Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a highly transmissible and pathogenic coronavirus that emerged in late 2019 and has caused a pandemic of acute respiratory disease, named ‘coronavirus disease 2019’ (COVID-19), which threatens human health and public safety. In this Review, we describe the basic virology of SARS-CoV-2, including genomic characteristics and receptor use, highlighting its key difference from previously known coronaviruses. We summarize current knowledge of clinical, epidemiological and pathological features of COVID-19, as well as recent progress in animal models and antiviral treatment approaches for SARS-CoV-2 infection. We also discuss the potential wildlife hosts and zoonotic origin of this emerging virus in detail. In this Review, Shi and colleagues summarize the exceptional amount of research that has characterized acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease 2019 (COVID-19) since this virus has swept around the globe. They discuss what we know so far about the emergence and virology of SARS-CoV-2 and the pathogenesis and treatment of COVID-19.

/pdf/characteristics-of-sars-cov-2-and-covid-19-eeqnaghpho.pdf

Characteristics of SARS-CoV-2 and COVID-19

The COVID-19 pandemic has had widespread effects across the globe, and its causative agent, SARS-CoV-2, continues to spread. Effective interventions need to be developed to end this pandemic. Single and combination therapies with monoclonal antibodies have received emergency use authorization1-3, and more treatments are under development4-7. Furthermore, multiple vaccine constructs have shown promise8, including two that have an approximately 95% protective efficacy against COVID-199,10. However, these interventions were directed against the initial SARS-CoV-2 virus that emerged in 2019. The recent detection of SARS-CoV-2 variants B.1.1.7 in the UK11 and B.1.351 in South Africa12 is of concern because of their purported ease of transmission and extensive mutations in the spike protein. Here we show that B.1.1.7 is refractory to neutralization by most monoclonal antibodies against the N-terminal domain of the spike protein and is relatively resistant to a few monoclonal antibodies against the receptor-binding domain. It is not more resistant to plasma from individuals who have recovered from COVID-19 or sera from individuals who have been vaccinated against SARS-CoV-2. The B.1.351 variant is not only refractory to neutralization by most monoclonal antibodies against the N-terminal domain but also by multiple individual monoclonal antibodies against the receptor-binding motif of the receptor-binding domain, which is mostly due to a mutation causing an E484K substitution. Moreover, compared to wild-type SARS-CoV-2, B.1.351 is markedly more resistant to neutralization by convalescent plasma (9.4-fold) and sera from individuals who have been vaccinated (10.3-12.4-fold). B.1.351 and emergent variants13,14 with similar mutations in the spike protein present new challenges for monoclonal antibody therapies and threaten the protective efficacy of current vaccines.

/pdf/antibody-resistance-of-sars-cov-2-variants-b-1-351-and-b-1-1-3jty1gsdai.pdf

Antibody resistance of SARS-CoV-2 variants B.1.351 and B.1.1.7.

Deep Mutational Scanning of SARS-CoV-2 Receptor Binding Domain Reveals Constraints on Folding and ACE2 Binding.

The Impact of Mutations in SARS-CoV-2 Spike on Viral Infectivity and Antigenicity.

The coronavirus disease 2019 (COVID-19) pandemic presents an urgent health crisis. Human neutralizing antibodies that target the host ACE2 receptor-binding domain (RBD) of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) spike protein1–5 show promise therapeutically and are being evaluated clinically6–8. Here, to identify the structural correlates of SARS-CoV-2 neutralization, we solved eight new structures of distinct COVID-19 human neutralizing antibodies5 in complex with the SARS-CoV-2 spike trimer or RBD. Structural comparisons allowed us to classify the antibodies into categories: (1) neutralizing antibodies encoded by the VH3-53 gene segment with short CDRH3 loops that block ACE2 and bind only to ‘up’ RBDs; (2) ACE2-blocking neutralizing antibodies that bind both up and ‘down’ RBDs and can contact adjacent RBDs; (3) neutralizing antibodies that bind outside the ACE2 site and recognize both up and down RBDs; and (4) previously described antibodies that do not block ACE2 and bind only to up RBDs9. Class 2 contained four neutralizing antibodies with epitopes that bridged RBDs, including a VH3-53 antibody that used a long CDRH3 with a hydrophobic tip to bridge between adjacent down RBDs, thereby locking the spike into a closed conformation. Epitope and paratope mapping revealed few interactions with host-derived N-glycans and minor contributions of antibody somatic hypermutations to epitope contacts. Affinity measurements and mapping of naturally occurring and in vitro-selected spike mutants in 3D provided insight into the potential for SARS-CoV-2 to escape from antibodies elicited during infection or delivered therapeutically. These classifications and structural analyses provide rules for assigning current and future human RBD-targeting antibodies into classes, evaluating avidity effects and suggesting combinations for clinical use, and provide insight into immune responses against SARS-CoV-2. Eight structures of human neutralizing antibodies that target the SARS-CoV-2 spike receptor-binding domain are reported and classified into four categories, suggesting combinations for clinical use.

SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies.

An outbreak of coronavirus disease 2019 (COVID-19)1-3, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)4, has spread globally. Countermeasures are needed to treat and prevent further dissemination of the virus. Here we report the isolation of two specific human monoclonal antibodies (termed CA1 and CB6) from a patient convalescing from COVID-19. CA1 and CB6 demonstrated potent SARS-CoV-2-specific neutralization activity in vitro. In addition, CB6 inhibited infection with SARS-CoV-2 in rhesus monkeys in both prophylactic and treatment settings. We also performed structural studies, which revealed that CB6 recognizes an epitope that overlaps with angiotensin-converting enzyme 2 (ACE2)-binding sites in the SARS-CoV-2 receptor-binding domain, and thereby interferes with virus-receptor interactions by both steric hindrance and direct competition for interface residues. Our results suggest that CB6 deserves further study as a candidate for translation to the clinic.

/pdf/a-human-neutralizing-antibody-targets-the-receptor-binding-1b4zi4uq0h.pdf

A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2.

Dear Editor, By targeting the programmed cell death 1 (PD-1) pathway with monoclonal antibodies (mAbs), immune checkpoint therapy (ICT) has achieved unprecedented clinical success in the treatment of multiple tumors. Cancer cells evade the host immune system via both the tolerance of T cells and functional suppression of innate immune cells. CD47 provides a “do not eat me” signal by binding to signal regulatory protein alpha (SIRPα) to prevent innate immune cells from attacking host cells. Recently, macrophages were found to restore antitumor reactivity by blocking the interaction between upregulated CD47 on tumor cells and SIRPα on innate immune cells. However, it remains unknown whether there are blocking “hotspots” on CD47 for mAb-based anti-CD47 therapy or additional blocking hotspot regions within CD47 for therapeutic mAb development. Although it is overexpressed on tumor cells, CD47 is also expressed in many normal cells, including red blood cells and platelets. Some validated CD47-blocking mAbs under clinical investigation induce hemagglutination and anemia. Thus, designing an engineered CD47-blocking antibody to exert a therapeutic effect with limited hemagglutination is needed. To overcome this obstacle, we identified a CD47-targeting mAb and evaluated its enhanced macrophage-mediated phagocytosis capacity and tumor suppression efficacy. We screened a panel of 50 murine mAbs from mice immunized with human CD47 and identified an efficient CD47/SIRPα-blocking mAb, m4C1 (Fig. S1a). m4C1 was subsequently humanized based on sequence homology and computational prediction (Fig. S1c). Humanized 4C1 (h4C1) blocked the CD47/SIRPα interaction as efficiently as m4C1 (Fig. S1e). Surface plasmon resonance (SPR)-binding assays demonstrated that h4C1 bound to the human CD47-ECD antigen with a KD of 0.85 nM (Fig. S1d), which is comparable to that of m4C1 (Fig. S1b, Table S1). To investigate the macrophagemediated phagocytosis activity of h4C1, a tumor cell phagocytosis assay was conducted with bone marrow-derived macrophages (BMDMs). We found that h4C1 exhibited significant phagocytosis in Raji cells, with an EC50 of 7.30 ng/mL (Fig. 1a). We further assessed the in vivo tumor suppression efficacy of the CD47-targeting antibody, h4C1, in