Life
in a Hurry

The Whole Thing in One Page

Life is easy to recognise and difficult to define. A bacterium is alive. A pebble is not. Then the border gets crowded. A virus evolves but cannot reproduce without a host cell. A dormant seed can spend years doing almost nothing. A mule is alive but sterile. A flame consumes fuel, grows and spreads, yet nobody mistakes it for an organism. The difficulty is telling us something useful: life is not one property hidden inside matter.

All known independent life is cellular, and every cell solves the same linked problems. It keeps an inside different from the outside. It obtains matter and usable energy. It controls reactions with catalysts. It stores information that can influence future copies. It repairs enough damage to remain organised. It produces descendants in which inherited differences can alter success. These functions can exist separately outside biology. Life is what happens when they become one continuing system.

That is why the origin of life is harder than making life's ingredients. Carbon chemistry is common. Amino acids can form without organisms. Fatty molecules can assemble into vesicles. Nucleotide precursors can arise through plausible prebiotic routes, and increasingly capable RNA catalysts can perform parts of the copying problem. Each result closes part of the gap. None by itself turns geology into a self-sustaining lineage.

The unsolved problem is integration. A replicator that cannot secure feedstocks runs down. A reaction network without heredity cannot preserve an improvement. A membrane that encloses useful chemistry but cannot grow or divide has no descendants. A compartment that divides but does not pass on the chemistry responsible for its success loses the thing selection would need to favour. Before modern cells existed, crude versions of these functions had to begin helping one another.

The clock is badly blurred. Earth formed about 4.54 billion years ago and liquid water was present surprisingly early. Strong geological evidence places life by about 3.5 billion years ago. Older claims exist, but they are harder to interpret. Molecular reconstruction reaches further back in a different way: a 2024 analysis placed the last universal common ancestor, LUCA, near 4.2 billion years ago and reconstructed it as a surprisingly elaborate prokaryote-grade organism. That estimate is model-dependent, but the larger conclusion is secure. LUCA was already the product of a long history. It was not the first living thing.

So there may never have been a dramatic instant at which dead matter crossed a bright line. The transition could have passed through systems that were partly lifelike: autocatalytic chemistry, evolving polymers, leaky compartments and networks that borrowed physical cycles from their surroundings before internalising them. Once inherited variation began changing which systems persisted, chemistry acquired memory. From then on, successful arrangements could accumulate rather than vanish with the molecules that first produced them.

The same uncertainty follows us into space. Liquid water marks opportunity, not life. Organic molecules are ingredients, not organisms. Oxygen, methane and other possible biosignatures have non-biological routes. The strongest detection will be a contextual pattern that is difficult to maintain without a process continually rebuilding it.

Life is matter organised strongly enough to maintain itself, reproduce consequential information and enter history. Its beginning was probably a transition, not a spark. That is the book.

Why You Should Care

A living cell is less permanent than it looks. Its ATP is turned over rapidly. Proteins are made and degraded. Ions cross membranes and are pumped back. DNA is repaired. Lipids move. Food and oxygen enter; carbon dioxide, heat and waste leave. The matter changes continuously while the organisation persists. You are therefore not a static object made from biological material. You are a maintained process.

Death exposes the distinction. Immediately after a cell dies, its DNA may still be readable, its membrane may still be present and many enzymes may still work in isolation. What has failed is the coordinated ability to restore gradients, repair damage, regulate reactions and rebuild components. A dead cell can contain the same kinds of molecules as a living one. The difference lies in what the system can still do with them.

That is why biological border questions are so stubborn. A virus has a genome and evolves, but outsources translation and much of its replication machinery to a host. A bacterial spore can suspend most measurable activity and later restart. A sterile worker ant cannot found a lineage alone, but is part of a reproductive colony. An embryo stored at low temperature can pause development without ceasing to belong to a living lineage. The question 'is it alive?' often becomes clearer when replaced by two others: what is the unit, and which functions are active, retained or borrowed?

The origin of life is the largest version of the same problem. Popular accounts often give it one decisive ingredient: a lightning strike, a warm pond, a vent, RNA. Those images survive because events are easier to picture than systems. Yet the hard steps sit between the celebrated experiments. A molecule must be made, concentrated and kept from destructive side reactions. A polymer must copy with enough fidelity to preserve information but enough error to vary. Useful products must remain associated with the sequences or compartments that produced them. Energy must drive synthesis without merely breaking things apart. Descendants must inherit enough of the successful arrangement for selection to work on it.

Seen this way, abiogenesis stops looking like a demand for a miracle and becomes a set of engineering constraints imposed by chemistry. That does not make the answer known. It makes ignorance more precise. We can ask which environments solve concentration, which molecules can carry heredity, how primitive membranes exchange nutrients, how gradients can be coupled to synthesis, and when replication becomes open-ended evolution.

The field also changes quickly enough that a fixed textbook picture is risky. Recent work has joined functions once studied separately and made RNA copying more capable than older summaries allowed. The pattern of progress is revealing: an apparent barrier falls, then the supports still supplied by the laboratory expose the next barrier more cleanly. Origin research advances by turning one large mystery into smaller engineering problems that can be attacked in sequence.

The same discipline matters when we search for life elsewhere. Mars preserves evidence of ancient rivers. Europa and Enceladus probably contain deep oceans. Exoplanet atmospheres can be studied for gases that living systems might maintain. None of that licenses a shortcut from habitability to habitation. Water is common. Organic molecules are common. Geological processes make methane. Photochemistry can make oxygen. A convincing detection will need independent observations that support one biological explanation better together than apart.

By the end of this hour, you should be able to distinguish a definition from a diagnostic, an ingredient from a system, replication from reproduction, LUCA from first life, and a plausible origin pathway from a laboratory trick. More importantly, you should know why nobody yet possesses a complete account of how chemistry became biology, while also knowing why the problem is scientifically tractable rather than mysterious.

The Core Ideas

1. Life Is Better Described by What It Does

Definitions become interesting at the edge. Nobody needs a committee to decide whether an oak is alive. Trouble begins with things that possess some biological properties and lack others. Viruses carry genes and evolve but depend on cells for translation. Viroids are smaller still: naked RNA molecules that exploit plant cells and can evolve without encoding proteins. Spores can become almost metabolically silent. Sterile organisms remain alive despite never producing offspring. A single trait will either exclude an obvious member or admit something we do not want to call living.

This has produced hundreds of proposed definitions. NASA often uses an operational one in astrobiology: life is a self-sustaining chemical system capable of Darwinian evolution. It is compact and experimentally useful, especially because heredity and evolution distinguish life from many merely complicated chemical processes. It is not a universal verdict. 'Self-sustaining' is awkward for obligate parasites. Darwinian evolution belongs to populations over generations, not to an individual cell considered alone. Definitions depend partly on what the classifier is trying to do.

For this book, the better approach is to separate definition from description. All known independent cellular life performs a recognisable cluster of functions. A cell maintains a boundary. It exchanges matter and energy with its environment. It catalyses reactions. It stores and expresses information. It repairs or replaces damaged components. It produces descendants or contributes to a lineage that does. Across generations, inherited differences alter which variants persist.

No single item is exclusive to biology. Soap molecules form membranes. Fire consumes fuel. Crystals grow. Computer code can be copied. Chemical networks can be autocatalytic. What distinguishes a cell is that the functions are causally tied together. Metabolism supplies the materials and energy needed to maintain the boundary and copy information. Information influences catalysts. Catalysts build and repair the machinery that keeps metabolism going. The boundary keeps enough of the products near the system that produced them. Reproduction passes the arrangement into new units.

The unit matters. A worker ant is sterile, but the colony has a reproductive lineage. A virus is inert outside a host but participates in Darwinian evolution across infections. Mitochondria were once free-living bacteria and now reproduce only inside eukaryotic cells. Biology contains nested dependencies, so arguments over whether one component is 'alive' can become disputes over where to draw the system boundary.

This is not an excuse to make the word meaningless. There are clear exclusions. A flame does not preserve a heritable sequence that accumulates adaptations to new fuels. A crystal can copy structural order during growth, but its defects do not normally support open-ended cumulative evolution. A thermostat regulates temperature but does not rebuild the machinery that measures and controls it. The important contrast is not organic versus inorganic matter. It is organisation capable of maintaining and transmitting consequential information.

That distinction also clarifies death. Killing a bacterium does not instantly erase its genome or dissolve every enzyme. The parts can remain while the network loses the ability to restore its own operating conditions. Life is therefore not an inventory. A freezer full of purified proteins, lipids and DNA contains more biological material than many living microbes and is not alive.

For origins, the consequence is decisive. There need not have been a first object possessing every modern criterion at once. Chemistry could move through partial systems: an autocatalytic network without templated genes, an RNA population without cells, a vesicle that grows and divides but has little heredity, or a compartment whose internal chemistry begins to influence its own persistence. The transition becomes biologically important when successful variations can be inherited strongly enough for cumulative selection to improve the system.

A definition gives us a label. The origin problem asks how a label became necessary.

2. Energy Keeps It Away from Equilibrium

Leave a living cell without an energy source and it does not settle into a quieter version of life. It dies. The reason is physical. Chemical systems tend towards equilibrium, a state in which available differences have been spent and no net work can be extracted. Life persists by remaining away from that state. It continually takes in concentrated energy or reactive chemicals, uses them to build and maintain structure, and exports lower-quality energy and waste.

This does not evade the second law of thermodynamics. A refrigerator keeps its interior cold by releasing more heat into the kitchen. A cell maintains local order by increasing disorder beyond itself. The accounting works because the cell is open. Sunlight arrives, food is oxidised, ions flow, heat leaves. The system can become more organised while its surroundings pay the larger bill.

Modern cells handle this through coupling. A reaction that releases usable energy drives another that would not proceed on its own. ATP is the best-known intermediary. Cells make ATP by attaching phosphate to ADP, then spend it by breaking that bond and linking the release to movement, synthesis, transport or signalling. ATP is not a storehouse filled for years. It is working currency turned over continually. A resting human uses an amount each day comparable to body mass because the same small pool is recycled again and again.

