DNA Replication & Repair

DNA Replication for the MCAT

Part 1 of 4 — DNA Structure & the Semiconservative Logic

The Molecule You Are Copying

• Nucleotide = nitrogenous base + deoxyribose + phosphate(s). Base + sugar alone = nucleoside. Purines (Adenine, Guanine) are two-ringed; pyrimidines (Cytosine, Thymine, and uracil in RNA) are one-ringed. Mnemonics: "Pure As Gold" (purines = A, G); "CUT the PY" (pyrimidines = C, U, T).
• Base pairing: A pairs T via 2 hydrogen bonds; G pairs C via 3. Consequence: GC-rich DNA has a higher melting temperature ($T_m$) — more heat to separate strands. Every pairing is purine + pyrimidine, keeping helix width constant (~2 nm).
• Chargaff's rules: in double-stranded DNA, %A = %T and %G = %C. A virus with %A different from %T must have single-stranded genetic material — a classic trap.
• Directionality: the sugar-phosphate backbone links the 3' hydroxyl of one sugar to the 5' phosphate of the next (phosphodiester bonds). The two strands are antiparallel: one runs 5' to 3', its partner 3' to 5'. Sequences are written 5' to 3' by convention.
• Why DNA and not RNA for storage: deoxyribose lacks the 2'-OH that makes RNA alkali-labile and prone to self-cleavage; thymine (methylated uracil) lets repair systems recognize deaminated cytosine (which becomes uracil) as damage.

Helix Geometry & Packaging

• B-DNA: right-handed double helix, major and minor grooves; sequence-specific proteins (transcription factors) mostly read the major groove.
• Eukaryotic packaging: DNA wraps around a histone octamer (2 each of H2A, H2B, H3, H4) forming a nucleosome; H1 seals the linker. Histones are lysine/arginine-rich (positively charged) to bind the polyanionic backbone — acetylation neutralizes that charge and loosens packing (regulation lesson crossover).

• The wrap is about 1.7 turns (~147 bp) per octamer.

• Heterochromatin = dense, transcriptionally quiet; euchromatin = open, active. Replication must transiently displace and then restore this packaging.

The Central Question of 1958: How Does Copying Distribute Old Strands?

Three hypotheses were on the table:

Meselson and Stahl settled it with density labeling — the worked example below walks the actual logic, because the MCAT re-runs this experiment constantly with new labels.

Worked Example — Meselson-Stahl, Read Like a Passage

Setup. E. coli are grown for many generations in medium containing heavy nitrogen ($^{15}N$), so all DNA is "heavy." Cells are then switched to light $^{14}N$ medium, and DNA is extracted after each generation and spun in a cesium chloride density gradient, where DNA bands at the position matching its density.

Generation 1 result: a single band at intermediate (hybrid) density.

Question 1 — Which model does generation 1 eliminate? Conservative replication predicted TWO bands (one heavy, one light). One hybrid band kills conservative — but note it does NOT yet distinguish semiconservative from dispersive, since both predict all molecules contain old and new material after round one. Recognizing what an experiment has NOT yet excluded is a signature MCAT move.

Question 2 — Generation 2 shows two bands: one hybrid, one light, in equal amounts. Which model survives? Semiconservative. Each hybrid molecule from generation 1 separates into one old ($^{15}N$) and one new ($^{14}N$) template; the old template rebuilds a hybrid, the new template builds an all-light molecule. Dispersive predicted a SINGLE band drifting progressively lighter — never two discrete bands.

Question 3 — Predict generation 3. Hybrid molecules stay constant in number (the original heavy strands are immortal templates, one per molecule) while light molecules multiply: 1/4 hybrid, 3/4 light. In general after n generations the hybrid fraction is $1/2^{n-1}$ — the two original strands never disappear, they just get diluted among descendants.

Transfer test: a passage that pulse-labels with BrdU or radioactive thymidine and asks which strands carry label is the same puzzle in new clothes. Always draw the strands, mark old vs new, and count.

