"Unlocking the Secrets of LUCA: Tracing Our Ancestors and Understanding Protein Science"

• You have an ancestor that lived billions of years ago named LUCA.
• LUCA is the last universal common ancestor to all modern life on Earth.
• Researchers painted the most complete picture of LUCA using various scientific disciplines.
• LUCA was likely complex, similar to modern bacteria, living around 4.2 billion years ago.
• The vagus nerve connects the brain and the body's immune system, regulating responses.
• Recent advances in AI, particularly AlphaFold, have revolutionized our understanding of protein folding.

You have an ancestor that lived billions of years ago who goes by the name of LUCA. LUCA is the last universal common ancestor to all modern life on Earth, including us, bacteria, frogs, fish, trees, fungi—everything that is alive that has a cell.

In a 2024 paper, researchers peered back in time to paint the most complete picture of LUCA yet. Using evidence from a wide range of scientific disciplines, they developed a holistic understanding of when LUCA existed and how it interacted with the environment.

So what was LUCA like? Scientists realized that all life on Earth shares certain characteristics suggesting a common ancestor. Darwin, in fact, mentioned in one of his letters that tracing evolution back to the very beginning implies the existence of some common ancestor. Thus, it was really Darwin who first started thinking about what that universal ancestor could have looked like.

LUCA was not the origin of life on Earth but marked the emergence of life as we know it today. Additionally, LUCA was not alone; we can think of LUCA more as a population rather than a single individual. Different types of organisms likely lived at that time, whose descendants died out soon thereafter. However, LUCA survived and evolved into all modern life.

Today, there are two domains of cellular life descended from prokaryotes and eukaryotes. Prokaryotes include bacteria and archaea, which are comparatively simple cells, while eukaryotes are more complex and include all forms of complex multicellular life.

Most of us are paleontologists interested in early evolutionary history; however, we can only get back to this ancestral point by examining modern organisms to unveil our evolutionary history. Edmund Moody, Phil Donahue, and their team sought to reconstruct LUCA’s genome—or collection of genes—and then build a working metabolic network from it, which helped them understand what LUCA would have been like.

Their approach involved filling out an evolutionary family tree based on the genomic relationships among 700 living bacterial and archaeal species. They aimed to infer a tree of these 700 genomes, requiring genes that had evolved slowly and been conserved over billions of years. Using software and immense computing power, the team constructed a probabilistic gene tree leading back to LUCA.

This work led them to estimate that LUCA's genome likely encoded 2,600 proteins. Their main conclusion was that LUCA was a complex organism, very similar, perhaps a bit smaller, than a modern-day bacterium. For instance, LUCA would have had a simple phospholipid membrane and the molecular machinery necessary for maintaining a genome and building proteins.

It could metabolize hydrogen gas and carbon dioxide. Interestingly, they found something surprising: this CRISPR-Cas system, a basic immune system that cells use to combat viral attacks. Some might be surprised that viruses existed that long ago. LUCA may have thrived on non-living sources such as hydrothermal vents or atmospheric gases, or perhaps on chemical waste from other microbes. It seems more likely that LUCA participated in a complex ecosystem, exchanging metabolic products with various entities.

For a complete picture, the team also needed to determine when LUCA lived. Thus, they used molecular and paleontological methods to estimate its age, arriving at around 4.2 billion years old. This timing is relatively soon after Earth cooled enough to be habitable, which is surprising given LUCA's complexity.

The vagus nerve is a long bundle of neurons that serves as a two-way information superhighway connecting the brain to many internal organs. An array of stimuli from the body can evoke various behavioral and physiological responses, while the brain can also regulate the body.

Recently, a team of scientists published research revealing a surprising new connection between the brain and the body's immune system. It's clear that diseases previously thought to be confined to the body will emerge as diseases of body-brain communication.

In 1921, physiologist Ottol Lovi discovered that stimulating a frog's vagus nerve slowed its heartbeat. He named the signaling chemical Vagusof and later won the Nobel Prize for this discovery. The vagus nerve contains some of the longest neurons in the body, connecting the brain stem to vital organs, including the stomach, lungs, and heart. It comprises both sensory neurons, which transport information from the body to the brain, and motor neurons, which go the opposite direction.

The brainstem regulates basic survival systems such as breathing, heartbeats, and hunger. One term often used with the vagus is homeostasis. Any deviations from that set point in the body will prompt signals from the body-brain to restore equilibrium.

In 2020, Charles Zuer's lab stumbled upon a remarkable brain-body connection. Their studies revealed that in mice, the sensation of sweetness comes not only from the tongue but also from the stomach. This opened new avenues for exploring body-brain signaling. The immune system is a fantastic example of this communication.

Previously, there was no evidence that the brain communicated directly with the immune system. However, experiments by neurosurgeon Kevin Tracy in the 1990s pointed to a possible connection. His groundbreaking work involved coarse electrical stimulation of the vagus nerve, revealing that this stimulation dampened the immune response, a phenomenon he termed the anti-inflammatory reflex.

Inflammation marks the immune system's first response to injury or infection, with pro-inflammatory molecules released to combat pathogens. Yet, anti-inflammatory molecules also need to be released to manage this response, preventing damage to the body's own tissues. An exaggerated inflammatory response has been associated with numerous diseases and disorders, including multiple sclerosis, type 1 diabetes, lupus, and various metabolic disorders.

