15 Oct 2021 Episode 182 : B cell memory and maturation during and after COVID and after vaccination

Dear colleagues,

Today’s episode dives into the immunology behind Spike-specific antibodies.  In previous episodes, the emphasis was usually on neutralizing antibodies, measured either with live virus, pseudoviral constructs or in the ACE-2 binding inhibition ELISA. As discussed in Episode 179, these assays have a different sensitivity, but correlate well with each other and with ELISA of antibodies binding to either S- and RBD (receptor-binding domain).  Evidence is accumulating that  neutralization tests reasonable correlate with protection (further confirmation in the last paper of this Episode.

There is, however, another more basic aspect to B cell immunology: the phenomenon of B cell “maturation”, a antigen-driven and genetically determined program that occurs in the germinal centers of lymphoid organs and is described in Ep 182-1.  B cell maturation is the consequence of interaction between antigen-presenting dendritic cells, the follicular T helper or Tfh (a very special CD4+ T subset) and B cells.  Maturation entails somatic mutations, which gradually increase the affinity (or avidity) of the B cell receptor (and the secreted antibodies) for the antigen.   Obviously, the higher the avidity, the better the in vivo neutralization. Importantly, a higher avidity could compensate to some extent for decreasing antibody concentrations.  

So, the first question is: will S-specific antibodies acquire enough avidity after either natural infection or vaccination to provide durable protection?  

Two other questions are equally important:

• Is there any broadening of the antibody responses with time i.e. will viral variants of concern (VOC) such as alpha-beta-gamma-delta be neutralized better with time?
• To what extent will not only short-lived effector B cells (producing antibodies for a limited time), but also B cell memory be induced to ensure a “boost” of high quality antibody production upon a first or new encounter with the virus?      

Clearly, we are witnessing how a new virus is spreading in the human population and how our immune system is dealing with it.  As always, the outcome depends on the genetic flexibility of the virus and of the individual and common “clonotypes” of human T and B cells.  

1. The basics in review

Ep 182-1: A very nice review by Baumjohann in Eur J Immunol explaining the role of follicular T helper cells (Tfh) in B cell responses and memory.  Fig 1 gives a didactic overview of the normal process and Fig 2 illustrates some key findings in SARS-CoV-2 infection and vaccination:

• Strong activation in non-severe COVID cases with Th1 polarization
• Decreased germinal centers and increased cytotoxic Tfh (potentially killing instead of activating B cells) in severe cases may lead to low affinity antibodies.
• Mixed IFNg/IL-4 response to mRNA vaccination enhances the efficacy of vaccination
• Persistence of memory Tfh and B cells for several months after mild/moderate disease and vaccination may prevent (re)infection.

Ep 182-2: The hypothesis of “original antigenic sin” explained by Eric Brown in mSphere, with the consequence that “cross-reactivity” to other human coronaviruses may either be beneficial or deleterious by facilitating or blocking neutralizing responses.  (see also attachment for didactic overview).

2. Characteristics of antibody responses during COVID disease

Ep 182-3 A:   Juanjie Tang (USA) in Nat Comm (Feb) starts with the recognition that predictive value of neutralization titers with disease outcome has not been established, since severe COVID-19 patients show much faster  and stronger neutralizing antibodies compared with mild cases. 

Therefore, they followed 3 groups of hospitalized COVID patients who (1) survived outside ICU versus ICU patients who either (2) survived or (3) deceased.

First they confirmed that more severe cases displayed higher pro-inflammatory cyto/chemo-kines, while neutralizing Ab titers varied in all 3 groups and failed to predict progression. In fact increase in Ab binding and neutralization was faster in more severe (group 2 + 3) patients.   

Moreover:

• In the severe deceased patients there were some holes in the IgG and IgA repertoire against Spike epitopes.  Remarkably, IgA to spike binding antibodies was higher and longer-lasting in most of the severe patients (2+3) compared with the non-ICU patients (1)
• The most striking finding was that affinity maturation of prefusion spike-binding antibodies clearly increased in survivors and not in non-survivors.

This finding refers to the observation of decreased germinal centers in severe cases (see Ep 182-1)

Ep 182-3 B: Moura (Brazil) in Sci Reports (Sept) conforms that:

• Patients who presented severe COVID-19 had higher anti-RBD IgG levels compared to patients with mild disease.
• Most patients displayed low antibody avidity, with 64.4% of the samples of patients who recovered from the disease and 84.6% of those who died in this low avidity range.

