Researchers in Japan have developed an electrical method that tells mirror-image forms of amino acids apart one molecule at a time, a step toward compact instruments that could one day help search for signs of life beyond Earth.
The work, from Osaka University, pairs a nanoscale gap between gold electrodes with artificial intelligence.
Why handedness matters
Amino acids, the building blocks of proteins, come in two mirror-image forms called L and D. Living organisms use almost exclusively L-form amino acids, while their sugars are in the D-form. Nonliving chemical and physical processes produce both amino acid forms in equal amounts, making the L/D ratio a potential biosignature.
The authors stress it is one clue and should be weighed with other evidence. Earlier reports of excess L-forms in meteorites are now generally considered contamination from Earth, they note, and NASA and JAXA asteroid sample-return missions found no significant excess.
Counting molecules one at a time
Traditional methods analyze large groups of molecules. The new approach counts them individually. As a molecule passes through a gap between two gold nanowires, roughly half a nanometer wide, it creates a tunneling current. Each amino acid form produces a different current waveform, which machine learning then classifies.
The method needs no optical measurements or chemical reagents, and the researchers say electrical detection could simplify instrument design and reduce sensitivity to vibration.
“By combining our nanogap tunneling technique with artificial intelligence, we were able to distinguish between the L- and D-forms of amino acids with over 80% accuracy,” said lead author Takahito Ohshiro. “This is the first discrimination of amino-acid chirality at the single-molecule level and constitutes a fundamental advance in chemical sensing.”
Meteorite and desert tests
A blind test of 39 forms, covering 19 chiral amino acids in L and D versions plus glycine, which has no mirror image, identified both identity and chirality with accuracy above 50%, well above chance. Mixtures of four amino acids with different chiral makeups were quantified with accuracy above 66%.
Because samples of interest typically contain many molecules, the team analyzed extracts from the Murchison meteorite and soil from two Atacama Desert sites in Chile. “Our method was comparable to traditional methods, as both were capable of capturing the major features of amino acid composition,” said senior author Masateru Taniguchi.
The paper describes a targeted approach rather than full profiling, focused on a preselected panel of 11 forms of six amino acids. Glutamic acid was not confidently identified in the desert samples. Agreement with liquid chromatography-mass spectrometry varied: ratios matched well for aspartic acid in the Murchison sample but differed for alanine and valine.
A long way to flight
The authors say practical use in space missions will require more validation with contamination-controlled tests and software trained on common contaminants. In calibration tests, the sensor detected tryptophan at 0.03 nanomolar, below the roughly 1 nanomolar target proposed for life-detection measurements, though they warn that figure does not reflect performance in complex samples.
The chip tolerates temperatures above 572 degrees Fahrenheit (300 degrees Celsius), so it could travel on a spacecraft without cooling.
The work was published Oct. 5 in Nature Communications.