Key points
- Every standard amino acid except glycine is chiral. Proteins are built almost exclusively from the L forms; their mirror images, the D forms, are rarer but far from absent in biology.
- D-amino acids appear in bacterial cell walls, in nonribosomal antibiotics, in animal venoms and skin peptides, and as free signalling molecules in the brain.
- A D-residue has exactly the same mass, charge and hydrophobicity as its L form. Mass spectrometry cannot see it, and every calculator on this site gives identical results for both.
- Proteases are built to recognise L-peptides, so D-residues are a standard tool for making peptide drugs last longer in the body.
Chirality and the L/D convention
The α-carbon of an amino acid carries four different groups: the amino group, the carboxyl group, a hydrogen and the side chain. Such a carbon is a stereocentre, and the molecule exists in two forms that are mirror images of each other and cannot be superimposed, much as a left and a right hand. Glycine, whose side chain is a second hydrogen, is the only standard amino acid without this property.
The L and D labels come from an old convention based on the configuration of glyceraldehyde, not from the direction in which the molecule rotates polarised light. A simple mnemonic for L-amino acids is the "CORN" rule: looking down the H–Cα bond from the hydrogen, the groups CO (carboxyl), R (side chain) and N (amino) read clockwise.
The modern Cahn–Ingold–Prelog system uses R and S instead. Most L-amino acids are S, with one exception: L-cysteine is R, because the sulfur atom in its side chain changes the priority order. The two systems describe the same molecules; they simply assign names by different rules.
Two amino acids have a second stereocentre in the side chain: isoleucine and threonine. Each therefore has four stereoisomers. The natural forms are L-isoleucine and L-threonine; their side-chain epimers are called allo-isoleucine and allo-threonine.
In sequences, D-residues are often written in lowercase in one-letter code, for example RGDfV for a cyclic peptide containing D-phenylalanine, or with a "D-" prefix in three-letter code. The tools on this site convert input to upper case, which is harmless: a D-residue has the same mass and ionisation as the L form, so the results are correct either way.
Where D-amino acids occur in nature
The ribosome uses only L-amino acids, and its proofreading actively rejects D-aminoacyl-tRNAs. D-residues therefore reach peptides by other routes.
Bacterial cell walls
Peptidoglycan, the mesh that gives bacterial cells their shape, contains D-alanine and D-glutamate in its peptide cross-links. The terminal D-Ala-D-Ala motif of peptidoglycan precursors is the target of vancomycin, which binds it tightly and blocks cell-wall assembly. Vancomycin resistance arises when bacteria replace the terminal D-Ala with D-lactate, which reduces binding by about a thousand-fold.
Nonribosomal peptides
The nonribosomal peptide synthetases of bacteria and fungi contain epimerisation domains that convert an L-residue to its D form during assembly. As a result, D-amino acids are common in peptide antibiotics and related natural products: gramicidin S contains two D-phenylalanines, cyclosporin contains a D-alanine, and the lipopeptide daptomycin contains several D-residues. Many of these peptides are cyclic, and D-residues help the ring close, as explained in the guide to cyclic peptides.
Animal peptides made by the ribosome
Some animals make peptides on the ribosome with an L-residue and then convert it to D with a dedicated isomerase. The first example, discovered in the early 1980s, was dermorphin, a potent opioid peptide from the skin of South American Phyllomedusa frogs, with the sequence Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH₂. The all-L version is almost inactive. Related frog peptides, the deltorphins, carry D-Met or D-Ala at the same position.
Later examples include ω-agatoxin IVB from funnel-web spider venom, a D-tryptophan-containing contryphan from cone snails, a C-type natriuretic peptide in platypus venom, and neuropeptides in crustaceans and molluscs. In each case the D-residue changes the peptide's activity or its resistance to degradation, and in each case it is invisible to the gene sequence.
Free D-amino acids
Free D-serine is present at high levels in the mammalian brain. It is made from L-serine by the enzyme serine racemase and acts as a co-agonist at NMDA-type glutamate receptors, where it binds the so-called glycine site. Free D-aspartate is found in neuroendocrine tissues. These discoveries overturned the long-held view that D-amino acids play no role in mammalian physiology.
Racemisation with age
In long-lived proteins, aspartate residues slowly convert to D-aspartate, through the same succinimide intermediate that causes deamidation and isomerisation, described in the guide to peptide stability. The D-Asp content of tooth dentin and eye lens proteins increases steadily over a lifetime and has been used in forensic science to estimate age.
Why D-residues resist proteases
Proteases are chiral catalysts. Their active sites are built from L-amino acids and are shaped to hold an L-peptide substrate with its side chains, backbone carbonyls and scissile bond in precise positions. A D-residue at or near the cleavage site places its side chain where the enzyme expects a hydrogen and vice versa, and cleavage slows dramatically or stops.
This is exploited widely in peptide drug design. Replacing a single residue at a known cleavage site is often enough to extend a peptide's half-life from minutes to hours:
| Peptide | D-residues | Use |
|---|---|---|
| Leuprolide | D-Leu at position 6 | GnRH receptor agonist |
| Octreotide | D-Phe, D-Trp | Somatostatin analogue |
| Desmopressin | D-Arg at position 8 | Vasopressin analogue |
| Icatibant | D-Arg and non-natural residues | Bradykinin B2 receptor antagonist |
| DAMGO, DADLE | D-Ala at position 2 | Research tools: stable enkephalin analogues |
In GnRH analogues such as leuprolide, the D-residue at position 6 does two things: it blocks a cleavage site, and it stabilises the β-turn conformation that the receptor recognises. The substitution increases potency as well as stability.
