Genetics Errors in Supreme Court Decision
Earlier today, my “in” box began to fill with info from everyone I’ve ever met letting me know that the Supreme Court had ruled on the Myriad case about patenting the breast cancer genes BRCA1 and BRCA2. I also received a dozen pitches from PR people offering me all manner of instant interviews with lawyers, doctors, bioethicists, and health care analysts.
No one offered me an interview with a geneticist – a person who knows something about DNA. So being such a person myself, I decided to take a look at the decision. And I found errors – starting right smack in the opening paragraph.
“Scientists can extract DNA from cells to isolate specific segments for study. They can also synthetically create exons-only strands of nucleotides known as composite DNA (cDNA). cDNA contains only the exons that occur in DNA, omitting the intervening introns.”
The definition is correct, the terminology, not. “cDNA” does not stand for “composite DNA.” It stands for “complementary DNA.” The document gets it right a little farther on. And I rather like the term “composite DNA,” but it’s confusing to introduce a new term in science when we already have one.
cDNA came into fashion when I was in grad school, circa 1977. Like many genetics terms, it has a very precise meaning, something I pay attention to because I write human genetics books, including 10 editions of a textbook. Changing a term in a textbook takes a lot of thought and feedback.
A cDNA is termed “complementary” because it is complementary in nucleotide base sequence to the messenger RNA (mRNA) that is made from the gene. Enzymes cut from the mRNA the sequences (introns) that do not encode amino acids and retains those (exons) that do encode protein. So a cDNA represents the part of a gene that is actually used to tell the cell to make protein.
A cDNA is created in the laboratory, and it is not a DNA sequence that occurs in nature, although its parts do. Hence, the Supreme Court’s part 2 of the decision, which acknowledges Myriad’s right to use a test based on a complementary, or cDNA.
I did a google search for “composite DNA” and just found the media parroting of today’s decision, and a few old forensics uses with what I think is a different meaning. So a caveat: my conclusion that the term is incorrect and invented is based on negative evidence. If I’m wrong, mea culpa in advance and I will feel like an idiot.
But cDNA isn’t the only questionable term. On page 16, footnote #8 discusses a pseudogene as resulting from “random incorporation of fragments of cDNA.” That’s not the definition I recall or use in my book.
A pseudogene in a classical sense results from a DNA replication error that makes an extra copy of a gene. Over time, one copy mutates itself into a form that can’t do its job. The pseudogene remains in the genome like a ghost of a functional gene. The mutations occur at random because the pseudogene, not being used, isn’t subject to natural selection. The globin gene locus on chromosome 11 is chock full of pseudogenes.
And how is the Supreme Court’s definition of a pseudogene supposed to happen, in nature or otherwise? A cDNA exists in a lab dish. A gene exists in a cell that is part of an organism. How does the cDNA “randomly incorporate” itself inside the cell? Jump in from the dish? Part of the footnote states, “… given pseudogenes’ apparently random origins … ” Pseudogenes’ origins aren’t random at all. They happen in specific genes that tend to have repeats in the sequence, “confusing” the replication enzymes. (See comments for corrections on this. I didn’t know that the viral versions are called pseudogenes too.)
Today’s decision is undoubtedly a wonderful leap forward for patients, their families, and researchers. It is long overdue. And some may think I am nitpicking, not seeing the forest for the trees. But these two questionable terms jumped right out at me — I’d troll for more but I want to post this. What else is wrong? How can we trust the decision if the science is not quite right? And what is the background of the people who research the decisions?
I know nothing about the law, zero, which is why I’m not writing about that. But the science in something as important as a Supreme Court decision should accurately use the language of the field under discussion.
There are(at least) two kinds of pseudogenes. In addition to the mechanism you describe, the SCOTUS opinion sounds like it is referring to the random integration of reverse transcription products, which is pretty common in vertrebrates, IIRC. Disclaimer: I work in E. coli, where we argue about whether we even have pseudogenes.
