Thread by @MomsenReincke

Excited to share our new preprint together with @menz45, @vmcorman and many others in which we show that #SARSCoV2 Beta elicits potent lineage-specific and cross-reactive antibodies! 1/n
biorxiv.org/content/10.110…

Why are we interested in Beta? RBD Beta can be considered „antigenically most distant“ from the original strain and it contains mutations reducing neutralization of many #SARSCoV2 antibodies. But how do antibodies elicited by Beta look like? 2/n

Thousands of monoclonal antibodies against the original RBD isolated in the last 2 years provide a reference frame to look at differences and similarities, e.g.: Are potent and abundant VH3-53/VH3-66 antibodies that are normally sensitive to mutations in Beta still elicited? 3/n

We first looked at serum of patients infected with Beta and found - perhaps expectedly - a reduced neutralization of the original strain vs. Beta. This already shows at the serum level that there are many antibodies which are specific to RBD Beta. 4/n

Interestingly, just two days ago @AllieGreaney from @jbloom_lab published a preprint in which they look at polyclonal sera from patients after Beta infection and show that these are more focused on the S309 epitope (class 3 mAbs). 5/n biorxiv.org/content/10.110…

We next isolated monoclonal abs after Beta infection. Comparing our sequences to Cov-AbDab (database of published #SARSCoV2 abs), we could at the sequence level predict that some abs are crossreactive to the original RBD - e.g. VH1-58 mAb CS44 in blue. 6/n

Many others monoclonal antibodies (~45%) were specific to the Beta RBD with binding depending on all of the three mutated amino acids: 7/n

We were particularly interested in VH3-53 mAbs. These mAbs with short CDR3 form an important class of RBD antibodies and are frequently elicited by the original strain, but binding typically depends on the unmutated K417. Would we still find those mAbs after Beta infection? 8/n

The answer is yes! To our surprise, Beta-specific VH3-53 antibody CS23 acommodates mutation K417N in the canonical binding mode just like VH3-53 antibodies previously isolated from non-VOC patients! 9/n

But what about cross-reactive antibodies elicited by Beta? We found that roughly half of the RBD Beta elicited antibodies are cross-reactive, and often also bind and neutralized other variants including Delta! 10/n

One cross-reactive antibody (CS44) was especially interesting, as we earlier found mAbs with similar sequences, e.g. CV07-287 which we isolated from a different patient 1 year ago! X-ray structures show them binding almost identically (and very similar to other VH1-58 mAbs) 11/n

VH1-58 antibodies are also elicited after infection with the original strain and vaccinations – see e.g. science.org/doi/10.1126/sc…. They are very potent and bind to the RBD ridge quite far from all currently observed mutated residues. 12/n

These mAbs contribute ~3% of all antibodies in our dataset and therefore are the most abundant cross-reactive clonotype. They also strongly bind and neutralized Delta. They therefore constitute one very important part of our humoral defense against all ciruclating variants! 13/n

Beta has declined worldwide – but are these observations relevant only to Beta? No, the same RBD Beta mutations appear in sublineages of Delta. And with increasing immunity through natural infection or vaccination, antibody escape might become more relevant in emerging VOCs. 14/n

Thanks to everyone involved, in particular @menz45, @vmcorman, Hans-Christian Kornau, Jakob Kreye, and many others from #ChariteVirology, @ChariteBerlin, @scrippsresearch, @DZNE_de ! 15/n