A growing polypeptide (red) binds puromycin (yellow) and becomes anchored to the glass chip.
After a diet rich in nucleotides, sometimes you just want a little protein with your chips. Although cell signaling and protein binding are clearly important in neuroscience, high-throughput analysis of proteins has lagged behind microarray analysis of gene expression due to a lack of accessible tools. Now Tao and Zhu report two new techniques to custom synthesize protein chips in a recent article in Nature Biotechnology.
The antibiotic puromycin terminates active translation. It resembles aminoacylated tRNA, which converts nucleotide codons into amino acids, and therefore enters the ribosome. There, it covalently attaches to the building polypeptide chain, which is then released. Ribosomes briefly stall when they reach double-stranded RNA or RNA-DNA hybrids, allowing enough time for a nearby puromycin to enter and attach to the growing polypeptide, according to the authors. They used the puromycin tag to anchor polypeptides to glass chips in two different ways.
The authors generated mRNAs encoding proteins of interest with ribosome-binding sites at the 5' end and biotinylated common primer-binding sites at the 3' end. They also synthesized oligonucleotides with puromycin at one end and biotin at the other. Incubation of the mRNA with an RNA oligonucleotide complementary to the primer-binding site generated double-stranded RNA. Because of their biotin labels, both the puromycin-labeled oligonucleotide and primer-bound mRNA attached to the streptavidin-coated glass slides. The authors used rabbit reticulocyte lysate, which contains protein translation machinery, to induce in vitro translation and subsequent peptide-binding to the puromycin-labeled oligonucleotide immobilized on the chip.
The authors tested their method with three mRNAs, encoding Hisx6, FLAG and StrepTagII. They detected all three proteins with specific primary and fluorescently conjugated secondary antibodies, suggesting that the proteins were both properly synthesized and firmly attached to the chip. However, optimal performance required that the mRNA be longer than the puromycin-labeled oligonucleotide.
The authors therefore devised a second strategy that placed the puromycin closer to the mRNA. They attached both the biotin and the puromycin label to a DNA oligonucleotide complementary to the common primer-binding site on the RNA. Incubation of the mRNA with the DNA oligonucleotide generated an RNA-DNA hybrid containing both puromycin and biotin labels on an mRNA encoding a protein of interest. The authors printed this mRNA molecule onto the slide. They tested this second method with mRNAs (and resulting proteins) of various sizes, including transcription factors, protein kinases and glutathione-S-transferase, which were detected equally well regardless of size. They also detected Mcm1 and Cbf1 yeast transcription factors using biotin-labeled DNA, indicating that protein binding could be successfully detected.
The authors calculated that each spot on the chip contained 0.8 fmole of synthesized protein, which is similar to the roughly 2.0 fmole/spot for existing arrays of pre-spotted proteins. Therefore, these methods, which are both cheaper than purchasing pre-spotted arrays and easier than methods requiring subcloned DNA, produced similar results to existing platforms and thus expand the proteomics tools available to neuroscientists.