30 Nov Episode 90: Role of antibodies, type1/3 interferon in children’s response to SARS-CoV-2

Mon, 11/30/2020 - 13:50

Episode 90: Role of antibodies, type1/3 interferon in children’s response to SARS-CoV-2

In episode  89 my provisional conclusion was that children may be equally susceptible and infectious,  then why are children  so much less symptomatic than adults, while that is not the case in many other respiratory infection (e.g. respiratory syncytium virus or RSV)? Obviously, I did not find a final answer to this intriguing question, but let us explore the recent literature together.

  1. Let’s start with two reviews on the topic
  • In SARS-CoV-2 infections in children and young people (CYP) Felsenstein (Clin Imm  Aug 2020 Ep 90-1) proposes the following hypotheses:  

Role of antibodies against SARS-CoV-2 and common CoV: Antibodies directed against seasonal CoV can provide some protection against novel CoV, including SARS-CoV-2. Waning antibody titres in adults (particularly the elderly) and/or recall antibody production can contribute to damage and inflammation.

ACE-2 expression: higher in CYP and lower in elderly: remarkable inverse correlation of ACE-2 expression levels with disease severity. Reason: even after infection CYP have more ACE-2 “available” for conversion of the pro-inflammatory angiotensin II into anti-inflammatory angiotensin 1, 3-7. (See ref Ep 90-1b and slide 2)

Recent vaccination and immune senescence: Various vaccinations (BCG, other childhood vaccines have been suggested to contribute to non-specific (innate) or cross-reactive immunity that are protective against SARS-CoV-2. On the contrary, the immune memory of elderly is dysfunctional and of low affinity; hence cross-reactivity will rather contribute to the hyper-inflammatory pathology in older COVID-patients.  


  • The second review “Why is SARS-CoV-2 infection milder amongst children?” (Ep 90-2) doesn’t provide additional hypotheses, but points to some additional questions. In fact, it is accepted that elements of the innate response, such as type 1 IFN and NK cells, as well as CD8 cytolytic T cells are weaker in young children, which at first view wouldn’t help to protect them against COVID-19?

In my further analysis, I will mainly focus on antibodies and interferon, but I will go beyond the question on children only, as you will see. I will end by mentioning some papers, who propose variations on the theme of “trained immunity”, as an explanation for observed differences in pathology across ages and geographical areas.   

  1. Let’s have a look at the antibodies first:
  • A letter from 2015 by Gao et al (Ep 90-3 and slide 3) shows indeed that antibodies against common CoV show an age-dependent pattern in China: low in young children; highest in 15-45 aged adults and low in elderly again.  If cross-protection has a role, this pattern could explain why “young” adults suffer less from COVID than elderly, but not really why CYP are generally asymptomatic…
  • A very recent paper in the JID by Henss et al from Germay (Ep 90-4 and slide 4) confirms that older adult COVID patients with severe disease have the highest levels of neut Ab against SARS-Cov-2 and that those decrease over time, but, remarkably that severe disease is also associated with the lowest neut Ab against the common CoV NL63 (an alpha CoV)!!  Thus indirectly suggesting that neut Ab against NL63 are associated with protection against severe disease.  
  • A few days later came the study by Ng from UK in Science (Ep 90-5 and slide 5):
    • Pre-existing SARS-CoV-2 IgG in some SARS-CoV-2 naïve subjects (in 2019), with a clearly higher frequency in children
    • These naïve Ab recognize the S2 subunit, while the Ab induced by infection with SARS-CoV-2 recognize both S1 (with RBD) and S2 (fusion domain).
    • SARS-CoV-2 IgG from both infected and naïve were able to neutralize the virus in vitro and did not show a tendency to enhancement.
    • These naïve SARS-CoV-2 reactive Ab most probably induced by beta CoV (OC43 and HKU1), since the S2 more homologous to SARS-CoV-2

While the latter paper is very appealing, it does not prove that these cross-reacting Ab also protect in vivo, since the persons who were harboring these Ab were not infected withg SARS-CoV-2.

It is also remarkable that the age distribution of cross-reacting Ab is apparently  not consistent over China, German and UK.   

