The Achilles Heel of the Coronavirus | Hacker News Reader

The Achilles Heel of the Coronavirus

ethz.ch

204 pointsoumua_don1782 comments

For me sounds scary to disable a mechanism in all our cells that is "rarely used".

Negative knowledge covers what we don't know and what we don't know that we don't know, and that rarely could be less rare than we thought, doing maybe not do frequent but anyway essential things, for all or some cells.

The article doesn’t say it’s “rarely used”. It says it “rarely happens”. It is not supposed to happen:

> Frameshifting almost never happens in our cells. It would lead to dysfunctional cellular proteins.

So the drug would disable an aberrant behavior that coronaviruses depend upon to replicate. This doesn’t sound like a bad thing. It’s also why trials exist.

There is a level of confidence in your post which is wholly unwarranted given how complicated the human body is. There is no design committee at work here, nobody is telling any body what the bits are supposed to do. It is all evolutionary equilibriums. The point of this drug is to shut the coronavirus down to the point where it cannot restart; it is not at all out of the question that something obscure but important in the human body uses the same mechanism and would similarly be permanently disabled.

New technologies are all very well, but this is not the sort of approach that should be rolled out in an emergency. If someone has 10 years of safety data from some trial and can put their hand on their heart to swear there are no suspicions of a problem then that is one thing, but otherwise there is no way this approach is safe for an emergency mass deployment. It probably won't be relevant for the current pandemic.

Au contraire, this is exactly the thing that we should look into in an emergency! RNA-slippage is a well-known mechanism that a healthy cell has safeguards against. The corona virus exploits this slippage, so disabling it using a drug is a novel and smart approach as this presents a likely very specific attack surface (meaning side-effects will likely be few). Ignoring this mechanism would be very imprudent. What makes you think the usual precautions (cell lines, animal models, phase I, II, III trials) will not be taken as per usual?

> What makes you think the usual precautions (cell lines, animal models, phase I, II, III trials) will not be taken as per usual?

You've skipped one of the usual phases, which is Phase IV trials. Cell lines -> Phases III trials aren't enough to catch long term side effects, because they are all done quickly.

An experimental mechanism passing Phase III trials is insufficient to prove it is safe. It proves that any negative events aren't evident quickly. The drug will probably prove to be safe in time, and I don't mind if people want to take that risk, I'm pretty free market. But it is, nevertheless, risky. We can't be confident that the body isn't using the same trick somewhere or somehow for something delicate that breaks quickly and shows symptoms slowly - biology is complex.

If the mechanism is novel then it wouldn't be suitable for a mass rollout to deal with this pandemic.

> But it is, nevertheless, risky.

But this assumes there is no risk to not doing it, which is false. Getting sick, suffering long term damage to your body, or dying are all very real risks of covid-19 that we already know about. We also have no idea what all the long term less evident negative effects are from covid-19. We have no idea if an asymptomatic covid infection does some sort of long term damage. So we are weighing the risk of treatment against the risk of the disease. People are naturally biased against active risk when compared to passive risk, so it is natural allow a high passive risk while worrying about a low active risk, but they are both just as real.

There’s also the second-order problem of normal people generally being bad at statistics[0], and therefore a significant number (even when a minority) assuming that any report that “one specific vaccine turns out to be harmful” implies “the general concept of vaccination is bad”.

We need people to be confident in the long-term, as well as to have a solution quickly in the short-term.

[0] hence, for example, the use of the word “obscure” in the headline “The obscure maths theorem that governs the reliability of Covid testing”: https://www.theguardian.com/world/2021/apr/18/obscure-maths-...

But I’m not sure the hyper risk averseness actually helps normal people to have a more realistic perspective on statistical risk. When people see a drug or vaccine that’s put on pause out of an abundance of caution, they don’t think, “oh, look how careful they’re being! I think I’ll trust their process more.” They think, “Wow, look how inconsistent and unreliable this is! I don’t know if I can trust the FDA or vaccines at all!”

So I’m skeptical of folks using 2nd order arguments about why going beyond rational approaches to risk is a good idea. People see overly risk-averse behavior by the FDA as evidence that the risk is actually high (relative to other stuff in their lives).

