Contents
01 — The chemistry and cell biology you actually need Part 0 — Orientation

01 — The chemistry and cell biology you actually need

Before this: Ch 00 · Time: ~30 min

This chapter is deliberately incomplete. It covers the chemistry and cell biology that the rest of the curriculum leans on, and stops. You are not going to need orbital or the citric acid cycle. You are going to need to know why a molecule with a negative charge behaves differently from one without, and why nothing in a cell is ever called — only collided with.

If you already know first-year biology, skim to §5. That section is the one that matters most and is the one most often taught badly.

What you'll be able to do

The core idea

Cells are bags of water containing an extraordinarily crowded mixture of molecules, at a temperature where everything is in constant violent motion. There is no scheduler, no addressing, and no calling convention. Every interaction happens because two molecules randomly collided and stuck together for a while.

Biological specificity — the fact that a finds one site among three billion — is not achieved by lookup. It is achieved by making the correct partner stick a few thousand times longer than the incorrect one, and then relying on enormous numbers of collisions per second. Specificity is statistical.

Once that clicks, a great deal of molecular biology stops seeming magical and starts seeming like a rate problem.


1. Atoms, bonds, and the only chemistry that matters here

Biological molecules are built from a small alphabet: carbon, hydrogen, oxygen, nitrogen, phosphorus, sulfur — plus ions like sodium, potassium, magnesium and calcium.

Two kinds of connection matter:

Covalent bonds — atoms sharing electrons. Strong and long-lived at body temperature, and they form the backbone of every macromolecule. When you read that DNA is a stable archive, this is why: its backbone is covalent. Breaking one usually takes an enzyme — but not always, and the exception that matters most is a bond inside DNA itself, which §3 comes to.

Non-covalent interactions — weak, transient, individually feeble, collectively decisive:

InteractionStrengthWhere you'll meet it
Hydrogen bondweakBase pairing; protein secondary structure; almost all molecular recognition
Ionic (electrostatic)weak in waterProtein–DNA binding — the backbone's negative charge
Van der Waalsvery weakClose-packing;
—Protein folding; membranes; the interior of the double helix

The whole logic of molecular biology: covalent bonds store, non-covalent interactions decide. Anything that has to be reversible — a protein binding DNA, two strands separating, an enzyme releasing product — is non-covalent. Anything that has to persist is covalent.

2. Water is not the background — it's a participant

Water is polar: the oxygen pulls electron density from the hydrogens, leaving the molecule with a slightly negative end and a slightly positive end. Consequences that recur constantly:

Polar and charged molecules dissolve; non-polar ones don't. Water forms favourable interactions with the former and not the latter.

The . Non-polar molecules in water clump together. Not because they attract each other — because water molecules cannot form their usual hydrogen-bond network around them, so burying non-polar surface frees up water. It is driven by water's entropy, not by any force between the greasy bits. This counterintuitive point does an enormous amount of work:

The double helix is, to a first approximation, what you get when you put a charged polymer with flat greasy rungs into water. Structure follows from solvent.

Hydrogen bonds are directional and specific, which is what lets A pair with T and G with C rather than promiscuously. Base pairing is a hydrogen-bonding geometry problem.

pH decides what is charged. Water dissociates slightly into H⁺ and OH⁻, and pH is just a log scale for how much H⁺ is around — 7 is neutral, lower is more acidic, and cytoplasm and blood both sit near 7.4. It matters because whether a chemical group carries a charge depends on the pH surrounding it: at cellular pH every phosphate along the DNA backbone is ionised, which is where DNA's negative charge comes from, and a handful of amino acid side chains are charged while the rest are not.

and — the two roles in every reaction in this book. Because electron density is unevenly shared, some atoms have a lone pair or a negative charge to spare and others are short of one. The electron-rich partner is the nucleophile — "nucleus-loving" — and it does the attacking; the electron-poor partner is the electrophile, and it is what gets attacked. Nearly every covalent bond made or broken in this curriculum is one such attack, and usually the same one: an oxygen carrying a lone pair attacking a phosphorus atom flanked by electron-hungry oxygens. That single reaction extends DNA, extends RNA, cuts either of them, and joins two fragments back together. So when Ch 02 labels a nucleotide's 3'-hydroxyl the nucleophile and an incoming nucleotide's phosphate the electrophile, that is the whole of what it means — and it is why a polymerase can extend only one end of a strand.

