DNA Not The Same In Every Cell Of Body

July 16, 2009 4:13 PM

DNA Not The Same In Every Cell Of Body."...calls into question one of the most basic assumptions of human genetics: that when it comes to DNA, every cell in the body is essentially identical to every other cell... if it turns out that blood and tissue cells do not match genetically, these ambitious and expensive genome-wide association studies may prove to have been essentially flawed from the outset"
posted by GuyZero (49 comments total)

Wow. I wonder how broadly that applies—does it only matter for individually tailored medications, or does it throw DNA identification into question as well?
posted by klangklangston at 4:19 PM on July 16

Yeah, whoa. Do we fling wide the prison doors now or what? And how many genetic researchers are watching their careers flash before their eyes over this? Questions.
posted by Devils Rancher at 4:22 PM on July 16

Not to mention that blood-based disease screening may be a lot less useful than hoped.
posted by GuyZero at 4:23 PM on July 16

Well, that is definitely the major consequence, and explains why blood-based genetic research has been so much less successful than we hoped it would be 10 years ago. The reason we haven't found the cause of every disease already is precisely because the genetics behind disorders are usually far more complicated than a basic cause-effect relationship, and this finding is one reason why.
posted by mek at 4:28 PM on July 16

Fortunately this will have no effect on my project to fix and identify every software bug by grepping through machine code.
posted by fleacircus at 4:33 PM on July 16 [10 favorites]

Do we fling wide the prison doors now or what?

Wouldn't it be the opposite? I mean, if you found a match, it's still almost impossible that anyone else was the owner of that DNA.

On the other hand, _not_ finding a match could be much more inconclusive than it's generally thought today, if the DNA can be different from the same person.
posted by wildcrdj at 4:33 PM on July 16

Crazy. I cannot fathom how this will change our understanding of the building blocks of life. Seems we learn something new and revise our understanding of the world all the time.

I'd say this is an exciting time to be geneticist, if anything.
posted by defenestration at 4:39 PM on July 16 [2 favorites]

The article is in the "Science News" section of "Science Daily" and yet there's zero science information in it. "Major differences" equals what exactly?
posted by DU at 4:40 PM on July 16 [7 favorites]

And how many genetic researchers are watching their careers flash before their eyes over this?

From what I know (which, admittedly, is very little), I would guess that this isn't coming as a huge surprise to many genetic researchers. Several years ago, I attended a talk by Margaret Lock, who studies genetic researchers among other things, and she told us that there is a fair amount of stuff that they were researching that is going to start changing the way that we think about the relationship of genes to whole, living organisms. My impression is that at the research level, there are a number of small findings that are chipping away at basic assumptions about how genes work.
posted by carmen at 4:42 PM on July 16

The article is in the "Science News" section of "Science Daily" and yet there's zero science information in it. "Major differences" equals what exactly?

I have the same question. I think it's way, way too early to start applying this very focused research finding to every single current use of DNA testing. From my cursory reading, they're talking about using DNA markers to identify risk factors for diseases, which is different than using DNA to identify individuals or parentage or what have you.
posted by muddgirl at 4:44 PM on July 16

That's some interesting shit. Abstract.
posted by Dr. Send at 4:45 PM on July 16

This "Human Mutation" journal, is it good? Google and Wikipedia don't tell me much about it.
posted by Dumsnill at 4:46 PM on July 16

Whoa, whoa, wait a second.

First off, we've known for some time now that the DNA is each cell can be different, at least in terms of epigenetics. We've also known that somatic mutation is out there in non-cancerous disease, e.g. the JAK2 V617F mutation. And everyone who's watched CSI knows about chimerism, where a person has different sets of chromosomes in different parts of the body.

That said, the authors haven't really done anything to explain what they found. Why didn't they sequence the genomic DNA from the aneurysms, and rely on cDNA only? (I think they address this in the discussion, but it wouldn't have been that much more work to do this...) Why didn't they sample some normal aorta from the patients, as opposed to relying on normal aorta from other sources? I've seen this operation done, it wouldn't be a big deal to get a little piece of normal aorta while you're fixing the aneurysm.