Conclusion:

Low affinity, rather than titer of spike antibodies, is associated with poor outcome in severe COVID

3. Long-term evolution of antibody responses after COVID (convalescence)

Ep 182-4: Gaebler (Michel Nussenzweigs group USA) in Nature (March) is the  first to show that despite waning antibodies titers and neutralizing activity in the plasma between 3 and 6 months after infection, memory B cells still display clonal turnover after 6.2 months, and the antibodies that they express have greater somatic hypermutation, resistance to RBD mutations and increased potency. 

Very remarkably also, they show in intestinal biopsies of 7/14 recovered patients persistence of SARS-CoV-2 RNA after 4 months.

Clearly, these data suggest that low level persistent SARS-CoV-2 drives evolution and maturation of B cell responses  at the clonal level, including broadening of recognition of for instance the “immune escape” E484K mutation !!

Ep 182-5 A: Moriyama (Japan) in Immunity (Aug) essentially confirms and extend these observations towards polyclonal responses from over 350 COVID patients in 3 clinical groups (mild, moderate and severe).  Although the overall neutralizing titer of convalescent sera decreased over up to 8 months, the “quality” of the antibodies increases in that affinity as well as breath improves:

At early times after infection, plasma from these patients infected with the early strain of the virus failed  to neutralize the beta and gamma variants of concern, but after several months they were neutralized.

Remarkably, this was the case for both mild and severe convalescent patients.  (Obviously, these are all “survivors” in the terms of Ep 182-3).

Ep 182-5 B: In the same issue of Immunity, these findings are independently confirmed by Frauke Muecksch from USA.  This paper is an extension of Ep 182-4, mainly focusing more in depth on  monoclonal responses.

Ep 182-5 C: Mrunal Sakkarkar from Texas in Sci Imm (June) provide similar data on increased somatic hypermutation, binding affinity, and neutralization potency over time, of memory B cells at the clonal level. Further interesting conclusions:

• Cross-reactive B cell populations, likely re-called from prior endemic beta-coronavirus exposures, comprised only a small but stable fraction of the repertoires and did not contribute to the neutralizing response.
• The neutralizing antibody response was dominated by public clonotypes that displayed significantly reduced activity against gamma and beta variants (This is in contrast with Ep 182-5 A and Ep 182-7)

Ep 182-6: Wan Ni Chia (Singapore) in June edition of Lancet Microbe provide a slightly different perspective.  They followed neutralizing capacity (measured as ACE-2 binding inhibition by a recombinant Wuhan Spike) in 266 convalescent patients over a similar time frame as Ep 182-5 A.  Instead of disease severity, these authors subdivided according to dynamics of neutralizing responses in 5 groups: negative, rapid waning, slow waning, persistent and delayed response (Fig 1 e243 p. 5).

Persistent response is associated with higher avidity, but remarkably, also with various clinical and lab parameters of more severe disease, including higher age, tendency for more co-morbidities, more pneumonia, oxygen and ICU sustained pro-inflammatory  milieu.   (but of course, we look at “survivors” here)

SARS-CoV-2 specific T cells were detected regardless of waning patterns of neutralizing antibodies.     

They rightly conclude that it remains to be seen whether this persistent pattern will be associated with less reinfection and wonder how this post-disease pattern relates to post-vaccination immunity…..

Ep 182-7:  Most (if not all) patients in the previous studies were infected with the “wave 1” (= Wuhan strain with minor variations).  Liane Dupont in medRxiv (June 2021) compares longevity and cross-neutralization after infection with wave 1 and wave 2 alpha (B1.117) viruses in UK.

• Neutralizing antibody titers decline from the initial peak response, but robust neutralizing activity can still be detected in a large proportion of convalescent sera at up to 10 months.
• The largest decrease in neutralization potency for both wave 1 (overall average 4.8-fold) and B.1.1.7 sera (overall average 5.7-264 fold) was observed against B.1.351 (beta VOC)
• Nevertheless, cross-neutralization against several VOC and WT increases over time, both after wave 1 and alpha infections. (Fig 2 C p. 21 and Fig 3 D-E p. 23).
• Severe disease results in higher neut titers against WT, alpha and gamma (but not beta) 10 months after wave 1 infection (Fig 2 E p. 21).  

Conclusion:

• Neutralization titers tend to decline in most patients (but there are exceptions or persistence).
• All these studies point to further maturation of antibody responses during convalescence, with increase in affinity and breath: improved recognition of variants.

