Hunting the elusive D-amino acid

MS spectra showing that (a) before digestion with alanyl aminopeptidase, NWFa/NdWFa peaks (m/z 465) are barely detectable above the noise, but (b) after digestion, the NdWFa peak is clearly visible. Tandem MS experiments confirmed the identity of both peaks.
View larger image

Sometime during the history of life on this planet, L-amino acids became the enantiomer of choice for building proteins. There doesn't seem to be a particular reason this configuration was selected; scientists have built proteins from exclusively D-amino acids and found that they fold just like their natural counterparts, albeit as a mirror image of the native structure.

Although proteins made entirely from D-amino acids do not exist in nature, some proteins and peptides do contain a small number of D-amino acids. For many peptides, this enantiomeric switch is thought to occur as a late-stage posttranslational modification of an L-amino acid in the peptide.

The true number of D-amino-acid-containing peptides (DAACPs) is not known, because distinguishing peptides that contain only L-amino acids from those that contain D-amino acids is not trivial; each peptide of interest must be examined individually, which can be extremely time-consuming. Now, in work published in AC (2008, 80, 2874–2880), Jonathan Sweedler and colleagues at the University of Illinois Urbana–Champaign have devised a faster and simpler MS method to scan for potential DAACPs in a complex biological sample.

In a peptidomics experiment, a sample (in this case, brain tissue) is fractionated, and hundreds of peptides are isolated and sequenced. A few selected peptides are then synthesized and tested for bioactivity. The catch is that any of the native peptides could contain a D-amino acid—which wouldn't be obvious from typical MS data—and the researchers could be unknowingly synthesizing the wrong molecule.

To further complicate the experiment, not every peptide will have biological activity, so DAACPs are easily overlooked. "Some of these putative neuropeptides may be working by targeting a particular receptor, others may have very subtle effects, some may do nothing," Sweedler explains. "If you don't have potent bioactivity, it's hard to know if you made the wrong molecule."

The researchers wanted to find a way to pick out the DAACPs before they went to the trouble of synthesizing the peptides. "Our goal here was to make an approach that would be general—that would allow you to take a complex sample of brain homogenate and say that [a particular peptide] is more likely to have a D-amino acid in it," says Sweedler.

To do that, the scientists took advantage of the specificity of alanyl aminopeptidase, an enzyme that digests proteins with a D-amino acid modification much more slowly than those made completely of L-amino acids. Sweedler and colleagues postulated that if they added the aminopeptidase to a mixture of peptides, it would digest away the sequences that contain only L-amino acids and leave the peptides that contain a D-amino acid.

The researchers first tested the enzyme's activity with representative peptides, some of which had a single D-amino acid substitution. The rates of digestion for the L-amino acid peptides ranged from a few minutes to an hour (influenced by certain L-amino acids or posttranslational modifications that are degraded more slowly than others), but the peptides with a D-amino acid were essentially resistant to degradation by the aminopeptidase.

As a proof of principle for a more complex mixture of peptides, they used a sample from the abdominal ganglion of the mollusk Aplysia kurodai, commonly known as the sea hare. The researchers compared a sample that had been digested with the aminopeptidase with one that had not undergone enzymatic treatment. On comparing spectra from the two samples, they found a peak in the treated sample consistent with NdWFa, a peptide containing a D-amino acid that is a known constituent of cells from this system; this confirmed that DAACPs could be isolated from a complex mixture.

Sweedler says that his group is now hunting for previously unknown DAACPs. Although the enzyme that carries out the conversion from an l- to a D-amino acid is not known, the position of the D-amino acid two to three residues from the N-terminus is consistent across peptides from mollusks, arthropods, and vertebrates; this suggests that a common enzyme is making this modification. "Most enzymatic pathways that are this well represented may be found everywhere," says Sweedler. "So it's an interesting thought to see if these are found in the mammalian brain." Now that searching for DAACPs has become simpler, it's definitely worth a look.

—Jennifer Griffiths