The deeper mechanism lies in membranes. Cells pump charged particles, often protons, across a membrane, creating an electrochemical difference. When the particles flow back through ATP synthase, a molecular rotary machine, the flow drives ATP production. Plants and many microbes use light to establish gradients. Animals harvest energy from food through respiration. The fuels differ; the electrical logic is widely shared. Across cellular life, electrochemical gradients across membranes are a fundamental part of energy handling.

That universality makes gradients central to origin research. Alkaline hydrothermal vents create natural differences in pH and redox chemistry where warm, hydrogen-rich fluids meet a more acidic ocean. Mineral pores can separate the fluids and provide catalytic surfaces. In one family of hypotheses, early reaction networks exploited geochemical gradients before cells learned to make their own. The first membranes may have converted an environmental battery into chemistry; evolution later internalised the battery with pumps.

The idea is powerful and incomplete. Natural vents do not reproduce the exact membranes, enzymes or ion ratios used by modern cells. Reactions in water face dilution. Some useful molecules are unstable under heat or in the presence of metals. A gradient can supply a direction for chemistry without supplying heredity. Energy is necessary, but energy alone produces storms, flames and crystal growth as readily as life.

Metabolism is the organised answer. It is the network of reactions by which a living system acquires materials, extracts energy, builds components and disposes of products. Modern metabolism contains thousands of enzyme-guided steps. Early metabolism must have been smaller and less precise, perhaps using minerals and short peptides as catalysts. The relevant transition was not from no reaction to reaction. The young Earth was chemically busy. It was from reactions that happened to a network that helped recreate the conditions for its own continuation.

This is why an origin scenario must provide more than ingredients. It needs a sustained source of free energy, a way to couple that source to useful synthesis, and a route by which improvements to energy handling become heritable. Until those three are joined, there is chemistry with motion but no system able to learn from what works.

3. A Boundary Creates an Inside

Every independent cell on Earth is enclosed. The boundary may be wrapped in a wall, coat or outer membrane, but beneath the variations lies the same requirement: life controls an inside distinct from the environment. This is not defensive packaging added after the important chemistry. Without an inside, useful molecules disperse, incompatible reactions interfere and benefits cannot be assigned to the system that produced them.

The raw material for a simple boundary has a useful habit. Amphiphilic molecules possess one part attracted to water and another that avoids it. In water they self-assemble. Depending on their shape and conditions, they form films, droplets or closed bilayer vesicles. No gene instructs the assembly. The arrangement follows from molecular interactions, which makes compartments one of the easier lifelike structures to obtain without life.

A vesicle solves several origin problems at once. It concentrates molecules that would otherwise be diluted. It protects some reactions from inhibitors. It allows the interior chemistry to differ from the surroundings. It can grow by absorbing additional fatty molecules. Under physical stress it can divide. Laboratory model protocells have demonstrated each behaviour, and some admit activated nucleotide building blocks while retaining longer genetic polymers. That combination permits copying inside a container without requiring the sophisticated transport proteins of modern membranes.

Yet a sealed bag is not alive. A boundary must be selectively permeable. If nothing enters, the interior starves. If everything crosses freely, there is no stable interior. Modern membranes solve this with protein channels, pumps and receptors. Primitive membranes made from simple fatty acids are leakier, which may have been an advantage. Small nutrients can diffuse in while larger polymers remain trapped. Later, as membranes became more robust and less permeable, evolving transport machinery became necessary.

Compartments also create individuality. Consider two replicating RNAs in an open pool. One catalyses a useful reaction; the other copies faster while contributing nothing. The useful product spreads through the pool, so the parasite receives the benefit without paying the cost. In separate vesicles, the outcomes diverge. Compartments containing cooperative chemistry grow better or survive longer. Parasite-heavy compartments fail. Selection can operate among vesicles even if competition within each vesicle favours selfish molecules.

This changes the level at which inheritance matters. A protocell need not divide with the precision of a bacterium. It needs enough continuity that daughter compartments inherit a biased sample of the parent’s chemistry. If beneficial combinations are frequently broken apart, improvement cannot accumulate. If they remain associated often enough, the population can explore variants while preserving successful partnerships.

Boundaries introduce new problems too. Replicating molecules need feedstocks. Waste must leave. Osmotic pressure can burst a vesicle. Magnesium ions help many RNA reactions but can destabilise fatty-acid membranes. Experiments in which catalysts modify membrane chemistry show how one subsystem can begin solving another’s difficulty. A catalyst that strengthens its own compartment gives that compartment a competitive advantage. The important advance is mutual support rather than perfection.

The first living boundaries may not have been modern lipid membranes. Mineral pores, coacervate droplets and other phase-separated structures can also concentrate chemistry. Some could have preceded free cells, with lipid compartments taking over once self-contained movement offered an advantage. Whatever the material, the logical job is constant. A boundary turns a location into a candidate individual. It says which reactions belong together, which products are retained, and which successes can be inherited.

4. Information Must Be Able to Do Chemistry

DNA is an excellent archive and a terrible origin story on its own. A DNA sequence does little unless a cell can copy it, transcribe it and translate parts of it into functional molecules. Those jobs are carried out by proteins and RNA. Proteins, in turn, are specified through nucleic-acid information. The modern cell therefore contains a loop: information depends on catalysts, and catalysts depend on information.

RNA weakens the loop because one molecule can perform both roles. Its sequence can be copied by base pairing, giving it an information function. Its folded shape can also catalyse reactions, giving it an enzyme-like function. Natural ribozymes are not laboratory curiosities. RNA catalyses peptide-bond formation in the ribosome, and other RNAs cut, splice and process nucleic acids. The modern cell still contains fossils of a world in which RNA carried more of the chemical burden.

That is the attraction of the RNA-world hypothesis. It does not claim that a naked modern RNA molecule was the first life. It proposes that an earlier stage relied more heavily on RNA, or a related genetic polymer, before DNA became the main long-term store and proteins became the dominant catalysts. A molecule that can both inherit a sequence and affect its own reproductive success gives natural selection something to work on without requiring the full DNA-protein translation system first.

The experimental progress is substantial. Polymerase ribozymes have been evolved in laboratories to copy RNA templates. In 2016, one could synthesise functional RNAs. In 2020, another system produced fragments that assembled into a functional ancestral ligase. In 2024, RNA-catalysed cycles supported evolution of catalytic RNA. Then a 2026 result changed the scale of the problem. Researchers discovered QT45, a polymerase ribozyme only 45 nucleotides long, from random sequence pools. In mildly alkaline eutectic ice it could synthesise its complementary strand with a reported per-nucleotide fidelity of 94.1 per cent and, using defined substrates, make a copy of itself.

That is close enough to the old dream to require careful wording. QT45 did not establish a self-sustaining RNA organism. The reactions took about seventy-two days and yielded roughly 0.2 per cent full-length product. They used prepared triplet substrates, selected conditions, primers and laboratory manipulation to address strand separation. The result shows that a surprisingly small RNA motif can perform the central polymerase chemistry. It does not show how such substrates were supplied on early Earth or how a population would complete repeated autonomous replication cycles.

The remaining problems are connected. RNA building blocks have to be made and activated. Long strands have to form rather than hydrolyse. Complementary strands have to separate after copying. Error rates must stay below a threshold set by how much information needs preserving. Catalysis must produce benefits that remain associated with the sequence responsible. A successful replicator must also survive parasites: shorter sequences can often copy faster while contributing less.

This is where proteins become attractive. Even short peptides can broaden catalytic chemistry and stabilise nucleic acids. Once an RNA system could recruit amino acids, selection could gradually transfer catalytic jobs to peptides and proteins. DNA offers another later advantage because its greater chemical stability suits long-term storage. The familiar division of labour, DNA stores, RNA mediates, proteins act, is likely an evolved arrangement rather than a starting condition.

The origin of genetic information was therefore not a transition from meaningless molecules to a code book. It was a transition from sequences whose shapes affected chemistry to sequences embedded in a system that could copy, use and improve those effects. Information becomes biological when differences in sequence can change what survives next.

5. Imperfect Copying Opens Darwin's Door

Chemistry can make patterns, cycles and self-amplifying reactions without becoming life. The decisive addition is cumulative selection. A system must produce descendants that resemble it, vary in consequential ways, and differ in their success. Once that occurs, useful accidents can be retained and combined. The population acquires a history that no individual reaction possesses.

Perfect copying would prevent innovation. Hopelessly inaccurate copying would erase it. Between them lies an error threshold. A replicating sequence must be copied faithfully enough that a beneficial arrangement survives, yet not so faithfully that variation disappears. Longer sequences demand higher accuracy because they provide more places for damaging errors. Modern cells solve this with polymerases, proofreading and repair. Early replicators had none of these at first, so their information capacity must have been limited.

This creates another bootstrapping problem. Better catalysts require more sequence, but more sequence requires better copying. One escape is cooperation among shorter molecules. Different RNAs can perform parts of a cycle. Compartments can keep them together. Another is redundancy, where several crude catalysts support the same process. Selection need not leap from a short replicator to a genome. It can improve populations of interacting molecules and compartments, gradually increasing the amount of stable information.

The first copying may not have resembled cell division. Template-directed chemistry can produce a complementary strand because matching bases pair. The strands must then separate and serve as templates again, a serious difficulty because the same pairing that enables copying holds the product to its template. Temperature cycles, changes in salt, mineral surfaces, freezing and drying can help, but each creates side effects. A credible origin must run a cycle, not one favourable reaction observed once.

Once compartments reproduce, selection works at several levels. Molecules compete within vesicles. Vesicles compete within a population. A fast-copying parasite can defeat cooperative molecules locally while destroying the compartment it depends on. Bottlenecks during division may purge parasites by chance. Compartments with balanced chemistry may grow faster. The evolutionary unit is not automatically the smallest replicator. It is whatever persists with enough continuity for differences in performance to affect descendants.

Darwinian evolution does not explain how the first copying system appeared. It explains what became possible after heredity, variation and differential success were linked. Before that threshold, chemical change can be elaborate but improvements are not reliably remembered. After it, selection can optimise functions no chemist planned. Catalysis becomes faster, membranes become better controlled, error rates fall and resource use expands because variants are repeatedly tested by survival and reproduction.

The threshold was probably broad rather than ceremonial. Weak selection may have acted on autocatalytic networks before templated genes. Chemical systems can preferentially produce stable products. Vesicles can grow at different rates without genomes. Such processes are often called selection, but open-ended biological evolution requires heritable variation rich enough to keep generating new functions. The label matters less than the increase in historical memory.

