16 Feb Episode 110 Integrated view on Ab, CD8 T cells and myeloid cells in COVD disease and vaccination

Tue, 02/16/2021 - 15:20

Dear colleagues,

As I realize that some of the papers that I recently discussed and will discuss in the next episodes, are getting more specialized, I will try and present in each episode some more general as well as more specialized papers.

  1. A comment by “good old” Jon Cohen in Science (Ep 110-1) about the tendency to mix various vaccines.  Two examples are given:  (1) the Ad26 prime of Sputnik V, followed by the Ch Ad Ox1 of Astra-Zeneca; (2) Prime by A-Z and boost with mRNA of Pfizer.  Obviously for both schedules, one rationale is that a prime-boost with AZ alone seems sub-optimal, most probably due to a  “sensitization” effect: the A-Z Adenovirus elicits antibodies against the vector that decrease the effect of the boost, which also explains why A-Z is more effective if the dose of the prime is halved and/or the boost is given after 3 months instead of 1 month after the prime.  So, there is a clear rationale to set up a well-controlled trial and then decide of this is a good strategy.  The problem, however, is that due to the shortage of vaccines and the policy of some countries (e.g. Denmark) to give the first dose, without knowing if the second dose will arrive in time, it is highly likely that all kinds of “wild uncontrolled” experiments will be done with mixing, postponing etc…. If such policies result in suboptimal immunity and protection, that is very good news for the partly resistant viral variants that are ready to concur Europe….    


  1. De paper from the group of Dan Barouch (Ep 110-2) on correlates of protection against SARS-CoV-2 in the macaque model is interesting: they show that:
  • Convalescent purified IgG at 250 mg/kg complete protects against challenge in naïve animals (Fig 1)
  • The same convalescent IgG brings viral load down by  max 2 logs if given 1 day after infection. (Fig 3)
  • The importance of CD8 T cells was studied in a re-challenge study: convalescent animals were depleted of CD8+ cells or sham treated at about 7 weeks after the first infection and re-challenged 3 days later.  Fig 4d shows that all except 1 sham animal did not show viral load after re-challenge, while all CD8-depleted animals became positive, but less so than the naïve-challenged animals.

Clearly, these “simple” experiments are based on “infection-induced protective immunity”, but may also be relevant for vaccine-induced immunity.    

You can follow a very interesting seminar by Dan Barouch on this and other aspects of vaccine development at https://videocast.nih.gov/watch=41236


  1. A cross-sectional study by Dan et al (Ep 110-3) on immunological memory over 8 months after symptomatic infection shows:
  • A slight decline in spike-specific antibodies, including neutralizing Ab
  • A rather stable memory B cell pool, with even some evidence of increase
  • Slowly declining SARS-CoV-3 specific CD4 and CD8 T cells

These kinetics are rather similar to the immune  memory towards Yellow Fever vaccine, which is known to be strong and durable.  There is, however, quite some heterogeneity and it is a cross-sectional study (not a cohort)

The authors conclude that the immunity may not be absolute and that reinfection is not excluded, but that the memory B and T cells will “kick in” in a matter of a few days and therefore prevent severe disease (which typically takes more than one week to develop).


  1.  A comment in Nature Briefings (Ep 110-4a) and an associated paper by Donal Skelly (Ep110-4b) discuss the very debated issue of how good the present vaccines (in this case the Pfizer vaccine) will protect against the new variant, especially the South-African and Brazilian.  In this paper the neutralizing antibody responses as well as T cell responses of either convalescent or vaccinated individuals are compared
  • Neutralizing activity of convalescent serum and after a single dose of the vaccine was much weaker against the South-African variant, but the neutralization capacity after two doses was much better (Fig 2)
  • Vaccination against SARS-CoV-2 S also induced strong binding antibodies against the other beta-coronaviruses (the common cold ones, as well as SARS-CoV-1 and MERS)
  • The T cell responses (measured in ELISPOT) of vaccinated subjects were equally strong against S peptides of wild-type and variant viruses.