Sigh. Phase IV trials are not part of the approval process; they follow formal FDA or EDA approval and noone is doubting they are important. Regarding risky: it is a well managed risk, and one that you simply cannot not take in the face of a debilitating condition, let alone a deadly pandemic. How do you propose new drugs can ever be introduced at all? And again: the Moderna and Pfizer (and also Janssen, somewhat) vaccines use a novel method of delivery, have been mass-rolled out. You seem to be saying that this should never have been done.

You can't prove safety.

You can demonstrate a high degree of safety.

But demands to prove it can always ask for more more more.

Phase IV trials are only undertaken after formal approval by FDA and EDA; they are not relevant in the current context. The current vaccins (including Pfizer, moderna which clearly use a novel mechanism for delivery) also don't have phase IV data yet. Thank God they have been approved though, they are currently clearly the way out of a pandemic that so far has claimed 3.4 million deaths. I really don't understand what you on about, sorry.

You are allowed to raise concerns without being labeled an antivaxxer.

Not disagreeing on the caution angle.

But, let's imagine we find, by the usual trial and error, a chemical compound whose effects are that of curing patients from the coronavirus. Trials show that it has some rare side effects, like many other drugs.

How much do we need to "understand" its first, second and third (and fourth) order effects before deciding to use it?

If it cures COVID-19 patients, you can use that drug on exactly those people whose case is severe enough to justify the risk. Many chemotherapy drugs have terrible side-effects. We break people’s ribs from CPR all the time.

No one argues we should stop breaking those ribs as that’s a fairly easy balance to strike versus a vaccine that you give prophylactically to healthy people. Far more caution is warranted there. (This article is about a therapeutic drug concept, meaning it’s the former case, but I’d still argue caution and where traditional supportive and therapeutic measures are working, I’d not rush to use this mechanism on mild cases.)

In this case it is using an unproven mechanism that requires proof of a negative ("nothing important in the body uses this mechanism"). It cannot be done safely and quickly.

Just because someone survives 12 months and feels fine doesn't tell us anything about the safety of that mechanism. We could literally be injecting people with HIV and it'd only be showing "rare side effects" for the first 12 months as far as symptoms go.

Corona virus exploiting frame shifting using a pseudoknot is not a at all unproven, has been known and researched since 2005

my point was: in this case we know this is an unproven mechanism; because we are designing a drug that makes use of a mechanism.

But we're using empirically discovered drugs since forever, where we had no idea on how the thing even supposed to act, let alone the consequence of this action.

Empirical drugs also came from an age where the alternative was cutting the part that hurt and hoping for the best.

You have to do better than the alternatives, and it was easier at a time when alternatives had 60% mortality.

That said, this could be a good alternative than straight up dying for those in critical conditions. It would however tell us nothing about its safety, as people in critical condition have some deal of damage taken by the coronavirus itself and it'll be hard to control for each secondary effects.

I don't think that the idea is to disable it forever.

If we disable oxygen to the lungs for 30 minutes then switching it back on doesn't help. The lung owner is probably dead. Some of this stuff causes permanent changes.

There is reason to be optimistic, because if for it to be really bad then there'd need to be:

* Permanent change

* Causing negative symptoms

* Which took a long time to be symptomatic

(which, coincidentally, HIV/AIDS exposure is a great example of)

Now if frameskipping supressing drugs are something we expose people to as a matter of course then very well, probably would work out great. But if this is new then it is absolutely not safe to do a mass rollout in the next decade. There are obvious risks.

There are always risks, and ignoring a novel approach is also a risk. The issue is to gauge risk intelligently, based on extensive prior medical, biological and physicochemical knowledge and modelling, and on sound clinical testing using randomized controlled trials (the kind of trial that proved dexametasone and disproved ivermectin, hydroxyquinoline etc. as a therapeutic for covid).

And the final step, declaring that disabling a cell mechanism cannot possibly have long term effects.

> There is a level of confidence in your post which is wholly unwarranted given how complicated the human body is.

Thank you for saying this, there is a great deal of over confidence bias in the approach to this virus.

Could be a possible treatment that would have only temporary effect. So after the treatment is done the things would return to normal, this could be tested for and give you peace of mind.

