Biochemistry

In the small hours the lab's red indicator light blinks every thirty seconds; the graduate student watches the PCR machine as the target fragment in that 0.2 mL tube is amplified a billionfold over thirty cycles. A thought strikes him: this machine can run without stopping only because someone once fished a bacterium out of a hot spring in Yellowstone, and that bacterium's DNA polymerase can survive at ninety-five degrees Celsius. Every test point in this chapter that looks scattered actually grows out of the very same chemical fact — the rule that extension can only proceed from a 3'-OH.

Start by getting Chargaff straight. In double-stranded DNA, A equals T and G equals C simply because A embraces T with two hydrogen bonds and G embraces C with three — that is merely a byproduct of pairing, not a separate rule. So when a question gives you T at thirty-one percent and asks for C and G, the method is like dividing candy: pair A with T first to use up sixty-two, and split the remaining thirty-eight evenly between G and C, so C and G are both nineteen — never impulsively write C as thirty-one. The conformation questions are also just an ID-number game: B-DNA is right-handed, ten bases per turn, with a base-pair rise of three point four angstroms; writing it as three point six angstroms or as left-handed is always wrong. The left-handed one is Z-DNA, which likes to haunt alternating GC sequences in transcriptionally active regions, while A-DNA is the conformation seen under dehydration or in RNA-DNA hybrids.

Next comes the true star of the show: the 3'-OH rule. DNA polymerase can only add a new nucleotide onto an existing 3'-OH, always in the 5' to 3' direction. That sounds like a plain chemical fact, but it explains several things at once. First, why replication needs a primer: the polymerase cannot initiate from scratch, so something must first provide a short tail carrying a 3'-OH for it to extend, and that is why primase synthesizes a short RNA primer as a starting scaffold — the primer in DNA replication is therefore RNA, not DNA. Second, why AZT and dideoxynucleotide triphosphates are chain terminators: their 3' position carries an azido group or a bare hydrogen instead of an OH, so once incorporated, nothing further can be attached. Third, why Sanger sequencing works at all: it exploits dideoxynucleotides to stop the reaction randomly and then reads out fragment length. Fourth, why the leading strand is continuous while the lagging strand is not: the two strands run in opposite directions, but the polymerase only works 5' to 3', so the lagging strand can only be built in the reverse direction as a series of Okazaki fragments. Fifth, proofreading activity and primer-removal activity must be kept separate: proofreading is a 3' to 5' exonuclease, while primer removal is a 5' to 3' exonuclease, each running in the opposite direction to do its own job. Taq lacks proofreading activity and therefore has low fidelity, but that is a side effect, not the main reason PCR cannot use E. coli polymerase — that main reason is always heat resistance.

The gap left behind once Okazaki fragments are finished is sealed by DNA ligase; the exam often swaps this role for polymerase as a trap, but sealing the gap is always ligase's job. Do not misremember the target of AZT either: it locks onto HIV reverse transcriptase, not the host's DNA polymerase alpha, not protease, and not RNase H, because reverse transcriptase's affinity for AZT triphosphate far exceeds that of host enzymes, allowing it to selectively lock down the virus while sparing the host.

Repair systems are not something to memorize by name; what matters is recognizing what the damage looks like. When a single base is damaged — deamination, oxidation, the appearance of uracil — the job calls for base excision repair, in which DNA glycosylase excises the damaged base and hands it off for further processing. A UV-induced pyrimidine dimer is a large distorting lesion that base excision repair cannot handle, so nucleotide excision repair takes over instead, which is why patients with xeroderma pigmentosum have a nucleotide excision repair defect. When a newly replicated strand incorporates the wrong base, that is called a mismatch, and it calls for mismatch repair, in which MutS recognizes the mismatch, MutL mediates, and MutH nicks the unmethylated new strand, since the old strand is already methylated and can serve as the reference; Lynch syndrome is a mismatch repair defect. Remember that DNA glycosylase belongs to base excision repair alone — the exam loves to plant it inside mismatch repair as a trap. The SOS response is the emergency measure triggered by massive damage, and its logic resembles a coup d'état. Under normal conditions the LexA repressor suppresses the repair genes, including uvrA, uvrB, recA, and others; damage exposes large stretches of single-stranded DNA, RecA is activated into a co-protease, and it promotes the autocleavage of LexA — once the repressor collapses, all of these repair genes are derepressed and begin transcription. It is the repressor LexA that gets cleaved, not any repair-gene product, and not RecA itself; RecA is the catalyst, not the one broken down. Remember one line: LexA falls, repair rises.