This is the sense in which chemistry became historical. A crystal can reproduce a pattern, but each crystal does not inherit a long record of innovations encoded in sequence. A flame can spread, but its descendants do not carry accumulated adaptations for new fuels. Cells do. Their chemistry contains the results of an unbroken contest extending back beyond the oldest rocks. The origin was complete only when successful organisation could outlive the matter that first embodied it.

6. The Environment Had to Do Work Before Cells Could

A laboratory begins by cheating in ways the experimenter can see. Reagents arrive purified and concentrated. Temperature is controlled. The pH is chosen. Products can be moved to another flask. If one step needs ultraviolet light and the next is destroyed by it, the scientist changes the conditions. Early Earth had no scientist. A plausible origin pathway has to explain who performed those chores.

This is why the setting matters. In an open ocean, useful molecules are easily diluted. Polymer-forming reactions can be hindered by abundant water. Reactive intermediates decay. Side products accumulate. Before enzymes and pumps existed, geology had to provide concentration, energy, separation and cycling.

Surface environments offer one family of solutions. Evaporation in ponds or geothermal pools can concentrate solutes by orders of magnitude. Dry phases can favour condensation reactions that join smaller molecules; rewetting can redistribute products and form vesicles. Ultraviolet light can power particular synthetic pathways. Mineral surfaces can adsorb molecules and alter reaction rates. Repeated heating, cooling, wetting and drying provides a timetable that chemistry can exploit without needing a molecular clock.

Cold environments solve different problems. As ice crystals form, dissolved material is pushed into narrow liquid channels, concentrating reactants. Low temperatures slow many destructive reactions. Freeze-thaw cycles can alter strand pairing and compartment structure. The 2026 QT45 ribozyme, for example, performed its long copying reactions in eutectic ice at about minus 7 degrees Celsius. That does not place the origin in an ice sheet. It shows how a physical environment can supply functions that later biology performs with enzymes.

Hydrothermal systems offer a contrasting package. Water reacting with rock can produce hydrogen and reduced compounds. In alkaline vent models, warm alkaline fluid meets more acidic ocean water across mineral barriers, creating persistent pH and redox differences. Those gradients can drive chemistry, and the pores provide natural compartments. Because all modern cells use electrochemical gradients across membranes to make usable energy, the possible continuity is attractive: biology may have internalised a geochemical energy system.

None of these settings has won. Surface photochemistry struggles with continuity into deep-water models. Vents supply gradients but can dilute products into the ocean, and some useful prebiotic chemistry depends on conditions absent at the seafloor. Drying helps polymerisation but can damage fragile compounds. Minerals catalyse desired and undesired reactions. Salts that stabilise one molecule can disrupt membranes. Every setting solves some constraints and introduces others.

The false choice is to demand one birthplace that performs every step. Molecules move. Rivers carry solutes. Aerosols move material through the atmosphere. Pores connect reaction zones. Wet-dry landscapes create sequences of environments. A precursor might form under ultraviolet light, become concentrated during evaporation, enter a vesicle during rehydration and later encounter a mineral catalyst. Geography can separate incompatible chemistry before biological regulation exists.

A credible origin scenario therefore needs a process map, not a postcard. Where do the feedstocks come from? What concentrates them? What powers bond formation? What prevents immediate destruction? How are products transferred to the next stage? Which cycles repeat without an experimenter resetting the apparatus? How does any successful chemistry become associated with a compartment or replicator that can inherit the advantage?

The origin of life happened in a place, but the scientific value of the place is the work it can do.

7. Once Chemistry Could Remember, Contingency Took Over

Every known cell shares a deep molecular inheritance. DNA carries most genomes. RNA connects genes to protein synthesis. The genetic code is nearly universal. ATP is a common energy currency. Ribosomes translate nucleotide sequences into proteins. Membranes define cells. These shared features are among the strongest evidence that all extant cellular life descends from a common ancestral population.

They also create a trap. Because every organism we can inspect inherited the same ancient machinery, we cannot tell by comparison alone which features are universal requirements for life and which are historical choices that became difficult to replace. Earth gives us immense biodiversity but only one known origin.

The genetic code illustrates the distinction. The mapping between most codons and amino acids is shared across bacteria, archaea and eukaryotes. Some assignments may reflect chemical or evolutionary constraints. Yet once translation machinery, genomes and proteins all depend on one mapping, changing it becomes costly. An early convention can harden into infrastructure. Universality then records common ancestry rather than inevitability.

The same caution applies to DNA and proteins. DNA is a superb storage molecule, but an unfamiliar life form might use another polymer capable of heredity. Carbon and water have strong chemical advantages, so alternatives face serious constraints, but strong advantages are different from logical necessity. Oxygen is certainly not required. Much of Earth's biosphere lived without an oxygen-rich atmosphere, and many organisms remain anaerobic.

This is why the origin question eventually becomes an astrobiology question. If life is defined by Earth-specific ingredients, missions risk overlooking unfamiliar systems. If the definition becomes so abstract that any persistent disequilibrium counts, it admits volcanoes and weather. The practical solution is layered evidence: look for environments that can support organised chemistry, then search for patterns that are hard to sustain geologically and that reinforce one another.

Synthetic-cell research provides the reverse experiment. Instead of stripping a living cell down and asking what remains, researchers build selected functions inside artificial compartments and ask what is still missing. A 2024 system supported adaptive evolution of self-replicating DNA in liposomes using a recombinant gene-expression system and controlled cycling. In 2026, another synthetic-cell study combined DNA replication with gene-directed phospholipid synthesis in the same liposomes. Only a minority of vesicles expressed both modules, and the amount of new lipid was far too small to support membrane growth, but two major cellular processes had been made compatible inside one genetic programme.

Another 2026 study attacked a different circularity: modern lipid membranes are normally made using enzymes associated with pre-existing membranes. Researchers showed that soluble enzymes and simple metabolites could generate lipids that spontaneously assembled into new bilayer vesicles without a pre-existing membrane template. The system still relied on modern enzymes and laboratory inputs, so it is not an abiogenesis pathway. Its significance is conceptual. Apparent circular dependencies in modern cells can sometimes be broken by a cruder route.

That returns us to the beginning. The first Core Idea said that life is better described by a linked set of functions than by one substance. The final consequence is historical. Once a system could transmit variations that affected its own continuation, natural selection began preserving accidents as well as necessities. A useful local solution could become universal because every later lineage inherited it. The longer evolution ran, the harder the original boundary became to reconstruct from the finished machinery.

Life is therefore both chemistry and history. Chemistry constrains what can work. History determines which workable path became ours. The origin problem is the attempt to separate those two after four billion years of entanglement.

How It Actually Works

A wet planet before a living one

The oldest surviving piece of the origin story is a zircon crystal from Western Australia. Some Jack Hills zircons formed more than 4.3 billion years ago, within a few hundred million years of Earth itself. Ratios of oxygen isotopes in them indicate that their parent rocks had interacted with liquid water near the surface. The early planet was violent, hot and repeatedly struck, but it was not a permanent ocean of magma. Crust and water appeared early enough to give prebiotic chemistry a long runway.

Earth supplied carbon, hydrogen, nitrogen, oxygen, phosphorus, sulphur and metals. Volcanoes released gases. Water reacted with rock. Sunlight and lightning provided energy. Meteorites delivered organic compounds, including amino acids and nucleobase-related material. None of these sources needs life. The universe makes organic chemistry readily because carbon forms stable bonds in many arrangements.

The abundance of ingredients did not make the result automatic. Most organics are useless to biology. Useful ones can be destroyed, diluted or trapped in tar-like mixtures. Modern cells rely on selected molecular forms, activated building blocks and enzymes that direct reactions away from dead ends. They also choose one handed form of many molecules, while ordinary chemistry often makes equal left- and right-handed versions. The early Earth had to concentrate, sort and connect its chemistry without those products of evolution. It had perhaps hundreds of millions of years, but no guarantee that time alone would solve the organisation problem.

The spark in the flask

In 1952 Stanley Miller, working with Harold Urey at the University of Chicago, built an apparatus containing water and gases then thought to resemble the early atmosphere. Water boiled through the system. Electrical sparks crossed the gas chamber. Condensate returned to the flask. Within days the liquid darkened. Analysis revealed amino acids, the building blocks of proteins. Miller published the result in 1953 in two pages of Science.

The experiment changed the field because it made chemical evolution testable. Organic building blocks did not need to arrive fully formed or require a guiding force. Energy applied to simple molecules could generate compounds used by life.

It did less than legend says. The apparatus did not make proteins, genes, membranes, metabolism or a cell. Its gas mixture was strongly reducing, rich in methane, ammonia and hydrogen. The bulk early atmosphere is now often modelled as less reducing, dominated more by nitrogen and carbon dioxide, which makes some spark chemistry less productive. Local volcanic plumes, impact-generated atmospheres and other environments could still provide reducing conditions. Later experiments with varied gases and energy sources have produced a wide range of organics. The durable result is not one recipe. It is that prebiotic feedstocks can arise by several routes.

The next difficulty is selection from the mixture. An amino acid in a flask is a part without an assembly line. It still has to be concentrated, activated and joined into a useful polymer while side reactions are restrained. The origin problem begins where the famous photograph of Miller beside his apparatus usually ends.

Chemistry needs a timetable

Water is indispensable to modern life and awkward for making many of its polymers. Joining monomers often releases water, so abundant water favours the reverse reaction. Drying can push condensation forward. Rehydration can release and reorganise the products. This gives wet-dry cycles a possible role as crude machinery.

In volcanic pools or shore settings, evaporation concentrates solutes that would vanish in an ocean. Heating, cooling, drying and rewetting alter which reactions are favoured. Mineral surfaces can adsorb molecules, bring them close and provide catalytic sites. Ultraviolet light can power specific synthetic routes. Freezing performs a different concentration trick: growing ice excludes solutes into narrow liquid channels, where reactions occur at far higher local concentrations.

No single cycle is benign. Heat, light and drying can destroy what they help create. Salts and metals assist one reaction and inhibit another. This is why origin chemistry increasingly looks staged. Products may move between zones or experience conditions in sequence. A credible route has to specify the order, not merely list reactions that work under mutually incompatible conditions.

