Cunning bacteria and antibiotics

Bacteria are masters of genetic recombination. They have a very short life cycle, large offspring, billions of attempts at mutations, and a very short feedback loop. This alone is enough to, if not brute force, then find a solution to the problem using Monte Carlo algorithms.

Plus, they have this nice thing called horizontal gene transfer. I mean, a bacterium can take and distribute its code. For example, in the form of plasmids — DNA storage facilities carried outside the chromosome. This means that you don’t even have to go through reproduction cycles — just one adapted bacterium that will start distributing. And it distributes it to different species, not just its own.

Therefore, antibiotics began to create a sharp evolutionary pressure within 15 minutes of their appearance.

Penicillin was first taken from mold. They grew it, cut it, dipped it in a solution, and got flasks with antibiotics. And for some reason, bacteria for which this environment was deadly began to grow on the edges of these flasks. But then they had to do some work and decided that these were some kind of wrong bacteria that made the wrong honey. Especially since when the dose was increased, they seemed to die.

In general, there are a couple of super interesting stories about how cunning and thoughtful bacteria are. It's time to figure this out. Today we are doing this with an expert – Denis Kuzmin, PhD, Director of the Phystech School of Biological and Medical Physics at MIPT.

What is an antibiotic?

This is a thing that can kill bacteria. Basically, molten metal is also a kind of antibiotic, but if you take it internally, it will end sadly and somewhat predictably for the main organism. That is, an antibiotic is something that kills bacteria without killing the rest of the person. This most often works on the difference between a bacterial cell and a eukaryotic one, or targets specific signatures of bacterial cells.

For example, it kills us and bacteria at different speeds. Actually, this is the most common case of practical application.

Industrial production of antibiotics changed humanity — people simply stopped dying from the most common causes at the time. Well, except for hunger. In general, antibiotics were discovered several times, but at first each time it seemed like a cool, but not very practical idea. In 1928, Flemming isolated penicillin, in 1938, Florey and Chain solved the problem of resistance, and only in 1943 did industrial production begin.

Here is a sketch from “All Creatures Beautiful and Wonderful” by the veterinarian Herriot, as it looked to him for the first time:

“All over the country, all over the world, veterinarians were seeing the first stunning results in those days, experiencing the same thing I was experiencing that day. Some with cows, others with dogs and cats, others with expensive racehorses, with sheep, pigs. And in the most varied settings. But for me it happened in an old wagon converted into a calf barn, among the rusting iron junk on Willie Clark's farm.
Of course, it didn’t last long – that is, the creation of miracles. What I witnessed in that calf barn was the impact of a completely new agent on an unprotected population of bacteria. Then things went differently. Over time, the microorganisms developed resistance, and new, more effective sulfonamides and antibiotics had to be created. So the battle continues. We are getting good results, but we are not performing magical cures, and I am very fortunate to belong to the generation that saw the very beginning, when one miracle followed another.”

In fact, he describes the golden age of antibiotics, when more and more new molecules were found. Penicillin is a beta-lactam antibiotic, meaning it disrupts cell wall synthesis. Generators of interference for DNA doubling (fluoroquinolones), agents for membrane destruction (linezolid, daptomycin) were found, we learned to stop the metabolism of bacteria's own vitamins, break protein synthesis with tetracyclines and macrolides. An arsenal of about 20 main groups appeared. All of them killed a healthy person more slowly than his bacterial population. They could be targeted more or less clearly, some, after modifications, did not kill our symbiotic bacteria at the same time or at least did it more carefully.

Bacteria have been evolving all this time

Penicillin turned out to be one of the simplest for most populations. Beta-lactam as a basis? Okay, there is a standard countermeasure – you can isolate beta-lactamase. Yes, according to the original architecture of bacteria, it was actually needed for something else and it was synthesized quite a bit. But, let me remind you, billions of quick attempts. Even a simple copy-paste of a DNA section with genes responsible for synthesis – and now more of it is isolated.

The problem here is not even in finding a countermeasure, but in refactoring it later, because the protection can eat up to 90% of resources, which is not very good for survival in a normal environment without antibiotics. Penicillin is easy, it quickly became clear that you can quickly and elegantly modify the code so that beta-lactams are no longer a problem.

Or macrolides and tetracyclines put a patch on the ribosome – they highlight a place where you can fix a little code and make it so that nothing else compiles normally. In response, bacteria change this landing site in the ribosome, rewriting the same functionality with another code. That is, they hide the signature of this area. Not purposefully, just under the evolutionary pressure of antibiotics, those who could do it do it. And those with whom they shared genes horizontally.