Why PCR absolutely requires Taq is, in fact, simple. The three-step cycle denatures at ninety-five degrees, anneals at fifty to sixty-five degrees, and extends at seventy-two degrees, and every round must return to that ninety-five-degree station; E. coli polymerase is irreversibly inactivated at that temperature and would have to be replenished every cycle, making the reaction unworkable — which is exactly why the heat-tolerant Taq, fished out of a hot spring, is needed. Taq lacks 3' to 5' exonuclease activity and so has relatively low fidelity, but that is a side effect, not the reason other enzymes cannot be used; the main reason is, in two words, heat resistance. Primer specificity also deserves careful thought: a single primer pair recognizes only one uniquely complementary sequence on the template and therefore amplifies only one segment; testing multiple sites at once requires multiplex PCR.

Distinguishing the two types of library is a frequently tested comparison. A genomic library is made by fragmenting the entire genome and inserting the pieces into vectors, so it contains introns and requires no reverse transcriptase; a complementary DNA library is made by reverse-transcribing mature mRNA into cDNA and then cloning it, so it contains no introns and absolutely requires reverse transcriptase — which is also why a cDNA library must be used to express a eukaryotic gene in prokaryotic cells, since prokaryotes lack splicing machinery and the introns must be removed beforehand. RFLP is an entirely different matter: it compares fragment lengths after restriction-enzyme digestion for paternity testing, linkage analysis, and DNA fingerprinting, and has nothing to do with building a cDNA library — the exam loves to slip it in as a decoy.

Up the ladder of vectors, capacity keeps growing: a plasmid holds about ten kilobases, phage λ about fifteen to twenty, a cosmid about forty-five, a BAC about three hundred, and a YAC can reach anywhere from a hundred kilobases up to several megabases — the YAC is the one that holds the largest insert, because it carries a eukaryotic origin of replication, a telomere, and a centromere, letting it persist stably the way a eukaryotic chromosome does. Remember that its advantage is sheer capacity, not the best transformation efficiency. A Type II restriction enzyme recognizes a palindromic sequence, where the 5' to 3' reading of the top strand matches the 5' to 3' reading of the complementary strand; if the sequence read across the complementary strand is asymmetric, it is not that enzyme's target. The standard method for plasmid transformation is preparing competent cells with calcium chloride plus a forty-two-degree heat shock; another route is electroporation, not low-voltage electrophoresis. Site-directed mutagenesis can be accomplished simply with a mutation-carrying primer plus a polymerase, requiring no reverse transcriptase — this direction is also a favorite on the exam. Finally, for the three blotting techniques, just remember what molecule is being detected: Southern detects DNA, Northern detects RNA, both by nucleic-acid probe hybridization, while Western detects protein using antibodies; the mnemonic is "snow," and only Western uses antibodies. One rule about the 3'-OH, combined with one rule that the type of damage determines the repair system, brings the whole chapter's test points to life.

A lump of lactose is placed in a Petri dish of E. coli. Half an hour ago these bacteria showed no response to lactose at all; now they have already begun synthesizing large amounts of beta-galactosidase to break it down. How did they decide to start working? The answer lies in the first gene switch ever deciphered, the lactose operon. Starting from this switch, we will look at two prokaryotic operons, promoter strength in eukaryotes, the epigenetic tug-of-war of methylation closing and acetylation opening, the division of labor among coenzymes, and the packaging of the chromosome itself. The whole chapter really has just one theme: how a gene decides whether it should be read out right now, and how metabolism decides whether electrons should go toward burning energy or building proteins.

The lactose operon is inducible. Normally, the repressor encoded by lacI sits on the operator, holding transcription down; once lactose appears, its metabolite allolactose is the true inducer, pulling the repressor off so the operon opens. But that is only negative regulation — the second layer is positive regulation: when glucose is low, cyclic AMP rises, and CAP, bound to cyclic AMP, strongly promotes transcription from just upstream of the promoter. So bacteria use glucose preferentially, and the lactose operon fires at full capacity, transcribing the three genes lacZYA, only when glucose is scarce and lactose is present. Remember that lacZ encodes beta-galactosidase, which hydrolyzes lactose; lacY encodes permease, which lets lactose into the cell; lacA encodes transacetylase; and lacI encodes the repressor — do not swap Z and Y.