And seriously, Human Mutation? You'd think this would at least make it into AJHG.
posted by greatgefilte at 4:54 PM on July 16 [12 favorites]

Might it just be the case that diseased tissue has different DNA from blood cells? If you're looking at abnormal tissue, you might expect to see abnormal DNA.
posted by empath at 4:55 PM on July 16

This may be a better like to the abstract as Dr. Send's link doesn't work for me.
posted by GuyZero at 4:55 PM on July 16

one study does not a paradigm shift make.... I want to hear some of the reactions from researchers working in the genetic field!
posted by hector horace at 4:58 PM on July 16

Does this have anything to do with Americans marrying Italians and Irish?
posted by Saxon Kane at 5:03 PM on July 16

My impression is that at the research level, there are a number of small findings that are chipping away at basic assumptions about how genes work.

Thomas Kuhn approves of this comment.
posted by YoBananaBoy at 5:03 PM on July 16 [7 favorites]

I read an article recently that showed there is a higher than expected number of genetically multiple individuals, than previously thought. Sometimes it is from in vitro fertilization, or the absorption of a fraternal twin.

Hereis an article about that thing. I will try to find the original work I saw some time ago.

I think the original article I read, came from Metafilter. The scientist discussed that there were many people living today, that have two different blood types, and some have entire organ systems, formed from different DNA when fraternal twins merge in utero.
posted by Oyéah at 5:09 PM on July 16 [1 favorite]

I was going to post how I couldn't believe that after decades of DNA research scientists discovered this only now?!...but on preview saw greatgefilte's informative post.

So, for scientists it isn't really "one of the most basic assumptions of human genetics" at all then?
posted by Glee at 5:09 PM on July 16

Well DUH! Cells are created and destroyed every second in our bodies. Inside many of those cells is a nucleus. Inside the nucleus are chromosomes...Each chromosome has TONS of genes. Gene replication isn't always perfect. BUT THERE IS A TEMPLATE!

THAT is why we can continue to do gene science, and not worry about mutations that occur in almost all the cells.

I'd be hard pressed to find a geneticist who can tell me that they can pluck a cell from X and another from Y in the same person and have the DNA match 100%. That just doesn't occur.


Its just sad that since no really reputable source is publishing it..."Human mutations" is publishing it.

Wait about 2 days. It will be all over CNN for the people that don't really like thorough news.
posted by hal_c_on at 5:25 PM on July 16 [2 favorites]

I read an article recently that showed there is a higher than expected number of genetically multiple individuals, than previously thought. Sometimes it is from in vitro fertilization, or the absorption of a fraternal twin.

Ya, but Tyler Hamilton just tested positive for a steroid too, so I still don't think it counts as a defense.
posted by Chuckles at 5:29 PM on July 16

Published in Vague Scientist.
posted by DU at 5:31 PM on July 16 [12 favorites]

I'm guessing that this applies heavily to this article.
posted by Midnight Rambler at 5:31 PM on July 16 [2 favorites]

posted by defenestration I'd say this is an exciting time to be geneticist, if anything.

I'd say this is an exciting time to be a criminal defense attorney.
posted by mattdidthat at 5:35 PM on July 16

OK, for the benefit of y'all, I have to explain further why this is a crappy paper. There may be something interesting underneath it all, but the way it's been couched is just horrible. (No offense, GuyZero, at least as long as you weren't one of the authors. ;)

Here's what they did. They got 31 samples of abdominal aortic aneurysms (AAA), those little bubbles in your aorta that can rupture when they get big and are usually due to a combination of heredity and atheroscleosis risk factors, like smoking. From the same patients, they also took blood samples. Then they got five samples of normal aorta from a tissue bank, without matching blood samples (obviously).

Then, they picked one gene -- BAK1 -- that's been associated with AAA (through the very same genome-wide association studies that are now "in trouble!") and said, "Hm, what if the BAK1 gene in these people's blood is different from the BAK1 in their aneurysms? A good question, to be sure.

They extracted DNA from the blood, and extracted RNA from the aneurysms and the normal aortas, which they converted to DNA. They sequenced everything. Here's what they found:

In the blood samples from the people with AAAs, everything was kosher. In the aneurysm samples _and_ the normal aorta samples, they found some genetic changes, changes which are present at a very low frequency in the population in general, and which are generally thought of as harmless variations. From this they conclude that the changes must have (a) happened independently of the rest of the body and (b) have something to do with the aneurysm.

