Influenza-binding antibodies immobilise influenza viruses in fresh human airway mucus

[tetano]

Eur Respir J. 2017 Jan 25;49(1). pii: 1601709. doi: 10.1183/13993003.01709-2016. Print 2017 Jan.
[h=1]Influenza-binding antibodies immobilise influenza viruses in fresh human airway mucus.[/h] Wang YY[SUP]1,[/SUP][SUP]2[/SUP], Harit D[SUP]3,[/SUP][SUP]2[/SUP], Subramani DB[SUP]3[/SUP], Arora H[SUP]4[/SUP], Kumar PA[SUP]4[/SUP], Lai SK[SUP]5,[/SUP][SUP]6,[/SUP][SUP]7[/SUP].
[h=3]Author information[/h]

PMID: 28122865 DOI: 10.1183/13993003.01709-2016
[PubMed - in process]

 

[Eight]

Plausible mechanism, but I doubt anyone alive has antibodies targeting the PR8 strain used in this study given it was isolated in 1934 and heavily passaged in eggs. It's interesting to speculate what's actually immobilizing these flu particles as they could never achieve mobility, even when attempting to quench the alleged antibody immobilization.

 

[sharon sanders]

Welcome Eight!

 

[Eight]

Hello! Thank you.

 

[JJackson]

Eight welcome, have you had a chance to read this in full? It is behind a payfirewall but I was interested in what they did, and found. I was under the impression that NA binding was a factor and that the URT mucus was rich in sialic acid residues that slowed passage and allowed clearance by both the immune response and mechanical flow of mucus by the ciliated epithelial cells.

 

[Eight]

It's a very short read--technically a 'research letter'--where most of their methods is really just pawned off in the form of a reference. I can provide the PDF if you would like. Essentially they tried two tactics:

1) Use influenza virions labelled with amine reactive fluorescent probe on several flu strains [PR8, A/Vic/72, compared with PEGylated latex beads];

2) Use H1 and N1 (from PR8) containing flu VLPs constructed with HIV gag-cherry reporter [with and without SA-binding domains] compared with HIV gag-cherry.

To account for the possibility of NA-altered virions through labeling or VLP expression, they do a MUNANA assay and find no differences between their VLP constructs and unlabeled PR8 H1N1.

All of this was observed for movement in 20s video periods in human mucus acquired from endotracheal tubes of elective surgeries "outside typical flu season months". They tried throwing 20X the amount of labeled virus to quench this putative antibody immobilization prior to adding their labeled virions. Problem is, there was never any 'moving' flu particles in either PR8 wild-type or the VLPs, even with SA-binding domain removal. Some of their analysis hints at a subset of flu particles out of the entire population that is mobile, but we don't see anything other than average movement (which will be terrible). And even this 'mobile fraction' is only ever creeping up to ~20% in a minority of their trials.

So the question is what is immobilizing their flu VLPs? I doubt it's antibodies stopping the PR8, and maybe even the Victoria72 as well.

 

[JJackson]

Many thanks. I wont bother with the .pdf for now as your explanation answers my questions. I had assumed the cilia's action on the mucus provided much of the transport between the mucal surface in contact with the air and that in contact with the epithelium, and this seems not to have been replicated. I very much doubt antibodies are playing much of a role and - as you say - PR8 exposure could only be via vaccine, and then only to the internal proteins.

 

[Eight]

What are you wishing they looked at more in depth? Sialic acid presence in mucous? Flu travel through the mucous?

It sort of comes across as an aborted study, but why they stuck to their guns about the antibody idea is a real head-scratcher.

 

[gsgs]

reminds me to this : http://www.virology.ws/2014/01/08/cutting-through-mucus-with-the-influenza-virus-neuraminidase/

 

[JJackson]

Eight: Flu travel through mucous. I never really had a clear mental model of how this would occur as the virion has no means of locomotion, apart from the the cilia moving the mucus. I do not see the adaptive immune system in this only the innate, in which the immobilisation, by binding, might put the virion in the path of a hungry macrophage.