[…] in the lab, the cDNA version of the BRCA genes lack so-called “junk” DNA, the pieces that don’t contribute to the gene’s production of proteins. This technical difference, according to the ruling, makes the genes unique enough to be […]
[…] https://blogs.plos.org/dnascience/2013/06/13/genetics-errors-in-supreme-court-decision/ […]
[…] Just to note, several friends have noted that aspects of the science in the ruling seem to have some howlers. That is not surprising (see Scalia’s admission of […]
Thanks Jim, I didn’t know the random incorporation of transcription products was called a pseudogene.
I’m wondering what the implications of cDNA patents would be. If researchers make a cDNA library could they be violating various patents if their library includes certain genes?
The ruling is also some what bizarre given it ignores homology. Brca1 can’t be patented in humans, but homologous genes in other organisms can.
Good point. It is a little like the problem of licensing several genes in a multi-gene test panel.
Now that I’m looking at the opinion and not just your comment on pseudogenes, the first paragraph of the syllabus contains a common error that I try to correct for my students:
Not all exons are coding. The distinction is based on what sequences end up in the mature mRNA, not whether it is coding or not.
Note that the cDNA error is not in Justice Thomas’ opinion. He correctly calls it “complementary DNA”. IANAL, but my reading is that homologs are not patentable either. The cDNA claim is more interesting because it sounds to me like a specific cDNA is patentable, but you may be able to get around it by adding one nt to one of the primers. If not, then the key biology error is in this statement:
As noted earlier in the Thomas’ own opinion, there is a naturally occurring exons-only molecule with the same information content, it’s just not DNA: the mRNA.
[…] Genetic Errors in Supreme Court Decision – cDNA terminology is misused (cDNA conventionally means complementary DNA, NOT composite DNA as the SCOTUS decision reads) […]
Your points about the Court’s errors in science terminology underscore something that both lawyers and science writers need to do better. Whether they’re addressing judges or the reading public, both groups need to communicate scientific and technical information clearly and accurately. Doing so helps people make the right decisions–whether it’s ruling on the case, acting on their health, or understanding the intersection of science and public policy.
Having a legal background, I do think the Court reached the right legal result. See my blog entry today at
http://summacumlatte.wordpress.com/2013/06/13/patently-obvious/ .
Some pseudogenes, like the ones you describe in your post, are formed by duplication at the genomic level. Others, like the ones Jim describes are formed by reverse transcription and intetgration of mRNAs. Pseuodogenes of this second type lack introns and are called processed pseudogenes.
The ruling is quite clear and precise: you can’t patent genes, because they are naturally-occurring. You can patent cDNAs, because they don’t normally occur in the cell, but have to be artificially created in the lab by reverse transcription.
Pseudogenes are only mentioned because (some) pseudogenes arise through reverse transcription of mRNAs by endogenous retroviruses. Potentially therefore if you could extract naturally-occurring cDNA from a cell and sequence it, it would get round the patent on the cDNA sequence – but why bother to do that when you can (now) just sequence the genomic DNA without infringing any patents?
[…] writer Carl Zimmer linked to a blog post pointing out some factual errors in the […]
The court is referring to what’s known as a “processed pseudogene”. And these do in fact arise from the insertion of cDNA copies of a gene into the genome.
http://gencodegenes.wordpress.com/2012/02/15/processed-pseudogenes-junk-or-gene/
They’re often called processed pseudogenes to distinguish them from the ‘standard’ ones that you describe above.
One interesting thing about processed pseudogenes is that they aren’t a random subset of the genome’s genes. There are two main reasons for this:
First, only genes expressed in the germ-line can give rise to processed pseudogenes, so they’re enriched for reproductive genes. Second, genes with higher expression levels are more likely to have transcripts reintegrate into the genome.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2764836/
“Composite materials” are combinations of substances with different properties. Concrete is a composite of cement and sand/gravel. Cermet is a composite of ceramics and metals. Perhaps the Supremes ( a composite item itself) meant to indicate hybrid chireras- manufactured things such as: “frogmen”, aquamen”, or “birdmen”. This would allow defense contractors to patent new types of troops.
