Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Collective Memory Discovered in Bacteria (sciencedaily.com)
137 points by nikolay on March 8, 2016 | hide | past | favorite | 55 comments


> salt stress causes a delay in cell division, leading to synchronization of cell cycles; secondly, survival probability depends on the individual bacterial cell's position in the cell cycle at the time of the second exposure.

So basically you thwack all the cells with something that pushes their lifecycles towards a similar schedule, and then later-on they tend to all still be on the same cycle?

Isn't that like saying two pendulums in the same house have a "collective memory" of an earthquake?


> So basically you thwack all the cells with something that pushes their lifecycles towards a similar schedule, and then later-on they tend to all still be on the same cycle?

Yeah, that much is pretty well understood. http://journals.plos.org/plosgenetics/article?id=10.1371/jou...

> Isn't that like saying two pendulums in the same house have a "collective memory" of an earthquake?

The important point isn't that there's memory (which seems pretty easy to explain), it's that it's collective. Using your analogy, it would be like finding that two pendulums in the same house have memory of an earthquake, but if there's only one pendulum then there's no memory.

This result might be regulated by quorum sensing (https://en.wikipedia.org/wiki/Quorum_sensing). Like the pendulums are talking to each other.


I think the point your parent comment is making is that "collective memory" implies more than just a common cause causing synchronization. If, as you say, "the pendulums are talking to each other" that would fit my definition of "collective memory", but if the pendulums stay synched just because they started at the same time, that doesn't seem like collective memory.


Totally aside, but funnily enough two pendulums can synchronize through weak vibrations, e.g. pendulum clocks hanging on the same wall. Observed by Huygens in 17th century:

http://www.nature.com/news/2002/020221/full/news020218-16.ht...


A cool thing to do is to hang two pendula next to each other and couple them via, say, a pencil (loop each pendulum's string once around the pencil, so the pencil is horizontal and is able to transfer energy from one to the other). Set one in motion. Eventually wonder where the rest of the day went :)


> So basically you thwack all the cells with something that pushes their lifecycles towards a similar schedule, and then later-on they tend to all still be on the same cycle?

Funny enough servers will do that as well. They will self synchronize and exhibit a thundering herd effect. They run the same software, power cuts our, or they get DoSed, maybe they restart, retry at the same schedule and so on.


Happened just last year! About exactly one year after heart bleed, all kind of websites, big and small, had their ssl certs expired! Apparently a lot of people forget to update the cert schedule


On of the biggest unresolved issue in biology is to explain observed speed of evolution. Random mutations just cannot do it. Collective memory can be a part of the puzzle if it allows effective propagation of changes helping survival.


Except memory is a loaded term implying some kind of intelligence. Here they're just synchronizing the lifecycles of many bacteria using a forcing event. It sounds spooky and mysterious when you call it memory, and in fact it is neither.


I think collective (or genetic) memory is too huge of an advantage to not exist in many species. I remember reading a paper about genetic memory in mices.


Mutations happen by repetition. In behavioural terms we call them habits


No, they do not. Giraffes did not get longer necks by straining to reach higher trees. Randomly longer-necked giraffe predecessors gained an advantage in reproduction.

The idea that we control our DNA through habits is newage bull promulgated by charlatans using the vocabulary of science to sell snake oil and collect speakers' fees.


I don't buy it; the Earth is really old, plenty of time for evolution to work by random mutations selected by environment.


Math for mutations leading to simultaneous advantageous traits doesn't seem to add up easily however: http://inference-review.com/article/haldanes-dilemma


I don't have time to read the whole article, but the sections I did read seemed to draw conclusions from an untenable chain of unjustified simplifying assumptions.


They didn't to me, but then am not an evolutionary biologist.

However, all natural sciences are simplified models approximating reality, the only question how much predictive force they can demonstrate. Either way, the article presents a better laid out theory than "it seems so to me".


Absolute age of life on Earth does not mean anything, it was just prokaryota 2/3 of the time. Compare it to age of Galapagos and Darwin's finches.


1/3 of a really big number is still a really big number. Endosymbiosis conveys a significant advantage.


We do not have an explanation for the speed bacteria develop resistance for drugs or pests overcome GMO.


It seems pretty straightforward to me. Variations exist in a population, hgt shuffles them around, and selection does the rest.


That explain how that happen, but not the speed. Even if we know the trait that bacteria will gain to overcome antibiotics, we still have no idea how to predict even with an error of an order of magnitude how fast that can happen. Models based on random mutations (that typically assume no memory of past performance) show way too slow evolution.


It puts 'bacterial cultures' in a completely different light.


It would if it weren't an overstatement of their results.


Humans also have a collective memory. Unfortunately, most of it is copyrighted by Big Media.


One can only wonder what collective memory might the more sophisticated species have (like Ants). Not sure why this whole thing reminds me about Azak Azimov - Foundation ..


Ants use stigmergy[0]. They lay down different pheremones in their environment as they walk. Intelligent behavior arises out of the combination of all these signals. Wikipedia, the law, and open source software are great examples of stigmergy in humans.

[0] https://en.wikipedia.org/wiki/Stigmergy


Thanks for the reference. Definitely an interesting read. Knew about the pheromones, but not about Stigmergy.


My daughter's suggestion: we should ally with the octopi.


Octopi are really interesting creatures. They're said to be about as smart as a three- to four-year-old human.

The disturbing part is, that's how long they live. What would a hypothetical twenty-year-old octopus get up to?


A twenty-year-old octopus would either be:

Switching majors, chasing tentacle and hanging in the quad

-or-

Getting their first $1B valuation as cofounder of an octocorn


Can't believe you went with "octocorn" when "unipus" was there for the taking :D


or "octoporn"?


