key: cord-0273211-oelpnf4u authors: Emig, Christopher J.; Mena, Marco A.; Henry, Steven J.; Vitug, Adela; Ventura, Christian John; Fox, Douglas; Nguyenla, Xammy Huu; Xu, Haiyue; Moon, Chaeho; Sahakijjpijarn, Sawittree; Kuehl, Philip J.; Revelli, David; Cui, Zengrong; Williams, Robert O.; Christensen, Dale J. title: AUG-3387, a Human-Derived Monoclonal Antibody Neutralizes SARS-CoV-2 Variants and Reduces Viral Load from Therapeutic Treatment of Hamsters In Vivo date: 2021-10-13 journal: bioRxiv DOI: 10.1101/2021.10.12.464150 sha: d40af0cd086d7974775dcf1ce79324804a543c65 doc_id: 273211 cord_uid: oelpnf4u Infections from the SARS-CoV-2 virus have killed over 4.6 million people since it began spreading through human populations in late 2019. In order to develop a therapeutic or prophylactic antibody to help mitigate the effects of the pandemic, a human monoclonal antibody (mAb) that binds to the SARS-CoV-2 spike protein was isolated from a convalescent patient following recovery from COVID-19 disease. This mAb, designated AUG-3387, demonstrates a high affinity for the spike protein of the original viral strains and all variants tested to date. In vitro pseudovirus neutralization and SARS-CoV-2 neutralization activity has been demonstrated in vitro. In addition, a dry powder formulation has been prepared using a Thin-Film Freezing (TFF) process that exhibited a fine particle fraction (FPF) of 50.95 ± 7.69% and a mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD) of 3.74 ± 0.73 µm and 2.73 ± 0.20, respectively. The dry powder is suitable for delivery directly to the lungs of infected patients using a dry powder inhaler device. Importantly, AUG-3387, administered as a liquid by intraperitoneal injection or the dry powder formulation delivered intratracheally into Syrian hamsters 24 hours after intranasal SARS-CoV-2 infection, demonstrated a dose-dependent reduction in the lung viral load of the virus. These data suggest that AUG-3387 formulated as a dry powder demonstrates potential to treat COVID-19. Since SARS-CoV-2 is primarily a pulmonary disease, early treatment or prophylaxis with neutralizing 66 mAb therapy targeted to the airways could improve disease outcomes. Initiating treatment early in 67 the disease cycle may alter the course of the disease, prevent the development of chronic 68 complications, reduce the hospitalization rate and decrease mortality. Only a fraction of a 69 systemically administered mAb is transported into the pulmonary compartment where viral 70 particles are released early in the disease; therefore, delivery of neutralizing mAbs directly to the 71 lung holds the potential to reduce the dose needed to achieve the same efficacy as systemically 72 delivered mAbs. Aerosolized delivery would have additional benefits including the potential for 73 patients to self-administer antibody therapy at home rather than in infusion centers, and the ability 74 to expand the supply of antibody to treat a larger population through dose-reduction. Furthermore, 80 a dry antibody formulation stable at ambient temperatures could be made available to patients in 81 geographic regions that lack suitable infrastructure for cold chain distribution of injectible antibody 82 formulations that require cold chain distribution. 83 Thin-film freezing (TFF) is a particle engineering technology that has been used to prepare dry 84 powder formulations of drugs that are administered to patients using a dry powder inhaler (DPI) 85 device. Powders produced by this process are currently in human clinical testing for the treatment 86 of pulmonary indications (NCT04872231 and NCT04576325) (8, 9) and have characteristic highly 87 porous brittle matrices with low bulk densities that can be delivered with good aerosol performance 88 (10). The TFF technology generates these powders by fast supercooling of drug-carrier solutions (11) 89 followed by lyophilization to remove water or other solvents and has been used to generate 90 powders of small molecule (12, 13) and biologic drugs (14). 