The tryptophan operon works by completely opposite logic — it is repressible, staying on by default to make tryptophan and shutting off only once tryptophan is sufficient. Its most elegant feature is attenuation: the leader sequence carries two adjacent tryptophan codons that act as a tryptophan sensor. When tryptophan is abundant, the ribosome moves so quickly that it fails to protect certain regions of the messenger RNA in time, so the mRNA folds into a terminator hairpin, RNA polymerase falls off, and transcription stops; when tryptophan is scarce, the ribosome stalls at the tryptophan codons waiting for the amino acid, the protected region shifts, and the messenger RNA instead folds into an antiterminator, so transcription continues. So the direction is that high tryptophan means termination and low tryptophan means continuation — this must never be reversed in memory — and the core mechanism is a switch in messenger RNA secondary structure, not the leader peptide interacting directly with RNA polymerase.

In the eukaryotic world, housekeeping genes are not expressed equally across the board — their expression levels differ enormously, and the main reason is the promoter's affinity for RNA polymerase. This direction is often wrongly answered as degradation rate or an inducing factor; remember that the root cause lies in promoter strength. The enhancer is a separate axis: it can sit far from the core promoter, lie upstream or downstream of the gene, or even sit inside an intron. What binds the enhancer directly is the activator, that is, a transcriptional activating factor, which then uses DNA looping to pull the basal transcription machinery onto the promoter and drive the initiation of RNA polymerase II. The coactivator does not bind DNA directly — it is merely a bridge — while the TATA-binding protein binds the TATA box and assembles with the other general transcription factors at the core promoter. The exam loves to swap these roles around; remember that it is the activator that binds the enhancer.

The two mechanisms of epigenetics really share a single core idea: the strength of electrostatic attraction determines how tightly DNA is packed. DNA carries a negative charge, and the lysines on histone tails carry a positive charge, so the two naturally attract each other; HAT adds an acetyl group to lysine, neutralizing the positive charge, loosening chromatin, and activating transcription; HDAC does the reverse, removing the acetyl group, restoring the positive charge, condensing chromatin, and repressing transcription. DNMT adds a methyl group to the cytosine of a CpG dinucleotide, recruiting repressor proteins that compact the chromatin, and the resulting transcriptional repression is called gene silencing. So, summed up in one line: methylation closes, acetylation opens, and HDAC shuts it back down again. CpG is the main methylation hotspot — not TATA, not the CAAT box, and certainly not the telomere — do not answer this direction incorrectly.

Of the four motifs in DNA-binding proteins, the one most often confused is which one mediates dimerization. The correct answer is the leucine zipper, because it has a Leu every seven residues, and its hydrophobic residues interlock like a zipper to form a dimer, with an adjacent basic domain then binding DNA. The zinc finger is used mainly to bind DNA, not to form dimers; the homeodomain is a helical structure that binds DNA directly and is encoded by Hox and homeotic genes; the beta-barrel is a membrane channel and does not count as a DNA-binding motif at all. Homeotic genes are expressed late in embryonic development and determine what structure each body segment develops into, things like antennae, wings, or legs; mutations scramble the identity of a segment, as in a fly growing legs on its head. They are not genes for specific organs, such as eye color or wing color — this direction is frequently confused.

NAD and NADP form another main thread within this same chapter. NAD is nicotinamide adenine dinucleotide: one side is adenine plus ribose plus phosphate, the AMP piece, and the other side is nicotinamide plus ribose plus phosphate, joined in the middle by a pyrophosphate bridge. The hydrogen-accepting site is C4 of the nicotinamide ring, which takes up one hydride ion — that is, two electrons plus one H-plus — at a time, so the reduced form is called NADH, not NADH-two; FADH-two is the one that actually takes up two hydrogens. NADP and NAD are almost identical in structure, and the only difference is one extra phosphate on the two-prime position of the ribose at the adenine end. It is precisely this extra phosphate that serves as a tag, letting different enzymes recognize and sort it, so NAD is routed toward catabolism to burn energy, while NADP is routed toward reductive biosynthesis and antioxidant defense, to build proteins or provide antioxidant protection. Fatty acid and cholesterol synthesis, glutathione regeneration, and cytochrome P450 reactions all use NADPH, while glycolysis, the citric acid cycle, and beta-oxidation all produce NADH, which is sent into the electron transport chain. The vitamin sources are also a favorite reversal on the exam: NAD and NADP come from vitamin B3 niacin, while FAD and FMN come from vitamin B2 riboflavin — do not get this direction wrong. The trick for structure-recognition questions is to look at the ring: two nucleotides plus a pyrophosphate bridge, with one end a six-membered nitrogen-containing amide ring, means NAD; if the ribose at the adenine end carries one extra phosphate, it is NADP; and if you see that three-ring isoalloxazine system, that is FAD or FMN.