One linked pathway reported in 2015 made precursors of ribonucleotides, amino acids and lipids from common feedstocks using ultraviolet-driven chemistry. It did not make life, but it weakened the assumption that genes, proteins and membranes had to originate in isolated chemical worlds. A shared environment can branch into the ingredients of several subsystems. Other work has found routes to both purine and pyrimidine nucleotide precursors and to chemical activation steps needed for polymer growth. Each route carries constraints about light, salts, feedstock supply and sequence. The bootstrapping becomes less improbable when the parts share a supply chain, but the supply chain still has to operate on a planet rather than on separate laboratory benches.

The molecule that can remember and act

Modern cells divide information and catalysis between several molecular classes. DNA is the durable archive. RNA carries messages and performs specialised structural and catalytic jobs. Proteins do most of the fast, selective chemistry. That division is effective now and awkward at the beginning, because each part depends on the others.

RNA offers a route through the circularity. A sequence of RNA can be copied by complementary base pairing, so it can carry heredity. The same strand can fold into a three-dimensional shape that catalyses reactions. Natural catalytic RNAs were discovered in the 1980s, and the ribosome supplied the most consequential example: the reaction that joins amino acids into proteins is catalysed by ribosomal RNA. Modern biology still contains machinery compatible with a more RNA-heavy past.

The RNA-world idea therefore asks whether an earlier stage of evolution used RNA, or a related polymer, for both information and chemistry before DNA and coded proteins took over their modern jobs. It is a framework rather than a complete origin story. It still requires plausible nucleotide chemistry, polymer formation, energy, compartments and repeated replication.

Experiments have steadily narrowed the replication problem. Polymerase ribozymes selected in the laboratory can extend RNA templates and make functional RNA products. A 2016 ribozyme copied a range of structured RNAs. A 2020 system made fragments that could assemble into a functional ancestral ligase. In 2024, ribozyme-catalysed replication supported repeated evolutionary improvement of catalytic RNA.

Then came a result that older summaries of the field could not include. In 2026, Edoardo Gianni and colleagues reported QT45, a 45-nucleotide polymerase ribozyme discovered from random sequence pools. In mildly alkaline eutectic ice, QT45 could synthesise its complementary strand with 94.1 per cent per-nucleotide fidelity using a random pool of trinucleotide substrates. With defined substrates and a strand-separation intervention, it could also synthesise a copy of itself.

The headline needs its denominator. Full-length yields were about 0.2 per cent after roughly seventy-two days. The experiments supplied prepared substrates, primers, controlled ionic conditions and a freeze-based regime. A complete autonomous cycle in which newly made RNA repeatedly generates more copies of itself was not demonstrated. The result matters because it shrinks the catalytic object required for RNA copying and shows that polymerase activity may occupy more of RNA sequence space than previously thought. It does not solve prebiotic feedstock supply or autonomous replication.

Strand separation remains a severe problem. Copying one RNA strand tends to make a complementary partner that binds to it. The resulting duplex is stable enough to protect information and stable enough to obstruct the next copying round. Heat can separate strands but also damages RNA. Ice, pH cycles and other environmental changes can help, which again shows geology performing work before enzymes exist.

Error is the second constraint. A replicator must vary, but if copying is too inaccurate, long useful sequences cannot be maintained. This creates an error threshold: the amount of heritable information a population can support depends on fidelity and on the selective advantage of the sequence. Primitive genomes were therefore likely short or distributed across cooperating molecules until better replication evolved.

The third constraint is access to chemistry beyond RNA. Short peptides can stabilise RNA and catalyse reactions RNA handles poorly. Once a genetic system could reproducibly associate certain sequences with certain peptides, selection could improve both sides of the partnership. DNA then offered more stable storage. The modern DNA-RNA-protein division of labour is best understood as an evolutionary destination built from earlier multifunctional components.

The important transition was not the appearance of a molecule called information. It was the appearance of sequences whose differences affected chemistry, whose chemistry affected copying, and whose successful variants could be inherited.

Bags of chemistry become competitors

Simple fatty molecules placed in water can form closed vesicles. This spontaneous assembly gives origin researchers something rare: a cell-like structure that does not require a cell to make it. The question is whether the bags can host chemistry and reproduce enough of their organisation for selection to act.

Experiments with model protocells have brought the pieces closer. Fatty-acid vesicles can encapsulate nucleic acids, take in some activated nucleotide building blocks, grow by incorporating extra lipids and divide when physically disturbed. In 2008 researchers demonstrated template-directed synthesis of a genetic polymer inside model vesicles. Other work showed that vesicles containing more nucleic acid can create osmotic pressure that drives membrane growth, linking internal copying to competition for membrane material.

A protocell does not need a modern division machine. If a growing vesicle becomes unstable and splits, both daughters may inherit some internal molecules. Most divisions will be untidy. Many daughters will fail. Physical processes such as shear, extrusion through pores or changes in surface area can produce division without dedicated proteins. A population can improve if the correlation between internal chemistry and descendant success is strong enough. Reproduction begins as statistical continuity, not faithful cellular choreography.

This is where compartments change the evolutionary logic. A catalyst that produces useful membrane material can help its own vesicle grow. A replicator that improves nutrient use can increase the descendants of the compartment carrying it. Products no longer dissolve into a common pool. Benefits become local and therefore selectable.

The membrane and chemistry must still cooperate. Magnesium ions useful for RNA folding can disrupt simple membranes. Fatty acids work across limited ranges of salt and pH. A 2016 experiment showed a route by which an encapsulated catalyst could make lipid derivatives that protected vesicles under otherwise damaging magnesium conditions. The experiment does not show that this exact reaction occurred on early Earth. It shows how coupled selection can solve conflicts which look fatal when subsystems are studied separately.

The energy problem below the sea

Surface chemistry offers light and cycling. Alkaline hydrothermal systems offer continuous geochemical power. When seawater circulates through certain rocks, reactions called serpentinisation can produce hydrogen-rich alkaline fluids. Where those fluids meet a more acidic, carbon-rich ocean, mineral barriers separate contrasting solutions. The resulting gradients can drive reactions.

This resembles a feature of all cells. Modern organisms maintain proton or ion differences across membranes and use the return flow to power ATP production. The similarity has inspired the idea that life began by exploiting natural gradients in vent pores. Mineral sulphides can catalyse carbon chemistry. Microscopic pores can concentrate molecules and hold reaction networks in place. Flow renews the supply. In this account, the universal dependence on ion gradients is not an arbitrary late invention. It is an inherited solution to an energetic problem present before free cells existed.

A vent model provides a persuasive route into metabolism. Hydrogen can reduce carbon dioxide, producing increasingly organic compounds under suitable conditions. Recent experiments continue to show that iron sulphides, thermal gradients and natural vent precipitates can support relevant reactions or convert gradients into electrical work.

The obstacles remain. A mineral pore is not a freely reproducing cell. Molecules must escape without losing the network. Early membranes must become independent of the geological barrier and generate their own gradients. Some leading prebiotic routes require ultraviolet light or drying unavailable at the seafloor. The ocean can dilute products. These are engineering problems rather than a disproof, but they prevent the setting from being declared the answer.

Surface and vent models may describe different stages. Photochemistry can make feedstocks; water-rock reactions can provide reducing power; porous settings can concentrate them; cycling environments can promote polymers and vesicles. The first evolving populations may have used a route through environments rather than a permanent address.

Crossing the threshold

No surviving record identifies the moment when a protocell population became life. A plausible sequence can nevertheless be stated without pretending it happened in that order.

Geochemistry produced and concentrated organic feedstocks. Some molecules formed catalysts or self-promoting networks. Compartments retained combinations. Template copying introduced heredity. Environmental cycles supplied repeated rounds of growth and division. Variants differed in stability, catalytic output, copying and resource use. Once those differences caused some lineages of compartments to leave more descendants, selection began improving the integrated system.

Autocatalysis may have provided a bridge before template copying dominated. In an autocatalytic network, products help accelerate reactions that produce more members of the network. Such chemistry can amplify particular combinations and compete for feedstocks. It still lacks the precise, open-ended heredity of a sequence, because composition is harder to copy than an ordered polymer. Yet networks could have supplied metabolism, maintained activated molecules or made components needed by primitive replicators. A partnership between compositional inheritance and templated inheritance is more plausible than either subsystem beginning in finished form.

The early populations would have been poor at everything. Their membranes leaked. Their copying was short and error-prone. Catalysts were weak. Their identities blurred through exchange of molecules. That weakness may have made integration possible. A modern bacterial membrane would starve without transport proteins; a leaky fatty vesicle can admit feedstocks. A modern genome requires accurate enzymes; short primitive sequences can survive lower fidelity. The earliest systems did not need to meet the standards imposed by their descendants.

They also had no obligation to resemble one modern species. Populations could merge, split and exchange components. Selection may have acted on communities of protocells before stable cellular lineages emerged. The familiar tree of life becomes appropriate only after descent was coherent enough for branches to remain separate.

As selection improved copying, longer sequences became maintainable. Catalysts expanded metabolism. Better membranes gave greater control but created demand for transport. Genetic coding linked RNA sequences to reproducible peptides, multiplying catalytic possibilities. The code itself probably emerged through intermediate associations rather than one complete table appearing at once. DNA later became a more stable archive. These transitions did not merely add parts. Each altered the selection pressures on the rest and created new dependencies that made reversal less likely.

At some point, the system no longer depended on one special pond, pore or imposed laboratory cycle. It could rebuild its boundary, regenerate catalysts, copy its information and reproduce using resources from its surroundings. Geology had produced a chemical population able to carry its own history forward.

From first life to LUCA

The first populations that qualify as living may be permanently unrecoverable. The oldest rocks have been heated, deformed, recycled and chemically altered. Even when ancient carbon or microscopic structures survive, non-biological processes can imitate parts of the signal. The geological record therefore becomes clearer only after life was already established.

By about 3.5 billion years ago, the evidence is strong. Archean rocks preserve stromatolitic structures, microbial textures, isotopic patterns and organic matter consistent with biological activity. Some individual claims remain contested, but the combined record makes a lifeless Earth at that date increasingly difficult to defend. Proposals for life at 3.7 billion years or earlier are scientifically important and much less secure.

