7 jan 2021 Episode 98 Antibodies and resistance

Dear colleagues,

During the last weeks a lot of information emerged on the significance and use of antibodies, both after infection (in convalescent plasma -CVP) or in therapy. In this episode, we will have a look at the evidence of therapeutic effect of either CVP or monoclonal antibodies, the evolution of neutralizing Ab after recovery and their effect on reinfection.  We also pay a lot of attention on the presence of potential escape mutations in naturally occurring SARS-CoV-2 viruses and novel strategies to overcome it.

1. Convalescent plasma (CVP):
   1. A recent review, summarizing some of the evidence of therapeutic effect (until Oct).  Both RCT and non-controlled trials provide arguments that CVP has a beneficial effect on mortality and some aspects of the pulmonary disease.  As could be expected, the evidence is stronger, if CVP is started earlier, if neut Ab levels are higher or in patients with underlying B cell defects. There are some arguments that not only anti-viral, but also anti-inflammation effects could play. No evidence for antibody-dependent enhancement or serum sickness-like side effects. But obviously no magic bullet…. (Ep 98-1; slide 2)

• 2. The evolution of neutralizing Ab titers after infection in Hong Kong (Eric Lau Nat Comm 2021 12:63 Ep 98-2; slide 3)
• Almost all individuals from day 15 to day 209 days after onset of illness had detectable PRNT90 (90 % neut titer) and PRNT50 antibody.  Titers peak at around 30–40 days after onset of illness followed by a bi-exponential waning of antibody titers. 99.1 % of sera collected ≥61 days after onset of illness remained antibody positive in both PRNT90 and PRNT50 assays.
• Patients with severe illness had higher peak titers of PRNT90 and PRNT50 antibody than did those with mild illness or those with asymptomatic infection.

It is likely that neutralizing antibody will be maintained over the first year after mild or severe disease with higher antibody titers.

1. 3. Despite the decline of serum antibodies, Turner et al, show that after 7-8 months, there are still a lot of bone-marrow progenitor cells (BMPC) that can be triggered to produce SARS-CoV-2 specific IgG and IgA. To a very similar extent as tetanus or even influenza.  They correlate with the serum Ab titers. (Ep 98-3 slides 4-5).

1. 4. The next paper in NEJM shows that the presence of anti-SARS-CoV-2 antibodies clearly protects against re-infection in HCW: with a relative incidence ratio of 0.11 over 6 months. (Ep 98-4 - slide 6).

2. Monoclonal antibodies:
   1. Monoclonal antibody LY-CoV555 (Eli Lilly) , when coadministered with remdesivir, did not demonstrate efficacy among hospitalized patients who had Covid-19 without end-organ failure (Ep 98.5 slide 7).

• 2. REGN-COV2 antibody cocktail reduced viral load, with a greater effect in patients       -      whose immune response had not yet been initiated (no Ab in serum)
• or who had a high viral load at baseline

These were symptomatic but non-hospitalized patients.  It is not very clear from the paper if the treatment alleviated symptoms or shortened the time to recovery. 

(Ep 98.6 slide 8)

1. 3. Prospective mapping of escape mutations for these Abs (Starr bioRxiv Ep 98.7):  Libraries of RBD mutants were created, expressed on the surface of yeast, and using fluorescence-activated cell sorting and deep sequencing to quantify how each mutation affects RBD folding, ACE2 affinity, and antibody binding.  

The map of all those mutations is shown in slide 10 (A + B).  Clearly, there are quite a number of mutations in RBD that can give rise to escape for individual Ab, but very few (just one E406W) to escape the Regeneron cocktail.  The validity of those mutations was confirmed in a virus neutralization test (slide 10 C). 

Some of those mutations could indeed be induced in vitro by dose-escalation studies and were found in a patient who escaped from the therapy (slide 11).

Many of these individual escape mutations were found at low frequency in the GISAID database with almost 200,000 sequences from naturally circulating viruses.

Conclusion: While the presence of escape mutations in the patient treated with REGN-COV2 is ominous, other viruses that typically cause self-limiting acute infections undergo extensive within-patient evolution only in long infections of immune-compromised patients and not in the broader population.

However, it is concerning that so many escape mutations impose little cost on RBD folding or receptor affinity, and that some of these mutations are already present at low levels among circulating viruses.

Ultimately, it will be necessary to wait and see what mutations spread as SARS-CoV-2 circulates in the human population.

1. 4.  One of the future solutions might be the creation of tetravalent (in stead of bivalent) monoclonal antibodies, which show increased potency of neutralization and increased resistance to mutations. An example is given in Miersch bioRxiv Ep 98.8 and slide 12.  Selected potent neut mAb 15033 and 15033-7 were produced as regular bivalent IgG (with a neut EC50 of 3000 and 500 pM resp.), were produced as tetravalent Fab-IgG or IgG-Fab.  Their EC50 decreased strongly (see Fig 5 A), indicating increased potency.  Most potential escape mutants were sensitive to the tetravalent mAb, with the exception of Phe486Ala mutant. Clearly a combination (cocktail) of different tetravalent neut mAb would be preferable.

1. 5. Function of non-neutralizing but cross-reacting SARS-CoV antibodies (Shiakolas bioRxiv Dec 2020 Ep 98.9): starting from a B cell repertoire from a SARS-CoV-1 convalescent patient, a number of monoclonal Ab were generated. 

