How the Genome "Decides" Where to Splice
Once upon a time, scientists believed that DNA that did not code for proteins was "junk" DNA. They believed that this junk DNA was a leftover from natural selection's trial-and-error process. This story is no longer the ruling myth it once was. Research from the 1990s to the present indicates that this "junk" is hardly junk at all.
Non-coding DNA is implicated in many important processes, including embryonic development, and regulating translation and transcription. Now we see that coding within the junk "non-coding" DNA helps the genome "decide" what parts to cut out and what parts to leave in before making the messenger RNA that code for proteins. Some genetic diseases may be due to mutations within this part of the genome.
Non-coding RNA segments (derived from non-coding DNA segments) are referred to as introns; the coding RNA segments are called exons. Eukaryotes (organisms whose cells have a nucleus) have a rather elegant procedure for removing the introns before making the messenger RNA that builds proteins. This procedure is known as splicing because not only are introns removed, but the exons are then spliced together.
There are genetic markers that tell the enzymes were to make the first cut. This marker is the 5' marker. Then there is a marker for what is called a branch point. This branch point will undergo a chemical reaction with the newly cleaved 5' site and create a looped, lasso-like structure called a lariat (see this figure). Finally, the other end of the lasso is cleaved at the 3' marker, and viola! The intron has been removed and the exons are ready to move on to the next phase in protein construction.
The leftover intron is now a lariat, a looped string of nucleotides. This lariat is unstable so it eventually breaks apart, never to be heard from again. Scientists have trouble studying these lariats in the body because they are so unstable. However, studying lariats could help scientists understand many aspects of protein production and possible causes for mutation-based diseases.
This nice, neat trick of the genome has another interesting feature to it. Sometimes alternative splicing occurs. When this happens the intron is not just removed. Sometimes an exon is also removed or sometimes parts of introns are left in place to be transcribed into messenger RNA. This totally changes the code for the protein, meaning one segment of DNA may actually code for multiple proteins. This vastly increases the complexity of the genome, and brings up a question:
How does the genome decide when to perform alternative splicing and when to do "normal" splicing?
A paper in Nature Structural and Molecular Biology highlights some extensive studies done in cataloging lariats using a method based on reverse transcriptase produces and modeling. This cataloging allows scientists to identify branch points. Researchers have found that these branch points affect the genomic decision to splice the genome.
Remember, the branch point is where the "knot" in the lariat (lasso) is formed, but the rope can be cut off at various locations away from the knot. The authors were able to deduce some general rules for where the enzymes cut the rope based on the 3' splice site location relative to the branch point. The rules are bit complicated because they are based on nuances in the genome sequence, but see the research article for some ways that the authors tested these rules on known branch points and splice sites.
Furthermore, studies show that when a mutation occurs within an intron and causes a disease, these mutations are usually found at branch points. Apparently the branch points are important pieces of code where one mutation can wreak havoc on protein assembly.
What does this have to do with evolution or intelligent design?
First, one of the predictions of intelligent design was that so-called "junk DNA" was not actually junk. Many scientists considered junk DNA part of the evolutionary narrative. Even after studies revealed that junk DNA had a functional purpose, many staunch Darwinists continued to support their view that the majority of the genome was useless "junk," an indication of their own evolutionary bias.
Secondly, the genome is an information-carrying system, so when we use terms like "makes decisions" we are using design language. The rhetoric in the article makes the program (the genome) the subject, performing the action of making decisions, but as with any computer program, these decisions are programmed into the code.
The authors of this article were likely not implying that the genome, in and of itself, is in any way sentient, but were looking for a mechanism for decision-making. However, this mechanism does not help us understand how the genome seems programmed to know when to code for what protein, given that more than one protein can be formed based on different splice site locations.
This sounds very much like computer programming, where options are programmed into the code, but the code is not itself making the decisions.