Mirror-image peptides and proteins
If a peptide is made entirely of D-amino acids, it is the exact mirror image of the L-peptide. In an achiral environment it behaves identically: same solubility, same melting behaviour, same mass. Against chiral partners, such as a receptor or an enzyme, it behaves as a completely different molecule.
In 1992 Stephen Kent's laboratory synthesised the enzyme HIV-1 protease entirely from D-amino acids. The mirror-image enzyme folded into the mirror-image structure, was fully active, and cleaved only mirror-image substrates. The experiment confirmed that all the information needed for folding is contained in the sequence and that chirality is simply inverted throughout.
This principle underlies mirror-image phage display. A library of natural L-peptides on phage is screened against a synthetic D-version of the target protein. The L-peptides that bind are then synthesised as D-peptides, which by symmetry bind the natural L-target, and are resistant to proteases. D-peptide inhibitors of HIV entry were among the first found this way.
A related strategy, the retro-inverso peptide, reverses the sequence and uses D-residues. The result presents its side chains in roughly the same spatial arrangement as the original L-peptide while its backbone runs in the opposite direction. It works well for some extended peptides and poorly for others, because the backbone hydrogen-bonding pattern is not preserved.
Detecting D-amino acids
Because D- and L-residues have identical masses and elemental formulas, standard analytical methods cannot distinguish them. This is why D-residues in natural peptides were overlooked for so long: a gene sequence says nothing about them, and a mass spectrum matches the L-peptide perfectly.
| Method | What it shows |
|---|---|
| Reversed-phase HPLC | Diastereomeric peptides, which differ in one residue's configuration, usually have different retention times. Comparing a natural peptide with synthetic L- and D-variants is the most common first test. |
| Hydrolysis and chiral derivatisation | The peptide is hydrolysed and the amino acids are reacted with a chiral reagent such as Marfey's reagent (FDAA), which converts D- and L-forms into separable diastereomers. |
| Chiral chromatography | Direct separation of enantiomers on a chiral stationary phase. |
| Enzymatic digestion | Resistance to cleavage at a site where an L-peptide would be cut suggests a D-residue there. |
| Ion mobility MS | Separates some diastereomeric peptides by shape in the gas phase. |
A complication affects the hydrolysis methods: acid hydrolysis itself racemises a small fraction of the amino acids. Hydrolysing in deuterated acid labels any residue that racemised during the procedure, so that it can be distinguished from D-residues present originally. The general HPLC approach is described in the guide to peptide HPLC.
D-amino acids in synthesis
For chemical synthesis, D-amino acids are simply other building blocks. Fmoc-protected D-amino acids are widely available and couple under the same conditions as their L counterparts in solid-phase synthesis.
The concern in synthesis is usually the opposite one: unwanted racemisation of L-residues during activation, which produces D-impurities. Cysteine and histidine are the most susceptible, particularly with strong bases or prolonged activation. The resulting diastereomers can be difficult to separate and are invisible to mass spectrometry, so they are a genuine quality concern for peptide drugs.
Frequently asked questions
Why does life use L-amino acids rather than D?
Nobody knows for certain. Any system built from one enantiomer works equally well as its mirror image, so the choice may have been an early accident, amplified because mixing the two disrupts folding. Proposed explanations for a small initial excess include circularly polarised light in space and asymmetric crystallisation; none is established.
Does "D" mean the molecule rotates light to the right?
No. D and L describe configuration relative to glyceraldehyde. The direction of optical rotation is labelled (+) or (−) and must be measured; L-amino acids include both dextrorotatory and levorotatory compounds.
Do D-amino acids change a peptide's pI or molecular weight?
No. Enantiomers have identical masses and pKa values, so the MW calculator and pI calculator give the same result. Diastereomeric peptides can differ slightly in measured properties, because their shapes differ, but not in calculated ones.
Are D-amino acids toxic?
Generally not at the levels found in food and in peptide drugs. Mammals have an enzyme, D-amino acid oxidase, that degrades most free D-amino acids. Fermented foods, aged cheeses and bacterially processed products contain small amounts of free D-amino acids.
References
- Montecucchi PC, de Castiglione R, Piani S, Gozzini L, Erspamer V (1981) Amino acid composition and sequence of dermorphin, a novel opiate-like peptide from the skin of Phyllomedusa sauvagei. International Journal of Peptide and Protein Research 17:275–283.
- Milton RC, Milton SC, Kent SBH (1992) Total chemical synthesis of a D-enzyme: the enantiomers of HIV-1 protease show reciprocal chiral substrate specificity. Science 256:1445–1448.
- Schumacher TN, Mayr LM, Minor DL, et al. (1996) Identification of D-peptide ligands through mirror-image phage display. Science 271:1854–1857.
- Wolosker H, Blackshaw S, Snyder SH (1999) Serine racemase: a glial enzyme synthesizing D-serine to regulate glutamate-N-methyl-D-aspartate neurotransmission. PNAS 96:13409–13414.
- Marfey P (1984) Determination of D-amino acids. II. Use of a bifunctional reagent, 1,5-difluoro-2,4-dinitrobenzene. Carlsberg Research Communications 49:591–596.