Saturday, December 08, 2007

Junk DNA in the Toronto Star

 
Cameron Smith is an an environmentalist who lives Gananoque, Ontario (Canada). He was written more that 500 columns for the Toronto Star. Most of them are about nature. His latest column in today's Idea section caught my attention [Getting to the root of plant life].

The column seems to be heading in the direction of promoting the "intelligence" of plants (see below). There will be a followup column soon, according to Cameron Smith.

The part that disturbs me is the following,
However, with their mapping of the human genome, they [molecular biologists] discovered that humans carry only about 25,000 protein-coding genes. This was startling, because the simple nematode worm has about 19,000 such genes – and the human body is immeasurably more complex than a worm's. So, why didn't humans have a lot more protein-coding genes – genes that instruct proteins what to do?
I've addressed this point several times [Facts and Myths Concerning the Historical Estimates of the Number of Genes in the Human Genome, SCIENCE Questions: Why Do Humans Have So Few Genes?]. It's simply not true that all scientists were surprised by the number of genes we have. For many, this result was anticipated and it poses absolutely no problems in understanding biological complexity. There's no pressing need to look for some magic bullet.
To find answers, molecular biologists had to revise their notions of the genetic code. They knew that a huge number of genes in the human genome, making up more than 98 per cent of the genome, don't code protein. These they had previously dismissed as evolutionary leftovers, or junk DNA.

In an enormous turnaround, they began looking at these non-coding genes more closely and discovered they were not junk after all.
There's a lot of nonsense in those few sentences. The most important flaw is that the basic message is completely wrong. It is simply not true that molecular biologists have discarded the concept of junk DNA. The vast majority of molecular biologists know the facts; namely that >90% (probably more) of our genome consists of junk DNA. Just because there are a few renegade scientists who don't know any better does not make these facts disappear.
They had an extremely important function. A key to the mystery lay in the nature of complexity. There was no doubt protein-coding DNA was capable of creating complexity.

It could issue instructions for creating the legions of proteins that, in the case of humans, make up half their dry weight. But regulating the process was another matter. Without regulation, the results would be mostly chaotic.

In addition, as the complexity of organisms increased, the amount of regulation that was needed increased exponentially.

Regulation, it turns out, is the job of RNA (ribonucleic acid), located in the nucleus of cells along with DNA. It's from the so-called junk DNA that RNA gets regulatory instructions.

This revelation opened the intellectual floodgates, and put to rest the notion that life was ruled by a robotic DNA ritually coding proteins, much like a machine stamping out widgets.
Here are the scientific facts in a nutshell.
  • Scientists have known about regulatory sequences for at least fifty years. They were never, ever, thought to be junk DNA by any competent molecular biologist.
  • There is no evidence to support the half-baked notion that the amount of DNA sequence required for regulation of "complex" organisms (i.e., humans) is exponentially more than that required for regulation of "simple" organisms (i.e., nematodes). All available evidence shows that gene regulation in all multicellular species is very similar.
  • We've known about regulation by small RNA's since the 1970's. Nothing new there. There is no solid evidence to suggest that regulation of typical human genes requires RNA and much theoretical and experimental evidence against such an idea. Active imagination doesn't count in science. Scientists need real data before jumping to the extraordinary conclusion that human gene regulation is fundamentally different than other species.
  • One of the many reasons for accepting that only 2% of our genome is functional has to do with the concept of genetic load [Facts and Myths ...]. If junk DNA is full of genes encoding regulatory RNAs then we're in big trouble because mutation rates are going to kill us off pretty quickly.
Cameron Smith seems to be quite taken with the ideas of Anthony Trewavas at the Institute of Molecular Plant Science, University of Edinburgh (Edinburgh, UK). Trewavas has been writing papers promoting the idea that plants have a form of intelligence. He doesn't mean they have a brain, of course, what he means is that plants have a complex form of regulation that responds to the environment.