Molecular reconstruction reaches backwards by another route. All extant cellular life shares the genetic code, ribosomes, ATP-based energy transfer and a large set of deeply conserved genes. Those common features point to a last universal common ancestor, LUCA. The name invites a mistake: LUCA is not the first life. It is the latest population ancestral to all sampled cellular lineages that survived to the present.

A 2024 analysis combined molecular clocks, gene-tree reconciliation and geochemical modelling to place LUCA around 4.2 billion years ago, with a 95 per cent interval of roughly 4.09 to 4.33 billion years. The same work estimated a genome near 2.75 million bases encoding about 2,657 proteins and inferred an anaerobic, acetogenic physiology. These numbers are model-dependent and differ from more conservative reconstructions. They should not be turned into a photograph of a cell we never saw.

What they do show is how far the origin lies below LUCA. Even the cautious picture of LUCA includes translation, heredity, energy handling and substantial metabolism. A population like that had already crossed many of the bootstrapping problems in this book. Other early lineages may have lived alongside it and vanished. Genes could also have moved between lineages before the tree of life settled into cleaner branches.

The record therefore has a missing first act. Geology shows that life was established early. Comparative genomics shows that the common ancestor of surviving cells was already complex. Between prebiotic chemistry and that ancestral population lies the transition researchers are trying to reconstruct in the laboratory.

How we know

The origin is reconstructed from three incomplete archives. Laboratory chemistry tests whether particular steps can occur, but experimenters choose reagents and conditions, so success establishes possibility rather than history. Ancient rocks preserve minerals, isotopes, textures and altered organic matter, yet the oldest crust has mostly been recycled and non-biological processes can imitate life. Living cells carry a third archive in shared genes, ribosomes, membranes and metabolism. Comparative genomics can infer features of LUCA, but later evolution, gene transfer and extinction obscure earlier stages.

Confidence is therefore uneven. The formation of organic building blocks, self-assembly of vesicles, catalytic RNA and selection among replicating systems are demonstrated. The exact environment, sequence of stages and identity of the first hereditary polymer are open. Dates older than the strong evidence near 3.5 billion years depend on contested signals or model-based molecular clocks. The decisive limitation is one surviving lineage: every organism studied descends from the same broad ancestry. We can recover a route that works in pieces and a descendant already complex. The crossing between them has not been observed or found in stone. The field advances by narrowing the possible bridges and making each one carry more of the system's load under conditions that could have existed on the young Earth for enough repeated cycles to matter.

What People Get Wrong

“Viruses prove that the definition is obvious”

Viruses are often used to settle the border of life: either they are alive because they carry genes and evolve, or dead because they lack cells and metabolism. Both verdicts reveal more about the chosen definition than about viruses. A virus outside a host cannot maintain itself or reproduce. Inside a cell, its genome redirects living machinery, makes descendants and evolves rapidly. The relevant system may be the virus alone, the infected cell, or a virus-host lineage. Giant viruses blur the picture further because some carry genes for functions once thought exclusive to cells. The correction matters because nature does not promise that categories created for horses and stones will remain binary at every scale. Viruses are not a trick question with a hidden official answer. The urge to force one comes from schoolbook classification, where every specimen belongs in one box. Research is better served by stating which capacities the virus has, which it borrows and which level is being judged. They show that life's capacities can be distributed across partners, a possibility that matters when considering parasites, symbioses and unfamiliar systems elsewhere.

“Life violates the second law of thermodynamics”

Living systems build order, while the second law says entropy tends to increase. The apparent clash disappears once the boundary of the calculation is drawn correctly. An organism is an open system. It takes in concentrated energy and matter, then releases heat and waste. A plant uses sunlight to build sugars while radiating heat. An animal breaks food down and warms its surroundings. Local organisation increases at the cost of a larger increase in entropy outside. Snowflakes, hurricanes and flames also form ordered structures while energy flows through them, so order alone is no proof of life. What biology adds is controlled coupling: energy is channelled into maintenance, repair, synthesis and heredity. The misconception persists because entropy is often translated as disorder, then treated as a ban on any spontaneous pattern. Thermodynamics is about the direction and accounting of energy dispersal, not a command that every corner must become visually messy at every moment. Invoking it neither creates a mystical exception nor explains the whole organism. It specifies the bill that every living process must pay and directs attention to the gradients and fuels that keep the process running.

“The Miller-Urey experiment made life”

Miller's apparatus made amino acids and other organic compounds from simple gases and electrical energy. That result was important because it showed that biological building blocks can arise without biology. It did not produce proteins, genes, membranes, metabolism or an evolving population. The strongly reducing gas mixture also does not represent the only or necessarily dominant model of the early atmosphere. Later work has explored less reducing gases, volcanic chemistry, impacts and extraterrestrial delivery. The durable correction is two-sided. Prebiotic organics are not a miracle; many form readily. Organising them into a system is far harder. The picture became persuasive because a sealed glass machine, a spark and a brown solution resemble an origin event in miniature. The visual story is cleaner than the chemical one. Calling Miller-Urey a failed attempt to make life understates what it proved. Calling it the creation of life erases the central problem it exposed: ingredients are cheap compared with integration. Modern origin research begins by asking how products survive, concentrate and enter repeatable cycles after the spark.

“The RNA world has solved the origin problem”

RNA has moved closer to the centre of the origin story, not to the finish line. It can carry sequence information and catalyse reactions. Natural ribozymes exist, the ribosome's catalytic heart is RNA, and laboratory polymerase ribozymes have become steadily more capable. The 2026 QT45 result is the sharpest correction to older scepticism: a tiny RNA catalyst can now perform both chemical halves needed to reproduce its sequence.

That still falls far short of an autonomous RNA lineage. The demonstration was slow, low-yield and dependent on prepared substrates and experimental assistance. It did not show repeated self-sustaining replication from geochemical feedstocks. Nor does RNA alone solve compartment formation, energy coupling, nucleotide supply or the transition to coded proteins and DNA.

The misconception persists because 'RNA world' sounds like the name of a completed historical era. It is better treated as a family of hypotheses about an early stage in which RNA or related polymers carried more heredity and catalysis than they do now. Peptides, lipids and metabolic chemistry may have been present from early on. The 2026 result strengthens the case that small RNA could perform more of the copying job than previously demonstrated. It also makes the remaining systems problem harder to ignore.

“Scientists know where life began”

Hydrothermal vents are often presented as the birthplace, while another tradition chooses warm little ponds or volcanic hot springs. Each setting solves real problems and creates others. Alkaline vents provide sustained chemical gradients, catalytic minerals and pores. Surface pools provide ultraviolet energy, evaporation and wet-dry cycles. Ice concentrates solutes and protects fragile molecules. Mineral surfaces, impact zones and beaches offer further mechanisms. No setting has carried plausible feedstocks through every necessary stage to an autonomous evolving protocell. The first life may also have used materials or stages from several environments. “Where” is not a prize awarded to the most cinematic location. The misconception survives because origin models are commonly named after places, and because a named setting is easier to picture than a sequence of chemical constraints. Supporters also emphasise the problem their setting handles best. A location earns confidence by supplying concentration, energy, cycling, compatible chemistry and a route to reproduction in one connected account. The winning explanation may be a route across settings rather than one address.

“LUCA was the first organism”

LUCA is the last universal common ancestor of all surviving cellular life. It is reconstructed from genes and systems shared across Bacteria, Archaea and Eukaryotes. Those shared features indicate a cell with ribosomes, a genetic code, metabolism and a substantial genome, not the first fragile replicator. Recent modelling places LUCA remarkably early, perhaps around 4.2 billion years ago, but the date and reconstructed details depend on assumptions. More important, “last common ancestor” is a position in a family tree. Other lineages may have lived before or beside it and vanished. Genes may have moved among early populations. Family diagrams make the mistake tempting because LUCA appears at the base of the visible trunk. The diagram omits extinct branches and the pre-cellular networks that cannot be reconstructed cleanly. Treating LUCA as the first life compresses an unknown interval of chemical and biological evolution into a single cell already equipped with machinery the origin must explain. LUCA is a destination for origin research, not its starting point, and the machinery it carried is evidence of how much earlier evolution had already occurred.

“Liquid water is evidence that life is present”

Water matters because terrestrial life uses it and because it is an effective solvent across a useful temperature range. Finding ancient rivers on Mars or subsurface oceans on icy moons establishes potential habitability. It does not establish habitation. Water is common in planetary systems and performs abundant non-biological chemistry. The same applies to organic molecules, methane and even oxygen: each can be produced without life under some conditions. A biosignature becomes persuasive through context and convergence. Several gases held far from expected equilibrium, isotope patterns, repeated structures and an environment able to support the inferred metabolism can reinforce one another. The correction changes mission design and public interpretation. The shortcut is reinforced by the phrase 'follow the water', which was an effective exploration strategy because water leaves geological traces and narrows the search. A strategy for finding habitable environments became confused with a test for life. The same discipline is needed for exoplanet atmospheres, where one gas can be produced by several planetary processes. A wet world is a place worth investigating, not a living world awaiting a press release, and habitability must never be reported as detection.

Use It

When someone asks whether something is alive, change the question

Borderline cases become clearer when you stop searching for one decisive trait. Write down the functions instead. Does the system maintain a distinct inside? Does it obtain and use energy? Can it rebuild or repair parts of itself? Does it contain heritable information? Can that information vary? Do those variations affect descendants? Which functions are internal, and which are borrowed from a host or environment?

This turns a sterile argument into a biological description. A virus has heredity and evolution but borrows translation and metabolic support. A spore retains the machinery and information required to restart despite low current activity. A prion propagates a protein conformation but lacks most of the organisation associated with cellular life. A synthetic protocell may reproduce one or two functions while depending on experimenters for the rest.

The verdict can remain disputed while the mechanism becomes precise. That is progress.

When you see an origin-of-life headline, ask what stage was crossed

'Building block of life created' can mean almost anything. Amino acids, nucleobases, lipids and simple sugars can be made without life. The harder question is what the result connects to next.

Classify the experiment. Did it make a precursor from plausible feedstocks? Did it concentrate or polymerise molecules? Did it produce a catalyst? Did it copy information? Did it couple copying to a compartment? Did the compartment grow and divide? Did variants differ in descendant success? Did the whole system regenerate its own components from simple inputs?