DNA Structure & Replication Models 🎯

Key Takeaways — Part 1

• Nucleotide = base + deoxyribose + phosphate; purines (A, G) pair with pyrimidines (T, C); A-T = 2 H-bonds, G-C = 3, so GC content raises melting temperature and origins are AT-rich
• Chargaff (%A = %T, %G = %C) holds only for double-stranded genomes — composition data diagnose strandedness
• Strands are antiparallel with phosphodiester backbones; sequences are written 5' to 3'; reverse-complement in two explicit steps
• Thymine instead of uracil lets cells recognize deaminated cytosine (uracil in DNA) as damage; the 2'-OH absent in deoxyribose makes DNA more stable than RNA
• Packaging: nucleosome = DNA around a histone octamer (positively charged lysine/arginine tails); acetylation loosens, heterochromatin silences
• Meselson-Stahl: one hybrid band at generation 1 kills conservative; two discrete bands (hybrid + light) at generation 2 kills dispersive; hybrid fraction = 1 over 2 to the (n-1) as original strands dilute
• Passage skill: state explicitly which models each observation excludes — early data often eliminate fewer hypotheses than they seem to

• Low-yield extras: DNA wraps about 1.7 turns (~147 bp) around each histone octamer

DNA Replication for the MCAT

Part 2 of 4 — The Replication Fork: Enzymes & Choreography

Opening the Helix

• Origin of replication: replication starts at defined, AT-rich sequences (easier to melt — only 2 H-bonds per pair). Prokaryotes: ONE origin on a circular chromosome, two forks moving in opposite directions. Eukaryotes: THOUSANDS of origins per linear chromosome — the only way to copy a huge genome in one S phase.
• Helicase unwinds the duplex at the fork, hydrolyzing ATP to break hydrogen bonds (not covalent bonds).
• Single-stranded binding proteins (SSBs) coat the exposed strands so they neither re-anneal nor fold into hairpins.
• Topoisomerase works AHEAD of the fork: unwinding piles up positive supercoils downstream, and topoisomerases nick (type I) or double-cut (type II/gyrase) the backbone, relax the strain, and reseal. Fluoroquinolone antibiotics poison bacterial gyrase — replication forks stall on their own tension.

Building the New Strands

Every DNA polymerase obeys two absolute rules:

1. Synthesis is 5' to 3' only — new nucleotides add to a free 3'-OH. The energy comes from cleaving the incoming dNTP's own triphosphate (pyrophosphate release, then hydrolysis, pulls the reaction forward).
2. No polymerase can start a chain — each needs a primer. Primase (an RNA polymerase, which CAN start de novo) lays down short RNA primers.

Because strands are antiparallel and the fork opens in one direction:

• Lagging-strand cleanup: RNA primers are removed (in prokaryotes by Pol I's 5' to 3' exonuclease), the gaps are filled with DNA, and DNA ligase seals the final nick (forms the last phosphodiester bond).

• RNase H, which degrades the RNA strand of RNA-DNA hybrids, also helps remove primers, especially in eukaryotes. Ligase is powered by ATP in eukaryotes and by NAD$^+$ in bacteria.

• Sliding clamp (beta clamp / PCNA) rings the DNA and holds the polymerase on — the source of processivity.
• Workhorses: prokaryotic Pol III (main synthesis), Pol I (primer removal + fill-in); eukaryotes split the same jobs among several specialized polymerases.

• The eukaryotic polymerases: Pol alpha (with primase, starts chains), Pol delta and Pol epsilon (extend the lagging and leading strands respectively).

• The whole assembly — helicase, primase, polymerases, clamps — moves as one replisome; the lagging-strand template loops so both polymerases travel together.

Worked Example — Diagnosing Mutant Forks

Passage-style problem. Four temperature-sensitive E. coli strains are shifted to the restrictive temperature; replication intermediates are analyzed on gels that separate short DNA fragments from long ones.

• Strain 1: replication halts almost immediately; no new strands of any kind appear.
• Strain 2: short fragments accumulate and are never joined into long strands; the fragments each begin with a stretch of RNA.
• Strain 3: short fragments accumulate and ARE all-DNA, but they are never joined.
• Strain 4: synthesis proceeds briefly, then forks stall; supercoiling ahead of the fork is dramatically elevated.

Question 1 — Assign Strain 1. No unwinding, no templates, no synthesis: helicase is dead. (A primase mutant would stop new priming, but forks already under way would keep extending their already-primed leading strands; a total shutdown of both strands from the start points to the enzyme every other enzyme waits on.)

Question 2 — Distinguish Strains 2 and 3. Both look like "Okazaki fragments never mature," but the fragments' composition is the tell. Strain 2's fragments still carry their RNA primers — primer removal (Pol I's 5' to 3' exonuclease function) has failed, and ligase cannot seal a nick against RNA. Strain 3's fragments are clean DNA with primers removed and gaps filled — only the final nick-sealing is missing: a ligase mutant. Order of operations is the discriminator: remove primer, fill gap, THEN ligate.

