Put intuition aside for a moment and imagine a scenario where E.coli knockout strains that have been deleted for conditionally essential genes are rescued by proteins taken from a protein library composed of >106 de novo designed sequences. The prevailing assumption- that functional proteins are constrained to a very small subset of possible sequences- would lead us to infer that finding them by a random search through sequence space would be tantamount to impossible. But a PLOS One paper published in early 2011 appears on the surface to have given us much room for thought. Scientists from Princeton’s Department of Chemistry and Molecular Biology used a combinatorial library of 102-residue long proteins to rescue non-viable E.coli knockouts. The functional losses in the knockout strains affected serine, glutamate and isoleucine biosynthesis and disabled the cells’ natural capacity for iron acquisition in iron-limited environments.
The E.coli knockouts were auxotrophic meaning that they exhibited a failure to grow on minimal (M9-glucose) media even after several weeks of incubation. But following transformation with the combinatorial library, several cases of successful colony growth were documented suggesting that the certain genes contained therein had successfully complemented the deletions. Sequence comparisons through formal BLAST searches showed that the rescue proteins involved, eighteen in all, were unlike any protein found in nature.
Truth be told these rescue proteins were not entirely random. Their sequences had been engineered to ensure that they would fold into a stable 3-D structure. And not just any structure. Molecular biologists are well aware that canonical sequence rules exist that must be adhered to if they are to maximize the chances of proteins folding correctly. In an alpha-helical fold, for example, polar and non-polar residues must be carefully ordered to make certain that the hydrophilic (‘water-loving’) and hydrophobic (‘water-hating’) faces of the fold emerge. The Princeton group adopted a binary code strategy of polar and non-polar residues to get 1.5x10exp6 four-helix bundles. In this singularly fundamental aspect they were designed.
What was the molecular basis that allowed rescue? From their own experiments the Princeton group ruled out the likely hood that novel pathways that bypass the adverse effects of the knockout genes had arisen since E.coli mutants that were deficient in other steps of the naturally occurring biosynthetic pathways could not be rescued. Also dismissed was the interpretation that “global alterations in metabolism had been induced by the mere expression of foreign genes” (a stress response of some kind) since none of the eighteen rescue proteins appeared to have been unfolded- a tell-tale symptom of such a global response. Mutations that minimally disrupted their structure abolished their rescue capabilities.
The evidence seemed compelling. These de novo sequences were exerting specific functional effects that served to avert an otherwise fatal outcome for their bacterial hosts. Still the ‘design’ point-of-departure raised above was without question central to this particular success story. In his book Signature In The Cell Stephen Meyer has noted how it is sequence specificity that ensures that amino acid chains fold into “useful shapes or conformations” (Ref 1). Without a library tailored for the formation of four-helix bundles, the Princeton study is unlikely to have yielded anything that would come close to salvaging the debilitated bacteria. To make matters worse, this study failed to consider in detail the cooperativity that so evidently characterizes the organismic molecular scheme. What we see here is akin to taking one piece out of a jigsaw puzzle and finding another to put in its place albeit with some considerable force of fit. Those who espouse blind evolution are still left reeling over how to explain the origin of the entire puzzle.
Importantly cells transformed with rescue proteins exhibited growth that was “significantly slower than those expressing the natural protein” (the non-knockout strains). Exponential growth occurred 24-144 hours later and reached culture densities that in some cases were as low as 12-15% of wild-type. The authors readily admit that the library proteins may “function by different mechanisms than the natural proteins they replace”. Indeed assays designed to test for the deleted functions failed to show that the de novo sequences exhibited comparable enzymatic activities. The evolutionary inference given by the authors- that billions of years of evolution have driven optimal activities for faster growth- therefore appears to be nothing more than a rehash of a positively stale Darwinian fairytale. After all, if the proteins function by different mechanisms, one cannot allege that they are in any sense on the way to becoming the more efficient naturally occurring protein entities we observe in E.coli today.
For the full PLOS One article see:
Michael A. Fisher, Kara L. McKinley, Luke H. Bradley, Sara R. Viola, Michael H. Hecht (2011) De Novo Designed Proteins from a Library of Artificial Sequences Function in Escherichia Coli and Enable Cell Growth, See http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0015364
Further Reading
1. Stephen Meyer (2009) Signature In The Cell: DNA And The Evidence For Intelligent Design, Harper Collins Publishers, New York, p.99