A result can be important without crossing all those stages. Miller's amino acids mattered because they demonstrated abiotic synthesis of useful organics. QT45 matters because a small RNA can perform both strands of the central copying chemistry under controlled conditions. A synthetic liposome that replicates DNA and makes some of its own membrane lipid matters because two modules can coexist. None should be described as 'scientists create life' unless the system's autonomy, heredity and reproduction justify the phrase.

The most useful extra question is what the researchers still supplied. If the answer includes purified enzymes, activated monomers, a genetic translation system, scheduled dilution, freeze-thaw handling or manual transfer between compartments, those supports belong in your mental model of the result. They are not reasons to dismiss it. They mark functions the experimental system has not yet internalised.

Inspect the hand-offs, not only the successful reactions

Origin research is full of impressive individual steps. The weak point often lies between them. A pathway makes nucleotide precursors under ultraviolet light. Another copies RNA in ice. A membrane system works at one salt concentration. A vent model supplies a proton gradient at a different pH. All can be good science while failing to form one plausible sequence.

So ask what happens between experiments. Can the product of one reaction enter the next without purification? Are the concentrations compatible? Does the temperature that helps one step destroy another? Can the molecules move between environments naturally? Does the sequence repeat? Who supplies activated feedstocks? Who removes waste?

A connected route deserves more weight than a collection of chemically possible events. The origin occurred once as a process, not as a bibliography.

This also helps compare rival models fairly. A vent model should not be rejected because it lacks ultraviolet chemistry if it can receive photochemical products from elsewhere. A surface model should not be credited with energy coupling merely because it produces monomers. Ask where each model is strongest, which transitions it explains, and whether its missing steps are compatible with movement through real early-Earth environments.

Distinguish replication from biological heredity

A pattern can copy without supporting open-ended evolution. Crystals replicate structural order during growth. Polymerase enzymes can copy DNA in a tube. Autocatalytic molecules can make more of themselves. None of those facts alone produces a biological lineage.

Look for three linked ingredients: variation, inheritance and differential persistence or reproduction. Variation creates alternatives. Inheritance keeps some relation between parent and descendant. Differential success changes which alternatives dominate later populations. The more faithfully functional differences survive across generations, the more chemistry acquires memory.

Then ask how much information the system can carry. A replicator with a high error rate may maintain a short motif and lose a longer one. A compartment that divides randomly may preserve a useful network only probabilistically. Early evolution did not need perfect heredity, but it needed enough continuity for improvement to accumulate rather than reset each generation.

Read origin environments as machines

A warm pond, ice sheet or hydrothermal vent is useful only for the operations it performs. Translate the setting into functions.

A wet-dry landscape concentrates solutes, drives some condensation reactions and imposes cycles. Ice concentrates dissolved material into liquid channels and slows degradation. Mineral surfaces hold reactants together and can catalyse chemistry. Alkaline hydrothermal systems supply hydrogen, catalytic minerals, natural compartments and electrochemical gradients. Ultraviolet light can energise synthesis but can also destroy products.

This prevents scenery from replacing mechanism. When two origin models compete, compare the full tasks they can perform and the transitions they leave unexplained. The best setting is not the one with the strongest photograph. It is the one that can support a credible sequence from geochemistry towards heredity and autonomous reproduction.

Treat a biosignature as a case, not a clue

Water, organics, methane, oxygen, pigments, isotope patterns and cellular-looking shapes can all be informative. None has a monopoly on biological origin. The correct question is how many independent observations fit one biological process and how well non-biological alternatives perform.

Start with environment. Is there a solvent, energy source and chemistry capable of supporting the proposed metabolism? Then examine the signal. Is it abundant enough? Is it spatially associated with relevant minerals or structures? Is it repeatedly produced? Does the atmosphere or ocean contain combinations that should react away unless replenished? Can photochemistry, volcanism, water-rock reactions or contamination explain the same pattern?

The aim is not maximum scepticism. It is discriminating evidence. A discovery becomes stronger when several observations require the same active process rather than when one observation receives a dramatic label.

The limits

Origin-of-life research does not yet provide one accepted historical sequence from simple geochemistry to the first autonomous cells. Laboratory plausibility is not geological proof. Molecular reconstruction of LUCA reaches the ancestry of surviving cells, not the first replicators. The oldest rocks preserve biological signals imperfectly. Every current model therefore contains sections with different confidence levels.

Nor can a biological definition settle questions about personhood, abortion, death law or moral status. Biology can describe fertilisation, development, neural function and organismal integration. Ethical and legal categories add purposes and values that chemistry cannot decide.

The same caution applies to artificial systems. A machine may be autonomous in one engineering sense without being a biological organism. A synthetic cell may satisfy selected operational criteria while borrowing most of its machinery from existing life. Definitions are tools. They should not be mistaken for discoveries.

The one thing to keep

Keep the transition from chemistry to history.

Before heredity, a successful reaction can vanish without consequence. A useful molecule is made, destroyed and forgotten. Once variants can reproduce differences that alter future success, the past begins to accumulate. Catalysts improve, boundaries become more selective, energy handling becomes more reliable and information can grow because previous successes are no longer lost with the molecules that achieved them.

That is the deepest difference life introduces into matter. A cell is chemistry carrying the consequences of earlier chemistry. Its genome, metabolism and membrane are not merely a set of clever parts. They are a record of solutions preserved because ancestral systems survived long enough to pass them on.

Life began when chemistry stopped having to start over.

Terms

Abiogenesis. The natural emergence of living systems from non-living chemistry. It names the origin problem without specifying a route and should not be confused with the old claim that maggots routinely arise from meat.

Amphiphile. A molecule with a water-attracting region and a water-avoiding region. Amphiphiles can self-assemble into films and vesicles, making them plausible raw materials for primitive compartments.

ATP. Adenosine triphosphate, the main transferable energy currency of modern cells. Cells continually make and spend it to couple energy-releasing reactions to synthesis, movement, transport and repair.

Autocatalysis. A reaction in which a product accelerates its own production, directly or through a network. Autocatalytic systems can amplify chemical organisation but do not automatically possess sequence-based heredity.

Biosignature. A substance, pattern or structure that may indicate life. A strong biosignature must be interpreted in environmental context and supported by converging evidence because abiotic processes can imitate individual signals.

Carbon fixation. The conversion of inorganic carbon, usually carbon dioxide, into organic molecules. Every ecosystem depends on fixation, performed today by several metabolic pathways in plants and microbes.

Catalyst. A substance that speeds a reaction without being consumed overall. Enzymes and ribozymes are biological catalysts; minerals and metals may have performed related jobs before evolved catalysts existed.

Cell. The membrane-bounded fundamental unit of terrestrial life. Cells maintain internal conditions, process energy and information, and reproduce, though multicellular organisms distribute some functions across specialised cells.

Chemiosmosis. The use of an ion gradient across a membrane to drive work, especially ATP production. Its universality makes gradient-based energy central to many origin models.

Compartment. A bounded region that concentrates and separates chemistry. Compartments allow products to remain associated with their makers, creating units on which selection can act.

Darwinian evolution. Change in a population through heritable variation and differences in reproductive success. It makes useful accidents cumulative and is a key threshold between recurring chemistry and biological history.

Disequilibrium. A condition in which matter or energy remains away from chemical or thermodynamic equilibrium. Life maintains disequilibria by continual energy use, but geological and atmospheric processes can do so as well.

Genetic code. The mapping by which nucleotide triplets specify amino acids during protein synthesis. Its near universality is evidence of common ancestry, although parts of the mapping may reflect frozen historical contingency.

Genome. The heritable genetic material of an organism or virus. Modern cellular genomes are DNA, but early heredity may have used RNA or related polymers.

Homochirality. The use of one molecular handedness, such as right-handed sugars in nucleic acids and mostly left-handed amino acids in proteins. How this asymmetry arose remains an origin question.

Hydrothermal vent. A site where heated, chemically altered water emerges from the seafloor. Alkaline vents provide gradients, minerals and pores that could support early metabolism, although no vent pathway is complete.

Lipid. A broad class of water-insoluble or amphiphilic molecules. Membrane-forming lipids create cellular boundaries, while simpler fatty acids are common ingredients in protocell experiments.

LUCA. The inferred ancestral population from which all sampled cellular organisms descend. It was already a complex cell or population and must not be mistaken for the first living system.

Metabolism. The connected reactions by which a system acquires materials and energy, builds components and handles waste. An origin model must explain how useful reactions became self-supporting rather than merely occurring.

Monomer. A small molecule that can be joined into a polymer. Amino acids are protein monomers; nucleotides are nucleic-acid monomers. Producing monomers is easier than assembling functional polymers.

Natural selection. The process by which heritable variants that leave more descendants become more common. Selection can act at molecular, compartment, cellular and organismal levels when heredity links traits to outcomes.

Nucleotide. The monomer of RNA and DNA, built from a base, sugar and phosphate. Prebiotic chemistry must form, activate and join suitable nucleotides under conditions compatible with other subsystems.

Polymer. A chain assembled from repeating or related monomers. Proteins and nucleic acids are polymers whose sequence permits complex structure, catalysis and heredity.

Prebiotic chemistry. Chemical processes relevant to life’s origin that occur without living organisms. Plausibility requires more than making a product: feedstocks, environment, concentration, cycling and downstream compatibility all matter.

Protocell. A cell-like chemical compartment that lacks the full machinery of modern cells. Model protocells test how membranes, internal reactions, growth, division and heredity might have become coupled.

Replication. The copying of a molecule, sequence or pattern. Replication becomes biologically powerful when copy differences are heritable and alter the success of the larger system carrying them.

Ribosome. The RNA-protein machine that translates genetic information into proteins. Its catalytic centre is RNA, supporting the idea that RNA once carried more of life’s functional burden.

Ribozyme. An RNA molecule that catalyses a reaction. Ribozymes prove that one polymer can combine information and action, although no known ribozyme yet sustains autonomous self-replication from plausible feedstocks.

RNA world. A proposed early phase in which RNA or related polymers carried heredity and catalysed reactions before DNA and coded proteins dominated. It is a framework with evidence and major unresolved steps.

Vesicle. A closed membrane sac formed by amphiphiles in water. Simple vesicles can grow, divide and retain polymers, making them useful experimental models of primitive cellular boundaries.