Thank you i needed that laugh today :-)


Your question seems to imply their intelligence increases at the same rate as a human and keeps doing so until they die. That would be very strange if it were the case.

I'm guessing they reach their maximum intelligence well before they die, and then they plateau, like most other animals. So living 20 years would make it more experienced but not able to apply to MIT.


Some scientists think they are quite intelligent animal and they can easily learn. The problem is they have to learn everything by themselves. I read that octopus doesn't transmit any knowledge from one generation to the next one. The mother put some eggs in a cave, protect them and then die once the babies hatch out.


Don't some octopus die young too? Around 3-4 years for certain species.

Well not young but more their life span isn't that long.


Ok don't get why I'm getting down voted but I did recall right at least: https://en.wikipedia.org/wiki/Enteroctopus_dofleini#Lifespan...

3-5 years and they're the longer lived octopus.


I am sure it is yours :p


So have we discovered how memory is actually stored here?

Also, how quickly can we implant new memories? I would like a nice ego trip to Mars.


Easy there McBain...


Apparently they have some sort of proof-of-work system going on.



Thanks, this is amazing.


I wonder if there is some Sussman's Propagator model like biochemical process going on underneath, resulting in the collective memory.


It is not a "memory". It might be merely a result of a selection pressure on population of bacteria.


As long as the storage mechanism is not explained, I'll assume a statistical fluke.


"Using computational modelling, the scientists explained this phenomenon in terms of a combination of two factors. Firstly, salt stress causes a delay in cell division, leading to synchronization of cell cycles; secondly, survival probability depends on the individual bacterial cell's position in the cell cycle at the time of the second exposure. As a result of the cell cycle synchronization, the sensitivity of the population changes over time. Previously exposed populations may be more tolerant to future stress events, but they may sometimes even be more sensitive than populations with no previous exposure."


> Previously exposed populations may be more tolerant to future stress events, but they may sometimes even be more sensitive than populations with no previous exposure

The last caveat undoes the whole research paper; it's not really memory, but a timing issue. All they did was exploit the synced the cell division cycles: they might have chosen the 2-hour mark (or got really lucky) as that was when the cell-division cycle was most (or somewhat) resilient.

This is similar to saying "batching all network writes to only happen in the first 30 seconds of every minute may make your network more resilient to random 5-second disconnections between data centers, or it make make it more sensitive". This is obviously dependent on which part of the cycle the disconnection happens! Maybe I should write that paper...


Yes but can we say human memory is not also (at least in part) a product of timing? Conceivably memory is mediated by synchronizing activity of certain neurons, after all, phase coupled neuronal signaling[0] is an element of brain functioning. Perhaps the bacteria encode information about their environment using low-level phase coupling. Probably not like that at all, it's just a thought.

[0] http://cercor.oxfordjournals.org/content/24/1/81.long


Why? We discover many phenomenon that we don't have a mechanism of action for.


Can you give an example? From a real science, not psychology or economy.


It's difficult to find an example that is not yet solved, but some discoveries had no good explanations for a time:

LGM (Little green men): https://en.wikipedia.org/wiki/LGM-1

High temperature superconductivity: https://en.wikipedia.org/wiki/High-temperature_superconducti...

And depending of the details about what you consider an explanation, we can include gravity https://en.wikipedia.org/wiki/Graviton It probably involves a family of alleged particle that nobody had seen, but with a few approximations you will probably get General Relativity and with more approximations Newtonian Gravity. Anyway, why do gravitons interact with quarks?


I'll try to answer your last question ("why do gravitons interact with quarks") and the implicit one ("why hasn't anyone seen a graviton?") as well.

You yourself are a pile of mostly quarks (and gluons) whose behaviour with respect to the surface of the Earth is extremely well described by General Relativity. There are plenty of large astrophysical objects which are mainly quarks-and-gluons and they follow General Relativity (GR) exactly.

The non-gravitational behaviour of all these objects is (within measurability) exactly described by the Standard Model, and if we divide them up into constituent parts -- right down to subatomic particles -- the description remains accurate. The description of the gravitational behaviour of these objects should also continue as we divide them into ever smaller parts. Quarks feel all four fundamental forces, so they must interact with the respective force carriers.

Gauge bosons are exchanged between elementary particles in a gauge theory (like the Standard Model), and carry the fundamental forces. Gravitons are directly comparable with photons, which mediate the other long range fundamental force; both should be massless because they are long-range. Classical light waves have spin-0 symmetry, and photons are spin-0; classical gravitational gravitational waves have spin-2 symmetry, so a quantization of them must preserve that.

The simplest graviton extension to the Standard Model works very well when gravitational effects are weak, but fails when gravitational effects are strong, largely because gravitons are self-interacting (unlike photons, at least at tree level).

This seen in classical theories: the Maxwell equations are linear; the Einstein Field Equations are not.

Schrödinger's equation is linear; quantum gravitational equations almost certainly will not be.

Weinberg's dimensional power counting of Feynman diagrams works for renormalizing field content described by Schrödinger's equation; that approach doesn't work for the gravitational field content for diagrams with more than one loop of gravitons.

Directly detecting an individual graviton will be difficult; you're interacting with enormous numbers of gravitational waves (even more than the number of light waves you are interacting with), but unlike the photon, the individual effect of each graviton is too weak to detect with anything close to current technology. (see for example http://arxiv.org/abs/gr-qc/0601043 )

Finally, we might arrive at another theory of gravity that doesn't have gravitons, although that would be a surprising result.



Maybe this is the mechanism behind homeopathic putrid water memory.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: