The reanimated zombie gene

August 19, 2010 12:42 PM

Noncoding "junk" DNA is a signature part of the genomes of eukaryotes. Scientists have now identified a case of such DNA causing a genetic disease (Facioscapulohumeral Muscular Dystrophy) in certain genetic backgrounds by stabilizing the messenger RNA of a gene.
posted by jjray (20 comments total)

It would be nice if people would stop saying "junk DNA."
posted by rxrfrx at 12:56 PM on August 19, 2010 [9 favorites]

"Comparative genomics reveals that some regions of noncoding DNA are highly conserved, sometimes on time-scales representing hundreds of millions of years, implying that these noncoding regions are under strong evolutionary pressure and positive selection."

Junk my ass.
posted by Orange Pamplemousse at 12:56 PM on August 19, 2010 [2 favorites]

Not to sound patronizing, but that "junk DNA" has function is not new news. It's very likely that there is an absurdly large amount of junk DNA in the human genome, but we are constantly finding that certain parts we thought were junk are not.
posted by battlebison at 1:05 PM on August 19, 2010 [1 favorite]

Obviously, the next step is to genetically engineer mice with their exons excised. They look perfectly normal, I know, but you haven't seen them staring at me from the cage, cleaning their whiskers thoughtfully. Perfectly healthy, every one I've sacrificed. I've rummaged through their rodent guts for hours looking for gross anatomical differences and nothing, nothing but for the way they look at me so calmly even as I give them the ether.

Mice born without souls.
posted by adipocere at 1:14 PM on August 19, 2010 [9 favorites]

I detest the term "junk DNA" but I wonder if the sequence has any utility in normal cell functioning? From my reading of the abstract it seems like the cell normally suppresses this gene product, making it more akin to toxic DNA.
posted by Hutch at 1:23 PM on August 19, 2010

Here's a metaphor for you: Noncoding DNA may be like rests in music. Literally, they do nothing - but without them, the music is just garbage.
posted by Xoebe at 1:23 PM on August 19, 2010 [2 favorites]

Another cool example of a "junk" piece of DNA with an important function.
posted by Hutch at 1:27 PM on August 19, 2010

Can we just say "introns"? Just because we haven't figured out what it does now, doesn't mean we wont figurre it out in the future.
posted by hal_c_on at 1:34 PM on August 19, 2010

I wish I could get my hands on the full article, because the Science Daily article makes the mechanism sound different:

That's because the DNA that codes for the gene is not as tightly coiled or elusive to the body's molecular machinery as usual when some copies are missing, and so the gene -- known as DUX4, which makes a protein harmful to muscle cells -- is more active than it should be.
...
It turns out that each of the D4Z4 repeats contains a copy of a gene known as DUX4, but scientists have not known until recently that DUX4 is a functional gene. When a critical number of copies are missing, the structure of the tip of chromosome 4 becomes more open, making the DUX4 gene more accessible for transcription.

When crucial pieces of DNA that introduce and conclude the repetitive string are composed of certain sequences, the ingredients for molecular mischief are in place, making the remaining copies of DUX4 much more stable than they normally are.

This doesn't sound like non-coding DNA stabilizing mRNA, but the number of D4Z4 repeats affecting how often the DUX4 gene gets transcribed. Fewer repeats means more DUX4 transcription. More DUX4 transcription means more DUX4 protein, leading to the dystrophy disease state.
posted by Mercaptan at 1:41 PM on August 19, 2010 [2 favorites]

Mercaptan, I had trouble with the descriptions of the mechanism as well. I think when Science Daily says "crucial pieces of DNA that introduce and conclude the repetitive string are composed of certain sequences, the ingredients for molecular mischief are in place, making the remaining copies of DUX4 much more stable than they normally are" they are referring to SNPs that stabilize the RNA (by promoting proper polyadenylation of the transcript). I assume these SNPs are only a problem if you have few enough repeats for DUX4 transcription to actually occur. You need both conditions (transcript stabilizing SNPs and fewer repeats) to occur to produce FSHD-causing DUX4 levels.
posted by Hutch at 1:52 PM on August 19, 2010

hal_c_on: "Can we just say "introns"? Just because we haven't figured out what it does now, doesn't mean we wont figurre it out in the future."

There are noncoding exons too, which make them just as junk-y as the noncoding introns.
posted by battlebison at 1:53 PM on August 19, 2010 [1 favorite]

Can we just say "introns"?

No, intron means something more specific, not the same as all noncoding DNA.

This doesn't sound like non-coding DNA stabilizing mRNA, but the number of D4Z4 repeats affecting how often the DUX4 gene gets transcribed.

From the abstract:
Transfection studies revealed that DUX4 transcripts are efficiently polyadenylated and are more stable when expressed from permissive chromosomes.

So it's the stability of the mRNA that improves, resulting in higher effective gene expression.
posted by jjray at 1:56 PM on August 19, 2010 [1 favorite]

Well, shoulda refreshed.
posted by jjray at 1:57 PM on August 19, 2010

Can we just say "introns"? Just because we haven't figured out what it does now, doesn't mean we wont figurre it out in the future.

There are noncoding exons too

Introns refer specifically to bits of coding sequence that are spliced out. This term wouldn't cover promoters, enhancers, silencers, pseudogenes, lincRNAs, ribozymes, retrotransposons, and gene deserts (notably, despite the name, gene deserts have been implicated in human disease -- the link is to an open access PLoS Genetics article).
posted by en forme de poire at 2:04 PM on August 19, 2010 [3 favorites]

Here's the abstract. To quote from the last paragraph of the paper:

Our study puts forward a plausible, genetic model for FSHD. In this model, two polymorphisms create a polyadenylation site for the distal DUX4 transcript, located in the pLAM sequence. In combination with the chromatin relaxation of the repeat, this leads to increased DUX4 transcript levels. FSHD may arise through a toxic gain of function attributable to the stabilized distal DUX4 transcript.

So Science Daily is partly right - you need a reduced number of copies of the DUX4 repeats to allow relaxation of the chromatin, presumably to allow transcription, but you also need to have the polymorphism which changes a region just downstream of the last repeat to match the DUX4 polyadenylation signal, so that the mRNA is stabilised rather than being rapidly degraded.

(On preview: Hutch got it already. My excuse is that I was battling with my university login so I could read the damn paper.)
posted by penguinliz at 2:10 PM on August 19, 2010

Hutch, jjray, penguinlliz: Okay that totally makes sense.

I had a mental image of the mRNA being stabilized because it bound to this non-coding region. Which would be pretty cool too.
posted by Mercaptan at 2:18 PM on August 19, 2010

Mercaptan: "This doesn't sound like non-coding DNA stabilizing mRNA, but the number of D4Z4 repeats affecting how often the DUX4 gene gets transcribed. Fewer repeats means more DUX4 transcription. More DUX4 transcription means more DUX4 protein, leading to the dystrophy disease state."

From reading the paper, I don't think this is correct, either. It's not the number of repeats, but where the repeats are located, the presence of specific mutations, and the stability of transcripts.

What is happening is that there is a set of SNPs (single base polymorphisms, which are base-pair-level differences between healthy patients and patients with muscular dystrophy) that is located outside this D4Z4 repeat region.

These SNPs may cause chromatin remodeling, changes to how DNA packs and unpacks itself, which in turn causes changes to the number and position of the repeats, relative to non-coding regions.

In the specific region or "locus" where repeats can cause the disease, those repeats are transcribed to mRNA. The transcripts are all unstable and are not made into protein, except for the transcript made from the last repeat.

The changes to DNA structure cause the transcript of the last D4Z4 repeat unit to gain a poly-A tail, a string of bases which stabilizes mRNA and allows subsequent translation to protein. The production of that protein leads to the disease state.

In general, without polyadenylation, an mRNA strand is unstable and is not translated to protein.

The "junk DNA" here is the region distal to the repeats, which contain two specific polymorphisms that affect DNA structure and lead to stabilization of the intermediate transcript.

The authors did classical genetic experiments along with so-called "in silico" (computational) analysis to isolate the pathway leading to the disease state, and to confirm the mechanisms along that path.
posted by Blazecock Pileon at 2:43 PM on August 19, 2010 [2 favorites]

Which all sounds reasonable, Blazecock. Again, I really need to get a hold of the complete paper. Stupid Georgia Tech and their lack of access to Science Express articles.

I'm beginning to feel like everyone's trying to describe an elephant to me and all I'm coming up with is "a big upside-down squirrel".
posted by Mercaptan at 2:57 PM on August 19, 2010 [2 favorites]

An enormous fraction of the human genome is comprised of transposable elements, sequences which encode the machinery (or hijack the machinery) to cut and splice themselves to new locations in the genome. Just one class of these, the Alu sequence, comprises around 10% of the human genome.

If these elements were strictly harmful, we would expect them to be selected against. Against this hypothesis stands the observation that evolution in classes of mobile elements has occurred among primates throughout our evolutionary history.

These elements may play an important role in biological development (and perhaps learning) as their ability to knockout genes (e.g. by splicing into them) encourages diversity and evolution in somatic cell lines. This has been studied in the context of the brain, where researchers discovered an astounding degree of genetic diversity and mobile element insertions among neurons.

I think that the link between the evolution of cell lines and learning is almost self-evident, but wouldn't disagree with anyone who suggests that more research is required to understand the importance of these (junk) systems. What I find fascinating is that the presence and behavior of mobile elements implies that we are much more fluid genetically than the popular, crystalline, interpretation of genetics would suggest.
posted by melatonic at 8:33 PM on August 19, 2010

As been stated earlier, the phrase "junk DNA" needs to go away, it was useful back when Phil Sharp introduced introns. It now is clearly wrong. There are a host of things going on in non-coding regions, we just didn't have the tools to understand what before. In a crude sense the expressed portion of our genome is not vastly different than that of worms. What is different is non-coding regions, which have been postulated to contains various elements of regulation, chromosomal, miRNA, etc. Once we have a better handle on those mechanisms we may have yet another biological renaissance of diseases cured.
/off soap box
/just my 2cp
posted by oshburghor at 7:18 AM on August 20, 2010