Go Deeper

Four books, four different jobs. None supplies a final origin story, because no such story exists. Together they move from the classification problem through experimental reconstruction and competing mechanisms to a full systems account.

The border of life

Carl Zimmer, Life’s Edge: The Search for What It Means to Be Alive (Dutton, 2021). Start here for the definition problem. Zimmer moves through viruses, hibernation, embryos, death and attempts to build lifelike systems, showing why a list of traits never quite becomes a theory. It is written for general readers and is the most inviting continuation of the first question in this book. Its strength is intellectual honesty rather than an origin model of its own. It also shows why practical decisions in medicine and law can require boundaries even when nature has not supplied a clean one.

The experimental route

David W. Deamer, Assembling Life: How Can Life Begin on Earth and Other Habitable Planets? (Oxford University Press, 2019). Deamer explains how researchers turn origin claims into experiments, with special attention to membranes, protocells and wet-dry cycling in volcanic settings. Read it to see what laboratory reconstruction can establish and where geological plausibility enters. He argues a clear case for surface hydrothermal environments, so treat the setting as a strong proposal rather than a settled birthplace. The value lies as much in the experimental discipline as in the location he favours.

The energy argument

Nick Lane, The Vital Question: Energy, Evolution, and the Origins of Complex Life (Profile Books, 2015). Lane builds the most forceful popular account of why ion gradients and chemiosmosis may connect alkaline vents to the basic architecture of cells. The book ranges beyond life’s origin into the rise of complex cells, sex and ageing. Its central argument is bold and partly contested, which makes it valuable: it shows how one mechanism can organise many facts without becoming established merely because it is elegant.

The systems textbook

Pier Luigi Luisi, The Emergence of Life: From Chemical Origins to Synthetic Biology, 2nd edition (Cambridge University Press, 2016). This is the demanding option. Luisi treats definitions, self-organisation, vesicles, minimal cells and synthetic biology through a systems perspective, then lays out the open questions. It is slower and more technical than the other three, but it makes clear how much must be coupled before a collection of molecules becomes a self-maintaining cell. Read it when the overview no longer feels sufficient and you want the open problems stated without the compression required here.

Notes and Sources

The notes follow the book in order. Origin-of-life research contains demonstrated chemistry, historical reconstruction and model-dependent inference in unusual proximity. The narrative therefore uses firm language for observed mechanisms, identifies leading proposals as proposals, and leaves unsupported sequences out.

The Whole Thing in One Page and Why You Should Care

Definitions and edge cases. The NASA Astrobiology Program uses “a self-sustaining chemical system capable of Darwinian evolution” as a working definition, while acknowledging that it was built from terrestrial experience. The National Academies’ astrobiology strategy and Carl Zimmer’s Life’s Edge both emphasise that no agreed definition cleanly settles viruses, dormant states, sterile organisms and collective systems. The treatment here follows a systems view: the relevant issue is the coupling of maintenance, boundaries, energy use, heredity and evolution, not the possession of one substance.

ATP turnover. The statement that a person turns over an amount of ATP comparable to body mass in a day is an order-of-magnitude illustration, not a measurement of a fixed pool. ATP is regenerated repeatedly. Whole-body estimates vary with energy expenditure and assumptions about efficiency. Stephan Wilkens’ biochemical overview gives the familiar estimate of about 50 kilograms per day for an average adult; current cell-biology literature continues to use the body-mass comparison.

Age, water and early evidence. The accepted age of Earth is about 4.54 billion years. Wilde and colleagues and Mojzsis and colleagues used Hadean zircons to infer crust and surface liquid water by about 4.3 to 4.4 billion years ago. Evidence for life near 3.5 billion years is supported by several independent geological and geochemical lines. Claims older than this are retained in the scientific literature but remain more dependent on interpretation.

Life detection. The National Academies advises that no single biomarker is infallible and that life detection must combine planetary context, chemistry and several observations. The book’s brief astrobiology treatment stops at that methodological consequence. The larger search, mission history and probability of life elsewhere belong to Life in the Universe in a Hurry.

Sources for the Core Ideas

Life as a system. Pier Luigi Luisi develops the cell and minimal life through a systems approach. Carol Cleland argues that lists of familiar traits may fail to identify a universal theory of life. The causal example involving compartments and local benefits draws on protocell experiments by Chen, Hanczyc, Mansy, Adamala and Szostak and their collaborators.

Thermodynamics and gradients. The account of life as an open system far from equilibrium is standard physical chemistry. Local order is compatible with the second law because organisms take in free energy and export heat and waste. Peter Mitchell’s chemiosmotic theory established the central role of ion gradients in modern bioenergetics. Lane, Allen and Martin connect the universality of chemiosmosis to origin models, while Sojo and colleagues present a detailed alkaline-vent hypothesis. The origin inference is contested; the modern mechanism is not.

Membranes and protocells. Amphiphile self-assembly is well established. Hanczyc, Fujikawa and Szostak demonstrated encapsulation, growth and division in primitive compartment models. Mansy and colleagues demonstrated template-directed polymer synthesis inside a model protocell. Zhu and Szostak showed coupled growth and division of model membranes. Adamala, Engelhart and Szostak tested cooperation between a soluble catalyst and a membrane under magnesium stress. These systems use designed experimental conditions and should not be described as reconstructed early organisms.

RNA and catalysis. Natural ribozymes and the RNA catalytic centre of the ribosome support an RNA-rich stage. Horning and Joyce evolved an RNA polymerase ribozyme capable of making structured functional RNAs. Tjhung and colleagues reported synthesis of an ancestral ligase ribozyme in fragments. Papastavrou, Horning and Joyce demonstrated repeated RNA-catalysed propagation and selection of a functional RNA. The text distinguishes these advances from autonomous self-replication from prebiotic feedstocks.

Prebiotic feedstocks. Miller’s 1953 experiment established abiotic amino-acid formation in a strongly reducing gas mixture. Johnson and colleagues’ reanalysis of Miller’s archived volcanic-discharge samples clarified the wider product range and the relevance of locally reducing environments. Patel and colleagues produced precursors of ribonucleotides, amino acids and lipids through linked cyanosulfidic chemistry. Powner, Gerland and Sutherland reported a prebiotically plausible route to activated pyrimidine ribonucleotides. These studies demonstrate routes to components, not a continuous origin pathway.

Selection and error. The account of an error threshold follows the long-standing logic developed in molecular evolution: copying fidelity limits maintainable sequence length. The discussion avoids assigning one exact threshold because it depends on sequence, fitness effects, population structure and compartmentalisation. The distinction between chemical selection and open-ended Darwinian evolution follows the need for sufficiently rich heritable variation.

Origin environments. Surface hydrothermal-field and wet-dry models are represented by Deamer and by Damer and Deamer. Terrestrial geothermal-field arguments are represented by Mulkidjanian and colleagues. Alkaline-vent models are represented by Lane, Allen and Martin and by Sojo and colleagues. Experiments on thermal concentration, mineral catalysis and recent vent chemistry show that each environment can solve defined subproblems. The National Academies’ conclusion that no single geochemical scenario is decisive governs the wording.

One lineage and universal claims. Shared genetic coding, ribosomal machinery, molecular chirality and ATP use support common ancestry of sampled cellular life. They do not prove that those features are universal requirements. The astrobiology discussion therefore distinguishes chemical advantages from logical necessities and uses converging biosignatures rather than a terrestrial checklist.

Synthetic-cell integration. Abil and colleagues’ 2024 liposome system sustained self-encoded DNA replication and adaptive evolution through repeated rounds using a recombinant PURE transcription-translation system and managed redistribution. Restrepo Sierra and colleagues’ 2026 liposomes integrated a six-gene programme, DNA self-replication and phospholipid synthesis; the reported phospholipid yield remained below 7 per cent of total lipid, insufficient for membrane expansion. Khanal and colleagues separately generated new bilayer vesicles from soluble metabolites using soluble enzymes and chemical ligation without a pre-existing membrane template. These systems probe module compatibility but still rely on machinery derived from existing life.

Sources for the Origin Sequence

Hadean Earth. The zircon evidence supports early water but does not supply a complete map of oceans or continents. The atmosphere, impact rate and amount of exposed land changed through time and remain active research areas. The book uses the evidence only to establish that surface and hydrothermal chemistry had early opportunities.

Miller’s apparatus. Miller began the experiment in 1952 and published in 1953. The gases were methane, ammonia, hydrogen and water vapour, with electrical discharge as an energy source. Modern views of the bulk atmosphere are generally less reducing, although reducing local environments remain possible. Archived experiments analysed with modern methods have yielded additional amino acids and amines. None produced a living system.

Staged chemistry. Wet-dry, freeze-thaw and thermal-gradient mechanisms are presented as concentration and cycling devices rather than one historical sequence. The linked 2015 cyanosulfidic network is important because it connects precursor families, but it requires specified feedstocks, ultraviolet light and redox conditions. Moving a laboratory network into geology remains a separate task.

RNA progress. The 2016, 2020 and 2024 polymerase-ribozyme results were checked against the original PNAS papers. The 2026 QT45 result was checked against the Science paper and PubMed record. QT45 is a 45-nucleotide polymerase ribozyme that synthesised its complementary strand at 94.1 per cent per-nucleotide fidelity and, with defined substrates, a copy of itself. Full-length yields were about 0.2 per cent after roughly seventy-two days in mildly alkaline eutectic ice. The experiment did not demonstrate an autonomous repeated self-replication cycle from geochemical feedstocks.

Protocell reproduction. The text deliberately uses statistical continuity. Model vesicles can grow and divide through physical processes, but current experiments do not provide a full autonomous cycle in which plausible feedstocks continuously regenerate membrane, metabolism and genome. The role of compartments in suppressing or localising molecular parasites is supported by experimental and theoretical protocell work.

Hydrothermal energy. Serpentinisation can generate hydrogen-rich alkaline fluids. Natural pH and redox gradients, catalytic minerals and pore structures motivate vent models. The book does not claim that a modern membrane, ATP synthase or exact cellular gradient existed before life. It states a proposed continuity between geochemical and biological energy coupling.