The SCOTUS has it right, although they didn’t use the correct full term. A “processed pseudogene” is a type of retroelement, in essence a cDNA created within a cell by reverse transcription of a cellular mRNA. Reverse transcriptase is encoded by every intact retrotransposon, and on occasion a cellular mRNA is erroneously reverse transcribed and, yes, randomly integrated. There are plenty of these in the human genome.
Hi,
Great news!
Regarding one of your comments:
I don’t know who wrote the syllabus, but its first paragraph is the only place where the “composite” error is found … even Scalia wrote the correct term: complementary DNA.
Best,
Antonio
They call it “composite DNA” in p. 1 and “complementary DNA” in p. 4, 6 & 22. Not bad for non-biologists 🙂
Yes, not bad overall, and of course I agree with the decision itself. Last week’s decision about the forensics was beautifully written. I wrote the blog today very fast under strange circumstances — I have to give the decision a better read. Thanks for your comment.
Thanks, yes, I read the decision fast under strange circumstances today — I had forgotten about the processed ones.
I like the term composite gene very much — it just stopped me in my tracks. I think the terminology should be consistent to avoid confusion in the media.
An error in your quotation of the decision:
You quote:
“Scientists can extract DNA from cells to isolate specific segments for study. They can also synthetically create exons-only strands of nucleotides known as composite DNA (cDNA). cDNA contains only the exons that occur in DNA, omitting the intervening exons.”
The actual Supreme Court states “introns” as the last word. Your quotation does not make sense.
Thanks, error from writing too fast.
So given an obvious mistake in the ruling… I’m wondering why Supreme Court decisions don’t have “erratum”
Just curious — is “nucleotide base sequence” redundant? Not a rhetorical question.
Yes, it is sort of redundant — but if you were referring to baseball, you’d have a different base sequence. “Base sequence” would have been better, you’re right. It is challenging to write for readers so varied in background. I usually run my genetics-babble past my husband, a chemist, but had no time for this one.
Yes they’re right about pseudogenes – the first gene I ever cloned was a pseudogene with no introns 🙁 – it was so long ago though that it even got published!
A question for someone – I’m wondering why Myriad stock went up 10% or more. What do they mean by cDNA – is it strictly related to that old fashioned stuff that we used to make from RNA with reverse transcriptase? Or ar PCR products also included in their definition of cDNA?
This seems to be an important point because to offer BRCA genetic testing most methods would have a PCR step and Myriad may claim they have that covered – do they?
This post seems to think that only NGS can get around it, http://bit.ly/12t1C8I
But this company has already started selling – http://on.mktw.net/14ACj42
I feel like both the courts and the genetic companies are completely missing the point.
DNA- cDNA, RNA, mRNA, tRNA, etc: these are all just the instruction set to make a biologically active molecule (proteins etc.). (Yes, I know RNA has some self-catalysis abilities, but I’m thinking of enzymes that do most of the work in living organisms. Or, the large scaffolding macromolecules that are the building blocks for cell membranes etc.)
If you design a totally new protein from scratch- or even a recombinant one- then you can patent that amino acid sequence. This avoids using something that is merely the instruction set all altogether, and sets the bar at the level where the “biological action” is taking place.
In any case, patents are supposed to be for a limited time only- enough to re-coup the investment costs and provide a reasonable return. After such a time, the idea ought of be released into the open market for free use. It feels like many companies nowadays are having difficulty with the ‘limited time only’ concept.
As a former physics student who now works in the software industry, I for one appreciate the extra clarifier — even if it is redundant. Particularly in this case, it speaks to a need highlighted in a comment upthread, for clear communication in science writing.