Now, the problems:

1. Their methodology didn't really match what they were trying to do. If they wanted to show that the DNA in the aneurysm cells are different, why did they extract RNA? Like a game of broken telephone, sometimes things change when the cell makes RNA from DNA. Maybe these changes aren't present in the DNA, but only in the RNA. They actually address this possibility, but don't have any good way of excluding it.

2. Technical point. Why the hell did they do restriction digests on the DNA? No explanation given.

3. They messed up the identities of the changes they found. In one table, it's "CGC changes to CAC," and in another table, it's "CAG changes to CAA," which doesn't match the expected amino acid change that they notate.

4. Worst of all, they don't give any numbers! They don't say how many changes were found in how many of the aneurysms and the normal aortas. Did all of them have all the changes they found? Some of them? None? Were the changes equal throughout the tissue, or was there a mixing of signals? Very sloppy.

5. Blatant misuse of commas, too.

6. They fail to take into the account the chimerism possibility -- maybe the DNA in the aneurysmal tissue is actually derived from the other parent than the blood DNA. This, to me, would be the most interesting possibility -- that diseases that show up in wonky heritable patterns are because of chimerism.


posted by greatgefilte at 5:47 PM on July 16 [29 favorites]

I'd say this is an exciting time to be a criminal defense attorney.

Not really, this has nothing to do with DNA fingerprinting, which doesn't depend on the sequence of the DNA so much as the number of times certain bits are repeated. It's already known that those bits (microsatellites) can be unstable, and the odds that all or even some of the markers would change is astronomically small.

Then again, it's always an exciting time to be a defense attorney, isn't it?
posted by greatgefilte at 5:51 PM on July 16

At first I thought this was about chimerism, but greatfefilte covers why this is a poor paper and even crappier mainstream reporting.

Comparing DNA from white blood cells to RNA, back to cDNA, from diseased tissue doesn't tell us a lot. I suspect either splicing differences (or since they're single nucleotide polymorphisms, changes in single nucleic acids which may lead to single amino acid changes), there could be a global defect in the transcription or error-checking machinery in those particular cells (which are diseased).

A brief check of Human Mutation's impact factor gives a 3-something, and press release saying how the editorial board is pleased that it's "now 6-something." Could just be a really prestigious but exceedingly niche journal (or since they work with human subjects, very incestuous) <shrug>. Never ran across it before, but that could easily because of my fields.

But yeah, that PhD comic linked upthread nails it, pretty much.
posted by porpoise at 6:09 PM on July 16

Haven't had a chance to check out the article... are the cDNA differences consistent with tissue specific RNA editing (ie. A to G for ADAR based editing)?
posted by Hutch at 6:12 PM on July 16

I also think that the most likely cause of their finding is different sequencing bias between the two techniques that they used for the two tissues. That's why this isn't in AJHG; it looks more like a screw up in their technique than a real finding. If this was true then they'd prove it by hybridizing a tagged RNA oligo and visualizing it. Since some of the substitutions are mis-sense, they could also extract the proteins and easily prove it with some easy blots.
posted by a robot made out of meat at 6:18 PM on July 16

This is not surprising. We have been hearing for years how a virus, gene therapy etc. can change DNA. Why would anyone believe that if these work that every cell has the same DNA?
posted by caddis at 6:19 PM on July 16

formed from different DNA when fraternal twins merge in utero

The sparrows are flying!
posted by maxwelton at 6:19 PM on July 16

Well this thread sure went 4 000 miles above my head in a hurry. I get dizzy toggling between tabs of this thread and Wikipedia pages explaining every third word you people write.

I now feel both more informed and less intelligent.
posted by Glee at 6:52 PM on July 16 [3 favorites]

Maybe I'm missing the point, but I don't understand the excitement. Of course the DNA of diseased tissues is going to be different, and the amount of difference vary enormously.

Take for example the brain tumor oligodendroglioma: in the diseased brain part of chromosome one and part of chromosome 19 are missing, but only in the tumor cells.