Question 3 — Assign Strain 4 and name a drug with the same phenotype. Torsional strain accumulating ahead of the fork means topoisomerase (gyrase) failure — helicase keeps injecting supercoils no one relaxes, and the fork grinds to a halt. Fluoroquinolones (e.g., ciprofloxacin) phenocopy this by trapping gyrase on DNA.

The general skill: match each phenotype to the FIRST step that fails, and use fragment composition (RNA-tipped vs all-DNA) to order the lagging-strand maturation steps.

Fork Machinery 🎯

Key Takeaways — Part 2

• Fork crew: helicase unwinds (ATP, breaks H-bonds), SSBs keep strands apart, topoisomerase/gyrase relaxes supercoils AHEAD of the fork (fluoroquinolone target)
• Two iron laws of polymerases: synthesis 5' to 3' only (3'-OH attacks incoming dNTP; pyrophosphate release drives it) and no de novo starts — primase makes RNA primers
• Leading strand: continuous, one primer; lagging strand: Okazaki fragments, primer each; maturation order = remove primer, fill with DNA, ligase seals the last nick
• Sliding clamp (beta/PCNA) = processivity, not catalysis; the replisome loops the lagging template so both polymerases move together
• Division of labor: Pol III synthesizes, Pol I removes primers and fills (prokaryotes); eukaryotes spread the same jobs over several polymerases
• One bacterial origin vs thousands of eukaryotic origins is a genome-size and fork-speed argument — be ready to do the arithmetic
• Mutant-phenotype logic: total shutdown = helicase; RNA-tipped fragments = primer removal; all-DNA unjoined fragments = ligase; runaway supercoiling = topoisomerase

• Low-yield extras: RNase H also strips RNA primers (especially in eukaryotes); bacterial ligase uses NAD+ rather than ATP; the eukaryotic polymerases are alpha-primase (starts chains), delta (lagging) and epsilon (leading)

DNA Replication for the MCAT

Part 3 of 4 — Fidelity, Telomeres & DNA Repair

Three Tiers of Accuracy

Replication is extraordinarily accurate because it stacks three filters:

1. Base selection: correct Watson-Crick geometry fits the active site; wrong pairs distort it.
2. Proofreading: replicative polymerases carry a 3' to 5' exonuclease — a mispaired 3' end will not sit properly for extension, the polymerase backs up, excises the wrong base, and retries. Note the direction: proofreading exonuclease is 3' to 5' (backwards over what was just made); primer removal is 5' to 3'.
3. Mismatch repair (MMR): AFTER the fork passes, mismatch-repair proteins find distortions, identify the new strand (in E. coli, the strand not yet methylated at GATC sites; in eukaryotes, nicks mark it), excise the error, and resynthesize. Inherited mismatch-repair loss causes hereditary colon cancer (Lynch syndrome).

• The numbers: base selection alone errs about once per $10^5$ nucleotides, proofreading improves that ~100-fold, and mismatch repair another ~100-1000-fold, for a final rate near one error per $10^9$-$10^{10}$ base pairs.
• The E. coli MMR proteins are MutS and MutL (MutH nicks the unmethylated new strand); humans use MSH and MLH homologs. Lynch syndrome is also called HNPCC, and its tumors show microsatellite instability.

The End-Replication Problem & Telomerase

• On a linear chromosome, removing the lagging strand's final RNA primer leaves a gap no polymerase can fill (no 3'-OH upstream). Chromosomes would shorten every division.
• Telomeres — short repeats at chromosome ends — are disposable buffers, capped by protective proteins so ends are not mistaken for breaks.

• The human telomere repeat is TTAGGG, and the protective cap is the shelterin complex.

• Telomerase is a reverse transcriptase carrying its own RNA template; it extends the parental 3' overhang with fresh repeats, letting standard priming and lagging-strand synthesis fill in the complement.
• Expression logic the MCAT tests: active in germ cells, stem cells, and most cancers (immortalization); low in most somatic cells, whose telomere erosion contributes to replicative senescence (the Hayflick limit).

Repairing Damage (Not Just Copying Errors)

Checkpoint tie-in: damage detected in G1/S or G2/M halts the cycle via p53 until repair completes — the bridge to the cell-cycle lesson.

Worked Example — Telomeres and Immortal Cell Lines

Passage-style problem. Researchers culture human fibroblasts: normal cells (line N), cells engineered to express telomerase constitutively (line T), and cells from a patient tumor (line C). Mean telomere length and population doublings are tracked.

• Line N: telomeres shorten ~50-100 bp per division; growth arrests permanently after ~60 doublings, though cells stay alive.
• Line T: telomere length is stable; the cells divide indefinitely.
• Line C: telomeres are SHORT but stable; the cells divide indefinitely.

