Back in 2003, I proposed that genomes were designed to be modular and mutable and that this was the probable explanation for the rapid species diversification immediately after the Flood. I also suggested that these genomic changes were the direct result of environmental stress. Barbara McClintock thought something similar about speciation and genomes. She called the genome a "highly sensitive organ" that can respond to new conditions by rearranging itself. In 2003, I had a few examples of eukaryotic retrotransposons that mobilized during periods of stress, but I knew of no examples of any rearrangements that could be linked to a useful phenotype.
The best examples of genomic modularity that I could give at the time were genomic islands. These are regions of bacterial genomes that are identifiable by a GC content or codon preference different from the rest of the chromosome where the island is found. They are often bounded by repeat sequences, and most importantly, they contain genes that code for a specific phenotype (and not always a nice one). The genes that allow rhizobia to colonize plant roots and fix atmospheric nitrogen are found on islands. The genes that let bacteria digest nylon or metabolize mercury are found on islands. The genes that let anthrax bacteria kill so quickly are also found on islands. Some genomic islands can be transfered from one bacterium to another using an enzyme called an integrase, which is usually genetically coded in the island itself.
Yesterday, PNAS published a paper by a group from the University of Lausanne in Switzerland called "Stochasticity and bistability in horizontal transfer control of a genomic island in Pseudomonas." This group studied transfer of an island called clc in Pseudomonas knackmussii strain B13. Clc allows P. knackmussii B13 to live on 3-chlorobenzoate and 2-aminophenol as its sole carbon source. They found that in a clonal population of P. knackmussii B13 bacteria, only about 3-5% of the cells express the integrase responsible for transfering the island. They also found that a second gene, named inrR (for "integrase regulator"), is required for transcriptional activation of the integrase gene.
Next, they used variants of green flourescent protein to track individual cells coexpressing inrR and the integrase. They created a strain of knackmussii with a GFP gene under the control of the integrase promoter and a variant gene of GFP (echerry) under the control of the inrR promoter. They grew this strain on fructose and 3-chlorobenzoate. Using only 3-chlorobenzoate as an energy source, there were approximately twice as many cells co-expressing inrR and integrase as cells grown with fructose.
Think about that for a second. Integrase is responsible for transfering the clc genomic island. Integrase regulator is somehow (they're not sure how) needed for transcription of the integrase. Both genes were upregulated when the cells were living exclusively on 3-chlorobenzoate. In this case, an environmental stress can be linked not just to genomic rearrangement but to environmentally relevant genomic rearrangement. It's as if the population of cells is trying to circumvent natural selection and directly adapt to the environment.
Their research leaves a lot of questions: What is the mechanism whereby integrase and inrR are upregulated in the presence of 3-chlorobenzoate? Is this a general mechanism found in other islands? Is this direct adaptation exclusive to prokaryotes or are their analogous mechanisms in eukaryotes? Whatever the answer to these questions, this paper would seem to demonstrate that Barbara McClintock was right, to a certain extent. There are cases where cells can detect environmental stress and remodel their genomes accordingly.
You know, Lamarck proposed (among other things) that organisms could directly adapt to their environment and pass those adaptations to their offspring. For 200 years now, people have laughed at that idea.
Who's laughing now?
McClintock, B. 1984. The significance of responses to the genome to challenge. Science 226:792-801. [PDF]
Minoia, M., M. Gaillard, F. Reinhard, M. Stojanov, V. Sentchilo, and J.R. van der Meer. 2008. Stochasticity and bistability in horizontal transfer control of a genomic island in Pseudomonas. PNAS 105:20792-20797.
Wood, T.C. 2003. Perspectives on AGEing, a young-earth creation diversification model. In Ivey RL, ed. Proceedings of the Fifth International Conference on Creationism. Creation Science Fellowship, Pittsburgh, pp. 479-489.