Finally, do not forget the packaging of the chromosome itself. The phosphate backbone of DNA carries a negative charge, and packing it into the nucleus requires positively charged proteins to wind it up. Histones are rich in the two basic amino acids lysine and arginine, so they carry a positive charge and can bind DNA electrostatically — this is neither covalent nor hydrophobic. The packaging hierarchy has DNA wound around eight histones to form beads-on-a-string nucleosomes, which H1 then helps fold into the thirty-nanometer fiber, which forms higher-order chromosome structure. While we're at it, let's demolish a frequently tested numerical trap: the human genome is about three billion base pairs, but the protein-coding exons make up only about one and a half to two percent of it, not forty. The vast remainder consists of introns, regulatory sequences, repetitive sequences, and transposon remnants, which is also why the story of gene regulation always circles back to seemingly minor modifications like which stretch is methylated and which stretch is acetylated. Holding onto the two main threads of this chapter, switches and batteries, connects prokaryote to eukaryote, NAD to NADP, and electrostatics to packaging into a single chain of causation.

A Black boy suddenly clutches his chest in agony and breaks into a cold sweat after strenuous exercise; under the microscope his red cells are twisted into crescents — this is deoxygenated HbS polymerizing and dragging the cells into a sickle shape. A single-base point mutation, codon six of beta-globin changing from GAG to GTG, glutamate replaced by valine, rewrites an entire lifetime of disease. If the same blood smear also shows target cells and microcytic, hypochromic cells, that points instead to thalassemia, which is not a wrong substitution but simply too little of one chain being made. One is a qualitative change, the other a quantitative change, and that contrast is a microcosm of this whole chapter, because this act has to string structural information and energetic matter into a single chain.

Let's keep the three types of recombination straight: homologous recombination requires a long stretch of homologous sequence and is used in meiotic crossing-over and the precise repair of double-strand breaks; site-specific recombination requires no homology and relies on enzymes that recognize specific signal sequences — V(D)J is exactly this type, with RAG1 and RAG2 recognizing the RSS and cutting and rejoining the V, D, and J segments of antibodies and T-cell receptors to generate immune diversity; transposition and viral integration likewise require no homology, and when a retrovirus's integrase inserts into the host, both LTRs are retained, so the LTR appears duplicated on either side of the insertion site while gag, pol, and env each occur only once. There is one more directional detail about antibody light chains: V to J happens at the DNA level, J to C happens by RNA splicing, and the exam loves to claim that J to C is also DNA recombination, which is wrong. The core contrast in the hemoglobinopathies is qualitative versus quantitative change: sickle cell disease comes from codon six of beta-globin changing from GAG to GTG, glutamate replaced by valine, one amino acid substituted wrong, and deoxygenated HbS polymerizes into fibers that pull the cell into a sickle shape, with patients able to survive into adulthood; beta-thalassemia comes from reduced or absent beta-chain synthesis, usually from splicing or promoter mutations, with the gamma chain compensating so fetal hemoglobin rises; alpha-thalassemia comes from deletion of the alpha-chain genes, with severe forms producing Hb Bart's or HbH. The telomere sequence is repeating TTAGGG, with a G-rich single-stranded overhang at the three-prime end; this strand forms a G-quartet among four guanines through Hoogsteen hydrogen bonds, and multiple stacked layers form the four-stranded G-quadruplex, not three-stranded and not five-stranded. Every round of replication shortens the telomere, serving as an aging clock; telomerase can re-lengthen it and shows high activity in stem cells, germ cells, and cancer cells, which is the molecular basis of cancer cell immortality. The cycle runs from G1 into S, then G2, then M, and terminally differentiated neurons and cardiomyocytes exit the cycle into G0, which is why damage to them is so hard to regenerate.

