per milliliter, and then 257

adjusted to the same concentration (3.8x104 copies in 50 μL) for the following 258

experiments. The virus infectivity was determined by a Luc assay expressed in 259

RLU. HEK293T cells expressing human ACE2 (293T-ACE2) was used test the 260

correlation between ACE2 expression and pseudovirus susceptibility. VSV-G 261

pseudovirus was used as a control. As shown in Fig. 3a, both HEK293T and 262

293T-ACE2 cells could be effectively transduced by VSV-G pseudovirus. However, 263



https://doi.org/10.1101/2020.06.20.161323


the entry of S-D614 and S-G614 pseudovirus is highly dependent on its cellular 264

receptor ACE2 expression. The 293T-ACE2 cells showed an approximately 265

250-fold and 530-fold increase in Luc activity compared to HEK293T cells, when 266

transduced by S-D614 and S-G614 pseudoviruses, respectively (Fig. 3a). We next 267

detected the inhibitory ability of ACE2-Ig, a fusion protein consisting of the 268

extracellular domain (Met 1-Ser 740) of human ACE2 linked to the Fc region of 269

human IgG1 at the C-terminus17. Both S-D614 and S-G614 pseudoviruses were 270

potently inhibited by ACE2-Ig, and the IC50 (the concentration causing 50% 271

inhibition of pseudovirus infection) was 0.13 and 0.15 μg/mL, respectively (Fig. 3b). 272

To further compare the viral entry efficiency meditated by S variants, we detected 273

the Luc activity at different time post-infection. With the G614 Spike variant, the 274

increase in viral transduction over the D614 variant was 2.2-fold at 48 h 275

post-infection. The highest transduction efficiency (approximately 2.5 × 104 RLU) 276

was observed 72 h post-infection with S-G614 pseudovirus, which was 277

approximately 2.4-fold higher than S-D614 pseudovirus (Fig. 3c). These data 278

suggested that the D614G mutation in S protein significantly promotes virus entry 279

into ACE2-expressing cells, and ACE2-Ig efficiently blocks both wild-type and 280

mutant S pseudotype virus infection. We next explored the mechanism with which 281

S-G614 increased pseudovirus infectivity. 282

283

Protease inhibitors blocked the infection of S-G614 pseudovirus 284

The proteolytic activation of S protein is required for coronavirus infectivity, and 285



https://doi.org/10.1101/2020.06.20.161323


protease-mediated cell-surface pathway is critical for SARS-CoV-2 entry4. Since 286

we observed that S-G614 could be more efficiently cleaved by host protease when 287

exogenously expressed in 293T cells, we assumed that host proteases may be 288

involved in the enhancement of S-G614 virus entry. Two clinically proven protease 289

inhibitors, camostat mesylate and E-64d, which block host TMPRSS2 and CatB/L, 290

respectively, were tested on S-D614 and S-G614 protein pseudotyped lentiviral 291

particles. As shown in Fig. 3a, in 293T-ACE2 cells lacking TMPRSS2, the serine 292

protease inhibitor camostat mesylate did not inhibit S-D614 or S-G614 pseudovirus 293

infection. While the cysteine proteases inhibitor E-64d significantly blocked these 294

two pseudoviruses entry, and the IC50 was 0.37 μM for S-D614 pseudovirus and 295

0.24 μM for S-G614 pseudovirus (Fig. 3b). As expected, these protease inhibitors 296

had no impact on VSV-G pseudovirus infection. Together, these results suggested 297

that S-meditated viral entry into TMPRSS2 deficient 293T-ACE2 cells is endosomal 298

cysteine proteases CatB/L dependent, therefore S-D614 and S-G614 pseudovirus 299

show similar sensitivity to the CatB/L inhibitor E-64d in 293T-ACE2 cells. 300

301

Neutralization effect of convalescent COVID-19 patient sera against S-D614 302

and S-G614 pseudoviruses 303

Neutralizing antibodies are important for prevention and possibly recovery from 304

viral infections, however, as viruses mutate during replication and spread, host 305

neutralizing antibodies generated in the earlier phase of the infection may not be as 306

effective later on18,19. To test whether D614G mutations could affect the 307



https://doi.org/10.1101/2020.06.20.161323


neutralization sensitivity of the virus, the neutralization activity of serum samples 308

from convalescent patients with COVID-19 against SARS-CoV-2 S-D614 and 309

S-G614 pseudoviruses were evaluated. To perform the neutralization assay, 50 μl 310

pseudovirus (3.8x104 copies) was incubated with serial diluted serum. As shown in 311

