key: cord-0949596-61rk7sto
authors: Neerukonda, Sabari Nath; Vassell, Russell; Herrup, Rachel; Liu, Shufeng; Wang, Tony; Takeda, Kazuyo; Yang, Ye; Lin, Tsai-Lien; Wang, Wei; Weiss, Carol D.
title: Establishment of a well-characterized SARS-CoV-2 lentiviral pseudovirus neutralization assay using 293T cells with stable expression of ACE2 and TMPRSS2
date: 2020-12-26
journal: bioRxiv
DOI: 10.1101/2020.12.26.424442
sha: 95174520adaf7e097ce5a379d1ab5c0e1c623e92
doc_id: 949596
cord_uid: 61rk7sto

Pseudoviruses are useful surrogates for highly pathogenic viruses because of their safety, genetic stability, and scalability for screening assays. Many different pseudovirus platforms exist, each with different advantages and limitations. Here we report our efforts to optimize and characterize an HIV-based lentiviral pseudovirus assay for screening neutralizing antibodies for SARS-CoV-2 using a stable 293T cell line expressing human angiotensin converting enzyme 2 (ACE2) and transmembrane serine protease 2 (TMPRSS2). We assessed different target cells, established conditions that generate readouts over at least a two-log range, and confirmed consistent neutralization titers over a range of pseudovirus input. Using reference sera and plasma panels, we evaluated assay precision and showed that our neutralization titers correlate well with results reported in other assays. Overall, our lentiviral assay is relatively simple, scalable, and suitable for a variety of SARS-CoV-2 entry and neutralization screening assays.

In December 2019 a cluster of atypical pneumonia cases appeared in Wuhan, China. The 35 etiological agent was later identified as severe acute respiratory syndrome coronavirus 2 (SARS-36 with ACE2 and TMPRSS2 genes for stable expression. Briefly, pseudoviruses with titers of 135 approximately 10 6 RLU/ml of luciferase activity were incubated with antibodies or sera for one 136 hour at 37 o C. Pseudovirus and antibody mixtures (100 µl) were then inoculated onto 96-well 137 plates that were seeded with 3.0 x 10 4 cells/well one day prior to infection. Pseudovirus 138 infectivity was scored 48 h later for luciferase activity. The antibody dilution or mAb 139 concentration causing a 50% and 80% reduction of RLU compared to control (ID50 and ID80 or 140 IC50 and IC80, respectively) were reported as the neutralizing antibody titers. Titers were 141 calculated using a nonlinear regression curve fit (GraphPad Prism software Inc., La Jolla, CA).

The mean 50% and 80% reduction of RLU compared to control from at least two independent We generated SARS-CoV-2 pseudoviruses bearing full-length S glycoprotein from the SARS-214 COV-2 Wuhan-Hu-1 isolate using a second-generation Fig 1A) . The A549, Caco-2, Calu-3, and Huh-7 cells also lacked infectivity above background 233 levels ( Fig 1A) . Transient 293T-TMPRSS2t or 293T-ACE2t cells had 5.7-and 40-fold signal 234 above background, respectively, while transient 293T-ACE2.TMPRSS2t cells resulted in a 3130-235 fold signal above background ( Fig 1A) . A stable 293T-ACE2s cell line displayed 144-fold higher 236 signal compared to background ( Fig 1A) . ACE2.TMPRSS2s cells for our future studies. We also confirmed S protein incorporation into 255 pseudoviruses and proteolytic processing of the S protein to generate S1 and S2 subunits that 256 migrate at 130kDa and 90kDa, respectively ( Fig 1B) . 258 We investigated several conditions for optimizing SARS-CoV-2 pseudovirus production in 293T 259 cells. When comparing S priming by TMPRSS2 during pseudovirus production to priming 260 during entry into target cells, we found that co-expressing TMPRSS2 during pseudovirus 261 production reduced pseudovirus infectivity, possibly due to TMPRSS2-induced premature 262 activation of S that promotes conformational changes to fusion-incompetent or post-fusion structures ( Fig 1C) . This finding is consistent with a previous report suggesting the importance 264 of tight regulation of protease cleavage of the S protein for preserving SARS-CoV-2 infectivity 265 (55). 266 We also investigated variant S proteins to further optimize pseudovirus production. We 

The higher infectivity conferred by the C-terminal cytoplasmic tail truncation of 19 amino acids 283 may be due to higher number of infectious particles (42).

Consistent with previous reports, we found that pseudoviruses bearing G614 and TR19 S 285 proteins displayed 0.5-and 0.2-log higher infectivity, respectively, compared to WT pseudovirus 286 compared to WT pseudoviruses ( Fig 1D) . However, in 293T-ACE2.TMPRSS2s cells the 288 pseudovirus bearing the G614 S and TR19 S were more similar to WT S pseudoviruses, but the 289 pseudovirus bearing the TR14 S displayed 0.5 log lower infectivity compared to WT pseudovirus 290 ( Fig 1D) . Infectivity titers of all pseudoviruses were 1-1.5-log lower on 293T-ACE2s cells 291 compared to 293T-ACE2.TMPRSS2s cells (Fig 1D) . Based on these studies, we used WT S to 292 qualify our pseudovirus neutralization assay because it represents the native, full-length spike, cells, but much less so to chloroquine treatment (Fig 3) . Thus, SARS-CoV-2 predominantly uses 323 TMPRSS2 for priming S during virus entry into the 293T-ACE2.TMPRSS2s cells.

