In the realm of biology, we are always finding exceptions to the rule. And now, we have just added another exception to the central dogma of genetics.
But first, let’s review what the central dogma is. The central dogma states that that the flow of genetic information goes from DNA to mRNA to protein. In other words, DNA makes mRNA and mRNA makes protein. Of course, there are already exceptions to this flow as DNA can be made from RNA, for example. The sequence of nucleotides (the A’s, G’s, T’s, and C’s) in the DNA gets transcribed to mRNA (A’s becomes U’s, G’s become C’s, T’s become A’s, and C’s becomes G’s). During translation, groups of three adjacent nucleotides (called codons) in the mRNA signify what amino acids are used to build a protein. Since there are four nucleotides that make up DNA, there are in total 64 unique three-letter sequences that encode a total of 20 amino acids.
Within this framework, only specific codons called start codons signal where protein synthesis starts. Only three of these codons could signal the start of protein synthesis: AUG (which accounts for 83% of start codons), GUG (which accounts for 14% of start codons), and UUG (which accounts for 3% of start codons). Or so we though…
Scientists from NIST and Stanford were researching something completely different when they stumbled along a path towards new start codons. They wanted to stop the synthesis of the enzyme dihydrofolate reductase (DHFR) in E. coli so they mutated its start codon. Seems like an easy step towards their goal, but after changing the AUG start codon sequence on four separate occasions, they still detected expression of DHFR. How could this be?
They took a detour away from DHFR to find out and constructed 64 variants of the green fluorescent protein (GFP) gene. By placing each of the 64 codons at the start position in front of the GFP coding sequence, they were able to determine which codons were capable of initiating translation. If the codon at the start position indeed acts as a start codon, GFP synthesis occurs resulting in the production of a green fluorescent signal. If the codon at the start position does not initiate protein synthesis, no green fluorescence will be observed.
What they found was that 47 codons were able to initiate protein synthesis above background levels. While the three canonical start codons (AUG, GUG, and UUG) initiated translation 10-100% of the time compared to AUG, the 47 codons initiated translation 0.07-3% relative to translation from AUG.
0.07-3% does not sound like much so could these initiation events be due to other errors during the process? Before protein synthesis occurs, a lot has to happen. DNA is replicated, DNA is transcribed into mRNA, and amino acids are brought to the mRNA and incorporated into the growing protein. Lots of opportunities for errors, right? The researchers took these factors into account and calculated the error rate. For 17 non-AUG start codons, the initiation rates were higher than the translation error rate.
To get back to the question: is translation initiation from these start codons “real?” That answer seems to be a “yes” for at least 17 of the non-AUG start codons. What we thought was a hard and fast rule in biology was broken yet again.