Fig.4a, the inhibition rate of serum from convalescent COVID-19 patients was 312

analyzed at a single dilution of 1:1000. Among the 41 tested sera, 38 of them 313

showed neutralizing activities against both S-D614 and S-G614 pseudoviruses with 314

comparable efficiencies. Notably, 3 serum samples (patients 1#, 7#, 40#) 315

decreased inhibition rate against the S-G614 pseudovirus. The serum from patient 316

1# failed to neutralize the S-G614 pseudovirus though it neutralized about 30% of 317

the S-D614 pseudovirus at 1:1000 dilution. Then the inhibition curves and ID50 for 318

serum samples from 5 convalescent patient and a health donor were analyzed. 319

Sera from patients 17# and 39#, which were able to neutralize both two pseudotype 320

viruses to similar extents, showed similar ID50 values (Fig.4b). However, sera from 321

patients 1#, 7# and 40# showed high neutralizing activity against S-D614 322

pseudovirus with ID50 ranging from 894 to 1337, but decreased neutralizing activity 323

against S-G614 pseudovirus with ID50 ranging from 216 to 367, indicating 3.6 to 324

4.7-fold reduction in neutralizing titers (Fig.4b). These data indicated that D614G 325

mutation changes the antigenicity of S protein, thereby decreasing neutralization 326

sensitivity to individual convalescent sera. 327

328

Discussion 329



https://doi.org/10.1101/2020.06.20.161323


Pseudovirus-based assays have been widely used for the study of cellular tropism, 330

receptor recognition, viral inhibitors, and evaluation of neutralizing antibodies 331

without the requirement of BSL-3 laboratories. The present study generated a 332

pseudotyped SARS-CoV-2 based on the viral S protein using a lentiviral system, 333

which incorporated a Luc reporter gene for the easy quantification of coronavirus 334

S-mediated entry. We study the major mutation in S protein at position 614, and 335

found that serine protease elastase-2 participate in the proteolytic activation of the 336

S-G614 protein, thereby enhancing viral entry into 293T-ACE2 cells. In addition, we 337

found that the S-G614 pseudovirus was more resistant to neutralizing antisera than 338

S-G614 pseudovirus. 339

340

In this study, we found that the entry efficiency of S-G614 pseudotyped virus was 341

about 2.4 times higher than that of the S-D614 pseudovirus when normalized input 342

virus doses, suggesting that D614G mutation promotes the infectivity of 343

SARS-CoV-2 and enhances viral transmissibility. Since pseudovirus was a 344

single-round of infection assay, this seemingly small increase in entry activity could 345

cause a large difference in viral infectivity in vivo. Hangping Yao et al reported that 346

a patient-derived viral isolate ZJU-1, which harbors the D614G mutation, has a viral 347

load 19 times higher than isolate ZJU-8 (harboring the S-D614) when infecting 348

Vero-E6 cells20. However, the ZJU-1 isolate contains other two non-synonymous 349

mutations in ORF1a and envelope (E) gene. Our results also provide some 350

explanation for the association of the D614G mutation in S protein of SARS-CoV-2 351



https://doi.org/10.1101/2020.06.20.161323


with increased entry efficiency. The S-G614 protein contains a novel serine 352

protease cleavage site, so it could be cleaved by host elastase-2 more efficiently. 353

Previously studies on SARS-CoV demonstrated that the protease-mediated cell 354

surface entry facilitated a 100- to 1,000-fold higher efficient infection than did the 355

endosomal pathway in the absence of proteases21. 356

357

The elastase-2, also known as neutrophil elastase, plays an important role in 358

degenerative and inflammatory diseases. Sivelestat, approved to treat acute 359

respiratory distress syndrome (ARDS) in Japan and South Korea, has a beneficial 360

effect on the pulmonary function of ARDS patients with systemic inflammatory 361

response syndrome22. About 10–15% of patients with COVID-19 progresses to 362

ARDS23. Since sivelestat may not only mitigate the damage of neutrophil elastase 363

to lung connective tissue, but also limit the virus spreading capabilities by inhibiting 364