According to Trewavas, systems biology and computer modeling have revealed a level of complexity that scientists never suspected (Trewavas, 2006). Naturally, this totally unsuspected level of regulatory complexity has been hidden from those scientists who have adopted a reductionist approach to science.
Systems approaches enable plant scientists to understand the structural stability of plants, their control and design structure, and how these lead to robust and resilient behavior. These capabilities are the result of a complex biological system in which control operates at many different levels (Figure 1). Complexity is a serious biological problem, and it is likely that biological systems are the most complex known. Increasingly, scientists are going to have to depend on computational biologists to construct models that can then be tested back in laboratory conditions. However, as indicated here, laboratory conditions are only one environmental circumstance among many in which plant systems develop. In 10 years, my own estimate is that plant molecular research groups will be half modelers and half wet investigators producing new data for modelers.
I'm getting a little tired of this sort of rhetoric. Systems biology, properly defined, can be a very useful approach to a problem but turning it into a religion isn't going to help. I'm content to wait and see whether the systems biologists are actually going to deliver something (other than rhetoric) before jumping on that bandwagon.

Trewavas believes in the power of information theory (IT). This faith him to conclude that plants have a form of intellignece (Trewavas, 2005a, 2005b, 2003).

This idea of intelligent plants should not be taken literally. Trewavas clearly means it to be controversial and clearly understands that it is a metaphor. However, the concept is based on an false premise, in my opinion. The premise is that there is a complex sophisticated (and largely undiscovered) regulatory circuit in plants that allows them to behave as though they were responding to the environment in an intelligent way. I don't think we need to go down that path. Yes, plants can control gene expression, just like bacteria, but I see no value in exaggerating that control to the extent that Trewavas does.
From the current rate of progress, it looks as though plant communication is likely to be as complex as that within the brain. (Trewavas, 2003, p.6).



Trewavas, A. (2003) Aspects of plant intelligence. Ann Bot (Lond). 92:1-20. [PubMed]

Trewavas, A. (2005a) Plant intelligence. Naturwissenschaften. 92:401-13. [PubMed]

Trewavas, A. (2005b) Green plants as intelligent organisms. Trends Plant Sci. 10:413-9. [PubMed]

Trewavas, A. (2006) A brief history of systems biology. "Every object that biology studies is a system of systems." Francois Jacob (1974). Plant Cell. 18:2420-2430. [PubMed]

15 comments :

  1. Larry, perhaps you could write in to the Star with a response to Smith? Do you think they'd print it? Perhaps they wouldn't as they may not be interested in serving as a debate forum for an issue that is not of much direct political importance...

    ReplyDelete
  2. Urgh... all that talk about undefined complexity hurts my brain.

    Can somebody here who has worked with C. elegans please explain to me exactly which part of it is so simple? My brief glances at nemotodes have suggested that they are rather tricky things with layers of interesting features and all sorts of fun things going on.

    ReplyDelete
  3. Good analysis. On an unrelated note, something you might find interesting if you haven't already seen it; a report in Cell that suggests a possible revising of our understanding of antibiotic action. It suggests that antibiotics (all of them) may kill bacteria by producing free radicals.

    "A Common Mechanism of Cellular Death
    Induced by Bactericidal Antibiotics"

    Cell 130, 797–810, September 7, 2007

    Also see:
    ACS CHEMICAL BIOLOGY • VOL.2 NO.11 pg. 708

    ReplyDelete
  4. There is no solid evidence to suggest that regulation of typical human genes requires RNA and much theoretical and experimental evidence against such an idea.

    I hope you're not talking about the non-importance of microRNA here...

    ReplyDelete
  5. anonymous asks,

    I hope you're not talking about the non-importance of microRNA here...

    I certainly am.

    Some micoRNAs are real and have real biological functions. I doubt very much that those real microRNAs make up more than a small fraction of the genome.

    ReplyDelete
  6. Larry, you said:

    "We've known about regulation by small RNA's since the 1970's. Nothing new there."

    I for one would like to see some pointers that are pertinent to this claim.

    You also said:

    " There is no solid evidence to suggest that regulation of typical human genes requires RNA and much theoretical and experimental evidence against such an idea. Active imagination doesn't count in science. Scientists need real data before jumping to the extraordinary conclusion that human gene regulation is fundamentally different than other species."

    There is abundant evidence that posttranscriptional control, even in human cells, can be accomplished with small RNAs. And I don't quite know what to make of the implication that someone is claiming that regulation by small RNAs in humans in some way differs from the same in other organisms.

    I think you need to revise your original assertion, as it is making a claim that is wrong.

    As for Trewavas, I think it's as easy to see how he is tweaking common-accepted (if perhaps poorly thought out) notions of intelligence as that he's pushing some way-out nonsense.

    ReplyDelete
  7. Art says, referring to regulation by small RNAs,

    I for one would like to see some pointers that are pertinent to this claim.