Question 1 — Why do line N telomeres shorten a fixed amount per division? Each round of lagging-strand synthesis fails to replace the terminal RNA primer's position — the end-replication problem — and end-processing trims further. Without telomerase, the loss is cumulative and roughly constant per division, making telomere length a "mitotic clock."

Question 2 — What is the arrested state of line N, and what triggers it? Replicative senescence (Hayflick limit): critically short telomeres lose their protective cap, are recognized as double-strand breaks, and fire a p53-mediated checkpoint arrest. The cells are metabolically alive but permanently out of the cycle — distinguish senescence (arrest) from apoptosis (death).

Question 3 — Reconcile line C: short telomeres, yet immortal. The tumor reactivated telomerase LATE, after substantial shortening had already occurred; telomerase needs only to MAINTAIN length, not restore it. Short-but-stable is the signature of telomerase reactivation during tumor progression — most human cancers do this. It also explains why telomerase inhibitors are explored as therapy, and why line T (activated while telomeres were long) shows long stable telomeres instead.

Data-reading habit: on telomere plots, the SLOPE tells you telomerase status; the INTERCEPT tells you the history before measurement began.

Fidelity & Repair 🎯

Key Takeaways — Part 3

• Fidelity stacks three filters: base-pair geometry, 3' to 5' proofreading exonuclease (mutator strains when lost), and post-fork mismatch repair with strand discrimination (unmethylated GATC in E. coli; nicks in eukaryotes; inherited loss = Lynch syndrome, hereditary colon cancer)
• Directions matter: proofreading = 3' to 5'; primer removal = 5' to 3'
• End-replication problem: the last lagging-strand primer's position cannot be filled; telomeres (repeat buffers) absorb the loss
• Telomerase = reverse transcriptase with internal RNA template, extending the 3' overhang; active in germ/stem cells and most cancers; its absence in somatic cells sets the Hayflick limit via p53-triggered senescence (arrest, not death)
• Telomere data: slope = telomerase status; short-but-stable = late reactivation in a tumor
• Match lesion to pathway: uracil/oxidized base = BER; bulky UV dimer = NER (xeroderma pigmentosum); replication mismatch = MMR; double-strand break = HR (faithful, sister template, BRCA) or NHEJ (error-prone, any phase)
• Damage arrests the cycle through p53 checkpoints — repair and cell-cycle control are one system

• Low-yield extras: the error-rate arithmetic (about 1 in $10^5$ from base selection, ~100-fold from proofreading, ~100-1000-fold from MMR, ending near 1 in $10^9$-$10^{10}$); MMR protein names (MutS, MutL, MutH; human MSH/MLH); Lynch syndrome's other name (HNPCC) and its microsatellite instability; the TTAGGG repeat and its shelterin cap

DNA Replication for the MCAT

Part 4 of 4 — Replication in Context: Cell Cycle, Drugs & MCAT Integration

Once and Only Once — Licensing

• Replication is confined to S phase, and every origin must fire at most once per cycle; re-replication means amplified, unstable DNA.
• The trick is temporal separation: licensing (loading inactive helicases at origins) is allowed only in late M/G1 when CDK activity is LOW; firing (activating those helicases) requires the HIGH S-phase CDK activity that simultaneously forbids new licensing. The same signal that starts replication prevents relicensing — an elegant mutual-exclusion design the MCAT phrases as "why can't an origin fire twice?"
• Checkpoint coupling: stalled forks and damage activate checkpoint kinases that halt the cycle (DNA damage also acts through p53); S does not hand off to M until replication completes (G2/M checkpoint).

• The sensor kinases are ATR (stalled forks, exposed single-stranded DNA) and ATM (double-strand breaks).

Prokaryote vs Eukaryote Scorecard

• Table details: forks move roughly 1000 nt/s in bacteria vs about 50 nt/s in eukaryotes; the eukaryotic polymerases are Pol alpha-primase (starts chains), Pol delta (lagging) and Pol epsilon (leading).

Drugs That Target Replication — Mechanism Classes

1. Chain terminators (nucleoside analogs): look like nucleosides but lack a usable 3'-OH. Acyclovir (activated by viral thymidine kinase — selectivity!) and AZT (azide at the 3' position; inhibits HIV reverse transcriptase) get incorporated, then nothing can be added. Same chemistry as Sanger's dideoxynucleotides in biotechnology.
2. Topoisomerase poisons: fluoroquinolones (bacterial gyrase) and several anticancer drugs (human topoisomerases I and II). These trap cleavage complexes, converting a normal intermediate into double-strand breaks.
3. Antimetabolites: methotrexate (dihydrofolate reductase) and 5-fluorouracil (thymidylate synthase) starve the cell of dNTPs — replication stalls for lack of building blocks. Rapidly dividing tissues (marrow, gut, hair) take collateral damage: the classic chemotherapy side-effect triad.
4. Crosslinkers/intercalators: crosslinking drugs bond bases together into bulky lesions (an NER substrate); intercalators distort the helix and block fork passage.

