Slashdot Mirror
RNA May 'Run' Genetic Coding
leonbrooks writes "First a Stanford Medicine Magazine article speaks about RNA 'produced by plants that turn genes on and off', and now a Science Magazine issue says 'For a long time, RNA has lived in the shadow of its more famous chemical cousin DNA and of the proteins that supposedly took over RNA's functions in the transition from the 'RNA world' to the modern one. The shadow cast has been so deep that a whole universe [of RNA] has remained hidden from view, until recently' and speaks of 'an order of magnitude more transcripts than genes', suggesting that more actual coding is done through RNA than DNA. Is everything we know about genetics off-base? (no pun intended)"
This is how science evolves. One theory revises another. At least they're willing to say they they were wrong, unlike hundreds of years ago.
No existe.
ok as someone that works in this field let me say this:
RNAi is a very useful tool, but this is definitely several years behind the curve. RNA has been shown to regulate much more than previously thought. However talk about "the secret world of RNA" is pretty much like claiming that there is a "secret world of open source software." Neither one is very secret or very new.
The biggest contention I have is this quote from the article: "This knack of completely eliminating a protein makes RNAi a valuable research tool." This is wrong, because RNAi does not work like this at all. This is actually one of the drawbacks to using RNAi to eliminate proteins. It does not eliminate, it reduces. To get rid of a certain protein, the classic method is to completely remove the DNA that codes that protein from the organism studied. This is referred to as a "knock out" because the organism has no ability to make that proteind from the removed DNA. RNAi however, provides only a "knock down" because the DNA is still there and no matter how much RNAi is used there is still some expression of the protein. Also, many RNAi protocols are transient supressors not permanent knock outs of protein.
So basically this is an exciting new field but don't necessarily believe all the hype because this is no miracle answer. The article is good, but oversimplified.
It's worth noting that the field of "genetics" precedes even the identification of DNA and RNA. It may be that what we now know about gene regulation is wildly incomplete (although even that is unlikely, although possible) but Mendelian genetics is completely agnostic as to whether "genes" are protein-coding or not.
What I'm listening to now on Pandora...
The machinery which is doing the splicing and dicing is effectively a form of coding, too, even if it doesn't have its official coding license from an appropriate scientific authority.
DNA isn't the only thing inherited; in the case of mitosis, you get half a cell packed with stuff, and in the case of sexual reproduction the egg constitutes a whole cell packed with stuff. Complicated stuff, including RNA.
Its interesting that you mention HCV because that is exactly what my lab is working on. The problem with RNAi is that there no effective delivery method for humans.
As someone who has thought about this seriously, I basically agree with much of what you said, but it needs clarification. A basic reference is "Darwin's Dangerous Idea", by Daniel Dennett, and his basic slogan is "Cranes but no skyhooks".
One basic idea is that hereditary variation occurs on many different time scales simultaneously, e.g. cutlural (mimetic), epigenetic (DNA methylation), as well as regular DNA mutations which themselves fall into several classes each with a different frequency of occurrence ( simple sequence DNA, duplications, point mutations). My own thinking is that this non-uniformity of time scales of hereditary variation, when viewed in retrospect, can appear pseudo-Lamarkian, i.e. locally goal directed.
This multiple time scale search is a known strategy of search in certain genetic algorithms, and in some circumstances speeds up the search considerably. So in some sense one could regard Darwinian evolution itself as a process of distributed design (Dennett), with the pseudo-Lamarkian "goals" and different time scales of search partially giving the retrospective appearance of intelligence.
Whence we reconcile the major apparent discrepancy between darwinian evolution and "intelligent design" (as it would be understood by a rational person).
Not so many realize this, so I'm glad to meet you.
OK, I think I see where this is going and where the confusion might lie.
The grandparent (and myself) wanted to point out that mRNA (messenger RNA) is a very old discovery. It was found back in the 1950s. Back then, no one knew how the information got from DNA to protein. Elliot Volkin, among others, found that there was a rapidly renewing (degraded/resynthesized) DNA-like RNA product in the cells. Meanwhile, at the Pasteur Labratories, Jacob and Monod were building a model of the lactose operon that was perfect except for the absence of "X", a short-lived, rapidly renewing intermediate between DNA and protein.
An informal conversation, a Eureka moment and a Nobel prize for Jacob and Monod soon followed.
So, mRNA has been around forever. But, what we're learning about now are the various methods of controlling mRNAs and mRNA levels besides transcriptional (making more/less). For instance, these new siRNAs (small interfering RNAs) and the process of RNAi (RNA interference) are apparently important during, at the very least, development (by turning off genes quickly and specifically) and may be adaptable for medical purposes.
It's now become a hunt to find these short (21-23 nucleotide) sequences, the genes they control and how the whole process occurs, so that we might be able to predict how and why RNAi works.