"Dysfunctional proteins" is a vast understatement. Too much frameshifting leads to cancer.

I don't see how it could lead to cancer. We are talking about a translation error, not a replication error.

it could, in theory lead to cancer, but you'd have to be super unlucky. (frameshift -> accidentally leads to some sort of oncogene agonist)

Even then, frame shifts are one off events, so you’d end up with a single instance of an oncogene that would quickly get recycled.

exactly. Don't know about quickly. Half lives of single proteins can be highly variable. You can have a single protein live long enough to destroy an entire cell (ricin); it evades degradation by not having many lysines and never being ubiquitinated.

Doing it just long enough in people with a severe infection so they don't die of the infection seems like a reasonable trade off.

Nobody's suggesting we use this as a daily supplement in healthy people.

This is not actually a mechanism in our cells. The virus RNA folds in a particular way that interferes with how the ribosome usually works. Imagine it like a machine that takes in a an RNA strand, and you make a knot in the strand. The machine can untangle the knot with some force, but slips while it is doing this and the alignment of the strand shifts and the machine continues on a different position on the strand than before. What they propose is to target the knot with a drug, not our cells.

Translational frameshifting is a feature of the human PEG10 mRNA, https://www.genecards.org/cgi-bin/carddisp.pl?gene=PEG10

http://recode.ucc.ie/search?q=homo+sapiens lists a number of other recode events.

I wouldn't want to do it permanently but doing it temporarily, to treat a severe acute infection, seems like a pretty good tradeoff. Especially after you've done your clinical safety trials to be sure.

Yeah I always laugh for the same reason when people describe organs as "vestigial."

We can say that confidently because plenty of people have lived long lives of average health without those organs.

Frameshifting might play a role in healthy gene regulation https://doi.org/10.1002/bies.201500131

Pretty interesting how well the article explains such a complex concept.

Does anyone have a sense of why the virus evolved to use this strange "frameshift" mechanism? Is this "hack" somehow advantageous enough for the virus that it would evolve it, rather than using more standard mechanisms?

edit - oh, wikipedia claims this is a compression mechanism, that's amazing! https://en.wikipedia.org/wiki/Ribosomal_frameshift#Function

“In the case of a translating ribosome, a frameshift can either result in nonsense (a premature stop codon) after the frameshift, or the creation of a completely new protein after the frameshift. In the case where a frameshift results in nonsense, the NMD (nonsense-mediated mRNA decay) pathway may destroy the mRNA transcript, so frameshifting would serve as a method of regulating the expression level of the associated gene.”

Nonsense-mediated mRNA decay. Love this term! :)

Correct, you can encode two different proteins with the amount (length) of DNA usually used for one protein. This is not really relevant for higher life forms, as DNA is cheap. But for a virus this 'cost saving' can really matter when trying to replicate quickly and cheaply.

I'd say the achilles heel is most likely in the genomic structure, which unlike influenza, is less likely to recombine.

Not true, the genome of viruses is very variable/evolvable due to its high mutation rates.

I wasn’t talking about mutation rates. Influenza has mini-chromosomes which it can swap between strains, leading to more variants since they can swap those mini-chromosomes. SARS-CoV-2 has a single linear genome - only recombination on the RNA level or accumulation of mutation will allow for evolvability. Also, viruses massively differ in their own mutation rates.

So, yes, true.

Remarkable work, I wasn't aware that it was possible to image transcription at the ribosome like that.

Yep, I've already seen that kind of animations. That's super cool !

> one of the two reduced viral replication by inhibiting ribosomal frameshifting, while the other might act through a different mechanism.

Just another random different mechanism there.

Wonder what it 'might' be doing.

Probably wasn't important.

Not confidence inspiring that their drugs are doing things that they don't fully understand.

I'm sure we can put this in humans without any adverse effects that 'might' happen.

> Not confidence inspiring that their drugs are doing things that they don't fully understand.

That's really common.

Yup. From hazy chemistry degree knowledge, we don't know how most drugs actually work. Designing drugs based on the shapes or proteins etc is pretty cutting edge stuff.