Bacteria have many other countermeasures. For example, Acinetobacter baumannii has a cool cell wall covered with a lipid membrane. It's like a space station with a force shield. There are several ports through which the bacterium can exchange substances with the outside world, and as soon as a threat appears, it closes most of the ports and waits. And that's only half the problem. Such colonies have a quorum sense – a chemical coordination mechanism that allows them to interact collectively. In the simplest case, each bacterium secretes a certain amount of a marker and can count this marker in the external environment – when there is enough of it, the colony understands that it can move on to the next phase of the plan and form the same bacterial films. This is what they have instead of more advanced cell coordination mechanisms in multicellular organisms. So, you can download plasmids that add a “urgent pupation” signal to the quorum sense, and not only do all the bacteria close their ports, they also launch emergency transporters and start pumping out everything unidentified from the inside, which washes out even the antibiotic that got inside.

Here is a beautiful one Jobwhich shows how it works in a complex. Conjugation (creation of a direct connection between bacteria) – 2 minutes, complete copying of information – 10 minutes, an hour after entering the population, about a third of the population receives the F-plasmid. TetA synthesis begins immediately after DNA enters the cell, even in the presence of tetracycline in the medium in concentrations that suppress division. A pump is turned on, which removes tetracycline from the cell. Division is still impossible, but the concentration decreases enough to ensure the synthesis of countermeasures. When the pump genes were removed, the bacteria could not resist tetracycline even in the presence of plasmids with resistance genes in the population.

It is clear that they cannot live in this mode for years, but usually the patient cannot either.

In general, they have something to say in response to various threats.

This story has affected, to one degree or another, every antibiotic that has been tested.After 5-10 years, an effective countermeasure was found that spread through bacteria across the planet.

And “about the whole planet” – I’m serious. Here. here describes how quickly resistance spreads with the help of plasmids. The blaNDM-1 gene was found in 2008 in India (that is, it appeared a little earlier in bacteria of this region), and already in 2013 – in the soil in the Arctic on the Spitsbergen cape. Actually, similar cases of spreading have even occurred in Antarctica. Why is this so – because as soon as bacteria get up to distribute something useful, it will be shared around the world. Perhaps, this happens so quickly due to the development of aviation. And yes, this gene could only have appeared as a result of anthropogenic influence, that is, it is not some kind of external ancient factor that scientists so luckily came across in their study.

Here is a 2 minute tutorial video:

What's the problem?

The fact that people are very attached to antibiotics. Firstly, they do not allow them to die in frequent situations. Secondly, they give them the opportunity to eat in principle. Because the same chickens live in such hell in terms of amplification of infections that only horse doses of antibiotics help to produce them at the required speed. By the way, while you were reading the last sentence, about 4 thousand heads died and were born. And if you read slowly, then from 6 to 8.

Well, and the lovely plants are not very happy when bacteria eat them, especially when you are into monocultures. That is why antibiotics are poured in huge doses into the fields.

Extrapolation of known sales data here research shows that currently about 99 thousand tons of antibiotics are used per year in livestock farming alone. About plant growing Here.

^{Find China}

And if all this stops working, first there will be nothing to eat and surgery will fail, and then something will have to be done about it. The prognosis is still very optimistic – 300 million deaths by 2050. It only sounds terrible, in reality – it is not the extinction of humanity or a superbug taking over the world.

On the other hand, in 2019 it was higher reportaccording to which antibiotic resistance is the third cause of death after ischemic heart disease and stroke. We are talking about 5 million deaths per year on the planet with a chance of reaching 10 million per year by 2050. For comparison, about 1.3 million people die in road accidents every year.

I mean, to exaggerate, you're afraid of cancer, but you'll die stupidly and unromantically either from overindulging in pizza, or from someone deciding to cure a cold with macrolides and developing a polyresistant version of E. coli.

In general, it’s somehow very stupid to die like this, but the chances are growing.

What is possible was do?

The first idea

The idea was not to let the “trained” bacteria out of the labs. The idea was that more and more antibiotics could be discovered, and then, as new epidemics occurred, it would be possible to surprise the naive bacteria. Given the clinical trials on hundreds of people—and the fact that the rate at which new antibiotics are discovered is slowing—this sounds like imbecility.

But we don’t have anything better yet.