Another example could be somebody infected with HPV 16 or 18. The virus is incorporated in the DNA of vaginal cells, thus making that DNA different from the cells in the rest of the body.

In some leukemias, the DNA of the blood and bone marrow has been changed due to a translocation between chromosome 9 and 22, and the more the leukemia progresses the wilder the changes, with some chromosomes missing and others presents in multiple copies.

In some breast cancers there is amplification of the HER2neu gene, in neuroblastomas there is amplification the NMYC gene, in sarcomas there are translocations involving chromosomes 11 and 14.

There is a lot happening in diseased tissues, and a lot of the changes do involve changes in the DNA.
posted by francesca too at 7:05 PM on July 16 [1 favorite]

Wouldn't it be the opposite? I mean, if you found a match, it's still almost impossible that anyone else was the owner of that DNA.

On the other hand, _not_ finding a match could be much more inconclusive than it's generally thought today, if the DNA can be different from the same person.

Ahh, good point. I wasn't thinking right -- I'm cooking in my own juices down here in the third level of hell, and the brain doesn't function reliably above about 95º F. I blame global warming for my inability to parse concepts -- I wrote that while outside of the nominal operating temperature thresholds.
posted by Devils Rancher at 7:06 PM on July 16

Dammit DU, I came here to post that link to Vague Scientist!
posted by exogenous at 7:09 PM on July 16

5. Blatant misuse of commas, too.

Yeah!

This I can get behind, because the rest of that fantastic post flew over my head like an XR-71 flies over a Yak-drawn wooden-wheeled wagon on a rutted dirt path in Siberia.

Thanks a ton, greatgefilte, for bringing some actual science to the thread, though!

also. Metafilter: the absorption of a fraternal twin.
posted by Devils Rancher at 7:18 PM on July 16

I love science (used to do it) so I appreciate the differences between scientists and religious folks. Back in tha day, anything that wasn't coding sequence or in a promoter, or maybe and enhancer, was 'garbage'. All of a sudden, someone discovered siRNA etc. and the paradigm changed. All our assumptions were (kinda) wrong!

So how did scientists respond? "Cool!"

Leading to work such as that referred to in this post. How would religious people respond to some revelation that turned their system of beliefs upside down? I don't want to speak for them, but I suspect it would provoke much consternation and defensiveness. (Perhaps that's unfair, as, after a few scant centuries, they forgave Gallileo.)

Science wins.
posted by Turtles all the way down at 8:07 PM on July 16 [3 favorites]

I can't speak to the strengths or weaknesses of this particular paper but I will note that among working scientists, the credibility ofScience Daily is somewhere between the National Enquirer and the Weekly World News, as filtered through The Daily Mail.
posted by dogrose at 8:21 PM on July 16

There is a lot happening in diseased tissues, and a lot of the changes do involve changes in the DNA.

Likewise, in normal lymphocytes the genome does change too. V(D)J recombination occurs, giving rise to unique T cell receptors or immunoglobulins that recognize an extremely wide range of epitopes. If these recombination events do not occur, you'd be in big trouble.
posted by extramundane at 8:39 PM on July 16

They based their findings on cDNA instead of DNA? No wonder why it's only published in a journal nobody's ever heard of.
posted by emd3737 at 8:58 PM on July 16

Maybe I'm missing the point, but I don't understand the excitement. Of course the DNA of diseased tissues is going to be different, and the amount of difference vary enormously.

My guess is that the DNA changes in cancer are somewhat expected -- you get a natural-selection like process where cells that multiply faster, better, and/or stronger because of a mutation have a survival advantage, hence they into a tumour. But if one cell in your aorta gets a mutation that makes it, say, less likely to hook up to other cells and maintain the strength the wall of the aorta, how exactly does that change get passed on to other cells? I guess one possibility is that maybe the mutation occurs in some sort of stem cell or progenitor cell that eventually gives rise to the aortic wall. Maybe that's what they were trying to show when they checked the DNA of the normal aortas, and that would've been pretty neat had they not done a piss-poor job of experimentation.
posted by greatgefilte at 9:13 PM on July 16

BTW, for the biologically-inclined, here's an interesting, though somewhat older, paper on somatic mutations in diseases other than cancer.