And also the nature of organic chemistry is that even if a drug does work a certain way, it probably has several other pathways it can go down. That is, only a certain (probably low) percentage of the drug will actually have the effect you want, the rest will take part in other reactions.

So common it would blow the average patient’s mind. The mechanism of action for Tylenol is still not entirely understood.

So the idea based on other threads is to eradicate covid with this method.

I can't wait to see how the public would react to this tech, given the response to mRNA vaccines, trying to blow their minds on this one.

We can't eradicate Covid by making all the birds wear masks, we can't eradicate it by vaccinating all the birds, and we certainly can't eradicate it by asking all the birds to go to the doctor to get treated when they get Covid.

Wow, you still think it's going to be eradicated

No, I don't.

if we targeted this, would the coronavirus simply be able to figure out a way to regulate its proteins some other way?

If we got rid of all living organisms entirely, it stands to reason that the virus wouldn't be able to go on living by itself. Therefore, between the current situation and now, there's a point where it's too hard for the virus to keep existing in its current form.

There's no reason to suspect that the virus will be able to tackle anything we throw at it without just dying out. It might not be the vaccine, but it'll be something (assuming it doesn't just turn into a common cold eventually and we lose interest.

There’s only one way to know for sure...

>Frameshifting almost never happens in our cells. It would lead to dysfunctional cellular proteins; however, certain viruses, such as coronaviruses and HIV, depend on a frameshifting event to regulate levels of viral proteins. For example, SARS- CoV-2 - the virus that causes COVID- 19 - is critically dependent on frameshifting promoted by an unusual and intricate fold in the viral RNA.

God is a hacker.

https://en.wikipedia.org/wiki/Viral_evolution#:~:text=Bubble...

you might find this interesting. bubble theory imagines that interacting molecules, let's call them 'scripts', formed simple self replicating metabolisms inside the cavities of hydrothermal vents. At a script level there are two successful strategies: participate in a co-operative network or exploit a co-operative network to self replicate you. At a community level, excluding non-contributors is beneficial and this pre dates and perhaps leads to the development of cell membrane. In this view cells and viruses co-evolve from the same origin by pursuing different strategies.

It's game theory and tragedy of the commons from day one.

Piast, Radosław W. ‘Shannon’s Information, Bernal’s Biopoiesis and Bernoulli Distribution as Pillars for Building a Definition of Life’. Journal of Theoretical Biology 470 (7 June 2019): 101–7. https://doi.org/10.1016/j.jtbi.2019.03.009.

https://www.sciencedirect.com/science/article/pii/S002251931...

Your reply got flagged, I’m curious whether that was auto or human. I was talking with my daughter last night about ambiguity in speech and interpretation last night, and about the space of true things well intentioned people avoid, because bad faith actors with an unspoken partisan agenda use that space as area of tenuous ambiguity into which they implausibly retreat when challenged on the hideous ideas they are trying to ship. I read your comment as a reflection in the hideous cruelty of life, anyone studying biology or medicine beyond high school confronts the cold indifferent ugliness of nature, and the enormity and absurdity of the task we have to build a kind compassionate society on this substrate. But once you factor in ambiguity, your comment keyword matches the crypto-misogyny leading comments of trolls claiming to know something about social Darwinism. Perhaps others read your comment in that light. On the whole I enjoy how this site is moderated, I think it’s close to optimum, but general optimization means specifics are sacrificed, and this is one where I would like to have participated in a quite backstage conversation to understand the perspective of the participants

I'm completely surprised by the negative reaction to my comment. I think people are misinterpreting my use of the word hacker, and I don't even mean to imply that God exists - only that the universe is intricately clever and whether by intent or by emergence this frameshift exploit on which some viruses rely for reproduction is an absolutely brilliant trick. My comment was made in reverence. Or perhaps it's the implication of intelligent design that upset the downvoters/flaggers?

>I read your comment as a reflection in the hideous cruelty of life, anyone studying biology or medicine beyond high school confronts the cold indifferent ugliness of nature

Perhaps my comment came off callously then. I made no value judgement, it was only the design that I meant to describe. Nuclear weapons are terrible things, but they are still technical works of art.