    I don't have specific references at hand. When I used to teach this stuff in the early 1980's I used the examples of anti-sense RNA in bacteriophage lambda and the small RNAs that regulated the initiation of DNA replication at bacterial origins.

    We also covered attenuation in the trp operon.

    There is abundant evidence that posttranscriptional control, even in human cells, can be accomplished with small RNAs. And I don't quite know what to make of the implication that someone is claiming that regulation by small RNAs in humans in some way differs from the same in other organisms.

    I think you need to revise your original assertion, as it is making a claim that is wrong.


    I agree that there is good evidence for regulation of some human genes by small RNAs.

    But the claim is much broader than that. The claim that you see being promoted by some people is that regulation by small RNAs is widespread. Most genes are supposed to be regulated in this manner.

    In fact, the number of small RNAs involved in regulation accounts for a large amount of noncoding DNA in the genome, according to some. That's the claim that seems ridiculous to me.

    The article by Cameron Smith picks up on the idea that there's something special about mammalian genomes that can account for the complexity of mammals (humans) with only a few more genes than nematodes. One of the claims to specialness is abundant regulatory RNAs that aren't seen in other species. These small RNAs have to be much more important in mammals compared to nematodes otherwise the claim doesn't make sense.

    ReplyDelete
  8. I wouldn't be surprised for it to wind up that most human genes are significantly regulated by microRNAs. Note that this is a whole different beast from the 30+ year old antisense RNA field.

    Here's a database of microRNAs under current investigation:
    http://microrna.sanger.ac.uk/sequences/
    The idea that microRNAs are critical for both human development and pathogenesis is very recent and worth tracking. See for example:
    Let-7 expression defines two differentiation stages of cancer. Proc Natl Acad Sci U S A. 2007 Jul 3;104(27):11400-5

    That being said, microRNAs are certainly not going to make up a significant fraction of human bulk genomic DNA. I also don't understand the point of debating the relative complexities of humans and nematodes.

    ReplyDelete
  9. anonymous says,

    I wouldn't be surprised for it to wind up that most human genes are significantly regulated by microRNAs.

    You wouldn't? I'd be shocked.

    Most human genes are quite well conserved. They are present in all eukaryotes. Are you saying that sometime in the recent past (200 myr) most of these genes in mammals simultaneously acquired a whole new level of regulation not seen in their ancestors?

    What's the point? Why would there be such strong selective pressure to acquire a new form of regulation by micro RNAs?

    Are the standard form of regulation not good enough for humans?

    ReplyDelete
  10. anonymous says,

    I also don't understand the point of debating the relative complexities of humans and nematodes.

    There are a great many people—including, unfortunately, many scientists—who are upset that the "obvious" superiority of humans isn't reflected in their genome. That's part of the motivation behind the search for some special mechanisms that sets humans apart from the other species.

    The leading contenders for this "specialness" are:

    1. Lots of extra regulatory sequences hiding in junk DNA.

    2. Alternative splicing: humans may not have more genes than the "lower" species but they can make far better use of what they've got by producing multiple proteins from each gene.

    3. MicroRNAs: humans have a extra layer of complexity not seen in other species. The microRNAs fine tune the human genes so that they can produce a more complex organism that that seen in the "lower" species.

    I call this the "Deflated Ego Problem." You can read the complete list of silly excuses at The Deflated Ego Problem.

    ReplyDelete
  11. L.M. There are a great many people—including, unfortunately, many scientists—who are upset that the "obvious" superiority of humans isn't reflected in their genome.
    Possibly so, but I find this debate uninteresting. If humans need to demonstrate their specialness, I'd prefer it be through creative endeavors like art and science.

    L.M. Most human genes are quite well conserved. They are present in all eukaryotes. Are you saying that sometime in the recent past (200 myr) most of these genes in mammals simultaneously acquired a whole new level of regulation not seen in their ancestors?
    No, the microRNAs are ancient as well and by no means restricted to humans. MicroRNAs are found in flies, frogs, chickens and nematodes, as well as humans. In fact, when the Nobel prize was awarded to Fire and Mello for microRNAs, some in the plant community were upset that they hadn't gotten appropriate credit for the discovery first. The cellular machinery by which microRNAs function must have been present in the relevant ancestral species. In some cases, examples of highly conserved (between species) microRNAs themselves have been found (such as let-7 from worms to humans), which is interesting because it requires sequence conservation at the nucleotide level rather than at the codon level and/or conservation of transcribed but non-coding sequences.