Nucleotides have two routes: making them from scratch, and salvage recovery. Salvage saves energy by attaching a ready-made purine onto PRPP to regenerate a nucleotide; three enzymes each handle a different base, and losing any one produces a distinct disease. HGPRT salvages hypoxanthine and guanine, and its deficiency is Lesch-Nyhan syndrome: once the salvage route is broken, purines are all pushed toward degradation and oxidized by xanthine oxidase into a large amount of uric acid, producing hyperuricemia, intellectual disability, and self-mutilating hand-biting clinically. One directional trap deserves special mention: de novo synthesis is intact and even accelerates because PRPP accumulates and feedback from IMP and GMP weakens, so the core issue is uric acid excess, not purine deficiency. APRT deficiency lets adenine be oxidized into insoluble 2,8-dihydroxyadenine, causing kidney stones; overactive PRPP synthetase lets PRPP accumulate, accelerating de novo synthesis and raising uric acid to produce gout. Xanthine has no corresponding phosphoribosyltransferase, so it cannot be salvaged and can only be oxidized into uric acid; for treatment, allopurinol and febuxostat inhibit xanthine oxidase to reduce uric acid production, while an acute flare is treated with NSAIDs and colchicine or steroids, never started with a urate-lowering drug, or it will provoke an attack. On the pyrimidine side, orotic aciduria lacks UMP synthase, so orotic acid accumulates and UMP, UTP, and CTP become insufficient, producing megaloblastic anemia that does not improve with folate or B12; the treatment is to supplement uridine, because it is already a nucleoside and, downstream of the defect, can be phosphorylated into UMP to bypass it while simultaneously providing feedback to suppress orotic acid — supplementing thymine, adenosine, or allopurinol is all ineffective. dUMP becomes dTMP through thymidylate synthase using 5,10-methylene-THF as the methyl donor, which is why thymine has one extra methyl group at the 5-position compared with uracil; 5-FU inhibits thymidylate synthase, and methotrexate inhibits DHFR, cutting off THF — both prevent dTMP from being made. Among the deamination reactions, 5-methylcytosine deaminating into thymine is the most dangerous, because the product is a normal base that the repair system cannot tell is wrong, which is why 5-methylcytosine is a mutational hotspot. Purine breakdown gets stuck at poorly soluble uric acid, so trouble follows; pyrimidine breakdown ends in the water-soluble beta-alanine and beta-aminoisobutyric acid, so it stays clean.

