The transcript weaves together three major advances in modern biology.
**LUCA (the Last Universal Common Ancestor)** – By analysing the genomes of 700 bacteria and archaea, researchers reconstructed a probabilistic gene tree that suggests LUCA lived about 4.2 billion years ago, shortly after Earth became habitable. LUCA was not a single cell but a diverse population possessing a phospholipid membrane, basic genome‑maintenance and protein‑synthesis machinery, a metabolism that could use H₂ and CO₂, and—surprisingly—a CRISPR‑Cas antiviral system, indicating it already faced viral pressure and was part of a complex microbial ecosystem. Its descendants split into the prokaryotic (bacteria/archaea) and eukaryotic lineages that give rise to all life today.
**Brain‑body immune communication** – Work on the vagus nerve revealed a bidirectional “inflammatory homeostat”: specific brainstem neurons act like a volume dial, sending pro‑ and anti‑inflammatory signals that modulate peripheral immunity. This anti‑inflammatory reflex, first hinted at by Kevin Tracey’s vagus‑nerve stimulation experiments, links neural activity to diseases such as multiple sclerosis, lupus, diabetes and obesity, opening avenues for neuro‑immune therapies.
**Protein‑structure prediction and design** – DeepMind’s AlphaFold 2 (and later AlphaFold 3) used deep‑learning transformers to predict 3‑D protein structures from amino‑acid sequences with near‑experimental accuracy, effectively solving the decades‑old protein‑folding problem. Complementary tools such as RoseTTAFold All‑Atom now model protein‑small‑molecule assemblies. These AI‑driven breakthroughs enabled rational protein design for medicine, energy, sustainability and new technologies, earning the 2024 Nobel Prize in Chemistry for Demis Hassabis, John Jumper and David Baker.
Together, these studies illuminate life’s deepest origins, reveal how the brain directly tunes immunity, and show how artificial intelligence is reshaping our ability to read, understand and engineer the molecular machines of life.
1. LUCA is the last universal common ancestor of all modern life on Earth.
2. LUCA lived billions of years ago.
3. LUCA is not the origin of life but represents the emergence of life as we know it today.
4. LUCA likely existed as a population rather than a single individual.
5. Many different types of organisms lived at the time of LUCA; LUCA survived and gave rise to all modern life.
6. Modern cellular life is divided into two domains descended from LUCA: prokaryotes and eukaryotes.
7. Prokaryotes include bacteria and archaea, which are comparatively simple cells.
8. Eukaryotes are more complex and include all forms of complex multicellular life.
9. A 2024 study produced the most complete picture of LUCA using evidence from many scientific disciplines.
10. The study estimated LUCA’s genome likely encoded about 2,600 proteins.
11. LUCA was a complex organism, perhaps slightly smaller than a modern bacterium.
12. LUCA possessed a simple phospholipid membrane and the machinery to maintain a genome and build proteins.
13. LUCA could metabolize hydrogen gas and carbon dioxide.
14. LUCA had a CRISPR‑Cas system, a basic immune defense against viruses.
15. LUCA may have obtained energy from hydrothermal vents, atmospheric gases, or the chemical waste of other microbes.
16. LUCA likely participated in a complex ecosystem, exchanging metabolic products with other organisms.
17. LUCA’s age is estimated to be around 4.2 billion years old.
18. This age places LUCA shortly after Earth cooled and became habitable.
19. The vagus nerve is a two‑way bundle of neurons connecting the brain to many internal organs.
20. The vagus nerve contains sensory neurons that carry body‑to‑brain signals and motor neurons that carry brain‑to‑body signals.
21. Via the vagus nerve, the brainstem regulates basic survival functions such as breathing, heart rate, and hunger.
22. Homeostasis is a term often associated with vagus nerve activity; deviations trigger signals that restore equilibrium.
23. In 1921, Otto Loewi showed that stimulating a frog’s vagus nerve slowed its heartbeat and identified the signaling chemical “vagusstoff,” later earning a Nobel Prize.
24. The vagus nerve includes some of the longest neurons in the body and links the brainstem to the stomach, lungs, and heart.
25. In 2020, Charles Zuker’s lab found that mice perceive sweetness not only from the tongue but also from the stomach.
26. In the 1990s, Kevin Tracey demonstrated that electrical stimulation of the vagus nerve reduced inflammation, calling this the “anti‑inflammatory reflex.”
27. Inflammation is the immune system’s first response to injury or infection, releasing pro‑inflammatory molecules.
28. Anti‑inflammatory molecules later curb the response to prevent tissue damage.
29. Excessive inflammation is linked to diseases such as multiple sclerosis, type‑one diabetes, lupus, obesity, and diabetes.
30. Zuker’s team identified mouse brainstem neurons that control inflammation like a volume dial, able to increase or decrease the response.
31. This was the first demonstration that immune‑system control resides in the brain.
32. Proteins are microscopic molecular machines that carry out vital chemical reactions in cells.
33. Understanding protein function requires knowledge of their three‑dimensional structure.
34. For over 50 years, scientists have studied how proteins fold into functional shapes.
35. Determining a single protein structure can cost about $100,000 and take a couple years of a PhD student’s effort.
36. DeepMind’s AlphaFold 2 uses a neural network to predict a protein’s 3D structure from its amino‑acid sequence.
37. For many proteins, AlphaFold 2’s prediction accuracy is 99 %.
38. DeepMind’s AlphaFold breakthrough made biology more computational.
39. AI now enables solving problems previously inaccessible experimentally, such as designing novel proteins for global challenges.
40. A protein’s specific function is determined by its folded three‑dimensional shape.
41. Proteins fold into functional shapes in less than a second despite astronomical numbers of possible configurations; this is the protein‑folding problem.
42. Google DeepMind was founded in 2007 to advance deep‑learning AI.
43. After mastering Go, DeepMind took on the CASP (Critical Assessment of Structure Prediction) challenge.
44. Experimental chemist John Jumper led the DeepMind team that created AlphaFold 2.
45. AlphaFold 2 was trained on data from over 100 000 known proteins, including sequences, structures, and evolutionary information.
46. The network employs transformer architectures and outputs confidence scores for its predictions.
47. In 2020, AlphaFold 2 entered CASP14 and produced the top‑ranking structure predictions.
48. By July 2022, DeepMind released structure predictions for 218 million proteins, nearly all known proteins.
49. Some scientists described the protein‑folding problem as essentially solved after this release.
50. At the University of Washington, David Baker has worked on software to solve the protein‑folding problem for three decades.
51. Baker aims to design new proteins not found in nature to address medicine, energy/sustainability, and technology challenges.
52. To create a novel protein, researchers first design the desired 3D shape, then use AI to derive the amino‑acid sequence before lab synthesis.
53. This approach enables the production of more sophisticated, precise, and safe proteins.
54. Beyond medicine, Baker’s lab works on better sunlight capture and toxic‑compound degradation methods.
55. The next frontier is predicting protein‑protein and protein‑molecule interactions inside whole cells.
56. In spring 2024, the Baker Lab released RoseTTAFold All‑Atom, which predicts 3D structures of protein assemblies and small molecules.
57. Shortly after, DeepMind released AlphaFold3, offering further improvements.
58. These advances are expected to unlock new scientific discoveries.
59. The combined work of DeepMind, Baker, and colleagues has revolutionized protein science.
60. In October 2024, the three groups were jointly awarded the Nobel Prize in Chemistry for this leap forward.