[…] may not be patented. But the text of the decision [PDF] is drawing complaints from people who know actual biology, and one justice admitted in his concurrence that he doesn’t understand the […]
[…] wanting to flex their nerd muscle, you can also read the whole SCOTUS decision here, though beware, as Ricki Lewis points out, not all the science is […]
Keith, can you or someone else who understands the law (which I don’t) explain how 23andme can sell BRCA tests? One of my bioethics students took the test herself when her mom tested positive, and one of my friends just took it. Do they license the 3 Ashkenazi mutations in a test panel from Myriad? Why doesn’t the media ever mention this fact? Through 23andme it is relatively cheap. And they are happy and quick to answer questions, at least from what my friend tells me.
Also, the oft-quoted $3200 or so price tag for Myriad’s test — isn’t that for the entire coding or gene sequence? If you are Ashkenazi with a family history of BRCA cancers (and they’re not just for breast and ovarian, another fact oft not reported) and you already know your family has a common mutation, you can be tested just for that mutation, and it is cheaper.
This is a highly nuanced situation that I think requires some background in genetics to thoroughly understand, and as you can see from my post, I write about genetics 24/7 and I made mistakes in terminology too! I think the decision should have been intensively reviewed by old and young geneticists, to get it absolutely correct. The media perpetuate genetics errors all the time — “human genetic code” for “human genome sequence,” for example. I could go on … thanks for responding.
From the comments at Volokh Conspiracy, AdamB responding to a comment about the incorrect composite vs correct complementary for the c in cDNA issue:
The BRCA test available covered only a few specific mutations – about 3 or 4 or so. The full BRCA examination covers many more mutations in these genes, well over 100. So the 23andme test is incomplete.
If you can get them on the telephone (which I did) they
will tell you this.
With this decision, hopefully they will offer a more complete testing option. Perhaps for $50-$100 rather than $3000.
[…] or “scientifically inaccurate” (Forbes) and filled with “genetics errors” (PLoS blog) and “ignorance” (New Republic). In this, they seemed to be building on the brief […]
[…] Ricki Lewis has more details on her blog about the science inaccuracies in the ruling. Unlike Lewis, however, I believe that if the ruling […]
[…] Ricki Lewis has more details on her blog about the science inaccuracies in the ruling. Unlike Lewis, however, I believe that if the ruling […]
Yes, I know that, 23andme covers the three Ashkenazi mutations. It isn’t really incomplete, because that and a few others are the population groups with the highest prevalence. As I wrote, if you are of this background and/or know what your family’s mutation is, testing the entire gene sequence is overkill. The risks are different in different population groups. The one on NPR yesterday was not only for the wrong cancer — they said 87% risk for ovarian — but the reporter did not state that this is the Ashkenazi risk, not everyone. (A few other groups are at high risk too.)
[…] Ricki Lewis has more sum on her blog about a scholarship inaccuracies in a ruling. Unlike Lewis, however, we trust that if a statute is […]
[…] Ricki Lewis has more details on her blog about the science inaccuracies in the ruling. Unlike Lewis, however, I believe that if the ruling […]
[…] homework, though the descriptions are complicated and heavy. Still, the science is mostly right. (Here are a few examples where their explaining goes […]
[…] A gene’s DNA sequence minus the non-protein-encoding parts (the introns) renders it patentable, argued Myriad, for it’s no longer a “product of nature.” The court agreed that this complementary DNA becomes a novel “composition of matter.” It’s called “complementary” because it corresponds to the messenger RNA sequence that excludes the introns, not “composite DNA” as in the initial version of last week’s decision. […]
I think most people has an unquestioning belief in the DNA test results.
Let me give you an example. Gary Leiterman is languishing in prison now because his DNA was present in the specimen (panty hose) of the murder victim Jane Mixer. However there was another DNA present, that of a John Ruelas, a convicted murderer. John Ruelas was 4 years old when Jane Mixer was murdered. The DNA of Gary Leiterman, John Ruelas, and Jane Mixer were processed in the same lab.
Most folks hear DNA and the word “guilty” pops up in their head. The presence of John Ruelas’ DNA in that specimen should have establised reasonable doubt by itself.