91 We now report the isolation of a fully human mAb, AUG-3387, using the SingleCyte® system, which 92 binds to the SARS-CoV-2 spike protein with high affinity and binds to all SARS-CoV-2 variant spike 93 proteins tested to date including Delta, Lamda and Mu. In vitro neutralization was confirmed using 94 pseudovirus neutralization with the wild-type SARS-CoV-2 and the Delta variant (B. For this study, we profiled patients with low disease burden, rather than severely affected 114 individuals. We hypothesized that asymptomatic or weakly symptomatic patients might have had 115 previous exposure to a related antigen (e.g., other coronaviruses), providing a breadth of antigenic 116 coverage and driving their resistance to COVID-19. Informed consent was obtained from all patients, 117 and all patient samples were collected after a full recovery from illness and under IRB approval. 118 Patient samples were profiled to identify samples with binding to multiple SARS-CoV-2 proteins 119 including the S1-receptor binding domain (RBD), full length S1, S2, E and M protein. following antigens were conjugated: B.1.1.17 (Alpha) S1, B.1.1.28 (Gamma) S1+S2, 20H/501Y.V2 158 (Beta) S1, B.1.617 (Kappa) RBD, B.1.617.2 (Delta) RBD, S1+S2 S494P, S1+S2 V483A, S1+S2 159 R683A+R685A+F817P+A892P+A899P+A942P+ K986P+V987P, S1+S2 G485S, S1+S2 D614G, S1+S2 160 E484K, S1+S2 D614G+V445I+H655Y+E583D, S1+S2 L452R+T478K. Each antigen was conjugated with 161 the xMAP conjugation kit at ratio of 5µg protein to 1 million beads. Assays were performed in 162 multiplex, with each spectrally encoded bead having a separate antigen and run together in a single 163 well. Antibody was titrated over therapeutically relevant concentrations, mixed with the beads, 164 washed twice, labelled with a secondary antibody, washed twice and run on the instrument. Dry 165 powder versions of antibodies were resuspended in water before dilution for assay. 166 The ScFv version of AUG-3387, designated AUG-3705, was run on a Carterra LSA instrument at 167 multiple concentrations for determination of the single domain affinity against Wuhan SARS-CoV-2 168 S1 and RBD. AUG-3705 was attached to the LSA flow cell via interaction with its V5 tag and a surface 169 bound anti-V5 antibody. Wuhan-1 RBD was delivered to the flow cell at concentrations of 2.06nM, 170 6.17nM, 18.5nM and 56nM for calculation of Kd. 171 An ACE2 expressing HEK293T cell line ("LentiX ACE2.S4") was constructed by packaging pCMV-AC-174 GFP (Origene) into lentivirus and transducing HEK293T's (ATCC). The cells were enriched 4 times 175 until 97% of the cells showed signal above the negative control as read out by staining with anti-176 ACE-2 and secondary antibodies. On average enriched ACE2-HEK293T's had 50-fold higher signal 177 compared to the signal of non-transduced cells. 178 Two days prior to infection, LentiX ACE2.S4 cells were grown to 85% confluency, then seeded in a 180 96-well plate at 15k cells/well in 50µL media per well and held at 37 ℃ in 5% CO2 until infection. 181 Antibody mixes were created prior to infection by performing a 128-fold serial dilution starting at 182 40µg/µL. Dry powders prepared by the TFF process of AUG-3387 and soluble negative control V5 183 Tag monoclonal antibody were seeded in triplicate, and soluble AUG-3387 was seeded in duplicate. 184 SARS-CoV-2 pseudovirus (Genscript) was diluted in DMEM complete media to an IFU of 3.2e7/mL, 185 and 100µL of virus solution was mixed with 100 µL of diluted antibody. The virus/antibody mix was 186 incubated for 60 minutes at 37 ℃ in 5% CO2. Following incubation, 50 µL of each 187 pseudovirus/antibody condition mix was added to each well of seeded cells. After 24 hours, 50 µL of the supernatant was removed for TCID50 assays. Vero E6 cells were seeded 217 at 10,000 cells in 100 µL per well. Infected cell culture supernatant was diluted with 950 µL D10 218 media, and then serial diluted before 50 µL of each dilution was added to 8 wells of Vero E6 cells. 219 After 72 hours, wells with complete cytopathic effect were counted. After 96 hours, 100 µL 220 CellTiterGlo reagent was added to each well of the infected Calu-3 cells to assay for live cells. 