S protein processing, Mahmoud M. A. Mohamed et al advocated the use of 365

sivelestat to alleviate neutrophil-induced damage in high-risk COVID-19 patients24. 366

Our results showed that the S-G614 protein could be cleaved more efficiently, 367

which may be due to a novel elastase-2 cleavage site on the S1-S2 junction region 368

of S-G614 protein, so it might be an effective option for treatment of COVID-19 369

caused by SARS-CoV-2 harboring the D614G mutation. 370

371

Takahiko Koyama et al reported that D614G is located in one of the predicted B-cell 372

epitopes of SARS-CoV-2 S protein, and this is a highly immunodominant region 373



https://doi.org/10.1101/2020.06.20.161323


and may affect effectiveness of vaccine with wild-type S protein18. D614 is 374

conserved in S protein of SARS-CoV in 2003. Previous studies in SARS-CoV 375

suggested that the peptide S597–625 is a major immunodominant peptide in humans 376

and elicits a long-term B-cell memory response after natural infection with 377

SARS-CoV25. Regions between amino acids 614 and 621 of SARS-CoV-2 S 378

protein were also identified as a B cell epitope by different methods, and the 379

change of D614G may affect the antigenicity of this region26. Here, we observed 380

that 7% (3/41) convalescent sera showed markedly different neutralization activities 381

between S-G614 and S-G61 protein pseudotyped viruses, indicating that D614G 382

mutation reduces the neutralizing antibody sensitivity. Whether these patients were 383

at high risk of reinfection of S-G614 variant should be explored in further studies. It 384

will also be important to determine the breadth of neutralizing capacity of 385

vaccine-induced neutralizing antibodies. 386

387

Very recently, several groups also reported that the S-G614 enhances viral 388

infectivity based on pseudovirus assays27–29, but due to the small sample size, they 389

found no effect on neutralization sensitivity of the virus27,28. Given the evolving 390

nature of the SARS-CoV-2 RNA genome, antibody treatment and vaccine design 391

might require further considerations to accommodate the D614G and other 392

mutations that may affect the immunogenicity of the virus. 393

394

Our study has some limitations. First, the C-terminal 19 amino acids of S protein 395



https://doi.org/10.1101/2020.06.20.161323


were lacking in order to improve the packaging efficiency of SARS-CoV-2 S protein 396

pseudotyped virus, and this pseudovirus only recapitulates viral entry events; 397

therefore, additional assays with authentic SARS-CoV-2 are required. Second, we 398

only tested neutralizing antibodies against the S protein. However, previous studies 399

on SARS-CoV indicated that only a small fraction of memory B cells specific for 400

SARS-CoV antigens are directed against neutralizing epitopes present on the S 401

protein30. Third, in addition to D614G, further studies on other mutations in the S 402

protein are needed to evaluate their impact on SARS-CoV-2 infection, 403

pathogenicity and immunogenicity. Further studies are needed to determine the 404

impact of these mutations on the severity of COVID-19. 405

406

In summary, we established a SARS-CoV-2 spike-mediated pseudovirus entry 407

assay and studied the cell entry of S-D614 and S-G614 pseudotyped viruses. Our 408

study provided evidence that D614G mutation introduces an additional elastase-2 409

cut site in S protein, thereby promoting its cleavage and viral cell entry, resulting in 410

more transmissible SARS-CoV-2. Importantly, the D614G mutation reduced the 411

sensitivity of the virus to serum neutralizing antibody in individual convalescent 412

COVID-19 patients. Our study will help for understanding the SARS-CoV-2 413

transmission and the design of vaccines and therapeutic interventions against 414

COVID-19. 415

416

Acknowledgments 417



https://doi.org/10.1101/2020.06.20.161323


We would like to thank Prof. Cheguo Cai (Wuhan University, Wuhan, China) for 418

providing the pNL4-3.Luc.R-E- plasmid. We also grateful to Dr. Hongbing Jiang 419

(Washington University in St. Louis, St. Louis, MO, USA) for the suggestions. We 420

gratefully acknowledge the authors, originating and submitting laboratories of the 421