The second idea It was necessary to agree with all of humanity and not to release some antibiotics into widespread practice, but to keep them as a reserve. Everything is a little more tricky there. The thing is that any protection complicates the bacterial code and begins to take up extra resources where they do not exist. That is, in a normal environment, it is unprofitable to have this code – you will be forced out by less paranoid colleagues. It was assumed that it would be possible to alternate antibiotics so that for 10 years we do not show tetracyclines to anyone, bacteria get used to the fact that they are not there, optimize, stop maintaining the protective circuit against them, it falls apart – and after some time they find themselves defenseless again. The second agreement is not to release antibiotics into widespread practice that bacteria have never seen at all, that is, to keep reserve lines. The penultimate and the last for the Day of Judgement.

We know very well how people negotiate. The two best examples are Covid quarantines and vaccination.

The third idea was to complicate the antibiotics themselves. For example, we have amoxiclav. This is the same penicillin, only wrapped in clavulinic acid. Bacteria release beta-lactamase, and clavulinic acid neutralizes it, leaving pure penicillin, which already affects them. In general, bacteria can quite quickly, by evolutionary standards, find an agent that neutralizes clavulinic acid, or, what is happening now, simply start making other means of disassembling beta-lactam. For starters, by coding the production of beta-lactamase with other genes and assembling it according to other recipes, that is, exactly the same, only different.

This race of shield and sword leads to the complication of antibiotics and the complication of bacteria. Alas, but the complication of antibiotics occurs faster, and the limit of survival of a protected bacterium in a normal environment and the limit of the complexity of an antibiotic do not intersect yet. There are several reasons – starting from the fact that it is trivially long and expensive – and pharmaceutical companies do not want to dig in the direction of such antibiotics, when it is possible to produce almost 20 times faster paying off antidepressants – and the fact that the complication of an antibiotic leads to the fact that it becomes more and more difficult to test. That is, either it cannot be made according to modern rules, or it is even more expensive to develop.

On the other hand, here example stories where bacteria defend themselves against a new threat (viruses) and in doing so lose resistance to antibiotics.

What's important is that if a bacterium simply acquires chromosomal mutations, it will be an unsupported fork with crutches. The probability that this will be viable is low. But if you refactor it into a separate microservice that is compatible with almost everything, plus put it in a normal update package – a plasmid – then you can support such a plasmid almost for free, it just lies there, the container with it is turned off. The plasmids themselves can (and most often do) contain not only the tool itself, but also a set of sequences for their own copying and embedding. That is, they come with an API right away. And compatibility is also constantly updated as it is embedded into new bacteria.

It would seem that the situation is rather dull. The discovery of new antibiotics is slowing down, the ability of companies to modify them is becoming more expensive, and all that remains is, by and large, to clearly agree on how to use antibiotics and how not to use them. And this will solve most of the problems – or at least give a respite.

But there is a nuance.

People are idiots

Firstly, it is very difficult to limit the consumption of antibiotics. People gobble them up like vitamins. Here is my

previous post

about viruses and antibiotics, there the wonderful Victoria Valikova talks about the African practice of going to the market, buying one pill and swallowing it just in case. That's folk medicine. Fortunately, they don't have expensive antibiotics – simply because they are expensive. And you still have to wash them down.

Secondly, the doctors themselves prescribe in a row. This is also due to the fact that people are not the most rational. The patient may simply disappear and not come to the second appointment (usually they do so). Therefore, the doctor immediately predicts the most common variants of the course of the disease and complications and gives a set of pills for this case. Because it is bad if a pre-deceased person comes to the second appointment. This can spoil the hospital's indicators. Therefore, there is an excellent chance of getting a prescription right at the first appointment, even if it is a viral infection, and it is correctly diagnosed. Because in two weeks an associated bacterial infection may come, and it will not occur to the patient to show up in the hospital. The third feature is that it is generally correct to take a culture, look at a specific pathogen and treat it in a targeted manner, but by the time the culture is ready (especially in remote hospitals without laboratories), the patient will already be dead. Therefore, they usually start with a broad spectrum.

Although there is progress, here is a description opportunity greatly speed up antibiotic susceptibility testing, up to one and a half hours.

Thirdly, agriculture. There was a really amazing story with colistin. They discovered it in the 50s, they were very happy, but it quickly became clear that in about half of the cases it causes kidney complications. In the 70s, in the choice of “alive, but without kidneys” and “dead, but with kidneys”, a third option appeared to use other normal antibiotics, and colistin was put in the last reserve group. Now it is taken out for those patients who are infected with some hospital-acquired Pseudomonas aeruginosa, and a carpenter is already involved in discussing the treatment.