221 Following incubation of CTG reagent for 20 minutes, luminescence was measured with a Spectramax 222 1L with 1s integration time. 223 In the preparation of the solutions for TFF manufacturing, AUG-3387 was combined with a 225 mannitol/leucine or trehalose/leucine. The solution was applied as drops onto a rotating 226 cryogenically cooled drum cooled to -70 °C. The frozen solids were collected and stored in a -80 °C 227 freezer before lyophilization. The lyophilization was performed in an SP VirTis Advantage Pro shelf 13 lyophilizer (SP Industries, Inc., Warminster, PA, USA). The primary drying process was at −40 °C for 229 20 h, and then, the temperature was linearly increased to 25 °C over 20 h, followed by secondary 230 drying at 25 °C for 20 h. The pressure was maintained at less than 100 mTorr during the lyophilization 231 process. The AUG-3387 mAb was devliered by one of two routes for each animal, IT and IP injection. The IP 288 injection was performed with a 16 mg/mL formulation in saline. Intratracheal insufflation was 289 performed with animals under anesthesia (4-5% isoflurane with oxygen) until a deep plane of 290 anesthesia was reached. Dry powder for inhalation delivery was transferred to the ABSL-3 facility 291 and each individual device quantitiatively loaded for delivery. Doses were based on method 292 development to quantify the amount of material that exited the devices assuming 100% 293 presentation at the terminus of the canuale and the animals average body weight during dosing. involved 5 minutes at 50°C for reverse transcription, followed by an initial denaturation step for 20 317 seconds at 95°C and 40 cycles of 95°C for 3 seconds and 60°C for 30 seconds. 318 were expressed and assayed (Figure 2b) . We recovered many S1, S2, and RBD binders and ultimately 337 chose AUG-3387 as our lead compound due to its breadth of binding activity, affinity to the Wuhan-338 1 strain, and strength in viral neutralization. The timeline for isolation process is shown in 339 Supplemental Figure 2 . 340 We utilized the Carterra LSA platform to determine the single domain affinity of AUG-3387 343 expressed as an ScFv, designated AUG-3705. Auto-fitting of curves was performed in Carterra 344 Kinetics software, which returned a calculated affinity of 1.2 nM (Fig 3) . To assess the susceptibility of AUG-3387 to mutational escape, we profiled AUG-3387 against the 362 S1 and RBD portions of the original Wuhan-1 strain of SARS-CoV-2, RBD's corresponding to WHO 363 designated dominant strains of concern, and S1 mutants known to affect the potency of currently 364 approved therapeutic antibodies. AUG-3387 binds every S1 and RBD of SARS-CoV-2 tested with a 365 binding EC50<200ng/ml (Figure 4 ), but only very weakly to SARS-CoV-1 (binding EC50 > 100ug/ml, 366 not shown). These data demonstrate the high potential for resistance to mutational escape of AUG-367 3387. 368 Wuhan-1 strain of SARS-CoV-2, RBD's corresponding to WHO designated dominant strains of concern, and S1 mutants known to affect the potency 371 of currently approved therapeutic antibodies. AUG-3387 binds every S1 and RBD of SARS-CoV-2 tested with a binding EC50 <200ng/ml, but only 372 very weakly to SARS-CoV-1 (binding EC50 > 100ug/ml, not shown). 373 20000 30000 S1 Wuhan-1 S1 V483A S1 G485S S1 N501Y 0 10000 20000 30000 S1 R683A, R685A, F817P, A892P, A899P, A942P, K986P, V987P S1 S494P BSA We compared the ability of full length IgG1 formatted AUG-3387 and its ScFv formatted version, 375 AUG-3705, to neutralize live SARS-CoV-2 in a 24 hour TCID50 assay and a 96 hour infected cell 376 viability assay ( Figure 5 ). AUG-3705 demonstrated somewhat higher efficacy in these assays over 377 AUG-3387, indicating the improved avidity of the dimeric IgG1 did not improve neutralization 378 enough to compensate for the higher molarity of AUG-3705 at the same concentration. indicate an IC50 of ~2ug/ml or less. 384 We assessed the ability of AUG-3387 to neutralize the SARS-0CoV-2 Delta variant pseudotyped virus 386 ( Figure 6 ). AUG-3387 demonstrated the ability to neutralize Delta pseudovirus, although at a higher 387 IC50 (30-40 mg/mL) than for the Wuhan-1 strain (approximately 2 mg/mL in TCID50 assay). 