One of the biggest surprises of the genomics era is the discovery that every family, genus, and even species, has a shocking number of unique genes that are not found in closely related species. These so-called “orphan genes” have intrigued scientists for the past two decades and their diverse origins are beginning to come to light. In this presentation, I will present evidence that at least two families of human-unique micro-RNA genes have been created through random genomic rearrangements. These micro-RNAs tweak the expression of hundreds of genes in tissues throughout the body and may have played a role in the unique evolutionary trajectory of our species.
@Nlents gives a talk about his research at the BayAreaSkeptics.
@NLENTS , this was a great talk! I have a few questions:
For clarity, is the basic idea that instead of looking for differences between humans and other apes in protein coding regions of our genome (structural changes to proteins, for instance), it turns out that the “unique” parts of our DNA may have more to do with regulation and gene expression and that’s its things like de novo micro-RNA that might be a big part in explaining our biological uniqueness as a species?
I am curious about how we go from “let’s just transcribe everything and see what happens?” to “hey, that was useful, let’s keep it around!” on a scale that leads to speciation. In other words, I can kind of understand a bit of what’s going on at a molecular level (at least for a physical chemist) but I’m having a hard time translating that into phenotypic differences that are “selectable”.
I don’t yet understand how gene expression really works, but is the stuff you’re looking at epigenetic? I guess my crude picture of the genome is that there is there are protein coding genes and then there is a whole lot of what I envision the immune system to look like – lots and lots of variability happening in the hopes that something is functional. Is that at all relevant and/or accurate?
This is currently an open question. To date, only a very small number of “human-specific” proteins have been found and really no direct connection to the kinds of things that make humans unique. We know that regulatory regions are more likely to hold the key to human uniqueness but not a whole lot has been concretely demonstrated. miRNAs represent a fascinating possibility because I tiny variation - even a single nucleotide - could perturb the expression of hundreds of genes. Some would be slightly turned up because the miRNA no longer targets them well, while others would be turned slightly down because now it does target them. I would not be able to estimate how much human-specific miRNAs contribute to human uniqueness because we’re just getting started with this work.
You and everyone else. It’s a big leap, to be sure, but this isn’t the kind of thing that we can deduce from fossils or even phylogenetics. However, there are some good examples from fruit flies. I think these two papers are a great start on that very important question:
No, epigenetic is a different layer of modulation altogether. Gene expression control is incredibly complicated. You have the gene, but there are tons and tons of “dials” or “knobs” that control exactly when, where, and how much each gene is expressed. It’s like a giant graphic equalizer or sound board, and every gene has its own set of knobs. (with some master knobs that control lots of genes at the same time.) Epigenetics is covalent modifications to DNA itself or the histones around which the DNA is wrapped. But miRNAs target gene expression at a much a later step… the stability of the mRNA. A very stable mRNA will be translated many many times, resulting in lots of protein. An unstable mRNA degrades quickly and very little protein is made from it. That’s where the miRNA “knob” affects the final output (amount of protein).
To add another complication, we have long known that lots of transcription and even translation is not meaningful, because from mice we know that when and where mRNA and protein is expressed does not predict the phenotypes of mice lacking that gene (homozygous null, both natural and knockouts).
This makes no sense in the context of intelligent design, but perfect sense in the context of evolution, because there should be far more selective pressure to turn a gene on when/where it is needed than it is to turn it off when/where it isn’t.
Is that primarily because we lack the connection between gene and unique human traits (and so we just don’t know) or that it’s just not in the proteins?
Are there any other strong contenders other than uniquely human proteins?
This is super helpful!
Could you unpack this one just a little bit more? I can understand how transcription and translation could not be meaningful theoretically, but I’m not sure how we get that from the second part of that sentence.
Sure, from my own postdoctoral work: Homozygous null mutant Myo5a mice (formerly called dilute-lethal), with null alleles caused by an array of deletions (so we knew they were nulls) have no phenotype other than their coat color until two weeks after birth.
When I looked at expression of the RNA in embryos by in situ hybridization, the entire nervous system lit up at 7 days of gestation, a full month before any neurological phenotype was observable.
I also did more sophisticated assays, but could see nothing neurological despite the absence of that protein during that month.
It’s clearly more in the subtleties of how the proteins interact, when they are expressed, and where they are expressed. That fits with what we know about anatomy very well. Is there any organ in your body that is not also present in your dog’s?
That’s why reading the phrase “the gene for X” makes me want to bang my head on concrete.
I think that the tweaking of DNA sequences in the promoter regions of genes, and other regulatory regions, is likely the main driver of evolutionary change over shorter time periods and more modest diversifications.