Energy metabolism is a power plant. Glucose is broken down glycolytically in the cytosol into pyruvate, which enters the mitochondrion where pyruvate dehydrogenase converts it into acetyl coenzyme A, which enters the citric acid cycle and is burned into carbon dioxide, producing NADH and FADH₂; the electrons are sent into the electron transport chain, which pumps hydrogen ions across the membrane, and finally ATP synthase harvests it as ATP. The pyruvate dehydrogenase complex operates only under aerobic conditions in the matrix; its three enzymes and five coenzymes are TPP, lipoic acid, coenzyme A, FAD, and NAD — biotin belongs to the carboxylases and coenzyme Q belongs to the electron transport chain, and neither is a pyruvate dehydrogenase coenzyme. B1 deficiency disables E1, and pyruvate and lactate accumulate, producing Wernicke encephalopathy and beriberi. One turn of the citric acid cycle yields two carbon dioxides, three NADH, one FADH₂, and one GTP, with net OAA consumption at zero, just like a runway that gets handed back once used; the rate-limiting enzyme is isocitrate dehydrogenase. Substrate-level phosphorylation occurs in both the cytosol and the mitochondrion, while oxidative phosphorylation occurs only at the inner mitochondrial membrane. The electron pathway has NADH entering Complex I and FADH₂ entering Complex II, passing through CoQ, Complex III, cytochrome c, and Complex IV, which hands the electrons to oxygen to form water; the ones that pump hydrogen are Complex I, Complex III, and Complex IV, from the matrix toward the intermembrane space; current ATP yield is about two point five per NADH and about one point five per FADH₂. Inhibitors stop the electron transport chain, lowering both oxygen consumption and ATP: rotenone blocks Complex I, antimycin A blocks Complex III, cyanide and carbon monoxide and hydrogen sulfide and azide block Complex IV, and oligomycin blocks Complex V; uncouplers let hydrogen ions leak back, so oxygen consumption actually rises while ATP falls and heat production rises — 2,4-dinitrophenol, aspirin overdose, and UCP1 are all examples, and the direction is opposite, so do not reverse it in memory. Antioxidant defense has superoxide dismutase turning superoxide into hydrogen peroxide, glutathione peroxidase turning hydrogen peroxide into water, and the reductase using NADPH to turn oxidized glutathione back into the reduced form to recharge it; it is glutathione peroxidase, not superoxide dismutase, that clears hydrogen peroxide. Acetyl coenzyme A carries a high-energy thioester bond. Finally come the scripts for the two states: in the fed state insulin dominates storage, and in the fasted state glucagon dominates mobilization; a low-carbohydrate, high-protein diet raises beta-oxidation and raises ketone bodies, producing a high-anion-gap metabolic acidosis, not alkalosis, along with increased urea and decreased fat. Beta-hydroxybutyrate is the most abundant ketone, but the nitroprusside strip cannot detect it, so urine ketones in early diabetic ketoacidosis may be falsely negative; the liver can make ketone bodies but cannot use them, because it lacks thiophorase, so ketone bodies serve the brain, heart, and muscle. Hypoxia-inducible factor 1, once stabilized under hypoxia, upregulates PDK1, which inhibits pyruvate dehydrogenase so that pyruvate no longer enters the citric acid cycle and instead goes through glycolysis to produce lactate; oxidative phosphorylation falls and reactive oxygen species decrease rather than staying unchanged, which is one of the molecular bases of the Warburg effect. Glycosylation has two routes: N-linked glycosylation attaches to the amide nitrogen of asparagine, with the consensus sequence Asn plus X plus Ser or Thr, where X cannot be Pro, initiated by N-acetylglucosamine in the endoplasmic reticulum; O-linked glycosylation attaches to the hydroxyl of serine or threonine, initiated by N-acetylgalactosamine in the Golgi apparatus; cysteine is not a standard glycosylation site. From a single point mutation to an entire power plant, every test point in this chapter grows on the same chain of causation.

The entire exam logic of cancer genetics mostly falls into place once you fix the car metaphor in your mind. A normal gene that promotes growth is called a proto-oncogene; once a mutation makes it "too good at pressing," it becomes an oncogene, and at that point one damaged allele is enough, so at the cellular level it behaves as dominant, like the gas pedal stuck to the floor. Brakes work the opposite way: both brake pads must fail before the car can roll, so tumor suppressors are recessive at the cellular level and require two hits. The proofreader is the DNA repair gene; once the proofreader leaves, every part in the car starts failing faster, so it follows the mutator path, and most repair genes also require two hits.

Once you understand this causal chain, it becomes easy to see why retinoblastoma presents bilaterally in early childhood: the child inherits one damaged RB allele from a parent, meaning the first hit is present from birth, and any retinal cell later acquiring a second hit — often through loss of heterozygosity — will grow into a tumor; sporadic cases must wait for both hits to strike the same cell after birth, which is naturally later and unilateral. RB is precisely the prototype of the two-hit hypothesis. But the rule that all tumor suppressors need two hits is a generalization, not an absolute — TP53 is the big pitfall here. Its mutant protein drags the normal p53 down into an immobile tetramer, which is called dominant-negative, so Li-Fraumeni syndrome shows a clear increase in cancer with just a single germline mutation. This looks like it breaks the rule, but really the damaged copy drags the good copy down with it. That is also why, when the exam asks for "the most commonly mutated tumor suppressor," the answer is TP53, not RB.