388 However, the 30-40 mg/mL is still a clinically relevant dose that can be achieved by delivery to the 389 lung. A series of dry powder formulations of AUG-3387 were prepared by Thin-Film Freezing and 397 evaluated for retention of biological activity and optimal aerosol properties for delivery to the lung. 398 The powders contained AUG-3387 at a range of mAb concentrations from 5-20% (w/w) and various 399 excipients. The powders were tested for the presences or absence of subvisible aggregates under a 400 light microscope, for mAb aggregation or fragmentation using SDS-PAGE, and for their aerosol 401 properties using the NGI. We assessed two formulations of AUG-3387 in TFF powders, AUG-3387.11 402 prepared with mannitol/leucine (95%/5%) and AUG-3387.13 prepared with trehalose/leucine 403 (95%/5%), and compared them to the original AUG-3387 formulation in PBS. AUG-3387.11 and 404 AUG-3387.13 performed virtually identically to their soluble counterpart in gel electrophoresis 405 ( Figure 7A ), multiplexed bead assays ( Figure 7B) . Lung tissue was homogenized and active viral load was characterized by RT-qPCR for the E-gene. 452 Since the start of the COVID-19 disease pandemic caused by the SARS-CoV-2, a high burden has 454 been placed on the healthcare system to provide adequate care and treatment of the high number 455 of patients infected by the virus that require hospitalization. Patient care has improved as the 456 pandemic has progressed and the mortality rate has dropped as the care has improved. One key 457 area of improvement in patient outcomes occurred when anti-SARS-CoV-2 monoclonal antibody 458 therapies became available. While not highly efficacious for the treatment of severe COVID-19 459 disease, the current mAb therapies demonstrate up to a 79% reduction in hospitalization in patients 460 that are symptomatic and at high risk to develop severe disease that will require hospitalization. 461 These risk factors include age greater than 65 years, obesity or being overweight, pregnancy, 462 diabetes, chronic kidney disease, immunocompromised patients due to disease or 463 immunosuppressive treatment, cardiovascular disease, chronic lung diseases or sickle cell disease. 464 suggesting that AUG-3387 binds a conserved epitope that has notmutated in the variants of concern 483 or newly emerging Lamda and Mu variants. Furthermore, neutralization and pseudoneutralization 484 data demonstrate that AUG-3387 prevents the virus from infecting cells by blocking interaction of 485 this conserved region of the RBD with the hACE2 receptor of target cells. The retained activity 486 against all tested variants is in contrast to the reduced susceptibility of Bamlanivimab and 487 Etesevimab, which show greater than 250-fold reduced binding and neutralization activity against 488 the Beta and Gamma variants (16). 489 In order to differentiate AUG-3387 from the current mAb therapeutics that are currently being used 490 under EUA, the mAb has been formulated as a room temperature stable dry powder utilizing the 491 thin-film freezing process. The room temperature stability may allow for distribution to geographic 492 locations where SARS-CoV-2 continues to spread but that do not have the capability of distributing 493 injectible formulations that require cold chain distribution and storage. We demonstrated that the 494 formulations prepared using the TFF process retain full binding activity of the input mAb solutions 495 and have no evidence of protein aggregation or instability following reconstitution in water. 496 In addition to the dry powder storage at room temperature of the TFF formulated dry powder mAb, 497 the dry powders can be encapsulated and delivered to the lung using a standard dry powder inhaler 498 device. When tested with the Plastiape RS00 high resistance device, which is designed to provide 499 maximum shear and aerosolization of powders at lower airflow rates, the AUG-3387 powder 500 formulations had a fine particle fraction with greater than 50% of the powder in the 1-5 µm range, 501 which is ideal for delivery to the deep lung of humans using a device matched to the potential for 502 reduced lung function for mild to moderate COVID-19 patients. 