    Here's a recent review on the subject:
    Curr Opin Genet Dev. 2007 Apr;17(2):145-50. Epub 2007 Feb 20. The evolution of animal microRNA function.

    - I wouldn't be surprised for it to wind up that most human genes are significantly regulated by microRNAs.
    L.M. You wouldn't? I'd be shocked.
    That's what makes science fun.

    ReplyDelete
  12. anonymous says,

    Here's a recent review on the subject:
    Curr Opin Genet Dev. 2007 Apr;17(2):145-50. Epub 2007 Feb 20. The evolution of animal microRNA function.


    That's the review by Ryusuke Niwaa and Frank Slack at Yale. It's an excellent example of the kind of hyperbole I'm talking about. Here's the abstract.

    MicroRNAs (miRNAs) are a large class of small RNAs that function as negative gene regulators in eukaryotes. They regulate diverse biological processes, and bioinformatics data indicate that each miRNA can control hundreds of gene targets, underscoring the potential influence of miRNAs on almost every genetic pathway. In addition to the roles in ontogeny, recent evidence has suggested the possibility that miRNAs have huge impacts on animal phylogeny. The dramatically expanding repertoire of miRNAs and their targets appears to be associated with major body-plan innovations as well as the emergence of phenotypic variation in closely related species. Research in the area of miRNA phylogenetic conservation and diversity suggests that miRNAs play important roles in animal evolution, by driving phenotypic variation during development.

    Do you agree that microRNAs are the most wonderful thing since sliced bread? Do you agree that each miRNA controls hundreds of genes in almost every genetic pathway?

    Do you agree that miRNAs are responsible for major body plan innovations as well as phenotypic differences between closely related species? Don't you think that's a case of having your cake and eating it too?

    Possibly so, but I find this debate uninteresting. If humans need to demonstrate their specialness, I'd prefer it be through creative endeavors like art and science.

    I think you'd better start paying attention. Science is being abused by promoting exaggerated claims about miRNA and you seem to have fallen for it.

    ReplyDelete
  13. Let's turn the question around: the microRNA machinery (drosha, dicer etc) is conserved from plants, to worms, to flies, to fish, to frogs, to chickens, to humans. Why do you suppose that is, unless it does something really important? What else do you know that is so widely functionally conserved that really doesn't matter that much?

    I don't know if microRNAs are going to turn out to be the next greatest thing since sliced bread, but I'm willing to stay tuned to see how things develop. I've wondered for a long time why bacteria have such excellent systems for negative gene regulation (cf the classic lac operon) and "higher" (generally meaning multicellular) eukaryotes don't. There's heterochromatinization, but that's such an all-or-nothing kind of process. Experimentally, a common difficulty in commercially available mammalian inducible expression systems is that they are leaky in the uninduced state, because of the lack of really efficient mammalian transcriptional repressors. There's no such difficulty in commercial bacterial inducible gene expression systems. However, if there's an additional layer of control post-transcriptionally, as microRNAs could potentially provide, then this may not be such a gap in the capabilities of eukaryotic gene regulation.

    Are the claims for microRNAs exaggerated? We'll know the answer to that in a couple of years. If keeping an open mind is "falling for it", then I guess I've fallen for it.

    ReplyDelete
  14. anonymous asks,

    Let's turn the question around: the microRNA machinery (drosha, dicer etc) is conserved from plants, to worms, to flies, to fish, to frogs, to chickens, to humans. Why do you suppose that is, unless it does something really important? What else do you know that is so widely functionally conserved that really doesn't matter that much?

    I presume it's conserved because it does something important.

    That's not the point. The point is how common is this type of regulation. There's no debate about whether it exists.

    Same thing with alternative splicing. We all know it exists but the real question is what percentage of genes are alternatively spliced.

    If keeping an open mind is "falling for it", then I guess I've fallen for it.

    You're not keeping an open mind if you've already concluded that most genes are regulated by miRNAs. Instead, you've already reached a conclusion that doesn't make a lot of sense.

    ReplyDelete
  15. We've known about regulation by small RNA's since the 1970's. Nothing new there.

    The earliest example I know is control of DNA replication in the E. coli plasmid ColE1. The first published ref I could find is this one from 1981.

    ReplyDelete