Chemists were taught that natural systems only use L-amino acids. Andy Extance finds out just how wrong that is proving

Two amino acids face each other. They are mirror images, isomers that chemists know as enantiomers. Yet, in an unfair twist of science, one dominates the other almost entirely. In chemistry lessons, we learn that l-amino acid enantiomers rule the biological world, making up vital structures in every living thing. So are their d-amino acid mirror images absent from the biological world? Definitely not! Instead, scientists are learning that these downtrodden enantiomers are more important than they had previously realised.

A step towards their recognition came in 1958, when Jeff Watkins, working at the Australian National University in Canberra, made a crucial discovery in neurochemistry. Together with David Curtis and John Phillis, Watkins set out to hunt chemical neurotransmitters in the brain. Sodium glutamate was among the first substances they tested because they had a 500g bottle on their lab shelves. And they found that the amino acid anion glutamate and its close relative aspartate both stimulated neurons in mammalian and amphibian central nervous tissue. Watkins and colleagues wanted to work out whether these ions stimulated the neurons through specific neurotransmitter receptors. In the classic analogy, the receptors would be like locks, and glutamate and aspartate like neurotransmitter keys that trigger electrical nerve pulses.

The researchers expected that if this were the case, only l-amino acids would activate the receptor and trigger pulses. ‘The first substances we needed to test were therefore the d-forms of glutamate and aspartate,’ Watkins tells Chemistry World. ‘Disappointingly, these seemed to be similarly active’ to their l-isomer siblings, he recalls. So Watkins made different amino acids, changing their structure ‘to probe the surrounding molecular space of the sites of action’, he explains. In each case where there was a d-amino acid, it showed a stronger effect than the l-form. One, N-methyl-d-aspartate (NMDA), showed particularly high potency.

Even though d-amino acids might be the less common enantiomers, that scarcity enables them to play important roles in human bodies

This and other aspects of how these acidic amino acids excited neurons did not fulfil Watkins and colleagues’ expectations for a specific transmitter role. Many other scientists were also sceptical that glutamate was a neurotransmitter, partly because they assumed d-amino acids were unnatural. But gradually, scientists realised that the l-amino acid concentrations used in these early experiments rapidly fell, because tissues surrounding the neurons swallowed them up. d-amino acids lingered for longer, which was partly why they seemed more potent.

Eventually, in the 1970s and 1980s, Watkins and colleagues could directly measure how potently each amino acid bound to the receptor. ‘We were surprised to find stereoselectivity favouring the l form of glutamate and N-methyl glutamate, little stereoselectivity for either d- or l-aspartate, but selectivity mainly for the d-forms of the N-methyl and other derivatives of aspartate,’ he says. ‘l-glutamate had the highest affinity for the particular binding site being investigated and seemed the logical likely transmitter at this site.’

By then, this glutamate receptor had taken on the name that reflected its unexpected d-amino acid key, the NMDA receptor. It would later reveal further surprises showing that d-amino acids are far from ‘unnatural’. In the time since, scientists have increasingly realised that even though d-amino acids might be the less common enantiomers, that scarcity enables them to play important roles in human bodies. As researchers pay closer attention to this looking glass world, what they see is providing inspiration that might improve many people’s lives.

Benefiting from separation

At first, scientists thought d-amino acids were biologically unimportant or just trace impurities, explains Daniel Armstrong from the University of Texas, Arlington. In the late 1940s, researchers studying how the antibiotic penicillin killed Staphylococcus aureus realised that d-amino acids strengthen the bacteria’s defences. ‘The d-amino acids in bacterial cell walls act as a kind of armour,’ Armstrong says. Enzymes that plants or animals unleash on the bacteria when infection starts can’t chew their way through the unusual molecules that result.

D-amino acids

Reports of d-amino acids’ presence in plants followed, but ‘were largely ignored by the scientific community’, Armstrong adds. Eventually, researchers realised that some higher organisms had evolved to make molecules like the antibiotic vancomycin that specifically bind to bacterial cell wall d-amino acids and then kill them.

In the early 1980s, Armstrong developed new chiral chromatographic methods to separate enantiomeric amino acids. This approach passes dissolved mixtures down a column filled with a material that has a stronger physical attraction to either the l- or d-isomer. In Armstrong’s case this material was cyclodextrin. As Armstrong passed more solvent down the column, the varying attraction would mean that the dissolved isomers would exit the other end at different times. This approach enabled studies like one showing that urine has higher d-amino acid concentrations than blood. That demonstrated that kidneys selectively reabsorb l-amino acids from urine back into our blood.