LUCA. Moody and colleagues’ 2024 analysis infers an age near 4.2 billion years, a genome of at least about 2.5 million bases and around 2,600 proteins. Their method combines molecular clocks, horizontal-gene-transfer-aware reconciliation and geochemical modelling. Other reconstructions differ. The numbers are included as a current, model-dependent result; the stronger point is that LUCA was already a prokaryote-grade cellular system and not the origin itself.

Early life evidence. Lepot reviews signatures from the Archean record. Walter, Buick and Dunlop described stromatolites around 3.4 to 3.5 billion years old. Baumgartner and colleagues reported organic matter and pyrite structures in 3.5-billion-year-old stromatolites. Mißbach and colleagues found biologically relevant organics in fluid inclusions of similar age. Older proposed structures and isotope signals are not used as firm chronological anchors.

What People Get Wrong and Use It

The virus, thermodynamics, Miller-Urey, RNA-world, origin-setting, LUCA and water corrections are supported by the sources listed above. Quotation marks identify common claims rather than attributed quotations. The practical lenses distinguish definition from diagnosis, components from integrated systems, individual reactions from connected pathways, replication from heritable evolution, environments from mechanisms and possible biosignatures from converging detections.

Bibliography

Original research and scientific reports

Abil, Zhanar, Ana María Restrepo Sierra, Andreea R. Stan et al. “Darwinian Evolution of Self-Replicating DNA in a Synthetic Protocell.” Nature Communications 15 (2024): 9091. https://doi.org/10.1038/s41467-024-53226-0.

Adamala, Katarzyna P., Aaron E. Engelhart and Jack W. Szostak. “Collaboration between Primitive Cell Membranes and Soluble Catalysts.” Nature Communications 7 (2016): 11041. https://doi.org/10.1038/ncomms11041.

Baumgartner, Raphael J., Martin J. Van Kranendonk, David Wacey, Marco L. Fiorentini, Martin Saunders, Matteo Caruso, Cristian Pagès, Paul Homann and Peter Guagliardo. “Nano-Porous Pyrite and Organic Matter in 3.5-Billion-Year-Old Stromatolites Record Primordial Life.” Geology 47 (2019): 1039-1043. https://doi.org/10.1130/G46365.1.

Chen, Irene A., Richard W. Roberts and Jack W. Szostak. “The Emergence of Competition between Model Protocells.” Science 305 (2004): 1474-1476. https://doi.org/10.1126/science.1100757.

Damer, Bruce and David Deamer. “The Hot Spring Hypothesis for an Origin of Life.” Astrobiology 20 (2020): 429-452. https://doi.org/10.1089/ast.2019.2045.

Hanczyc, Martin M., Shelly M. Fujikawa and Jack W. Szostak. “Experimental Models of Primitive Cellular Compartments: Encapsulation, Growth, and Division.” Science 302 (2003): 618-622. https://doi.org/10.1126/science.1089904.

Johnson, Adam P., H. James Cleaves, Jason P. Dworkin, Daniel P. Glavin, Antonio Lazcano and Jeffrey L. Bada. “The Miller Volcanic Spark Discharge Experiment.” Science 322 (2008): 404. https://doi.org/10.1126/science.1161527.

Gianni, Edoardo, Samantha L. Y. Kwok, Christopher J. K. Wan, Kevin Goeij, Bryce E. Clifton, Enrico S. Colizzi, James Attwater and Philipp Holliger. “A Small Polymerase Ribozyme That Can Synthesize Itself and Its Complementary Strand.” Science 391 (2026): 1022-1028. https://doi.org/10.1126/science.adt2760.

Khanal, Satyam, Alessandro Fracassi, Alexander Harjung, Michael D. Burkart et al. “Breaking the Membrane Heredity Paradox through De Novo Protocell Formation.” Nature Communications 17 (2026): 6920. https://doi.org/10.1038/s41467-026-73667-z.

Horning, David P. and Gerald F. Joyce. “Amplification of RNA by an RNA Polymerase Ribozyme.” Proceedings of the National Academy of Sciences 113 (2016): 9786-9791. https://doi.org/10.1073/pnas.1610103113.

Lane, Nick, John F. Allen and William Martin. “How Did LUCA Make a Living? Chemiosmosis in the Origin of Life.” BioEssays 32 (2010): 271-280. https://doi.org/10.1002/bies.200900131.

Mansy, Sheref S., Jason P. Schrum, Mathangi Krishnamurthy, Sylvia Tobé, Douglas A. Treco and Jack W. Szostak. “Template-Directed Synthesis of a Genetic Polymer in a Model Protocell.” Nature 454 (2008): 122-125. https://doi.org/10.1038/nature07018.

Miller, Stanley L. “A Production of Amino Acids under Possible Primitive Earth Conditions.” Science 117 (1953): 528-529. https://doi.org/10.1126/science.117.3046.528.

Mitchell, Peter. “Coupling of Phosphorylation to Electron and Hydrogen Transfer by a Chemi-Osmotic Type of Mechanism.” Nature 191 (1961): 144-148. https://doi.org/10.1038/191144a0.

Mißbach, Helge, Jan-Peter Duda, Alfons M. van den Kerkhof, Volker Lüders, Andreas Pack, Joachim Reitner and Volker Thiel. “Ingredients for Microbial Life Preserved in 3.5 Billion-Year-Old Fluid Inclusions.” Nature Communications 12 (2021): 1101. https://doi.org/10.1038/s41467-021-21323-z.

Moody, Edmund R. R., Sandra Álvarez-Carretero, Tara A. Mahendrarajah et al. “The Nature of the Last Universal Common Ancestor and Its Impact on the Early Earth System.” Nature Ecology & Evolution 8 (2024): 1654-1666. https://doi.org/10.1038/s41559-024-02461-1.

Mojzsis, Stephen J., T. Mark Harrison and Robert T. Pidgeon. “Oxygen-Isotope Evidence from Ancient Zircons for Liquid Water at the Earth’s Surface 4,300 Myr Ago.” Nature 409 (2001): 178-181. https://doi.org/10.1038/35051557.

Mulkidjanian, Armen Y., Andrew Y. Bychkov, Daria V. Dibrova, Michael Y. Galperin and Eugene V. Koonin. “Origin of First Cells at Terrestrial, Anoxic Geothermal Fields.” Proceedings of the National Academy of Sciences 109 (2012): E821-E830. https://doi.org/10.1073/pnas.1117774109.

National Academies of Sciences, Engineering, and Medicine. An Astrobiology Science Strategy for the Search for Life in the Universe. Washington, DC: National Academies Press, 2019. https://doi.org/10.17226/25252.

NASA Astrobiology. “About: Life Detection.” Accessed 9 August 2026. https://astrobiology.nasa.gov/research/life-detection/about/.

Papastavrou, Nikolaos, David P. Horning and Gerald F. Joyce. “RNA-Catalyzed Evolution of Catalytic RNA.” Proceedings of the National Academy of Sciences 121 (2024): e2321592121. https://doi.org/10.1073/pnas.2321592121.

Patel, Bhavesh H., Claudia Percivalle, Dougal J. Ritson, Colm D. Duffy and John D. Sutherland. “Common Origins of RNA, Protein and Lipid Precursors in a Cyanosulfidic Protometabolism.” Nature Chemistry 7 (2015): 301-307. https://doi.org/10.1038/nchem.2202.

Powner, Matthew W., Béatrice Gerland and John D. Sutherland. “Synthesis of Activated Pyrimidine Ribonucleotides in Prebiotically Plausible Conditions.” Nature 459 (2009): 239-242. https://doi.org/10.1038/nature08013.

Restrepo Sierra, Ana María, Federico Ramirez Gomez, Mats van Tongeren et al. “A Synthetic Cell with Integrated DNA Self-Replication and Lipid Biosynthesis.” Nature Communications 17 (2026): 2727. https://doi.org/10.1038/s41467-026-69531-9.

Sojo, Víctor, Barry Herschy, Alexandra Whicher, Eloi Camprubí and Nick Lane. “The Origin of Life in Alkaline Hydrothermal Vents.” Astrobiology 16 (2016): 181-197. https://doi.org/10.1089/ast.2015.1406.

Tjhung, Katrina F., Maxim N. Shokhirev, David P. Horning and Gerald F. Joyce. “An RNA Polymerase Ribozyme That Synthesizes Its Own Ancestor.” Proceedings of the National Academy of Sciences 117 (2020): 2906-2913. https://doi.org/10.1073/pnas.1914282117.

Walter, Malcolm R., Roger Buick and John S. R. Dunlop. “Stromatolites 3,400-3,500 Myr Old from the North Pole Area, Western Australia.” Nature 284 (1980): 443-445. https://doi.org/10.1038/284443a0.

Wilde, Simon A., John W. Valley, William H. Peck and Colin M. Graham. “Evidence from Detrital Zircons for the Existence of Continental Crust and Oceans on the Earth 4.4 Gyr Ago.” Nature 409 (2001): 175-178. https://doi.org/10.1038/35051550.

Zhu, Ting F. and Jack W. Szostak. “Coupled Growth and Division of Model Protocell Membranes.” Journal of the American Chemical Society 131 (2009): 5705-5713. https://doi.org/10.1021/ja900919c.

Books and syntheses

Cleland, Carol E. The Quest for a Universal Theory of Life: Searching for Life as We Don’t Know It. Cambridge: Cambridge University Press, 2020.

Deamer, David W. Assembling Life: How Can Life Begin on Earth and Other Habitable Planets? New York: Oxford University Press, 2019.

Lane, Nick. The Vital Question: Energy, Evolution, and the Origins of Complex Life. London: Profile Books, 2015.

Lepot, Kevin. “Signatures of Early Microbial Life from the Archean (4 to 2.5 Ga) Eon.” Earth-Science Reviews 209 (2020): 103296. https://doi.org/10.1016/j.earscirev.2020.103296.

Luisi, Pier Luigi. The Emergence of Life: From Chemical Origins to Synthetic Biology. 2nd ed. Cambridge: Cambridge University Press, 2016.

Wilkens, Stephan. “ATP Synthesis, Chemistry of.” In Wiley Encyclopedia of Chemical Biology, edited by Tadhg P. Begley. Wiley, 2008. https://doi.org/10.1002/9780470048672.wecb648.

Zimmer, Carl. Life’s Edge: The Search for What It Means to Be Alive. New York: Dutton, 2021.