Lynch and FAP represent the causal divide between two faces of colon cancer. The mismatch repair genes MLH1, MSH2, MSH6, and PMS2 are responsible for fixing pairing errors; once damaged in the germline, every division leaves errors with no one to proofread them, and slippage errors appear in microsatellite regions — this is high microsatellite instability. Mutations accumulate rapidly, tumors emerge quickly, and onset is often before age 50; and because the errors accumulate drop by drop rather than as a field of polyps, Lynch tumors mostly grow one at a time, favor the right colon, and in women often coexist with endometrial cancer. FAP, by contrast, comes from a disabled APC tumor suppressor in the germline, leaving the Wnt pathway unchecked, so the colon becomes covered with hundreds to thousands of adenomatous polyps. The exam loves to mix Lynch and FAP together in one question; remember "Lynch = one, FAP = a field" and you won't go wrong. Lynch also connects to a newer test point in recent years: tumors with high microsatellite instability carry large numbers of neoantigens that the immune system ought to see, but the tumor counters using PD-L1, so an anti-PD-1 agent like pembrolizumab is especially effective once it releases that brake. Driver genes and their targets do not need to be memorized as a lookup table — they come alive once you go back to mechanism. KRAS is a GTPase locked in the "on" position, so an EGFR inhibitor hitting upstream is useless, and cetuximab being ineffective in KRAS-mutant colon cancer is a high-stakes question. HER2 is an amplified receptor tyrosine kinase, and trastuzumab binds directly onto its extracellular portion. BCR-ABL is the persistently active kinase produced by the Philadelphia chromosome, the hallmark of chronic myeloid leukemia, and imatinib is its nemesis. In Burkitt lymphoma, MYC is relocated under the immunoglobulin promoter by t(8;14) and never stops working. BRCA1 and BRCA2 are repair genes for homologous recombination; in a deficient tumor, adding a PARP inhibitor cuts off single-strand repair as well — this is synthetic lethality, killing the tumor while leaving normal cells unharmed.

The whole roster of B-vitamin coenzymes likewise comes back to the reaction itself. The active form of B1 is TPP, the coenzyme for pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase. In a patient with alcohol use disorder, thiamine has long since been depleted and pyruvate dehydrogenase can no longer run; if glucose is then infused aggressively, the sugar needs TPP to enter the citric acid cycle but none is available, and what little thiamine remains gets burned up too, so Wernicke encephalopathy explodes instantly from subclinical to full-blown. So giving thiamine before glucose is not superstition — it is reaction kinetics. The other coenzyme pairings likewise come back to the reaction: transamination needs PLP; B6 is also the coenzyme for ALA synthase, so its deficiency causes sideroblastic anemia, and INH, which depletes B6, must be co-administered with it; one-carbon transfer needs folate; methylation needs B12; carboxylation needs biotin, and avidin in raw egg white binds biotin and causes deficiency; coenzyme A comes from pantothenic acid, B5, not folate — a frequent trap. The three D's of pellagra from niacin deficiency are dermatitis, diarrhea, and dementia, and because niacin can also be synthesized from tryptophan, which requires B6, Hartnup disease, carcinoid syndrome, and INH can all produce pellagra-like presentations. Vitamin C is not merely an antioxidant — it is the cofactor for prolyl and lysyl hydroxylase, and without it collagen cannot stabilize its triple helix, which is scurvy; it also reduces intestinal ferric iron to ferrous iron to aid absorption.

The fat-soluble vitamins A, D, E, and K follow a different route. A and D are fat-soluble and can enter the nucleus, acting through the slow, long-lasting route of entering the nucleus as a transcription factor; A binds RAR and RXR, D binds VDR. Vitamin D's two-step activation is the exam's favorite directional question: it starts with cholesterol, the skin makes D3 via ultraviolet light, the liver uses CYP2R1 to make 25-OH-D3, the clinical monitoring marker calcidiol, and the kidney then uses CYP27B1 to make 1,25-dihydroxyvitamin D, the active form calcitriol. The first step is in the liver, the second in the kidney — memorize that order cold. Finally comes B12 and pernicious anemia. Pernicious anemia comes from autoimmune destruction of gastric parietal cells, intrinsic factor deficiency, and failure to absorb B12 in the ileum. Both consequences of B12 deficiency grow out of its mechanism: methionine synthase hands the methyl group of N5-methyl-THF to homocysteine, and this requires B12; without that key, the methyl group stays trapped on THF and cannot get out, so DNA synthesis has no raw material, producing megaloblastic anemia; at the same time, methylmalonyl-CoA mutase fails, methylmalonic acid accumulates, and myelin is disrupted, producing the neuropathy of subacute combined degeneration. Folate deficiency affects only the first pathway, with no rise in methylmalonic acid and no neurological symptoms, so the lab distinguishes them by methylmalonic acid: B12 deficiency raises both methylmalonic acid and homocysteine, while folate deficiency raises only homocysteine. Why can't you treat with folate alone? Because folate can bypass the methyl trap and correct the appearance of anemia, but the methylmalonic acid pathway depends solely on B12, so the nerve damage keeps progressing — this is a high-stakes question. The metal in B12 is cobalt, not iron or magnesium, and the zinc finger coordinates zinc through cysteine and histidine, not leucine — don't lose points on these details either.