At the same time, NMDA receptors threw up another d-amino acid surprise. In 1987, Jon Johnson and Philippe Ascher at the École Normale Supérieure in Paris, France, found that as well as glutamate or aspartate, another amino acid must be present at the same time to unlock NMDA receptors. NMDA receptors are the only receptor whose activation requires two molecules like this, explains Herman Wolosker from Technion, the Israel Institute of Technology in Haifa. This unique behaviour likely evolved from how NMDA receptors activate nerves by triggering them to let in calcium ions. If nerves let in too many calcium ions, they might die. ‘It may provide a fail-safe mechanism to prevent unwanted NMDA receptor activation and neurotoxicity, as if you need two different keys to launch a rocket so no one can launch it alone,’ Wolosker explains.

The very fact that d-amino acids are so uncommon may be why they can be useful

Ascher’s team first discovered that glycine was an NMDA receptor co-agonist – but d-serine followed shortly after. Again, received wisdom that d-amino acids are unnatural was an obstacle to exploring this finding further. But then, in 1992, researchers at the National Institute of Neuroscience in Tokyo, Japan including Toru Nishikawa and Noriko Fujii, developed a method to measure d-serine. They used chiral gas chromatography and mass spectrometry, passing mixtures down a column filled with a material attached to the amino acid l-valine. Studies had suggested that NMDA receptors work less well in people with schizophrenia. According to Wolosker, the Japanese scientists had intended to study whether d-serine could help remedy that. When they used the technique to analyse the serine naturally present in rat brains, however, they found about a quarter of it was the d-isomer.

Still surprising

Questions remained whether the rats had made d-serine in their bodies or consumed it in their food and drink. At the time, Wolosker was a post-doctoral researcher in Solomon Snyder’s team at Johns Hopkins University in Baltimore, US. He decided to try to figure out how the d-serine got into rats’ brains, publishing his findings in 1999. ‘I discovered serine racemase, an enzyme that makes d-from l-serine,’ he says. ‘That proved d-serine to be an endogenous amino acid.’

The very fact that d-amino acids are so uncommon may be why they can be useful, Wolosker adds. ‘d-amino acids have unique binding partners – most proteins do not recognise them,’ he says. For example, d-serine appears to be exclusively an NMDA receptor coagonist. By contrast, glycine is an NMDA receptor coagonist, but also binds to dozens of other biological targets. d-amino acids also last longer, a quality measured in terms of its half-life. ‘For example, the half-life of d-serine is at least 20 times slower than that of its L counterpart,’ says Wolosker. ‘d-serine could therefore mediate a longer-lasting signal.’

We still do not know how drugs targeting the increase of d-serine will perform

Wolosker’s Technion team has gone on to find more clues to d-serine’s importance in schizophrenia. The researchers found lower d-serine levels in patients’ cerebrospinal fluid associated with lower expression of serine racemase in the hippocampus. ‘Serine racemase has been identified as a risk gene for schizophrenia in large genetic studies of schizophrenia,’ Wolosker explains. ‘Several studies have shown that administration of d-serine to patients alleviates cognitive, positive and negative symptoms of the disease.’ Yet Wolosker urges caution about trying to treat schizophrenia by boosting NMDA receptor function. So far, large trials of drugs seeking to do this, for example by raising glycine levels, have failed. ‘Despite some success in small clinical trials, we still do not know how drugs targeting the increase of d-serine will perform in double-blind studies,’ Wolosker says.

NMDA receptors are also central to how we remember. That opens an opportunity for d-serine to help in post-traumatic stress disorder (PTSD). ‘In people with PTSD, trauma-related memories are reactivated, and they have deficits in erasing those memories,’ Wolosker notes. He explains that something similar is seen in mice genetically modified so that they can’t produce serine racemase. d-serine and d-cycloserine help mice to forget the fear that they’ve been trained to show, however.

Yet Wolosker still regularly talks to people who are surprised to hear that d-amino acids have such effects. ‘In neurobiology, especially people who study neurotransmitter channels, everyone has heard of d-serine,’ he explains. ‘But in biochemistry and chemistry, not much is known about it.’

Ageing concern

d-amino acids can also influence cancer because some cancer cells produce more NMDA receptors, explains Armstrong. Cancer cells also often have higher levels of d-amino acids than normal cells. ‘If one adds a drug that blocks NMDA receptors, cancer cell growth is inhibited or stopped,’ Armstrong explains. ‘However, we found that such drug-treated cancer cells immediately increased their production of d-alanine and d-aspartic acid thereby rescuing them from the effects of the drug.’