• Named examples: etoposide poisons human topo II and camptothecin analogs poison topo I; cisplatin crosslinks guanines.

Selectivity logic to carry into pharmacology passages: a good drug exploits a difference — a viral enzyme (thymidine kinase, reverse transcriptase), a bacterial-only target (gyrase), or merely a KINETIC difference (cancer cells divide more; antimetabolites hit them hardest).

Worked Example — An Antiviral Passage, Fully Dissected

Passage-style problem. Acyclovir is a guanosine analog whose "sugar" is an open chain rather than a ring, with no 3'-OH. It is inert until phosphorylated; the FIRST phosphorylation can be performed efficiently only by herpesvirus thymidine kinase (TK), after which host kinases finish making the triphosphate. Resistant herpes isolates from immunocompromised patients frequently carry loss-of-function mutations in viral TK.

Question 1 — Why does acyclovir stop viral DNA synthesis mechanistically? Once the triphosphate is incorporated by the viral DNA polymerase opposite a template C, the new 3' end has no hydroxyl. Polymerization chemistry requires that 3'-OH to attack the next dNTP — synthesis terminates absolutely. This is chain termination, identical in principle to a dideoxynucleotide in Sanger sequencing.

Question 2 — Explain the drug's selectivity for infected cells. The prodrug design puts the selective step FIRST: uninfected cells lack viral TK, so acyclovir is never activated in them; infected cells convert it to the active triphosphate, which additionally inhibits and is preferentially incorporated by the VIRAL polymerase. Two multiplied selectivity filters (activation + target preference) give a wide therapeutic index — contrast with antimetabolite chemotherapy, whose only filter is division rate.

Question 3 — Why do TK mutations cause resistance, and what does that predict about such strains? No viral TK, no first phosphorylation, no active drug — resistance by FAILURE TO ACTIVATE, not by target alteration. Prediction: TK-negative strains remain sensitive to drugs that do not need TK activation, such as ones that hit the viral polymerase directly — resistance mechanisms are pathway-specific.

The standard example is foscarnet, a pyrophosphate analog that blocks the viral DNA polymerase without any activation step.

Question 4 — Data twist. If a passage shows acyclovir-monophosphate accumulating but no triphosphate, the block is at the HOST kinase steps, not viral TK. Always locate the failed step in the activation cascade before naming the mechanism — the MCAT loves handing you the intermediate that pinpoints it.

Replication Integration 🎯

Key Takeaways — Part 4

• Licensing (low CDK) and firing (high CDK) require opposite CDK states, so each origin fires once per cycle; re-replication is structurally forbidden
• Scorecard: circular/one origin/fast (prokaryote) vs linear/many origins/slow with nucleosomes, telomeres, S-phase confinement (eukaryote)
• Chain terminators = nucleoside analogs without a usable 3'-OH: acyclovir (viral TK activates — prodrug selectivity), AZT (HIV reverse transcriptase), and Sanger dideoxynucleotides share one mechanism
• Topoisomerase poisons (fluoroquinolones for bacteria; anticancer drugs for human topoisomerases) freeze cleavage complexes into double-strand breaks — HR repairs them, and HR-deficient tumors are hypersensitive
• Antimetabolites (methotrexate, 5-FU) starve S phase of dNTPs; selectivity is purely kinetic, so fast-renewing tissues explain the side-effect triad
• Selectivity checklist for any drug passage: unique enzyme (viral TK, RT), unique target (gyrase), or unique kinetics (division rate)
• Integration habits: locate the failed step in an activation cascade from the accumulating intermediate; identify a compound's direct target by which process shuts down FIRST

• Low-yield extras: ATR (stalled forks) and ATM (double-strand breaks) are the checkpoint sensor kinases; fork speeds are roughly 1000 nt/s in bacteria vs 50 nt/s in eukaryotes; eukaryotic polymerases are alpha-primase (start), delta (lagging), epsilon (leading); etoposide (topo II), camptothecin analogs (topo I) and cisplatin (guanine crosslinks) are the named chemotherapy examples; foscarnet blocks the viral polymerase directly, so TK-negative herpes stays sensitive to it