3. The four macromolecules

Biological polymers, built by linking monomers with covalent bonds.

Nucleic acids — DNA and RNA

Monomer: a nucleotide = sugar + phosphate + base. Phosphates link sugars into a backbone; bases hang off it.

DNARNA
Sugardeoxyriboseribose (one extra –OH)
BasesA, C, G, TA, C, G, U
Strandsdoubleusually single
Stabilityhighlow — that extra –OH makes it self-cleaving
Rolearchiveworking copy, catalyst, regulator

RNA's chemical instability is a feature: working copies should be disposable. A single hydroxyl group is the difference between an archive and a message queue.

The four bases fall into two chemical classes, and the split does real work later. A and G are — two fused rings, so the larger bases. C, T and RNA's U are — one ring, smaller. Every pair in the double helix is one of each, which is what holds the helix to a constant width; Ch 02 derives the pairing rules from that rather than asking you to memorise them.

The purines are also the archive's weak point. The bond joining a purine to its sugar hydrolyses spontaneously — no enzyme needed and no damage required — so DNA loses purines continuously, on the order of 10⁴ per cell per day. That slow chemical attrition is where a large share of comes from (Ch 16), and it is why a stable archive still needs constant repair.

The backbone carries one negative charge per nucleotide. DNA is a strongly polyanionic molecule, and this matters more than it sounds:

Proteins

Monomer: amino acid — a common core plus one of 20 side chains. The side chains are the whole story: some charged, some polar, some greasy, one (cysteine) able to form covalent cross-links, one (proline) that kinks the chain, one (glycine) so small it grants flexibility.

Amino acids link into a chain, and the chain folds into a specific three-dimensional structure determined by the sequence — largely by burying hydrophobic side chains inside and leaving polar ones exposed.

Two intermediate shapes are worth naming now, because later chapters use them without stopping to explain them. Locally, the backbone's own hydrogen bonds fold it into one of two regular motifs: the α-helix, a right-handed spiral roughly 12 Å across, and the β-sheet, several extended stretches of chain lying side by side. These are secondary structure; how those helices and sheets then pack together into a whole folded unit is tertiary structure. The α-helix's width is not trivia — it is why a protein can push one into the wider of DNA's two grooves and read the sequence there (Ch 02 §5).

Shape is function. A protein works by having a surface complementary to its partner. This is why a single amino acid change can abolish function (if it disrupts the fold or the active site) or do nothing at all (if it sits on the surface pointing at nothing) — a fact that underlies the whole difficulty of variant interpretation (Ch 55).

Proteins do essentially all the work: catalysis (enzymes), structure, transport, signalling, and regulation of the genome itself.

Lipids and carbohydrates

Briefly, because they matter less here. Lipids are non-polar; in water they self-assemble into bilayers, which is where all cellular membranes come from — including the nuclear envelope that separates from in eukaryotes. Carbohydrates are sugars, used for energy and, attached to proteins, for molecular identity tags — including the ABO blood group system you'll meet in Ch 11.

4. The cell

Two architectures.

Prokaryotes (bacteria and archaea) — no nucleus. One circular sitting in the cytoplasm. and happen in the same space, simultaneously: latch onto RNA while it's still being transcribed. Small, fast, streamlined genomes.

Eukaryotes (everything else — animals, plants, fungi, protists) — DNA enclosed in a nucleus. Transcription happens inside, translation outside. That separation creates a processing step in between, and that step is where a great deal of regulatory complexity lives (Ch 06).

graph TB subgraph EU["Eukaryotic cell"] N["<b>Nucleus</b><br/>DNA, transcription,<br/>RNA processing"] C["<b>Cytoplasm</b><br/>translation, metabolism"] R["<b>Ribosomes</b><br/>make protein from RNA"] M["<b>Mitochondria</b><br/>energy — and their own<br/>small genome"] N -->|"mature mRNA<br/>exported"| C C --- R C --- M end

The parts that recur in this curriculum:

Mitochondria having their own genome is not a quirk. They are descended from free-living bacteria engulfed by an ancestral cell — endosymbiosis — and they have retained a remnant chromosome ever since. Chloroplasts in plants have the same origin.