a mouse model. We subcutaneously engrafted luciferase-labeled Raji cells into the backs of immune-compromised NCG mice. Seven days after the engraftment, the mice were administered h4C1, Hu5F9 (a validated CD47-blocking mAb under clinical investigation), or isotype control mAb daily for 2 weeks (Fig. S2a). The average luciferase intensity in the h4C1-treated group was significantly lower than that of the isotype mAb-treated group (Figs. 1b, c and S2b–d). Treatment with h4C1 also resulted in a substantially improved survival of the tumor-bearing mice (Fig. S2e). Taken together, these results suggest that h4C1 represents a potential therapeutic mAb for lymphoma. The binding characteristics of h4C1 to CD47 were subsequently explored through the determination of the h4C1/CD47 complex structure (Fig. S3a, b, Table S2). The overall structure reveals that h4C1 binds to CD47 with a buried surface area of 688 Å. All three CDRs of heavy chain (VH) and LCDR2 provide contacts with CD47ECD, while LCDR1 and LCDR3 are not engaged (Fig. S4). The findings indicate that h4C1 exhibits a VH-dominated interaction with CD47, while the binding of h4C1 is mainly located on three loops of CD47 (FG, BC, and C’C”). To analyze the mechanism of CD47 antagonism by h4C1, the structure of the h4C1-Fab/CD47-ECD complex was superimposed onto the CD47-ECD/SIRPα complex (PDB ID: 2JJS). Detailed comparison of reciprocal binding areas between the two complexes revealed multiple striking parallels. Specifically, the FG loop of CD47-ECD inserts into a groove on the surface of domain I of SIRPα and contributes more than half of the interfacial surface. Similarly, the FG loop of CD47-ECD dominates the interaction with VH of h4C1. The binding of h4C1 to CD47 displayed substantial stereospecific hindrance to the binding of h4C1 to SIPRα (Fig. 1d). There are five potential N-linked glycosylation sites on CD47ECD. In the complex structure of h4C1/CD47, glycosylation modifications were observed at all five of these sites in CD47 (Fig. S5a). Notably, the glycan chain on N32 substantially stretches toward the h4C1/CD47 interface and contacts h4C1 (Table S3 and Fig. S3c). Additionally, glycosylated N55 is also near the h4C1/ CD47 interface. The binding profiles of h4C1 to N32A-mutated or N55A-mutated CD47 proteins were further analyzed with SPR to investigate the glycosylation dependency of the h4C1/CD47 interaction. SDS–PAGE analysis revealed that N32A-mutated or N55A-mutated CD47 proteins with a reduction in molecular weight indicates the existence of the N-glycosylation modification at these sites (Fig. S3d). The SPR assays demonstrated that no substantial differences were observed in the binding affinity of WT CD47-ECD (Fig. S1d) and N32A-mutated or N55A-mutated CD47ECD (Fig. S5b, c). Therefore, the binding of h4C1 to CD47 is not affected by glycosylation modifications of CD47. Several reported complex structures of CD47-blocking agents enabled us to compare these binding characteristics and to elucidate sites for preferential blocking. Overall, these mAbs mainly bind to the FG loop of CD47, and the conformation of the FG loop displayed no substantial differences, indicating the structural conservation of the loop (Fig. 1e). In particular, the FG loop of CD47 dominates the binding to SIPRα and to all of the mAbs. The binding of Hu5F9 to the FG loop of CD47 is mainly mediated by HCDR1, HCDR3, LCDR1, and LCDR2. The complex

