Return of the not-so-junky DNA
by KrauzeOn evolgen, "RPM" comments on Peter Andolfatto's recent paper in Nature (and he even links to my post about it. Thanks, RPM! It's nice to know that some aren't so prickly). Evolgen makes some good remarks about the use of "junk DNA", which have prompted me to try to make some sense of what is some really confusing terminology. But first, a crash course in molecular biology.
Genes are encoded along the DNA molecule, and when a gene is expressed, it's usually transcribed into a string of RNA (since there are several types of RNA, this type is called "messenger RNA", or "mRNA" for short). The RNA is then translated into a string of amino acids, which folds up and becomes a protein.
When scientists started to make sense of the genome, they noticed that large stretches of DNA didn't result in any proteins. It was assumed that these areas of the genome were useless to the organism, so they were termed "non-coding DNA" or "junk DNA". Looking for the protein product of a gene made sense at the time, but it suffers from the same problem as using a coarse-masked fishing net to measure the size of fish in a pond: If your masks are 2 inch wide, you won't see any fish smaller than 2 inch. Similarily, only looking for proteins means you'll miss those genes which carry out a function without resulting in proteins.
Scientists have in recent years discovered that some RNA molecules are used to regulate the expression of other genes - when and where they're turned on or off (see this post for more details). So we have a situation in which non-coding DNA can code for something, and junk DNA doesn't have to be junk. Like RPM, I prefer "non-coding DNA" over "junk DNA", as the latter implies that the DNA in question is useless. However, "non-coding DNA" can also be confusing, as I was reminded when a reader sent me a message asking how selection could work on a gene that was non-coding (i.e. wasn't doing anything). I suppose the most accurate name would be "DNA that doesn't encode protein, but might encode RNA", but my brain cramps up, just thinking of typing it. Anybody got a suggestion for a good acronym?
November 1st, 2005 at 12:17 pm
NPCF DNA- Non Protein Coding Functional DNA- with the function TBD
OK it's only my first attempt…
Comment by Joe G — November 1, 2005 @ 12:17 pm
November 1st, 2005 at 6:39 pm
Here's a thought. Let's say Coding DNA contains gene sequence whose products directly participate in the process of translation, as translation is built around the genetic code (amino acid sequenced conventionally encoded in a nucleotide language).
The other class of DNA would be Regulatory DNA and comes in two types - cis and trans. Cis-regulatory DNA would be DNA sequence involved in regulating gene expression (promoters, enhancers, etc.) and trans-regulatory DNA would express RNA that further regulates expression of Coding DNA (small RNA, etc.)
Comment by MikeGene — November 1, 2005 @ 6:39 pm
November 2nd, 2005 at 10:43 am
When scientists started to make sense of the genome, they noticed that large stretches of DNA didn't result in any proteins. It was assumed that these areas of the genome were useless to the organism, so they were termed "non-coding DNA" or "junk DNA".
I suspect that the concept of junk DNA is also derived from early observations of the C-value paradox, and that is still a pretty good reason to infer that a lot of DNA is non-functional. Either way, I don't think anyone uses "junk DNA" a simple synonym of non-coding DNA.
I really think you guys are overcomplicating the terminology. It shouldn't be confusing. In molecular genetics, "coding" simply refers to DNA that encodes an amino acid sequence, nothing more, nothing less. Since coding DNA is easily recognizable, non-coding DNA can be unambiuously be identified. The person who asked how selection could act on non-coding DNA just misunderstood the definition and could be easily corrected.
The term "non- coding" contains no implications about functionality or the lack thereof. Geneticists have known for decades that some non-coding DNA is transcribed, although that does not necessarily mean that the sequence is functional or, if functional, has a function that is beneficial to the organism. They have also known that some non-coding DNA is conserved among different species, strongly implying function. Again, I doubt that anyone other than journalists uses "non coding" and "junk" as synonyms. I don't see any particular value in coming up with yet another catchall phrase. Why not simply refer to the various types of functional non-coding DNA by their specific names?
Whether a particular piece of non-coding DNA is functional or nonfunctional can be inferred in a variety of ways.
Comment by Nick — November 2, 2005 @ 10:43 am
November 2nd, 2005 at 3:32 pm
Nick:
In molecular genetics, "coding" simply refers to DNA that encodes an amino acid sequence, nothing more, nothing less.
I take it you are referring to m,r,t RNA
Nick:
Since coding DNA is easily recognizable, non-coding DNA can be unambiuously be identified. The person who asked how selection could act on non-coding DNA just misunderstood the definition and could be easily corrected.
I believe what Krauze proposes it to prevent any confusion from happening- therefore no corrections required.
Nick:
Whether a particular piece of non-coding DNA is functional or nonfunctional can be inferred in a variety of ways.
That is true, however determining that function requires more research. And then we can properly label those stretches of non-coding DNA by the function they perform.
So I can shorten my original to- NCF DNA Non Coding xFunctional DNA.
Comment by Joe G — November 2, 2005 @ 3:32 pm
November 2nd, 2005 at 5:51 pm
I take it you are referring to m,r,t RNA
No
As I said, coding DNA refers only to that DNA which encodes an amino acid sequence. DNA that is transcribed into tRNA or rRNA is noncoding. mRNA contains coding sequence, but it is not coding in its entirety. The 3' and 5' UTR of mRNA is noncoding.
I believe what Krauze proposes it to prevent any confusion from happening- therefore no corrections required.
Simply paying attention to the existing terminology would go a long way to preventing the confusion. For instance, in his third paragraph, Krauze states that scientists use "non-coding" and "junk" interchangably. They do not. Krauze's confusion seems to proceed from that misunderstanding. See, for example, his comments about "a situation in which non-coding DNA can code for something." That situation simply does not exist, so there is no reason to generate new terminology for it.
That is true, however determining that function requires more research. And then we can properly label those stretches of non-coding DNA by the function they perform.
Or lack of function. Don't forget that possibility. The C-value paradox strongly implies that a lot of DNA in many organisms is non-functional, even if we cannot state with certainty whether a particular segment of DNA is functional or not.
So I can shorten my original to- NCF DNA Non Coding xFunctional DNA.
If you are talking about a sequence with a currently known function, then why not just call it what it is instead of a new vague category? If the function is currently unknown, then you also have to consider the possibility that there is no function. Since the "x" could refer to either function or lack of function, your new term simply reduces down to good old "non-coding DNA."
Comment by Nick — November 2, 2005 @ 5:51 pm
November 2nd, 2005 at 8:16 pm
Nick:
Since the "x" could refer to either function or lack of function, your new term simply reduces down to good old "non-coding DNA."
It's too easy to confuse "non-coding" with "no function", and then "no function" with "junk". Even if we do find some stretches offer no function whatsoever we may find these these stretches are actually resource "stockrooms" for DNA repair (for example).
Perhaps the scientists who work with this stuff every day can keep it sorted but for us engineers proper nomenclature avoids confusion.
(And thanks for clarifying about the coding- noncoding)
Comment by Joe G — November 2, 2005 @ 8:16 pm
November 2nd, 2005 at 8:40 pm
Glad to be of help. I still don't really understand why noncoding is confused with nonfunctional, but perhaps it has something to with the word "code." There is a genetic code. DNA that is translated according to that code is "coding" DNA. DNA that is not translated according to the genetic code is noncoding. So, perhaps the problem is that genetics has a specific meaning for code, but in general conversation, people tend to use a looser definition.
Comment by Nick — November 2, 2005 @ 8:40 pm