Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Rolepgeek]

Isn't chemotherapy just poisoning the body and hoping the cancer dies first? And that's sorta the whole reason it wreaks a terrible toll on the body and doesn't always work, but is still the best option available?

[MetalSlimeHunt]

Not exactly. Chemotherapy is primarily the usage of drugs that disrupt cellular mitosis. Cancer being what it is, this is pretty effective, but drugs don't discriminate and high-replacement areas can suffer from cell loss. In addition, cancers more typically enter remission than are eliminated, and you'd probably see the latter more in cases of surgical removal than chemotherapy.

[Reelya]

Too optimistic. Mutation rate is not that high, and most of the mutations would not be directed toward ubp integration.  Would be significantly lower than those values, and would need sustained environmental pressure to select mutants favorable to the desired outcome.

ecoli is one of the bacteria that passes a plasmid though, right? that could have interesting amplification effects.

But the thing is, being sub-optimal is itself a strong selection pressure. Selection pressures are only 50:50 when your organism is already near the optimum. If you put an organism in an unfamiliar environment, adaptation happens a lot quicker than trying to eke out a more optimal solution vs one that's already near optimal.

Also, Human-pig 'chimera embryos' detailed:
http://www.bbc.com/news/health-38717930
Here's a picture:

Spoiler (click to show/hide)

Additionally, new research shows that nicotine reduces schizophrenia symptoms:
https://science.slashdot.org/story/17/01/25/2042206/nicotine-shown-to-reduce-symptoms-of-schizophrenia

A meta-analysis of worldwide studies conducted in 2005 definitively showed what many doctors had been anecdotally noting for decades. Schizophrenia patients were much more likely to become heavy smokers than than those in the general population. In fact some studies found over 80 percent of those diagnosed with schizophrenia were smokers. There were many social and psychological hypotheses proposed to explain this strange anomaly, but none were ever sufficient. A new study published in Nature Medicine has not only revealed how smoking can normalize the impairments in brain activity associated with schizophrenia, but unlocks an entirely new field of drug research to combat the disease.

Basically society currently treat smokers like pariahs, which is a very puritan way of looking at things. But they've known (or suspected) for a few decades now that there are strong correlations between smoking and a number of mental health issues. People are self-medicating, basically, because as the new research shows conclusively, that stuff works. So it's basically wrong to point out "eww your a smoker. get self control" the same as it would be to point out fat people the same way. The trade off is that cigarettes cut 10 years off your life, but they stop schizophrenics from having the symptoms. Lose 10 years off your life but be sane instead of crazy? Actually that sounds like a rational choice that many people would take. Perhaps we should rethink what we think we know about people who smoke. If you know 1 person in 100 who can't or won't quit cigarettes, they're probably schizophrenic (whether or not they've been diagnosed).

[Sheb]

Reelya, what would be the selection pressure in this case? I mean, we're talking about the genetic code, the thing is so stable that it's basically the same in bacteria and in humans: aka it barely changed in BILLIONS on years. But you seem to think that a few decades of evolution of that SSO will lead to a change.

[Solifuge]

On the Cancer Treatment front, how about generic Universal Immune Cells that can be used to treat cancer? Seems London doctors engineered modified T-Cells that aren't seen as a threat by the bodies they're administered to (thus the same treatment can be administered to multiple patients), and which know how to attack even sneaky cancers. So far, they've successfully been used to treat Leukemia in 2 infants, where other treatments had failed.

Linko: https://www.technologyreview.com/s/603502/two-infants-treated-with-universal-immune-cells-have-their-cancer-vanish/

EDIT: Though they're not ready for widespread use, other similar techniques that require immune cells tailored to each individual is hopefully going to be available later this year. Even so, engineered generic immune cells have kicked off tons of hopeful research in treating cancers and diseases of all sorts. Exciting stuff!

[Frumple]

Can... can we call them U cells or something instead? I can't help but think putting T cells in our bodies to cure disease is tempting the dark gods of narrative a bit too much.

[ChairmanPoo]

They are engineered T lymphocytes. Hence T cells /CAR-T cells.

On that front: my umderstanding is that the one getting approved soonish is Novartis', which is patient specific...

[Reelya]

Reelya, what would be the selection pressure in this case? I mean, we're talking about the genetic code, the thing is so stable that it's basically the same in bacteria and in humans: aka it barely changed in BILLIONS on years. But you seem to think that a few decades of evolution of that SSO will lead to a change.

The selection pressure is specifically the fact that creating something that's deeply sub-optimal itself is the pressure. Actually the fact that basic systems in normal cells haven't changed in billions of years is actually the point. Those systems took a relatively short time to develop, then once they optimized they stopped changing. If you veer an organism far from equilibrium then the time to get that into an optimal range is far less than the billions of years your talking about. The sheer lack of change for billions of years is the point. Those systems didn't take billions of years to develop, they took a much shorter time to develop, then stopped changing once they were in an optimal equilibrium.