Gs: I wondered if you would show up on this thread. I recall this coming up in the Cochrane NAI report discussion https://flutrackers.com/forum/forum/...al-usage/page2 in which you linked Vincent's post and I found an earlier discussion I had had with him on the topic of the relative speeds of HA binding and NA cleavage.

 

[Eight]

JJackson, you would probably very much appreciate this study then: https://doi.org/10.1371/journal.pone.0110026

The gist of it is this: the virus must somehow diffuse (i.e. passive, entropy-based Brownian motion) through the mucus amidst the false receptors present on the mucins. Unlike the above study, these authors were able to visualize its movement through mucus and it's pretty remarkable how far and fast it can travel, all things considered.

Note this is swine influenza (truly swine influenza, not swine-origin) in swine mucus. Also note this study used flu-naive baby pigs for the mucus source which effectively excludes the proposed antibody-mediated immobilization suggested above.

 

[JJackson]

Again thanks that helped considerably but without an understanding of the mucocilary churn effect on the mucus it is hard to know how much of the transport from gas surface to epithelial surface is due to diffusion and how much to the mucal flow. I am sorry to keep asking questions but you seem familiar with the literature and I never manage to read a paper without having more questions afterwards than before I started. So do you know of a paper that looks at the motion of particles within the epithelial mucus and how thick would the mucus typically be (they seemed to be getting ~100nm in 30mins is this a fair test)?

 

[Eight]

I never manage to read a paper without having more questions afterwards than before I started.

You're in very good company in that regard!

So do you know of a paper that looks at the motion of particles within the epithelial mucus

I know of none that look at both simultaneously. I only know of the original and this most recent paper that discuss actual influenza particle motion in mucus. It's a relatively unexplored area in general, from what I can tell. Recent look at creative drug delivery and gene therapies (e.g. Cystic Fibrosis treatments) have looked at various particle transport through mucus and even some viruses, but not flu specifically, which is a real shame.

how thick would the mucus typically be (they seemed to be getting ~100nm in 30mins is this a fair test)?

It will vary considerably depending on the location and state of the respiratory tract (e.g. hydration and disease), but around 10-50 micrometers in some spots is what I've seen mentioned. So anywhere from 100-500 times more than the distance they observed in a half hour, if we want a really crude estimate.

All of this is really dealing with what happens above the cilia in the gel mucus layer. I've not seen anything combining that with the layer below--the periciliary layer (PCL)--and certainly not anything that also looks at the entire system of both layers plus ciliary beating. The PCL should itself exclude objects as small as 40 nanometers, which is notably half the size of the smallest influenza particles, if not more so. How influenza can manage to diffuse through the gel layer is considerably less difficult to imagine then what happens once it reaches the PCL and I have no good papers or ideas really of how this happens, other than a lot of viruses are capable of doing it and we all get colds and flu regularly.

 

[Eight]

It will vary considerably depending on the location and state of the respiratory tract (e.g. hydration and disease), but around 10-50 micrometers in some spots is what I've seen mentioned. So anywhere from 100-500 times more than the distance they observed in a half hour, if we want a really crude estimate.

Looks like I made a napkin math error. 100nm is a very small distance to flu--about the diameter of one spherical particle. By 30 minutes the entire population averaged ~60 micrometers which is pretty far in my opinion. It's bimodal, so you're averaging zero with whatever the moving particles achieve. That's a very relevant distance in half an hour to an hour.

 

[JJackson]

Thanks for the clarification, it was my fault. I said "they seemed to be getting ~100nm in 30mins" due to a misread of the scale in Fig.5. it should have read ~100μm in 30mins. I look forward to your participation on the site in future and if you see any papers on the gap between the airborne/droplet and binding/fusion areas of research please post on them.

 

[Eight]

I can certainly do that! It's an area that I find interesting as well.