The question is how this maturation is induced: does it require continuous presence of the antigen?  Only Nussenzweigs group addressed this question and found evidence of persistent viral RNA in half of the patients.   

4. Characteristics of antibody responses after vaccination

1. 1. Naïve vaccinees

Ep 182-8: Pratesi et al (Italy) in Vaccines (June) show that the Pfizer (BNT 162b2) vaccine in HCW induces high levels of  antibodies of high avidity in vaccinated subjects, which can inhibit the binding of Spike to ACE-2. Mainly IgG, but also IgA and IgM are induced.  No longitudinal data.

Ep 182-9 (= Ep 186-13): Gonzales Lopez Ledesma from Argentina in medRxiv (Aug) shows in 88 Sputnik vaccinees that Spike-specific antibodies wane over 4-6 months (Fig 1 p. 5), but neutralization activity is rather stable in absolute terms (Fig 1 D) and actually relatively improves against variants of concern (Fig 2 p. 7). 

• 42 days after vaccination, neutralization against beta, gamma and delta VOC is resp. 19, 14 and 5 times lower than against Wuhan
• 120 days post vaccination the reduction is only resp. 10, 4 and 3.4 times.

1. 2. Naïve vs previously infected vaccinees

Ep 182-10: Sokal (France in Immunity (Dec 2021) compares mRNA vaccination in naïve vs convalescent subjects:

• In previous COVID-19 patients mRNA vaccine boosts high-affinity RBD memory B cells

→ retain their diversity and express potent variant-neutralizing antibodies

• Naïve subjects early after vaccination: low neutralization against VOC

→  however: maturation of memory B cells allows good neut against VOC

1. 3. Compare vaccinated with convalescent

Ep 182-11: Franziska Neumann (Germany) in Vaccine (June) compare antibody responses shortly after vaccination with either mRNA or Astra-Zeneca with those in recently (about 1 month) convalescent subjects (10/11 infected with alpha VOC). 

• Two doses of either Pfizer or Moderna result in a higher level of high avidity IgG and more potent neutralization against both wild-type and alpha VOC.   
• Astra-Zeneca was only studied after one dose: it has a lower neutralizing capacity than either convalescent serum or 1 dose of mRNA vaccine

Ep 182-12 A: Friedhelm Struck ( Freiburg Germany in J Med Virol (Sept):

• During COVID disease there is a rise in ant-RBD Ab titers and avidity, but it levels off.
• Both antibody titers and avidity rise much higher after 2 doses of mRNA vaccination in naïve subjects as compared to the evolution in COVID convalescents.  (Fig 4 p. 8).  

Ep 182-12 B: Georg Bauer (same group as in A): illustrates in more detail how avidity should be measured.  They also show that the serological responses towards seasonal coronaviruses neither have a negative nor positive impact on SARS‐CoV‐2 serology in general.

Ep 182-13: Ivanova (USA in medRxiv (Aug) observes that SARS-CoV-2 mRNA vaccine elicits a potent adaptive immune response in the absence of IFN-mediated inflammation, which is  observed in COVID-19.  This COVID-inflammation results in upregulation of cytotoxic genes in various lymphocytes and effector phenotype in B and T cells, whereas vaccination induces rather memory T and B cells  

Conclusion:

mRNA (but also Adenovirus) vaccination results in more pronounced B cell memory and  maturation than after COVID disease, also with broadening of the response towards VOC.

5. Bonus: breakthrough infections correlated with lower neut Ab

Ep 182-14: Bergwerk (Israel) in NEJM (14 Oct) shows that breakthrough infection after 2 doses of  Pfizer vaccination  in HCW are associated with:

• Lower peri-infection anti-S and Neut titers
• Higher neut titers correlate with lower VL (higher Ct)
• Most breakthrough mild, but 1/5 persistent symptoms for > 6 weeks
• Most breakthrough remained negative in the rapid viral antigen test and no secondary infections observed

GENERAL CONCLUSIONS:

Although neutralizing titers tend to wane with time during and after infection or vaccination, a memory B cell response of high quality is being build, displaying increasing avidity and breath towards variant viruses.  

Obviously, there is a genetically determined range, with some people showing poor responses and others better, which may be one of the determining factors in disease outcome and chances on (re)infection after vaccination.

It is evident that vaccination will provide a B cell response of better quality than infection and that advanced age and/or immune compromising conditions will decrease the quality of this response.

Provisionally, there is no clear evidence that cross-reactivity with endemic human coronaviruses has a major modulating effect in the B cell response to SARS-CoV-2