- They cross-react strongly between SARS-CoV-1 and SARS-CoV-2 and weakly with other beta-Coronaviruses (OC43 and HKU-1).  (Slide 13)

- Epitope mapping shows that 2 mAb bind to RBD (46472-12) and to N-terminal Domain (46472-6) of S1 and 4 others (46472-1,2,3,4- bind to S2 domain(slide 14)

- They are not neutralizing, but active in other Fc mediated functions, such as antibody-dependent cell phagocytosis and trogocytosis, involved in antigen presentation.

- In a mouse model, the RBD-specific 46472-12 and the S2 specific 46472-4 did not influence survival, but mitigated some pathological features (lung hemorrhage score)  Slide 15.

These data thus suggest that betaCoV cross-reactive and non-neutralizing Ab could have some disease modifying effects…  

3. Resistance mutations
   1. A nice overview of circulating mutants by Lauring (Ep 98-10) and CDC Ep 98.11

• The D614G mutation is circulating since March and there is good evidence of increased transmission.
• The Y453F mutation arose in minks and is transmissible to humans.  Associated mutations are del69_70, I692V, and M1229I.  Apparently decreased sensitivity to neut by convalescent plasma.
• Lineage B.1.1.7 (also called 501Y.V1) in South-East England with associated deletion 69_70, and P681H in the furin cleavage site.
• The South-African N501Y variant with K417N and E484K, which may further increase the binding to human cells and promote transmission.

1. 2. Two papers on induction of resistance to neutralizing plasma or monoclonal antibodies:
      1. Andreoni (Ep 98-12) uses a high tittered CVP and shows that after passaging SARS-CoV-2 in vitro with a serial dilutions, first a deletion of phenylalanine in position 140 of the N-Terminal Domain (NTD) of S1 (del F140) occurs, later emerges the mutation of glutamine acid to lysin in position 484 (E484K See South-African variant)) and finally an insertion in loop 5 of NTD (slide 16).  They show that this mutant has a similar “fitness” than the WT. The activity of various CVP and monoclonal Ab (mAb) against WT, D614G and the resistant mutant is shown in slide 17. While some other CVP also lose activity against D614G and even more against the mutant, the pattern of mAb is most interesting. As could be expected, some RBD-spec mAb lose activity, but those that target epitopes in S1 outside RBD and those that target the trimeric conformation-dependent epitopes completely lose activity !!  

1. 1. 2. Weisblum (Ep 98-13) performs a more extensive exercise with 4 CVP and 3 mAb, selected from these patients (slide 18).

• • • As can be seen in slide 19, the mAbs C121 and C144 induced overlapping resistance mutations in the RBD 484 and 493, while C135 targets a different are in 346 and 440 positions of S1.
    • The CVP, however, induced mutations in other positions, even in the case of the corresponding plasma COV47 and mAb C144 from the same patient, indicating that the mAb is not dominant in the polyclonal plasma.
    • Importantly, these mutations occur naturally in circulating SARS-CoV-2 viruses, including of course the E484K, which is part of the South-African variant!
    • As can be seen  in slide 20, the monoclonals lose activity against several naturally occurring mutations, either in the “heart” of RBD (mAb C121 and 144) or more at periphery (C135). However, some of are in a different part of the S1 sequence.
    • The frequency of the mutations that confer resistance to either CVP or mAb in naturally occurring SARS-CoV-2 viruses is low (slide 21)
•  
  • • Interestingly, emergence of resistance could be blocked by combination of two mAb with clearly distinct target i.e. either C121 or C144 combined with C135

• 3. The paper by Voss et al finds protective, and convergent IgG recognition of SARSCoV-2 spike epitopes, outside the RBD in COVID-19 convalescent plasma (Ep 98   14). 
• Indeed they show that the plasma response contains relatively few clones (is oligoclonal) and close to germline, meaning that antibodies have to undergo only a limited number of mutations to acquire high affinity for SARS-CoV-2 spike.
• Remarkably also > 80 % of targeted epitopes lie outside the RBD, rather targeting S1 NTD (N-terminal domain) and S2 (slide 22).
• Moreover, although the anti-RBD and one of the anti-NTD monoclonal antibodies were equally potently neutralizing in vitro, the anti-NTD antibody was sufficient for protection to lethal viral challenge (slide 23).

Some conclusions and perspectives:

1. Convalescent plasma and naturally occurring Ab

• Has some therapeutic effect, if high tittered and applied early.
• Neut antibodies in convalescent plasma slowly decline, but, even after many months there are bone-marrow thar can produce specific IgG and IgA.

• Convalescent Ab protect against reinfection in HCW
• Neut Ab are oligoclonal and not far from germline → much more quickly induced than in HIV f.i.
• Target = RBD, but also NTD of S1 and even S2 → it might be better to use whole S than RBD as vaccine.   

2. Resistance:  complex pattern: there is a focus on heart of RBD (480-510), but also outside

• Naturally circulating SARS-CoV-2 already contain escape mutations at low frequency

3. Monoclonal Ab therapy: first trials not very convincing.  Much room for improvement.

→ Best combination of neut mAb with different targets (1 mAb easy escape)

→ New concepts such as (combination of) tetravalent Ab and non-neutralizing Ab in therapy.

→ Has to be applied very early on.