There's also something really neat about the JAK2 V617F mutation mentioned earlier, the one that cause polycythemia vera (too many red blood cells, which can be a bad thing!). It's a somatic mutation, which means it happens spontaneously, and is not inherited from ma and pa. Yet, you can inherit a tendency to develop this particular mutation. Think about it -- out of a zillion letters of DNA, you can somehow inherit the ability to change one of those letters (for the worse, of course). I'd like someone to tell me how the hell that works.
posted by greatgefilte at 9:23 PM on July 16 [1 favorite]

But if one cell in your aorta gets a mutation that makes it, say, less likely to hook up to other cells and maintain the strength the wall of the aorta, how exactly does that change get passed on to other cells?

One mechanism that comes to mind would be through viral hitchhiking. A virus incorporates itself into a totally different chromosome in that particular cell that has a random mutation that would affect the aorta wall. It takes over cell replication in order to reproduce its genome: soon you would have a patch of mutated cells, not necessarily a tumor but a lesion. This is what happens to the cervix of the uterus when a woman is infected with the Human Papilloma virus. The body is able to heal most lesions caused by a lot of these viruses, but in some cases and with some viruses (HPV16 and HPV18 especially) the lesion evolves into a tumor.

Keep getting your yearly PAPs, ladies.
posted by francesca too at 4:08 AM on July 17

If I have a big stone sphere sitting on a big flat concrete parking lot, I could roll it around pretty easily, all things considered. Put it in a relatively shallow depression, though, and it takes a lot more work to move it. Put it in a foot deep hole I'm not moving it anywhere.

The JAK2 V617F mutation you describe greatgefilte may be like that. Most of us have a sequence that isn't going anywhere - the stone in a deeper hole, but for some of us that hole isn't as deep ergo things are more likely to get weird.

A lot of time it doesn't matter if you have a single base change - if you look at my favorite codon table for a while you'll realize that a lot of one letter changes don't matter at all. Or. as the pros say, are "silent mutations." A few change things, but it's a pretty conservative change so you would have a hard time finding the difference in the expressed protein* even if you were looking. And, sometimes, they try to kill you.

*I'm assuming protein expression because that's what I deal with day to day. YMMV.
posted by Kid Charlemagne at 5:36 AM on July 17

I'm sure my genetics lecturer would be appalled that I can't remember the figures but everyone walks around with shit loads of mutations in all different types of their cells every day. For the most part your body manages to catch them and sort it all out before the cell with the mutation has time to replicate and pass on the mutation. Or as Kid C. says some don't get fixed but are silent mutations so they don't cause any trouble.
The problem with inherited dispositions to particular diseases is that you already have a "bad" copy of one gene, and when your "good" copy is altered in one way or another and that cell replicates, all hell breaks loose. But then again it's only a predisposition so that doesn't mean it's going to happen, only that there is a higher chance of it happening.

Thanks to all who took the words out my mouth about why that was an "iffy" article.

Also on a side note it's threads like this that keep me reading metafilter. Cheers
posted by Gisela at 6:46 AM on July 17

Since I have nothing smart to say about genetics, let this suffice: in those Vague Scientist covers, that bit about cats in the army has cracked me up, what, seven times now, and I'm just going to do something else for a few minutes, then I'll click back and I will still laugh again. I mean, seriously:

ARMY CATS!

[picture of kind of chipper-looking cat in helmet]

Cats in the army

That's honestly the funniest thing I've seen all month. I may be a moron, but man am I easily entertained.

Army cats. Hehe.
posted by goodnewsfortheinsane at 7:58 AM on July 17

I may be a moron, but man am I easily entertained.

Me too. Those things may not be unrelated ...

And, Army Cat at The Onion.
posted by asusu at 8:51 AM on July 17

They were looking at cDNA in the AA tissue, and genomic DNA in the blood and gave a good rationale for that given the original intent of the study. So why didn't they go back and look at the genomic DNA in the AA tissue when they way this result? Or did they find RNA modification and ignore it because it wasn't as interesting a story?
posted by Mental Wimp at 12:38 PM on July 17