Why all living cells emit a faint glow

Hello Nature readers,
Today we see evidence that a ‘timer’ in the brain tracks sleep, explore why all living cells emit a faint glow and consider who’s accountable when physicians and AI work together.

If you enjoy this newsletter, please consider recommending it to a friend or colleague. Click here to forward it by e-mail. Thank you!

‘Timer’ in mouse brain tracks sleep

Chemical tags on proteins in a mouse’s brain serve as a marker of both the length of a sleep session and how likely an animal is to wake up at a given moment. An enzyme adds these tags to proteins on the surface of brain cells, and their levels follow a predictable rhythm, researchers found. The extent of protein tagging stayed high and steady when mice were awake and began a slow decline when they fell asleep, spiking if they experienced a brief interruption in their slumber.

Consciousness research ponders AI

The debate over whether artificial intelligence (AI) systems might become conscious has inspired some tech firms to hire philosophers. But researchers still haven’t agreed on what gives rise to consciousness in humans, let alone in an AI chatbot — and some say the issue is a red herring. “What we might see is a sort of capture of consciousness research by the AI sector,” says consciousness scientist Anil Seth. On the plus side, the influx of interest might help to validate a field that was not taken seriously as a scientific endeavour for years — and bring in all-important funding.

Alarm over ‘young mitochondria’ treatment

A small number of fertility clinics around the world are offering controversial mitochondrial replacement therapy (MRT): a procedure to furnish clients’ embryos or eggs with donated mitochondria. The theory behind the treatment — which remains unproven in rigorous clinical trials — is that the donor’s mitochondria might help the embryo to mature. “I’m pretty horrified,” says fertility specialist Richard Anderson, who reviewed the MRT offering of one of the clinics identified by Nature. He says that the technique is difficult, the risks are high and the benefits are uncertain.

The faint glow of life

All living cells emit a faint glow, thanks to ‘biophotons’ produced by their metabolic activity. In the past, researchers have theorized that cells might use this glow as a way of communicating with one another, but the idea was dismissed as pseudoscience. Now, the study of biophotons is making a comeback, with some researchers hopeful that a cell’s glow could be a subtle marker of health, or a way of sending signals to its neighbours. “It’s one of those fields that feels like it is on the edge of respectability,” says biochemist Nick Lane.

A kaleidoscope of scientific illustrations

In The Art and Science of Ernst Haeckel, science historian Julia Voss and zoologist Rainer Willmann bring together 450 of the polymath’s drawings and watercolours of the natural world. Over his long and varied career, Haeckel identified thousands of species, many of which he depicted in the detailed and colourful illustrations that the book contains. Although his racist and eugenic personal views should be deplored, “it is still possible to be impressed by the sheer breadth of Haeckel’s artistry”, writes science journalist Josie Glausiusz in her review.

Who’s liable when AI treats patients?

“With the benefits of AI tools come fresh patient-safety risks — and questions about who is responsible when things go wrong,” write a group that includes physician-scientist Eric Topol. This legal uncertainty could cause patients to fall into liability gaps when things go wrong, and hospitals to avoid useful tools because of concerns about murky legal risks. The authors define seven levels of AI capabilities — similar to the graded levels used by regulators for aircraft and self-driving cars — to help regulators, policymakers, governments, courts and health-care bodies to plan for future medical-AI systems.