Or you might blame SARS-CoV-2 and HIV-1 on the other guy.

Not very sporting to target your own exploits.

Hackers aren't athletes

Anyone else getting SSL_ERROR_UNSAFE_NEGOTIATION?

Its not using the new secure renegotiation. So its possibly vulnerable to https://blog.ivanristic.com/2009/11/ssl-and-tls-authenticati... but more likely they have renegotiation disabled entirely.

just means it's an old server.

A bit redundant.

We know how to “kill” the SARS-CoV-2 virus in vitro (in a test tube). Ammonia, bleach, UV light. Heck, even diluted mouthwash will do the trick.

The problem is that none of these approaches are safe in vivo (in the body). All of the approaches the article talks will require years of human testing. The drug EIDD-2801 does similar things but has yet to pass testing even after 12 months of studies.

This isn't about "killing" the virus, it's about inhibiting its ability to reproduce. Pointedly, it's about doing so without introducing mutagens like EIDD-2801, but instead by inhibiting a specific ribosomal behavior induced by a component of the virus. Finding a safe mechanism to do so is the next step in their research. I suspect your interpretation of the article might be off.

This work is a great example of the new insights we get from recent advances in cryo-electron microscopy. It's much easier to formulate responses to mechanisms when we can get clear pictures of the sites in question.

Suppose protein folding simulation and manufacturing can rapidly advance well beyond current levels. It seems like a hypothetical synthetic immune system that whitelists every configuration of every protein an individual can possibly generate and excludes everything else (if possible) would be a way to blanket eliminate unknown unknowns proactively like SELinux for the body.

Such a system would effectively halt evolution, though.

I don't think biological evolution is an important factor in modern humans.

That comment is spectaculary equally insightful, humorous and sad.

Within 20 years every newborn will be gene edited.

Likely, human intelligence and world view will end up being derived from a handful of genes, thus further blinding humanity to data outside of our worldview.

Imagine a world where everyone has the same level, and type, of emotional response to data, feels the same as a result of environmental stimuli, eg cold/warm/wet etc.

Imagine a world where if your next door neighbor cannot figure out something, no one ever can, for there are no geniuses or dullards.

The sameness of thought will make future society dull, compared to the Amish.

(Think it won't happen that fast? Laws will prevent it? Imagine your neighbour flies off-shore, and comes back with his child gene edited for appearance, intelligence, health. Now you are about to have a child. What do you do? Compared to the neighbour's child, yours will be dull, sickly, ugly. What will you do?

Now imagine the rich and powerful do this, bit you cannot afford it?

1%er indeed.)

I think you overestimate the importance of genes. Twins are not identical people.

Twins are identical platforms.

Some studies show political leanings as genetically derived. Each human views the world with different senses, yet imagine all sensory input being significantly less divergent. Same taste. Smell.

I think you significantly underestimate the importance of genes.

You should read Brave New World by Aldous Huxley.

I did a thought experiment about how the immune system must work and came up with a simple system using only mechanisms that I know are available molecularly.

If you could have a bunch of random protein sensors that undergo a training phase where they learn what "Host" is. Anything that makes a "detection" during training is discarded. Anything that doesn't is either never going to detect stuff, or might detect something "Foreign".

These then float around your blood stream and either do nothing until they are lost, or go crazy replicating and raising the alarm when something Foreign shows up.

A danger with this method is random sequences that are just really bad at detecting "Host" or a training period that is cut short. This might lead to autoimmune responses.

This thought might give some hints about how to combat autoimmune diseases without disrupting the greater immune system.

Perhaps an improved auto-recognition "self" system that uses something like public key crypto and a challenge-response.

The biggest advantage we have is that we might identify the responsible bad signaler, and then target it for removal somehow. I wouldn't want to mess with it more than that...for now.

The article's about a method they hope will be safe in the body.

Maybe they're right maybe they are wrong, but hardly seems redundant.

We also have known about this for a long while. A pity the article doesn't mention what the "compounds" are zinc based and are the mechanism of action of HCQ based treatments.

This trial says UV light can be safe. Take it for what you will.

https://www.medrxiv.org/content/10.1101/2021.03.05.21252997v...

Reply on news.ycombinator.com