5. The part that changes how you think: everything is stochastic

This is the section to read twice.

A protein does not go to its binding site. There is no address, no dispatch, no call. It diffuses — random-walks, driven by thermal motion — colliding with everything, and at every collision it either sticks briefly or doesn't. It finds its site because sticking there lasts longer.

Some numbers to calibrate against:

QuantityOrder of magnitude
Diameter of the DNA double helix2 nm
Length of DNA per human cell~2 m, packed into a ~6 μm nucleus
Typical protein3–10 nm
Molecular collisions~10⁹ per second per molecule
rate~1–2 kb per minute
rate~5–20 amino acids per second
Proteins in one cell~10⁹–10¹⁰
Copies of a typical protein10²–10⁵

The nucleus is extremely crowded — nothing like a dilute solution. Macromolecules occupy 20–30% of the volume. Diffusion is hindered, local concentrations are far from uniform, and there is a live argument — biomolecular and separation are the current framing — that some of the organisation once attributed to active machinery may instead emerge from crowding and weak multivalent interactions. How much is unresolved: the in-cell evidence that a given nuclear focus is a genuine liquid condensate rather than something that merely looks like one is actively contested.

Three consequences that will keep mattering:

Binding is a probability, not a state. A "bound to" a is actually bound some fraction of the time — associating and dissociating continually. Gene expression levels are the time average of an intrinsically noisy process. Two genetically identical cells in identical conditions express different amounts of the same gene. That noise is measurable, sometimes functional, and it's why single-cell measurements (Ch 48) show distributions where bulk measurements show a single number.

Specificity is relative, not absolute. A transcription factor that binds its target sequence a thousand times more tightly than random DNA sounds highly specific — until you remember there are three billion positions. It spends a great deal of time bound to the wrong places. Real specificity comes from combinations: several factors that must all be present, each individually sloppy, jointly precise. This is why regulatory elements contain clusters of binding sites rather than one.

Error rates are set by kinetics, and errors are managed rather than prevented. mis-incorporates at some rate; catches most of it; catches most of the rest. Those three imperfect filters in series — not one perfect mechanism — get replication down to roughly 10⁻¹⁰ per base per replication. The observed rate is ~1.1–1.3 × 10⁻⁸ per base per generation, a hundredfold higher, and the gap is itself instructive: a generation stacks hundreds of germline cell divisions on top of that filter stack, and much germline begins as spontaneous chemical damage that never passed through a polymerase at all (Ch 16). Every fidelity number in biology decomposes this way — and every rate needs its denominator checked.

For programmers. The mental model to discard is the function call. The one to adopt is something like a massively concurrent system with no locks, no ordering guarantees, and unreliable message delivery — where correctness is achieved statistically, by redundancy and by error correction downstream, and where the observable behaviour is an ensemble average over enormous numbers of unreliable events.

6. Energy, equilibrium, and rates

Reactions that don't happen spontaneously are driven by coupling them to one that does — usually the breakdown of a nucleoside triphosphate, most familiarly ATP. When you read that a "unwinds" DNA, that is ATP hydrolysis doing the work. The currency varies: polymerases are powered by the dNTPs and NTPs they are polymerising rather than by ATP, and when a polymerase "proofreads" the cost is not a triphosphate at all — the excising step is plain hydrolysis, and what gets paid is the nucleotide thrown away and added again. You don't need the biochemistry; you need to know that molecular machines consume energy to do directional work, and that this is what stops everything from simply drifting to equilibrium and stopping. A cell at equilibrium is dead — that is very nearly the definition.

You do need a small amount of the bookkeeping, though, because from the next chapter onward the curriculum quotes energies in kcal/mol and expects you to read them. Here it is, once.

Equilibrium and K. Take a reversible reaction, two molecules sticking together: A + B ⇌ AB. Left alone it does not stop — it reaches a state where the forward and reverse rates are equal and the concentrations stop changing. That state is summarised by a single number, the :

  K = [AB] / ([A][B])

Large K means the reaction sits mostly on the right, small K mostly on the left. K depends on the reaction and the temperature, not on how much you started with. (Concentrations are counted against a 1 M reference, which is what the ° superscripts below mean and what makes taking a logarithm of K legal.)