SARS-CoV-2 sequence data from GISAID. This work was supported by the 422

Emergency Project from the Science & Technology Commission of Chongqing 423

(cstc2020jscx-fyzx0053), the Emergency Project for Novel Coronavirus Pneumonia 424

from the Chongqing Medical University (CQMUNCP0302), and a Major National 425

Science & Technology Program grant (2017ZX10202203) from the Science & 426

Technology Commission of China. 427

428

Author contributions 429

A-L.H., N.T., and K.W. conceived the project and supervised the study. J.H., C-L.H., 430

Q-Z.G. and G-J.Z. performed most experiments. J.H. and K.W. performed serum 431

neutralization assay. H-J.D. and L-Y. H. performed SARS-CoV-2 genome analysis. 432

Q-X.L., J.C. and X-X. C. collected the serum samples. J.H. and Q-Z.G. contributed 433

to the statistical analysis. K.W. and N.T. wrote the manuscript. All the authors 434

analyzed the final data, reviewed and approved the final version. 435

436

Competing interests 437

The authors declare no competing interests. 438

439



https://doi.org/10.1101/2020.06.20.161323


Data availability 440

The data supporting the findings of this study are available from the authors upon 441

reasonable request. 442

443

References: 444

1. Zhou, P. et al. A pneumonia outbreak associated with a new coronavirus of probable 445

bat origin. Nature 579, 270–273 (2020). 446

2. Sanders, J. M., Monogue, M. L., Jodlowski, T. Z. & Cutrell, J. B. Pharmacologic 447

Treatments for Coronavirus Disease 2019 (COVID-19): A Review. JAMA (2020) 448

doi:10.1001/jama.2020.6019. 449

3. Becerra-Flores, M. & Cardozo, T. SARS-CoV-2 viral spike G614 mutation exhibits higher 450

case fatality rate. Int J Clin Pract (2020) doi:10.1111/ijcp.13525. 451

4. Hoffmann, M. et al. SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is 452

Blocked by a Clinically Proven Protease Inhibitor. Cell 181, 271-280.e8 (2020). 453

5. Du, L. et al. The spike protein of SARS-CoV — a target for vaccine and therapeutic 454

development. Nat Rev Microbiol 7, 226–236 (2009). 455

6. Wrapp, D. et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion 456

conformation. Science 367, 1260–1263 (2020). 457

7. Walls, A. C. et al. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike 458

Glycoprotein. Cell 181, 281-292.e6 (2020). 459

8. Zhou, Y. et al. Protease inhibitors targeting coronavirus and filovirus entry. Antiviral 460

Res 116, 76–84 (2015). 461



https://doi.org/10.1101/2020.06.20.161323


9. Bhattacharyya, C. et al. Global Spread of SARS-CoV-2 Subtype with Spike Protein 462

Mutation D614G is Shaped by Human Genomic Variations that Regulate Expression of 463

TMPRSS2 and MX1 Genes. bioRxiv 2020.05.04.075911 (2020) 464

doi:10.1101/2020.05.04.075911. 465

10. Connor, R. I., Chen, B. K., Choe, S. & Landau, N. R. Vpr Is Required for Efficient 466

Replication of Human Immunodeficiency Virus Type-1 in Mononuclear Phagocytes. 467

Virology 206, 935–944 (1995). 468

11. Long, Q.-X. et al. Antibody responses to SARS-CoV-2 in patients with COVID-19. Nature 469

Medicine 1–4 (2020) doi:10.1038/s41591-020-0897-1. 470

12. Ou, X. et al. Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and its 471

immune cross-reactivity with SARS-CoV. Nature Communications 11, 1620 (2020). 472

13. Geraerts, M., Willems, S., Baekelandt, V., Debyser, Z. & Gijsbers, R. Comparison of 473

lentiviral vector titration methods. BMC Biotechnol 6, 34 (2006). 474

14. Shang, J. et al. Cell entry mechanisms of SARS-CoV-2. PNAS (2020) 475

doi:10.1073/pnas.2003138117. 476

15. Song, J. et al. PROSPER: An Integrated Feature-Based Tool for Predicting Protease 477

Substrate Cleavage Sites. PLoS ONE 7, e50300 (2012). 478

16. Sandrin, V. et al. Lentiviral vectors pseudotyped with a modified RD114 envelope 479

glycoprotein show increased stability in sera and augmented transduction of primary 480

lymphocytes and CD34+ cells derived from human and nonhuman primates. Blood 481

100, 823–832 (2002). 482

17. Lei, C. et al. Neutralization of SARS-CoV-2 spike pseudotyped virus by recombinant 483



https://doi.org/10.1101/2020.06.20.161323


ACE2-Ig. Nat Commun 11, 2070 (2020). 484

18. Koyama, T., Weeraratne, D., Snowdon, J. L. & Parida, L. Emergence of Drift Variants 485

That May Affect COVID-19 Vaccine Development and Antibody Treatment. Pathogens 486

