Receptors, flies, and the fountain of youth

A 2007 paper in the Journal Nature Chemical Biology (William Ja et al.), describing the modulation of a particular receptor as a means of extending the lifespan of the model organism Drosophila melanogaster, received a lot of press attention. Much of this came in the form of misguided speculation as to how this strategy could be extrapolated to humans, with similar gain. Nonetheless, the paper is clearly interesting for other reasons, most notably, the repertoire of difficult techniques employed throughout. What follows is a summary of the work, highlighting those aspects that I found most compelling- with a focus on methods.

Background: GPCRs

I’ve posted quite a bit about G protein-coupled receptors (GPCRs) on this blog, but as a quick refresher, these receptors comprise the largest known superfamily of cell-surface proteins. They are central to eukaryotic signal transduction, and the target of 40-60% of all currently marketed pharmaceuticals. They transduce signals, in the form of extracellular ligands, into intracellular cascades with specific biochemical consequences to the cell. The general stimulus-response pathway is fairly simple. At the intracellular interface, GPCRs interact with G proteins (heterotrimeric guanine nucleotide-binding proteins). Agonists helps shift a receptor’s structural equilibrium towards the active-state conformation. This leads to the activation of G proteins, and the initiation of second-messenger biochemical cascades. A more detailed overview of the mechanism is shown below, meant mostly to capture the primary biochemical endpoints of GPCR signal transduction.

GPCR signal transduction pathways and biochemical endpoints

Background: Drosophila melanogaster

Drosophila is a model organism belonging to the phylum Diptera, or the “order of flies”. It is considered a classical model in the study of genetics and developmental biology. Its short lifespan, inexpensive maintenance, simple diet, and amenability to genetic manipulation, all help strengthen its role in laboraties dealing with all sorts of complex problems on a reductionist plane. Drosophila mutant and trangenic variants are easy to produce, and the 165 million base-pair genomic assembly of this organism was released in 2001.

Background: Drosophila lifespan

Quite a bit of work has been done on this front. For our purposes, we’re only interested in the role of Methuselah (Mth), a GPCR implicated in lifespan regulation.

• Mth identified in screens of age-prolonging Drosophila mutants (1)
• Crystal structure of large N-terminal Mth ectodomain (2)
• Mth controls synaptic efficacy at neuromuscular junctions (3)
• Endogenous Mth ligand identified: Stunted (peptide) (4)

1. Lin YJ et al. Science 1998, 282:943-946.
2. West AP et al. PNAS 2001, 98(7):3744-3749.
3. Song W et al. Neuron 2002, 36:105-119.
4. Cvejic S et al. Nat Cell Biol 2004, 6:540-546

Mth GPCR signal transduction pathway

What next?

We know that down-regulation of Mth extends Drosophila lifespan. We believe that the Mth ectodomain functions as the ligand binding site, and specifically, that an exposed N-terminal tryptophan residue might be crucial for binding. It’s been shown that two short-length peptides, expressed as two isoforms of a single gene, are the endogenous ligands with which our receptor couples. The authors go on to identify synthetic peptide Mth ligands, using a novel GPCR-screening strategy (mRNA display).

mRNA display

1. An initial DNA pool is randomly generated via PCR

Primers include T7 polymerase promoter site, and 5′ UTR.

2. In-vitro transcription of this DNA pool yields our primary mRNA library
3. mRNAs are then ligated to 5’-(Ps)UAGCGGAUGC(dA)16(S9)2(dCdC)Pu-3’

This is effectively a chemical ligation with a spacer and puromycin (Pu). The spacer provides the required minimal distance requirements for the function ascribed to puromycin. In short, puromycin enters the ribosomal A-site, and guarantees premature chain termination. The final product is a pool of mRNAs covalently bound to the peptides they individually encode.

4. In-vitro translation of mRNAs: mRNA-peptide fusion library
5. Reverse-transcription of fusions: mRNA/cDNA-peptide fusion library

This step confers added stability to the complex.

6. Biotinylated Mth ectodomain is expressed, purified, and immobilized

Biotin binds tightly to streptavidin, in turn, immobilized on agarose beads. There’s an inherent risk involved in deciding not to express full-length receptors.

7. 27-mer fusion library (from 5) incubated with immobile ectodomain
8. Bound fusions eluted and precipitated
9. Precipitated cDNA is PCR-amplified, resulting in a new DNA library

This entire process is repeated until the final pool is enriched with peptides that bind to the Mth ectodomain with incredibly high specificity.

Resulting peptide pool

DNA sequencing of the final (8th round) pool, allowed for the identification of an apparent peptide motif (-[R/P]XXWXXR-), dubbed the “RWR” motif. This motif is not shared by the native ligand, and the authors mention no “statistically significant” homology to the Drosophila proteome. I’m not sure what methods were used, but for certain, the BLAST algorithm is not meant for short-length matching of this sort. There are algorithms specifically suited for the task (including SVMs and HMMs), that I don’t believe the authors made use of.

Surface Plasmon Resonance (binding studies)

Binding studies for selected peptides using (SPR) showed a dissociation constant (Kd) < 30 nM for "RWR" peptides. The next important question is whether or not members of this synthetic peptide clade compete for binding at the same ectodomain interface. To answer this question, R8-01 (10 muM) was competed with radiolabeled R8-01 & R804. The result was a 96% and 94% reduction in binding, respectively. To explore whether or not synthetic RWR peptides and Stunted bind at the same time, N-Stunted (30-mer synthetic peptide) was competed with R8-01, and a 79% reduction in binding was calculated.

In conclusion, either RWR peptides and Stunted all bind to the Mth ectodomain in a similar fashion and at the same “hotspot”, or allosteric competition is taking place.

X-ray crystallography (structural determination)

The Mth ectodomian, in complex with R8-01, was captured in crystal array form. The resulting diffraction pattern, and the tertiary structure of the complex calculated from the former, establish that the binding site is closer to the C-terminus of the N-terminal domain. The authors then speculate that the full-length Mth binding site is the ectodomain-EC loop juncture. They strengthen this claim with the identification of a partial RWR motif in the EL2 (extracellular loop 2) helix, although a synthetic version of this peptide failed to compete against R8-01 in SPR binding studies.

Heterologous expression