Beyond NMDA, d-amino acids are also linked to the condition with perhaps today’s greatest unmet need, Alzheimer’s disease. Amyloid beta protein makes up the sticky globules found in Alzheimer’s patients’ brains. It contains more d-aspartic acid and d-serine than other proteins in the body, explains Armstrong. There are three possible sites where amyloid beta contains aspartic acid residues and two for serine residues. Until recently, where exactly the d-amino acids were more common was unknown, but Armstrong’s team has developed two different methods that allow them to locate and quantify each. Their study showed that the d-amino acids made amyloid beta more hydrophilic.

The team found a huge amount of d-aspartate in insoluble materials in old eye lenses

Other studies discovering d-aspartate in proteins used methods developed by Noriko Fujii, who died in August 2021, and her team at Kyoto University in Japan. After joining the team that discovered d-serine in rats’ brains in 1992, Fujii had switched to focus on d-aspartate. Her team had found that d-aspartate forms from l-aspartate inside people’s bodies during their lifetimes, Takumi Takata, one of her Kyoto University colleagues, explains. Proteins integrate d-aspartic acid building blocks as we age, which is especially important in causing cataract formation.

The change from l-aspartate into d-aspartate alters ‘protein structures and protein–protein interaction’, explains Takata. There is a ‘highly concentrated protein soup’ of proteins in the eye’s lens, he adds. Normally, the soup is clear and stable. But as it breaks down, protein solidifies and forms cataracts that scatter light, making it harder to see. The Kyoto University team found ‘a huge amount of d-aspartate’ in insoluble materials in old eye lenses, and much less in clear, young lenses.

Bizarre and hard to ignore

d-aspartate also seems to have an important but unknown function in our early lives, explains Silvia Sacchi, from the University of Insubria in Varese, Italy. Before a human embryo is 14 weeks old, scientists have found more d-aspartate than l-aspartate in some brain areas. Whatever the d-aspartate does, we need it only temporarily. After birth an enzyme called d-aspartate oxidase (DDO), which Sacchi studies, reduces its concentrations rapidly. Alessandro Usiello from the University of Campania in Italy, and colleagues including Sacchi, found that at first increased d-aspartate benefits young mice. The team genetically engineered the mice not to produce DDO. To start with, the mice’s brains functioned better.

That came together with a toxic effect. ‘d-aspartate has a gentle face, but it’s a bizarre molecule,’ comments Sacchi. ‘When its levels are persistently elevated, then you have the other side of the coin. Now you start to notice negative effects. It’s like the brain gets older more quickly. Performance starts to get worse due to these persistent high levels of d-aspartate.’ Working with Usiello and others, Sacchi showed that this is because d-aspartate raises glutamate release to levels that let in enough calcium ions to kill off neurons.

d-aspartate has a gentle face, but it’s a bizarre molecule

While we don’t yet understand its early role well, d-aspartate offers another route to help in schizophrenia. Genes that regulate DDO production have been linked to schizophrenia, Sacchi explains. Meanwhile, olanzapine is a widely used drug to treat schizophrenia, which interacts with many types of brain protein. Sacchi and her colleagues have found that olanzapine blocks DDO’s function. ‘It was really amazing, these antipsychotics, by inhibiting the enzyme that degrades d-aspartate could induce positive effects on patients,’ she says. ‘We found molecular mechanisms that explain the effect of the antipsychotics in new findings, which was previously not expected.’

Researchers need to do more work before we can exploit most such findings, but one advanced application sees drug companies borrowing ideas from bacteria. ‘Already, pharmaceutical companies are developing peptide drugs that incorporate d-amino acids,’ Armstrong explains. Our bodies can’t break peptide drugs down so easily, meaning we can take smaller doses. For example, two Chinese companies, Shanghai Duomirui Biotechnology and Chia Tai Tianqing Pharmaceutical Group, are developing antimicrobial peptide drugs containing d-amino acids.

It’s clear that far from being unnatural, as chemists once thought, d-amino acids are important in their scarcity. Armstrong doesn’t doubt that these long-overlooked enantiomers are starting to get the recognition they deserve. ‘Discoveries involving the presence, role and importance of d-amino acids in biological systems are increasing exponentially,’ he says. ‘It is becoming impossible to minimise or ignore this important biochemical area.’

Andy Extance is a science writer based in Exeter, UK