9, 324 (2020). 487

19. Shimizu, Y. K. et al. Neutralizing antibodies against hepatitis C virus and the emergence 488

of neutralization escape mutant viruses. Journal of Virology 68, 1494–1500 (1994). 489

20. Yao, H. et al. Patient-derived mutations impact pathogenicity of SARS-CoV-2. medRxiv 490

2020.04.14.20060160 (2020) doi:10.1101/2020.04.14.20060160. 491

21. Matsuyama, S., Ujike, M., Morikawa, S., Tashiro, M. & Taguchi, F. Protease-mediated 492

enhancement of severe acute respiratory syndrome coronavirus infection. 493

Proceedings of the National Academy of Sciences 102, 12543–12547 (2005). 494

22. Okayama, N. et al. Clinical effects of a neutrophil elastase inhibitor, sivelestat, in 495

patients with acute respiratory distress syndrome. J Anesth 20, 6–10 (2006). 496

23. Barnes, B. J. et al. Targeting potential drivers of COVID-19: Neutrophil extracellular 497

traps. J Exp Med 217, (2020). 498

24. Mohamed, M. M. A. Neutrophil Elastase Inhibitors: A potential prophylactic treatment 499

option for SARS-CoV-2-induced respiratory complications? 2 (2020). 500

25. Wang, Q. et al. Immunodominant SARS Coronavirus Epitopes in Humans Elicited both 501

Enhancing and Neutralizing Effects on Infection in Non-human Primates. ACS Infect Dis 502

2, 361–376 (2016). 503

26. Kim, S.-J., Nguyen, V.-G., Park, Y.-H., Park, B.-K. & Chung, H.-C. A Novel Synonymous 504

Mutation of SARS-CoV-2: Is This Possible to Affect Their Antigenicity and 505



https://doi.org/10.1101/2020.06.20.161323


Immunogenicity? Vaccines 8, 220 (2020). 506

27. Zhang, L. et al. The D614G mutation in the SARS-CoV-2 spike protein reduces S1 507

shedding and increases infectivity. bioRxiv 2020.06.12.148726 (2020) 508

doi:10.1101/2020.06.12.148726. 509

28. Ozono, S. et al. Naturally mutated spike proteins of SARS-CoV-2 variants show 510

differential levels of cell entry. bioRxiv 2020.06.15.151779 (2020) 511

doi:10.1101/2020.06.15.151779. 512

29. Daniloski, Z., Guo, X. & Sanjana, N. E. The D614G mutation in SARS-CoV-2 Spike 513

increases transduction of multiple human cell types. bioRxiv 2020.06.14.151357 (2020) 514

doi:10.1101/2020.06.14.151357. 515

30. Traggiai, E. et al. An efficient method to make human monoclonal antibodies from 516

memory B cells: potent neutralization of SARS coronavirus. Nature Medicine 10, 517

871–875 (2004). 518

519

520



https://doi.org/10.1101/2020.06.20.161323


Table 1. Mutations in spike protein of SARS-CoV-2. 521

Mutations in Spike Protein

No. of mutation No. of wildtype Total No. of sequences

Mutation (%)

D614G 2995 1637 4632 64.659 A829T 37 4602 4639 0.798

L5F 33 4614 4647 0.710 H146Y 26 4594 4620 0.563 P1263L 15 4631 4646 0.323 V483A 12 4387 4399 0.273 S939F 12 4626 4638 0.259 R78M 10 4623 4633 0.216 E583D 9 4639 4648 0.194 A845S 9 4632 4641 0.194

Top 10 abundant non-synonymous mutations observed in S protein of 522

SARS-CoV-2. 523



https://doi.org/10.1101/2020.06.20.161323


Figures and figure legends 524

525

Fig. 1. Phylogenetic analysis of SARS-CoV-2 sequences from GISAID. 526

Prevalence of S-D614G genomes over time produced by the Nextstrain analysis 527

tool using GISAID dataset (n = 2,834 genomes samples from January 2020 to May 528