503 Finally, we demonstrated that administration of AUG-3387 by either intraperitoneal injection or by 504 intratracheal insufflation of the dry powder into SARS-CoV-2 infected Syrian hamsters resulted in a 505 dose dependent reduction of the viral load in the lung tissue of the infected hamsters. The result 506 of this in vivo study is notable because we utilized a treatment paradigm that creates a high burden 507 for efficacy to be demonstrated. In our study, mAb treatment of the hamsters was not initiated until 508 24 hours after the hamsters were infected with SARS-CoV-2 by intranasal inoculation. By contrast, 509 sotrovimab administered by IP injection prophylactically at doses of 5 mg/kg or more when given 510 24-or 48-hours prior to viral infection resulted in improvement in body weight loss and decreased 511 viral load in the lung tissue compared to control animals. Likewise, the casirivimab and imdevimab 512 combination of mAbs administered to hamsters by IP injection 24 hours before viral inoculation (17) 513 resulted in a dose dependent viral load reduction in lung tissue. However, no change in viral load in 514 the lung tissue was reported when casirivimab and imdevimab were administered 24 hours after 515 viral inoculation in a manner similar to our study. For sotrovimab there was no report of therapeutic 516 treatment resulting in reduced viral load. Thus, to date, the demonstration that AUG-3387 517 administered by either IP or IT routes in a therapeutic mode resulted in a dose dependent viral load 518 reduction in the lung tissue represents the first report of a mAb therapy that works in the hamster 519 model in a therapeutic mode. Furthermore, the viral load reduction of the dry powder when 520 delivered by IT insufflation represents the first report of successful reduction of viral load using 521 inhaled delivery of a mAb therapeutic for COVID-19 disease. 522 Taken together, these data suggest that AUG-3387 is a potent mAb that has the potential to treat 523 all known variants of SARS-CoV-2 and that the powders produced by the TFF formulation process to 524 make room temperature stable powders has the potential to reduce the amount of mAb needed for 525 efficacy because of the local delivery to the lung. Furthermore, since the TFF AUG-3387 powder 526 does not require cold chain storage, it represents an opportunity to distribute the powder 527 formulation globally to reduce the human cost of the COVID-19 pandemic by facilitating delivery of 528 this therapy to locations lacking cold chain distribution capabilities. 529 530 531 trimeric SARS-CoV1/2 spike proteins and performing preliminary pseudoneutralization assays, Sarah 532 553 Coronavirus interactions with the cellular autophagy machinery A new coronavirus associated with human respiratory disease in China Angiotensin receptor blockers as tentative SARS-CoV-2 therapeutics. Drug 559 development research Novel 561 serological biomarkers for inflammation in predicting disease severity in patients with 562 COVID-19 High-throughput isolation of immunoglobulin genes from 565 single human B cells and expression as monoclonal antibodies Efficient 568 generation of monoclonal antibodies from single human B cells by single cell RT-PCR and 569 expression vector cloning Rapid cloning of high-affinity human monoclonal antibodies against 572 influenza virus Single 574 Ascending Dose and Multiple Ascending Dose Study of Voriconazole Inhalation Powder in 575 Healthy Adult Subjects Dry powder 580 insufflation of crystalline and amorphous voriconazole formulations produced by thin film 581 freezing to mice Processing design space is critical for 583 voriconazole nanoaggregates for dry powder inhalation produced by thin film freezing Enhanced Aerosolization of High 586 Potency Nanoaggregates of Voriconazole by Dry Powder Inhalation Using 589 thin film freezing to minimize excipients in inhalable tacrolimus dry powder formulations The Development of Thin-592 film Freezing and Its Application to Improve Delivery of Biologics as Dry Powder Aerosols A simple method of estimating fifty per cent endpoints Fact sheet for health care providers emergency use authorization (EUA) of bamlanivimab 597 and etesevimab REGN-COV2 antibodies prevent and treat SARS-CoV-2 infection in 601 rhesus macaques and hamsters Supplemental Figure 1. Serological profile of 7 patients against 6 SARS-CoV-2 antigens generated by diluting plasma 1:1000 in a 6 plex Luminex assay Samples from each patient were evaluated for binding to antigen proteins including the Receptor binding domain (RBD), Spike Protein 1 (S1) Supplemental Figure 2. Timeline for discovery of Augmenta's first SARS-CoV-2 antibodies The authors would like to thank the laboratory of Peter S. Kim for sharing