Commentary on Kritzer et al. 2009

In the current edition of Nature Chemical Biology, Dr. Lindquist's team reports a novel approach to identifying chemical inhibitors of protein toxicity.  This study has broad general relevance, but is particularly relevant to PD research because Dr. Lindquist's team used the method to identify novel inhibitors of a-synuclein toxicity.

The method basically begins with a genetic library that expresses a protein, termed dnaE intein.   Intein is a relatively new class of protein that has the ability to auto-splice itself out of a protein much like self-splicing RNA introns and DNA transposons can auto- splice themselves out of larger pieces of RNA and DNA.  The construct begins with dnaE, which contains an intein.  In the middle of the intein protein is inserted a linker region that randomly codes for 5 million different octamer peptides.  Expression of the chimeric intein protein leads to rapid proteolytic splicing of the octamer, which is then cyclized by the process to yield cyclic peptides.  These peptides are highly versatile for several reasons:

1. The sequence of the peptide can be rapidly determined because the peptides are derived from DNA libraries, which means that DNA sequencing can yield the identity of the peptide.
2. Structure function studies on any selected peptide can be readily accomplished using site directed mutagenesis of the parent plasmid.
3. The peptide is also functionalized to by addition of a His-Pro- Gly motif that is readily bound by streptavidin, which means that - in theory - it is possible to affinity purify the peptide ligand and identify its target of action.

Dr. Lindquist's group has applied this powerful technology to identify two cyclic peptides that inhibit a-synuclein toxicity in a yeast and also in C. elegans.  a-Synuclein toxicity is thought to derive from its fundamental tendency to aggregate.  This basic biophysical property means that the very simple assays can be used to identify compounds that inhibit a-synuclein toxicity.  Dr. Lindquist's group has used yeast to explore mechanisms of a-synuclein toxicity in multiple different studies with powerful results.  Earlier studies identified vesicular binding proteins, such as the rab5 protein as mediators of a-synuclein toxicity.  One might be skeptical of what yeast can tell us about the human brain, but in many cases the same genes that modify a-synuclein toxicity in yeast also have virtually the same effects in a-synuclein mouse models, as well as in simpler animals, such as C. elegans.  The strong results obtained from these studies provide the best argument to counter questions of skeptics.  

Given the ability of the yeast studies to translate to the mammalian brain, Lindquist teamed up with Guy Caldwell to use C. elegans as an intermediate system to validate the results in vivo because of the ease with which one can genetically alter C. elegans.

The current study identified two cyclic peptides, CP1 and CP2, both of which abrogate a-synuclein toxicity in yeast.  The actions of CP1 and 2 are quite specific because they do not inhibit toxicity in a polyglutamine yeast model and they do not cause toxicity on their own.  One might expect that the assay would identify compounds that directly inhibit a-synuclein aggregation, but the group found that the peptides do not inhibit the accumulation of a-synuclein.  They appear to inhibit downstream mediators of toxicity.  The actions show strong structural dependence, because deletion of two C-terminal amino acids blocks the protection offered by the peptide.

At this point in the essay, the reader is probably eager to know the identity of the targets that are inhibited by the peptides.  The group reported affinity purifying the targets, but then in one of the truly frustrating weaknesses of the article they did not reveal the identities of the proteins.  The group asked whether increasing or decreasing the protein targets, which I will name P-Susan and P-Guy, elicits the same protective effects as CP1 and 2.  Surprisingly, these genetic modifications did not alter toxicity, which leaves a large unanswered question regarding the mechanism by which CP1 and 2 act.  

Stay tuned....

Despite the absence of a mechanism, further studies clearly showed that CP1 and 2 were protective C. elegans.  Dr. Caldwell's group has developed C. elegans that strongly express a-synuclein in dopaminergic neurons, and used this in multiple studies to identify genetic modifiers of a-synuclein toxicity.  The team expressed CP1 and 2 in the C. elegans a-synuclein model, and used the same approach to investigate test whether these were protective.  The results were quite clear.  Expressing CP1 or CP2 prevented degeneration of the dopaminergic neurons, judged by loss of GFP fluorescence in the neurons or formation of degenerative morphology.  Since prior studies show that genetic protection in C. elegans translates to protection in mice, it is reasonable to expect that CP1 and CP2 might also be effective in murine a-synuclein models.

So what lies ahead?  First, we can expect this technology to be applied to many other targets to explore mechanisms of disease or mechanisms of biological function.  Second, we will eagerly await studies that identify the targets and mechanisms of action of CP1 and CP2.  An unexplored question is whether the cyclic peptide technology can be modified to allow exogenous application of the peptides and create a true pharmaceutical.  This technology appears to be very powerful and seems likely to become an important tool in the future.