/pdf/the-identification-of-a-cd47-blocking-hotspot-and-design-of-3giwu61y8m.pdf

The identification of a CD47-blocking “hotspot” and design of a CD47/PD-L1 dual-specific antibody with limited hemagglutination

Dear Editor, Monoclonal antibody (mAb)-based tumor immune checkpoint therapy (ICT) has gained particular interest in recent years. The molecular basis of binding between mAbs and PD-1 or PD-L1 has been reported, providing clear information of the binding “hotspots” for mAbs. Tumor suppression efficacy of PD-L1 specific mAbs relies on not only the blocking of PD-1/PD-L1 interaction to restore T cell reactivity, but also Fc-mediated tumor cell cytotoxity. PD-L1 antibody drug conjugate (ADC) for selective chemo-guided immune modulation of tumor has also been developed which has shown promising tumor suppression potency. MAbs that could bind to antigen in a pH-dependent manner would improve recycling of the antibodies and engineered IL-6R mAbs with pH-dependent binding properties have displayed increased lysosomal delivery and therapeutic potency. However, no PD-L1 specific mAb with pH-dependent binding property has been reported, and whether the binding to a specific region on PD-L1 would induce pH-dependent interaction remains unknown. Here we report the binding properties of a PD-L1 specific antibody JS003 with tumor suppression potency. JS003 is a humanized PD-L1 specific mAb which could block the binding of PD-L1 to PD-1 or CD80 and showed a binding affinity (KD) of 2.88 × 10 M in surface plasmon resonance (SPR) analysis (Fig. 1a, b; supplementary Fig. S1a). The ability of JS003 to promote T cell reactivity in vitro was investigated with mixed leukocyte reactions (MLR) assay. The results revealed that JS003 has substantially enhanced the allogeneic T cell response as measured by IL-2 and IFN-γ secretion (Fig. 1c and supplementary Fig. 1b). The in vivo tumor suppression efficacy of JS003 was examined in human PDL1 knock-in mice of the C57BL/6 background (C57/hPD-L1) with syngeneic MC38-hPD-L1 tumor cells. The results showed that inhibition of tumor growth was observed in a dose dependent manner with significant anti-tumor efficacy in 3 mg/kg and 10mg/ kg JS003 treatment groups compared with PBS group at the end of the observation period (Day 27) (p < 0.001), while the low dose group (1 mg/kg) showed no significant change in tumor size compared to negative control PBS group (p= 0.07) (Fig. 1d and supplementary Fig. 2). We next examined the ability of the PD-L1-specific JS003 to induce PD-L1 internalization using a pH-sensitive cyanine dye derivative CypHer5E in hPD-L1-expressing CHO-K1 cells. The results showed that JS003 has induced PD-L1 internalization with an EC50 of 0.42 nM (Fig. 1e). The binding kinetics of JS003 under different pH conditions were further investigated by using SPR. JS003 antibodies were captured on a biosensor and the association with PD-L1 proteins was monitored in a neutral buffer of pH 7.4, while the dissociation was followed at pH 7.4, 6.0 or 5.5. JS003 exhibited a dissociation rate 3.4-fold and 8.7-fold faster at pH 6.0 and pH 5.5 than that at pH 7.4, respectively (Fig. 1f and supplementary Table 1). These results indicate that JS003 has high affinity to PD-L1 at the neutral pH but may rapidly dissociates with PD-L1 at lower pH in the endosome. Internalized JS003 could be degraded in the acidic endosome. The rapid dissociation in lower pH would enable antigen dissociation in endosome and reduce the possibility of degradation in lysosome, and facilitate recycling of the antibody for reuse. To elucidate the binding mechanisms of JS003 to PD-L1, the complex structure of JS003-Fab/PD-L1 was determined at a resolution of 2.0 Å (Fig. 1g, supplementary Table 2). The structure analysis revealed that JS003 utilizes both of its heavy (H) and light (L) chain to interact with PD-L1. Of note, the binding of JS003 H chain is mainly attributed to HCDR2 that residues from HCDR2 (Y48, Y51, T52, S54, T55, and Y56) and frame region (K62) of JS003 H chain have formed hydrogen bond interactions with amino acids from C and C′ strands (Y56, E58, Q66, H69, and E71) and F strand (R113) of PD-L1 (Fig. 1g and supplementary Table 3). Additionally, Y92 and Y94 from LCDR3 have formed hydrogen bond interactions with I2, S117 and A121 of PD-L1. Superimposition of the structure of JS003/PD-L1 complex with PD-1/PDL1 complex (PDB code: 4ZQK) revealed that the binding of JS003 to PD-L1 shows a stereo clash with that of PD-1 and the binding surface of JS003 and PD-1 is substantially overlapped (supplementary Fig. 3). The pH-dependent binding properties of JS003 is of particular interest. Previous studies revealed that pH-dependent binding of the mAbs is usually mediated by the presence of histidine in the CDRs that interacts with the antigen. However, there is no histidine in the CDRs of JS003 that contacts PD-L1. Therefore, we further analyzed the residues in PD-L1 that interact with JS003. The analysis revealed that the only histidine involved in the interaction between JS003 and PD-L1 is H69, which has formed multiple van der Waals interactions with amino acids from HCDR2 (Y56, S57, N58, L59 and K62) (Fig. 1h and supplementary Table 3). Moreover, H69 also formed a hydrogen bond interaction with Y56 of HCDR2 (Fig. 1h). Therefore, we speculate that the interaction with H69 in PD-L1 is responsible for the pH-dependent binding properties of JS003. To further validate this hypothesis, we analyzed whether H69 is also involved in the interaction with other mAbs with complex structures available. The binding orientations of these mAbs and their binding areas on PD-L1 varied (supplementary Fig. 4). The complex structure of BMS-936559/PD-L1 showed that H69 of PDL1 has formed not only multiple van der Waals interactions with amino acids from HCDR3 (S106) and LCDR3 (S92, N93, and W94), but also hydrogen bond interaction with S106 of HCDR3 (Fig. 1i). On the other hand, no substantial interactions with H69 of PD-L1 were observed for other PD-L1 specific mAbs, i.e., avelumab, durvalumab, and atezolizumab (supplementary Fig. 5). Therefore, we speculate that the binding of BMS-936559 to PD-L1 may also

/pdf/identification-of-a-hotspot-on-pd-l1-for-ph-dependent-3vb1jecodk.pdf

Xiaoshan Bi

Papers

A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2.

The identification of a CD47-blocking “hotspot” and design of a CD47/PD-L1 dual-specific antibody with limited hemagglutination

Identification of a hotspot on PD-L1 for pH-dependent binding by monoclonal antibodies for tumor therapy