Optimization of SARS-CoV-2 pseudovirus inoculum for 325 neutralization assays 326 We next determined the inoculum range that would assure consistent neutralization titers 293T-ACE2.TMPRSS2S cells (Fig 4) . Although 100% neutralization was achieved at high serum 330 generation of a reliable curve for calculating 50% neutralization titers (Fig 4A and B) . However, 333 inoculums in the 4 x 10 5 -1.4 x 10 7 RLU/ml range generated overlapping curves with little 334 variation (Fig 4C and D) . These dose-response curves yielded 50% neutralization titers with 335 serum inhibitory dilution (ID50) and mAb inhibitory concentration (IC50) values that varied less 336 than 2-fold over all pseudovirus inoculums. Therefore, inoculums of 5 x 10 5 -1 x 10 7 RLU/ml 337 were used for the neutralization assay. increase in the number of ACE2 + /TMPRSS2 + cells (Fig 5A) as well as cell surface expression 347 (Fig 5B and 5C ). Neutralization assays performed with rabbit sera against RBD or S1 subunit, 

Neutralization titers (ID50 and ID80) segregated into high, medium, and low groupings, consistent 368 with reported titers (Fig 6A and 6B) . Plasma for patients hospitalized with acute COVID-19 369 showed a wide range of titers, consistent with previous reports (63). We note, however, that the were imputed to be 1:20. We consider these results to be acceptable for a neutralization assay 389 and adequate for most clinical studies. When titers <1:40 were excluded from the analysis, the 390 %CV was 27.5% and 20.7%, respectively. 392 We used the samples with reported neutralization titers as a benchmark for assessing accuracy of 393 our assay. Although the reported titers were generated using different neutralization assay 394 formats, we nevertheless found a strong correlation between our titers and the neutralization 395 titers reported for the focused concordance samples (Fig 7A and B) and several NIBSC reference 396 standards (Fig 7C and D) . These results provide assurance that our assay provides titers that 397 correlate well with titers measured in other assay formats. 

Evidence 445 favouring the efficacy of convalescent plasma for COVID-19 therapy

Hospitalized COVID-19 patients treated with 448

Convalescent Plasma in a mid-size city in the midwest

The 452 therapeutic effectiveness of Convalescent plasma therapy on treating COVID-19 patients 453 residing in respiratory care units in hospitals in Baghdad

Convalescent Plasma for COVID-19. A randomized clinical trial

A serological assay to detect SARS-CoV-2 seroconversion in humans

Diagnostic SARS-CoV-2

Quantification of SARS-CoV-2-Neutralizing Antibodies Using Propagation-Defective Vesicular 469

Serological assays for 471 severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)

Molecular, serological, 474 and biochemical diagnosis and monitoring of COVID-19: IFCC taskforce evaluation of the latest 475 evidence

SARS-CoV-2 antibody testing-questions to be asked

SARS-CoV-2 neutralizing antibodies and protection from disease in a small animal model

Infectivity on 293T-ACE2s and 293T-ACE2.TMPRSS2s of 600 pseudoviruses primed with or without TMPRSS2 during pseudovirus production. (D) Infectivity 601 on 293T-ACE2s and 293T-ACE2.TMPRSS2s of pseudoviruses bearing full-length, wildtype S 602 glycoprotein (WT), an S glycoprotein with the D614G substitution, or S glycoproteins with C-603 terminal truncations of 14 (TR14) and 19 (TR19) amino acids. Data are shown as means and 604 standard deviations from 3-4 independent experiments (panel A, C and D). The tests for two-605 group comparison were analyzed using GraphPad Prism software

Infection of Vero E6, 293T-ACE2s and 293T-ACE2.TMPRSS2s with replicating 608

Cells inoculated with 100 PFU/ml of virus were fixed and imaged at 609 24 hours post infection by confocal microscopy. The left column shows bright field images 610 (black and white), with the 293-ACE2.TMPRSS2 cells (bottom) showing a high degree of 611 cytopathic effect and syncytium formation, resulting in fewer cells

TMPRSS2 cells (bottom) showing positive signals in red. The right column shows merged 614 images of SARS-CoV-2/mNG (green), TMPRSS2/mCherry (red), and Hoechst dye (blue)

Compared to Vero (top) and 293-ACE2 (middle), the 293T-ACE2.TMPRSS2 (bottom) cells 616 show stronger SARS-CoV-2 nNG signals in green

Figure 3. pH-dependent and -independent pathways of cell entry of SARS-CoV-2 S 618 pseudoviruses. Camostat mesylate inhibits TMPRSS2 activity and chloroquine inhibits 619 endosomal acidification required for cathepsin activity