  • An hypothetical extension of these data on the effect of presumed respiratory infections with common CoV on COVID-19 is provided by Ravid Philip Rajkumar (Ep 90-5), who proposes that previous intestinal infections with common CoV could set the stage for relative protection against COVID severity upon infection with SARS-CoV-2. 
    • One of his arguments is that children, who more frequently experience diarrheal episodes than adults, indeed have milder disease
    • In addition, also the geographical distribution of more serious COVID in countries with higher hygienic standards is suggested as a relative argument, although the obvious confounder here is the population pyramid, with much more elderly in the “global North”.
    • Interestingly, the author does not suggest that intestinal SARS-CoV-2 infection itself could mitigate the course of COVID-19. Quite on the contrary, based on published evidence, he thinks  that GI symptoms are associated with more severe disease elsewhere in the body.  


  1. The role of type I and type III IFN: From episode 85, we know that type I IFN induction by CoV in general and even more by SARS-CoV-2 is delayed after infection and insufficient, which fuels viral replication, but also inflammation (see Ep 90-6 and slides 6-7 Margerida Sa Ribeiro). 

My question was: could either type I or type III IFN response in children be more efficient than in adults?  The short answer is: I could not find any evidence of that, certainly not in young children, but there is no specific data.   I just mention a few elements for thought.

  • A paper on RSV infection (which is more serious in young children), by Marr (Ep 90-7 and slide 8) shows that upon RSV infection type I IFN is mainly produced by plasmacytoid dendritic cells (PDC), via the cytoplasmic retinoid acid-inducible gene I protein (RIG-I) and is deficient in neonates and young children (> 5 yrs),  while the inflammatory IL-6 production via monocytes is equally elevated across age.
  • The review by Park on IFN response to SARS-CoV-2 (Ep 90-8 and slides 9-10) broadens the picture in several ways:
    • SARS-CoV-2 induces both type I and type III IFN via several pathways, including cytosolic RIG-I/MDA5, but also endosomal TLR 3/7/8 and even membrane TLR4.
    • The distribution of receptors as well as the effects of type I and III are different
      • IFNAR: ubiquitous; IFN-lambda receptor more restricted to epithelial cells (resp, GI, urogenital) and myeloid cells.
      • IFN III results in more sustained expression of Interferon stimulated genes (ISG) and less pro-inflammatory cytokines
  • This concept of beneficial immunomodulation by type III IFN in viral infection is further explained in a review by Zanoni of 2017 (Ep 90-9 and slide 11)  and worked out for respiratory and GI tract by Broggi in 2020 (Ep 90-10 and slide 12).

Therefore, the proposition is that while type I is beneficial only when administered prophylactically or early in infection, type III could still be used in later stages (slide 13).

  • Two papers Felgenstein (JBC Ep 90-12 slide 14) and Vanderheiden (J Virol Ep 90-13 and slides 15-17) show indeed that type I and III IFN can suppress SARS-CoV-2 infection in cell liners, but also in primary human airway epithelial cells, which upon SARS-CoV-2 infection do not produce either type of IFN.  
  • More good news came from Stanifer  (Cell Reports Ep 90-14 and slide 18-20), showing that SARS-CoV-2 infection could not only be modeled in colon carcinoma cell lines (AQco-2 and T-84), but also in ex vivo organoids.  Both type 1 and 3 IFN (pre)treatment could strongly suppress viral growth, but knock-out experiments in T-84 cells strongly suggest that type 3 IFN is really crucial in endogenous restriction of SARS-CoV-2 viral proliferation
  • Unfortunately, the Science paper by Major dampens our optimism (Ep 90-15 and slide 21).  They show that type I as well as type III IFN contribute to disrupt lung eptithelial cell repair in a mouse model after influenza A infection, both at the level of the alveolar type II (AT2) cells (which, remember, are the primary targets of SARS-CoV-2) and in mouse or human primary airway epithelial cells (AEC).  The authors found clear indication that IFN-lambda was more deleterious than ether alpha or beta.   This is likely due to increased IFN-λ production during infection combined with greater induction of antiproliferative pathways, compared to IFN-α/β, they state.

Unfortunately, I could not find any information on the capacity of children to produce type III IFN, but it is tempting to speculate that a predominant, early but time-limited type III IFN response in children could have a protective effect towards COVID-19 pathology ?   

  1. The hypothesis of trained immunity as an explanation why children and certain populations in the South seem are affected by COVID complications than adults resp population in the North, is still very popular. 
  • As you know, first BCG vaccination has been proposed.  While the epidemiological evidence has been controversial, clinical trials in HCW are ongoing to test the hypothesis (Ep 90 16a-c).
  • Alsulaiman extends the hypothesis to other live attenuated vaccines (Ep 90-16d).
  • Maternal lactoferrin and antibodies as well as pneumococcal vaccination is proposed by Root-Bernstein (Ep 90 17).

Not sure what to think about these concepts….