An example from real-world lab studies is some amoebas who somehow survived having a bacteria infection that should be fatal. The surviving bacteria were extremely weak and had a very slow dividing time compared to normal. But the researchers kept them alive. After about a month, they had amoebas who could breed about as fast as normal amoebas, but still had hundreds of the bacteria cells inside them. The cool thing was, these amoebas would now die if you removed the bacteria. They had become dependent on them being there. So yeah, no. Single-cell organisms can in fact adapt to things pretty darn fast. Generation time measured in hours or minutes and the fact that every individual cell is a unique competitor means the equation is completely different to the "evolution takes million of years" argument that would apply to humans. Think about two million-year old proto-humans compared to modern humans. if a generation is 20 years, that's 1 million generations. 1 million generations of a bug that splits every 20 minutes is a lot less, especially when the population size of a jelly plate of bugs is more than the historic total population of humans until relatively recently. Bacteria have both population size and rapid generations to provide feedback and have many, many different mutations tried at once. Also they have stronger competitive pressures than human-scale animals. Bacteria want to double every 20 minutes. So at the limit, you've got half your population being culled every generation / 20 minutes, or if not, extremely aggressive competition for limited food resources. So you ask where the selection pressure is coming from? From right there. Basically only half of any generation can possibly pass their genes onto the next one. Additionally, higher animals express high-level changes in phenotype. So you expect 2 million year old humans to have a certain amount of difference to modern humans. But bacteria are low-level organisms. They thus have more likelihood of low-level changes. Basically because there is no high level to be changed. So things like basic chemistry changes and dealing with biochemical threats/systems, that's where all the generational change is with bacteria, because that's the level they operate at.

[Frumple]

They are engineered T lymphocytes. Hence T cells /CAR-T cells.

On that front: my umderstanding is that the one getting approved soonish is Novartis', which is patient specific...

Cart cells would be good. We can go with that.

[Solifuge]

my umderstanding is that the one getting approved soonish is Novartis', which is patient specific...

Whoops! Did a quick reread, and edited the post accordingly. Still, I'm really interested to see how far we can push Engineered Immune Cells as a technology. I'd love to see widespread ways to teach people's immune systems to fight disease without needing individual vaccines, or more varieties of cells that are engineered to attack other hard-to-treat illnesses.

[Reelya]

BTW let me summarize my points about the synthetic DNA critters.

If you do something totally whacked-out to an organism, what's the likelihood that the results are anywhere near optimal? They're basically zero. Whacked-out big changes basically screw the organism, and mean the genome is now completely suboptimal.

If your organism is optimal, then a random mutation is very unlikely to make you a better organism. But that's not true of organisms that are far from optimal. Imagine a bacteria in which you'd engineered all the possible defects that make it worse, but leave just enough for it to stay alive. Say it's got 10000 defects which could be improved, vs an optimal bacteria, which has say 10 base pairs which could be improved. Given equal chance of mutations, the 10000 defects bugs has 1000 times the chance of a positive mutation compared to the 10 defects bug. So at that level, it will be picking up positive adaptations 1000 times faster than a "normal" bug that's already optimized. The rate of positive adaptations vs negative adaptations slows down as a bug approaches optimal performance.

Effectively, if you cobble something in the lab from spare parts, it's 100% comprised of defects, thus mutations have a far greater chance of being better than what it started with compared to a stable line of bacteria, probably thousands of times higher chance. you also see rapid change when organisms are introduced to a completely new environment, because their previous genome is no longer optimal.

[Sheb]

Great. So you have a selection pressure to get rid of the UBP (maybe by mutating out Cas9 or the guide RNA). Please tell me how that selection pressure will create 1) A new tRNA, B) a new AARS, C) incorporatet he UBP in the genome where it can do some good D) a new pathway to produce the new amino acid in quantities.



(Also, please, stop quoting the 20 minute doubling time when we're talking about sick bacteria like those, that's the absolutely optimal doubling time)

Don't get me wrong: long-term Lansky type experiment could make that bug fitter. It won't make anything interesting with the UBP on human timescale though.

[Reelya]

Great. So you have a selection pressure to get rid of the UBP (maybe by mutating out Cas9 or the guide RNA). Please tell me how that selection pressure will create 1) A new tRNA, B) a new AARS, C) incorporatet he UBP in the genome where it can do some good D) a new pathway to produce the new amino acid in quantities.

Not every base pair is a coding pair. The fact is, nature has a habit of finding novel uses for random things. What I wrote doesn't make the assumption that it's going to be the specific things you're talking about. A deficiency is often a trigger to come up with some novel way to workaround the problem, as in the amoeba example I gave.

But back to your main points. First you said it'll take "millions of years". But I countered that by pointing out that it's a parallel process. If a specific mutation that's needed for the next step is 1 in 1 million chance, then obviously, it's going to express in 1 cell per million in each generation.

But I really think you might be underestimating the speed at which parallel evolution can explore the search space of a genome. For example, insects that could digest DDT appeared a couple of decades after the stuff started to be used, and insects don't quite have the generation speed of microbes.

[Sheb]

You don't get the same degree of parallel evolution in bacteria because bacterial sex is much less efficient. Also, 1 chance in a million is ridiculously optimistic. The average rate of mutation in E. coli is 1 per thousand genome replication. Then you need that mutation to be one that does something nice.

But still, the point is that stuff like digesting DDT is relatively easy: you just need to tweak an enzyme. Same for the classical Lensky experiment: they manage to get E. coli to grow on citrate in a mere 20 years... But that involved just getting rid of the repression of one gene. This is much, much more complicated. The fact that you need specific Cas0 sgRNA just to keep one UBP in the plasmid means that you're unlikely to see that UBP popping up in other places (let alone the genome).

Full disclaimer: I actually worked in a lab for about 6 months where they did long-term E. coli evolution.

[Execute/Dumbo.exe]

They are engineered T lymphocytes. Hence T cells /CAR-T cells.

On that front: my umderstanding is that the one getting approved soonish is Novartis', which is patient specific...

Cart cells would be good. We can go with that.

Start doing that, though, and we lose the name 'Killer T-cells'

We can get some alliteration going on, though.
Not great alliteration, but 'Killer CAR-T' is an alright start.