But in agriculture, no one really cares about animal kidneys, because few people live to middle age there. In general, they threw tons of it at cattle (according to one estimate, 12 thousand tons per year is China's official consumption). Moreover, at some point it was even banned in agriculture, too, but not everywhere and not completely.

Naturally, chromosomal mutations with resistance were formed, but they usually did not go beyond a single farm and quickly died themselves due to the general curvature of the code. But in 2014, for some reason, 21% of meat samples during monitoring in China showed resistance. They started delvefound a new gene mcr-1, carefully packed into a plasmid. Bacteria feel fine in an environment without colistin, the plasmid lies and lies, almost does not require additional resources. Colistin appears – the package is immediately installed throughout the population. In general, now bacteria all over the planet have this patch, and we have one less antibiotic of last resort.

And here results 2019 on where and how much antibiotics are in rivers. In short – out of 711 samples, 65% were antibiotics. In some countries, the concentration was higher than the supposedly harmless one. In Africa, the water is rich in ciprofloxacin, in Asia instead of antibiotics – suddenly metformin, in the Danube and Thames – more than 5 different antibiotics. In Pakistan, water from one of the rivers could be considered as a raw material and antibiotics could be extracted industrially, given the concentration.

So what to do?

Displace

. Since protected bacteria are so complex and have such a hard time surviving in a normal environment, you can seed their habitats with conditionally safe bacteria that do not harm us, but devour or kill pathogens. This is already being done, this is a fecal transplant. In its pure form, without the word “feces”, you know it as prebiotics – when bacterial cultures are isolated from the biomaterial of astronauts, which no longer remember where they came from. But the original feces are also not bad, it works. So far, it is known that microbiota transplantation

turns out

more effective than antibiotics in specific cases. Also from the practically applicable bacterial colonies, BLIS K12 and M18 work very cool – microflora that competes with carious bacteria near the teeth and nasopharynx infections. If you go on arctic expeditions with a crooked nasal septum – this will be very useful.

There was an idea with viruses that kill bacteria. Viruses as nanorobots can process almost everything on themselves. There are bacteriophages, and they began to be issued as drugs. The problem with them is proven effectiveness. It is not particularly so. More precisely, they are difficult to target, and can kill flora, but do not really distinguish between the target and the useful. Therefore, in terms of overall usefulness, they are still worse than antibiotics, and that is why you do not see how pharmacies are bursting with them. Promising work there lies in the field of CRISPR / CAS – you can make viruses that will modify the code of bacteria. In essence, a new technological bacteriophage. Perhaps it will show fierce effectiveness.

Modern antibiotics may use new targets. For example, if earlier we broke the cell wall and the DNA of the bacteria itself, and the bacteria learned to resist this, now we can aim at the flagellum, which it uses to move. We can slightly break the proteins that rotate the flagellum, and here's hello NOGOMM.

This is what the new antibiotic looks like zosurabalpinit is highly specific.

Therapy can be complicated by combining methods.. Here story about attempts to introduce gold nanoparticles into double shells, which should lead to increased effectiveness of antibiotics, because they will be able to physically destroy bacterial walls.

Much of what is being researched now is rethinking.

Denis believes that it is possible and necessary to be more effective receive antibiotics. There are excellent studies on how a set of possible combinations is analyzed using neural network models (fortunately, protein coding is a language, almost like music or pixels). But there is a super problem with the fact that it should only work in theory. And in a mouse it is no longer a fact, and then when moving from a mouse to a person it is even less so. But candidates are found faster. It is like CAD software instead of a drawing board. Here example publications. On the other hand, the market is such that it is becoming more difficult to obtain permission, testing is more expensive, and the efficiency of antibiotic mining has been falling all the time. And the market has become more important than development, R&D costs are usually several billion dollars – and marketing can be 2-3 times higher. Plus, there are a lot of problems with antibiotics. Firstly, patients recover from them. This is very bad, because you can't sell them for years. Chronic, like antidepressants, is much more profitable. Secondly, the cost of the course is low, a maximum of a couple of thousand dollars for the upper segment. The same chemistry is an order of magnitude more expensive, so immunomodulators are more profitable. As a result, these are not the most profitable drugs from the market point of view.

In fact, the bottom line is that antibiotic resistance is not as big a problem in practice as it seems. On the one hand, E. coli with resistors kills where there is no modern medicine, hospital-acquired superinfections do not spread beyond hospitals, and there are no plans for a new pandemic. On the other hand, we continue to pour antibiotics into rivers and patch bacteria all over the planet, and at some point, we may have to become expert bacteriologists so that we can leave the house. Even those who do not have a dog.