Clearly, these data provide a level of confidence that Pfizer vaccination would protect against the new variants, potentially by collaboration between the neutralizing Ab and T cell responses.  There are, of course, limitations to this in vitro approach and we will have to wait for epidemiological data from South-Africa


  1. As an illustration of the relative weakness of “natural immunity” after infection, Cohen discuss 15 cases of well-documented reinfection (Ep 110-5).  As shown in Table 2, the second episode was usually after 2 months or more and in 7 cases it was worse (with one death).  This was probably before the variants were circulating. As far as I see, only one of these patients was clearly immunologically “at risk” receiving B cell depleting therapy for Waldenstrom macroglobulinemia.  So, clearly, convalescent people are not “immune”, should continue the preventive measures, certainly now that variants are circulating and should be vaccinated…


  1. Jordan provides an overview of innate and adaptive immunity in SARS-CoV-2 infection (Ep 110-6).  It is a nice narrative overview.  The only new aspect for me is the possible implication of complement in the pathogenesis with an emphasis on the strongly inflammatory C5a fragment and a potential role for C1 inhibitor.  Both elements could be targets for immunotherapy. To be followed up….


  1. The next review on “cartography of macrophage activation syndromes”  (MAS Ep 110-7) is very interesting, but you need a thorough knowledge about auto-immunity, inflammation and immunotherapy to really understand every detail.  The bottom line is that the immune pathogenesis is clearly different from other MAS.  Those are usually systemic syndromes, based on severe genetic deficiencies (e.g. perforino-pathies) or overall immune activation (such as CAR T cell therapy, hemoraghic or septic shock).   In most adult severe COVID cases the pathogenesis really starts locally in the lung, with a relative deficiency of type 1 interferon, which is partly due to the virus (suppressing the IFN system) and partly host-related (either old age, comorbidities or genetic defects in IFN pathways).  As a consequence of this early incapacity to contain the virus, the local inflammatory system is over-activated, with implication of epithelial cells, endothelial cells, immunocytes (pro-inflammatory macrophages and neutrophils) and activation of coagulation (with microthrombi).  Only in a second phase, the inflammation; coagulation etc becomes more widespread, with also potential implication of secondary infection sites (such as the heart).

The authors show that inflammatory markers such as IL-6 are of course elevated in severe  COVID, but much less so than in the above more systemic syndromes.  They believe that for this reason the beneficial effects of the various monoclonal antibodies against IL-6, Il-1 etc may be less effective in COVID than in severe auto-immune or CAR T cell diseases, but that corticosteroids could have at least some effect.  Clearly, this review was written before the result of various trials were  know, but those trials have indeed confirmed the prediction….


  1. The next very nice and comprehensive paper by Vanderbeke et al (Ep 110-8) elaborates on the local interaction between the various components of the immune system in the lung, showing that:
  • COVID-19 is characterized by an ‘atypical’ cytokine release with reduced type II interferon signaling, refuting the canonical cytokine storm paradigm,
  • Antigen presentation is impaired in critical disease
  • Neutrophils are important effectors of the resulting local (lung) and systemic tissue damage. 
  • So, it is NOT a T cell, but myeloid-driven atypical cytokine storm with mainly activated monocytes-macrophages and neutrophils that dominate.
  • There is a reduction in type I, II and III interferon, which reduces antigen-presentation to T cells, which are dysfunctional.
  • Importantly, CD8+ T cells were more affected, leading to an increased CD4+/CD8+ T-cell ratio especially in ‘critical’ disease. This distinguishes COVID-19 from other viral respiratory infections and bacterial sepsis, where lymphocytopenia is evident but CD4+/CD8+ ratios are decreased28
  • There is also vascular remodeling with elevated vascular endothelial growth factor (VEGF) , immunothrombosis and microcoagulopathy.   
  • Lung neutrophils have a highly activated phenotype and show upregulation of NET (neutrophil extracellular trap) formation related genes.


  1.  The final paper for today is by Sara Falck-Jones (Ep 110-9) is on the role of myeloid suppressor cells: these “immature” monocytic and neutrophilic cells were found to be increased in the blood according to disease severity.  They were indeed able to suppress type II IFN production by T cells and reduce the CD3ζ chain via an Arginase-1 dependent mechanism.  There was also a correlation with male sex and age. Unfortunately, the relations were not so clear in the lung

In summary, today we have explored a more complete picture of immune-pathogenesis, with a prominent role of myeloid cells, exerting both pro-inflammatory, pro-thrombotic, but also T c ell suppressive activities in severe COVID disease. 

The  good news is that vaccine-induced immunity may be stronger than “infection-induced immunity” to confront viral variants and that both neutralizing antibodies as well as T cells could play a role therein.


I will be off for a week or so….

Best wishes,