2020). 529

530



https://doi.org/10.1101/2020.06.20.161323


531

Fig. 2 Detection of S-D614 and S-G614 protein expression and cleavage. a, An 532

additional serine protease (elastase-2) cleavage site in S1-S2 junction was found 533

in S-614G protein of SARS-CoV-2. b, Detection of S protein expression in 534

HKE293T cells by Western blot using the anti-RBD monoclonal antibody. Cells 535

were transfected with pS-D614, pS-G614 plasmid or with an empty vector, and 536

incubated with sivelestat sodium or not. To compare the S1 and S ratio, integrated 537

density of S1/S was quantitatively analyzed using ImageJ software. Significant 538

differences were analyzed by one-way ANOVA. **P < 0.01. 539

540



https://doi.org/10.1101/2020.06.20.161323


541

Fig. 3 The S-G614 protein pseudotyped virus showed increased infectivity. a, 542

HEK293T and 293T-ACE2 cells were infected with lentiviruses pseudotyped with 543

vesicular stomatitis virus G (VSV-G) and SARS-CoV-2 S protein variants. Virus 544

titers were quantified by RT-qPCR and adjusted to 3.8x104 copies in 50 μL to 545

normalize input virus doses. The relative luminescence units (RLU) detected 72 h 546

post-infection (hpi). b, Inhibition of pseudovirus entry by ACE2-Ig. Pseudovirions 547

were pre-incubated with ACE2-Ig and added to 293T-ACE2 cells, RLU were 548

detected 72 hpi. c, Viral entry efficiency meditated by S variants. The RLU were 549



https://doi.org/10.1101/2020.06.20.161323


detected 24-72 hpi. The data are presented as the means ± standard deviations 550

(SDs) of three independent biological replicates. Significant differences were 551

analyzed by one-way ANOVA. **P < 0.01, ***P < 0.001. 552

553



https://doi.org/10.1101/2020.06.20.161323


554

Fig. 4 Detection of protease inhibitors against SARS-CoV-2 pseudovirus 555

infection. a-b, The TMPRSS2 and CatB/L inhibitors were evaluated by 556

pseudovirus cell entry assay. 293T-ACE2 cells were pre-incubated with camostat 557

mesylate (a) or E-64d (b), and subsequently inoculated with pseudovirions. The 558

VSV-G pseudovirus was used as the control. RLU were detected at 72 h 559



https://doi.org/10.1101/2020.06.20.161323


post-pseudovirus inoculation. Cell viability was examined by the methyl thiazolyl 560

tetrazolium (MTT) assay. The data are presented as the means ± SDs of three 561

independent biological replicates. Statistical significance was tested by two-way 562

ANOVA with Dunnett posttest. **P < 0.01. 563

564



https://doi.org/10.1101/2020.06.20.161323


565

Fig. 5 Detection of neutralizing antibodies in convalescent sera against 566

S-D614 and S-G614 protein pseudotyped viruses. a, The inhibition rate of sera 567



https://doi.org/10.1101/2020.06.20.161323


from 41 convalescent COVID-19 patients against S-D614 and S-G614 568

pseudoviruses. Serum sample from healthy individual was tested as a negative 569

control. Convalescent sera were collected at 2-4 weeks post symptom onset from 570

confirmed case patients (#1 - #41). Sera was analyzed at a single dilution of 1:1000. 571

RLU were detected at 72 hpi. Compared the RLU of serum neutralized sample to 572

control and calculated the inhibition rate. The data are presented as the mean 573

percentages of inhibition ± SDs of three independent biological replicates. b, The 574

inhibition curves for serum samples from 5 convalescent patient and a health donor. 575

The initial dilution was 1:40, followed by 4-fold serial dilution. Neutralization titers 576

were calculated as 50% inhibitory dose (ID50), expressed as the serum dilution at 577

which RLUs were reduced by 50% compared with virus control wells after 578

subtraction of background RLU in cell control wells. Representative data are shown 579

as percent neutralization (mean ± SDs of three independent biological replicates). 580

581



https://doi.org/10.1101/2020.06.20.161323


Documents

and decreases neutralization sensitivity to individual ......2020/06/20 · 15 Ai-Long Huang, Ni Tang, Kai Wang, Key Laboratory of Molecular Biology for 16 Infectious Diseases (Ministry