In a series of experiments, Zuer's team set out to determine if there was a mechanism in the brain that controls inflammation. They hypothesized that some form of homeostatic control must exist. Initially, they identified groups of neurons in mice activated by immunological challenges. They then wondered what would occur if they used genetic techniques to manipulate these neurons, either exciting or inhibiting them.

To their surprise, they discovered that these neurons in the brainstem functioned like a volume dial for inflammation. They could turn the inflammatory response up or down, signifying which neurons carried pro-inflammatory and anti-inflammatory signals. This was the first time control of the immune system was located in the brain. Further research on this inflammatory homeostasis in humans could lead to new treatments for various diseases linked to inflammation.

If we could find ways to tap into and control the activity of different neuron types, we might unleash profound effects, thereby opening new avenues to address these conditions.

These are microscopic molecular machines essential to life on Earth. They've evolved over millions of years to perform a vast array of vital functions. These are proteins—the molecules that perform work by interacting with other molecules, building new ones, breaking them down, and facilitating chemical reactions. To grasp this chemistry of life, one must understand the structure of these molecules.

For more than half a century, biologists have sought to unravel the enigma of how proteins fold to function—a monumental effort costing around $100,000 and taking years of a PhD student's time to obtain even a single structure.

During a recent grand challenge, a team at DeepMind leveraged artificial intelligence to address part of the protein puzzle. They trained a neural network to interpret the one-dimensional molecular sequence of a protein and predict its three-dimensional structure. For many proteins, AlphaFold2 boasts an accuracy of 99%.

DeepMind's breakthrough has ushered in a new era of biology. Many believe biology is transitioning to a computational science, and this is certainly true. The AI revolution is enabling researchers to tackle problems previously unsolvable through experimental means, including the engineering of human-designed proteins to confront some of humanity's biggest challenges.

A protein's specific molecular function results from its three-dimensional folded shape. Proteins fold into precise shapes, and this shape carries out biological functions. If we know the 3D structure, we can begin to understand how these molecules behave and function. This origami-like shape results from the sequence of its primary structural components, amino acids.

All proteins are constructed from 20 different flavors of amino acids connected in chains called polypeptides. Proteins are initially assembled inside cells as unfolded chains, their amino acids strung together like beads on a necklace. These amino acids can be arranged in countless configurations to form various proteins, with the specific sequence encoded within a cell's DNA.

In 1969, biologist Cyrus Leventhal observed a paradox: for any given protein, even small ones, there exists an astronomical number of possible folding configurations. Yet, proteins consistently fold into functional shapes in less than a second—the mystery behind this became known as the protein folding problem.

Google DeepMind was founded in 2007 to advance a nascent form of AI called deep learning. After the AI systems successfully mastered Go and other games, DeepMind's founder Demis Hassabis sought new challenges. Although we are early in AI development, many problems remain unsolved.

DeepMind entered what has been termed the Olympics of protein folding, known as the Critical Assessment of Structure Prediction Challenge (CASP). Participants strive to predict proteins’ three-dimensional structures from amino acid sequences using computer algorithms.

Experimental chemist John Jumper led the team that developed their protein structure prediction algorithm, AlphaFold. They trained a deep learning neural network with data describing over 100,000 known proteins, including both their sequences and their folded structures, plus evolutionary data about the proteins.

What's called inductive bias in machine learning allowed the algorithm to learn extraordinarily rapidly from this data. The data is processed by powerful neural networks known as transformers. After cycling through the entire algorithm, AlphaFold predicts a structure and assigns a confidence score to different parts of the protein structure.

In 2020, DeepMind entered AlphaFold in CASP14, and its predictions came out on top—remarkably so. By July 2022, DeepMind released the structure predictions for 218 million proteins, almost all known in the world. Some consider the protein problem essentially solved.

For many biologists, it took months or years to accept these results. However, at the University of Washington, David Baker had spent three decades developing software to solve the protein folding problem. At that point, they sought to take these concepts and apply them to protein design, which involves synthesizing new and novel proteins.

For their work on novel protein creation, researchers must make synthetic genes encoding these proteins. Baker's goal is to design new proteins that don't exist in nature to address tough challenges facing humanity.

Their research broadly falls into three areas: medicine, energy, and new technology for designing new proteins. Research teams run the protein folding challenge backwards; rather than predicting a 3D shape from a sequence, they design a novel protein's 3D shape and use AI tools to generate the corresponding amino acid sequence before synthesizing the protein.

In the lab, they can now design proteins that are much more sophisticated, precise, and safe. Beyond medicine, they're also working on improved methods for harnessing sunlight and degrading toxic compounds.

The next frontier in AI protein science entails predicting interactions among proteins within an entire cell. Proteins operate as the cellular machinery, interacting with many different molecules, including DNA, RNA, and metals.

Baker's team, DeepMind, and others have initiated the development of AI algorithms capable of predicting these intricate interactions. In the spring of 2024, the first generation of AI prediction tools debuted, with Baker’s lab releasing Rosetta Fold All Atom to predict 3D structures of protein assemblies and small molecules. Soon after, DeepMind introduced AlphaFold 3, enhancing capabilities and paving the way for new scientific breakthroughs.

The collaborative efforts of these researchers and their teams have revolutionized the study of proteins. In October 2024, they were jointly awarded the Nobel Prize in Chemistry to honor their remarkable advancements.

You always want to push the frontier during a technological transition. It's a very intense time filled with excitement. There remains so much yet to be understood. Truly, this is only the beginning.