. The quantity that predicts K is the standard free-energy change, ΔG° — the thermodynamic driving force, quoted in kcal/mol. It splits into two parts:

  ΔG°  =  ΔH°  −  T·ΔS°

ΔH° (enthalpy) is roughly the heat balance: whether the bonds and contacts formed are better than the ones broken. ΔS° (entropy) is the change in how many arrangements are available. T is absolute temperature in kelvin, so the entropy term matters more the hotter things get. A negative ΔG° means the reaction is favourable as written. This is the equation behind §2's : burying a greasy surface releases ordered water, so ΔS° is positive, −TΔS° is negative, and the process is driven by entropy rather than by any attraction.

The two are the same fact in different units:

  ΔG°  =  −R·T·ln K          R = 1.987 cal/(mol·K), the gas constant

At body temperature every 1.4 kcal/mol of ΔG° is about a factor of ten in K. So the thousandfold preference §5 describes — a holding its own site that much more tightly than random DNA — amounts to only about 4 kcal/mol, two or three hydrogen bonds' worth. Molecular recognition runs on very small energy differences, which is exactly why it is never absolute.

Rates are a separate question. ΔG° tells you where a reaction ends up, never how fast it gets there. Speed is set by the barrier between start and finish — the free energy of the highest point on the path — and lowering barriers is the whole of what an enzyme does. An enzyme cannot change K, and so cannot make an unfavourable reaction favourable; it can only bring a reaction that was already going to happen forward by millions of times. Every "enzyme X catalyses Y" in this book means precisely that: the barrier came down, the destination did not move.

The first place this gets cashed in is Ch 02 §7, where the temperature at which a DNA duplex comes apart is this equation applied to A + B ⇌ AB.


Common misconceptions

What people thinkWhat's actually true
Molecules in a cell are transported to where they're neededThe overwhelming majority arrive by diffusion and random collision. Active transport exists but is the exception
An enzyme and its substrate fit like a lock and keyBoth flex; binding is dynamic and often induces fit. "Lock and key" was a useful nineteenth-century approximation
Hydrophobic molecules attract each otherThey're pushed together by water reorganising around them. The driving force is water's entropy, not attraction between the non-polar parts
A bound is boundIt's bound a fraction of the time, cycling on and off continuously. Occupancy is a probability
Genetically identical cells behave identicallyThey don't. Expression is intrinsically noisy because it depends on small numbers of stochastic binding events
DNA is a rigid ladderIt bends, twists, breathes open locally, and is under constant torsional stress. Its flexibility is functionally essential

Worked example: why the double helix has the shape it does

Predict the structure from the chemistry alone.

Given: a polymer whose backbone carries one negative charge per unit, with flat, largely non-polar bases attached, dissolved in water.

  1. The charged backbone must face outward. Burying charge away from water is energetically expensive. So: backbone outside, bases inside.
  1. The flat bases will stack. Once in the interior, away from water, they pack face-to-face — favourable van der Waals contact and minimal exposed non-polar surface. Base stacking, it turns out, contributes more to helix stability than base pairing does.
  1. Two strands, because the bases have unsatisfied hydrogen-bonding groups. Buried in the interior with no water available, those groups must pair with something. The something is a base on a second strand.
  1. Pairing must be A–T and G–C to keep the helix a constant width: a large two-ring must pair with a small one-ring , and the hydrogen-bond donor/acceptor geometry permits only these two combinations.
  1. The result twists because the backbone geometry can't stack the bases at a constant spacing without rotation. Roughly 10.5 base pairs per turn.
  1. The two strands run antiparallel because the pairing geometry demands it.
  1. Grooves of unequal width follow from the backbones not being diametrically opposite. The wider exposes enough of each base pair for a protein to read the sequence without opening the helix — which is how recognise their sites.

Nothing here required knowing the answer in advance. The structure is close to forced by putting that polymer in water, and every feature turns out to be functionally load-bearing.

Connections