Handling amino groups is a three-stage project of collection, release, and packaging. Transamination collects the amino groups scattered across individual amino acids onto glutamate, requiring PLP — that is, vitamin B6 — as its coenzyme, which is exactly why ALT and AST need it. Oxidative deamination releases the amino group on glutamate as free NH₄⁺. The urea cycle then packages this free ammonia into urea for excretion. Free ammonia is itself toxic, so the blood mainly transports it in two nontoxic forms: the whole body and the brain rely on glutamine, and muscle relies on alanine. The brain's ammonia detoxification depends heavily on glutamine synthetase, so hyperammonemia depletes α-ketoglutarate and glutamate, driving glutamine up and the brain to swell — this is the core of hepatic encephalopathy.

The six steps of the urea cycle are memorized by location. The first two steps occur in the mitochondrion: CPS-I is the rate-limiting enzyme, requiring allosteric activation by N-acetylglutamate — a high-protein diet raises N-acetylglutamate and speeds ammonia disposal; it combines NH₄⁺ with carbon dioxide to make carbamoyl phosphate, the first nitrogen source; OTC then combines this with ornithine to make citrulline, and citrulline exits the mitochondrion. The last three steps occur in the cytosol: ASS combines citrulline with aspartate to make argininosuccinate, the second nitrogen source; ASL cleaves off arginine and fumarate, and fumarate rejoins the TCA cycle — not oxaloacetate, a high-frequency trap; arginase finally cleaves off urea and ornithine, and ornithine is recycled into the mitochondrion. A deficiency of any one enzyme causes hyperammonemia; OTC deficiency is the most common X-linked defect, characterized by hyperammonemia plus orotic aciduria, because carbamoyl phosphate spills over into pyrimidine synthesis.

One-carbon metabolism and the methyl cycle form another main thread. Folate carries one-carbon units for purine and dTMP synthesis; N5-methyl-THF is the only form in which it can hand off a methyl group, and B12 takes this key and passes the methyl group to homocysteine to make methionine. Without B12, the methyl group stays stuck on THF and cannot get off — this is the methyl trap, producing a functional folate deficiency and megaloblastic anemia. At the same time, methylmalonyl-CoA mutase also fails, methylmalonic acid accumulates, and myelin is damaged — this is why B12 deficiency harms the nerves while folate deficiency does not. Homocystinuria is a deficiency of cystathionine β-synthase, which requires B6; homocysteine cannot be cleared and accumulates, causing thrombosis, intellectual disability, and downward lens dislocation — the exact opposite of Marfan syndrome's upward dislocation — and some patients respond to high-dose B6. Phenylketonuria is a deficiency of phenylalanine hydroxylase or its cofactor tetrahydrobiopterin; phenylalanine cannot be converted and tyrosine cannot be replenished, so tyrosine becomes essential, and downstream precursors for melanin, catecholamines, and thyroid hormone are all affected — management requires newborn screening, a low-phenylalanine diet with tyrosine supplementation, and avoiding aspartame; the tetrahydrobiopterin variant also disrupts neurotransmitters, so restricting phenylalanine alone is not enough. Alkaptonuria is a deficiency of homogentisate oxidase, with urine that darkens on standing; albinism is a deficiency of tyrosinase, a completely different pathway from phenylketonuria; maple syrup urine disease is a deficiency of branched-chain α-ketoacid dehydrogenase, which requires B1. Glutathione is the tripeptide γ-glutamyl-cysteinyl-glycine and carries a γ bond; creatine is synthesized from glycine, arginine, and methionine, with SAM supplying the methyl group; osteogenesis imperfecta is a type I collagen mutation, usually a glycine substitution.

The two axes of lipid metabolism come down to one sentence: synthesis occurs in the cytosol, oxidation in the mitochondrion, and malonyl-CoA stands at the fork, simultaneously supplying synthesis while shutting the gate to β-oxidation, carnitine palmitoyltransferase I. Acetyl-CoA carboxylase combines acetyl-CoA with carbon dioxide to make malonyl-CoA, requiring biotin and ATP — this is the rate-limiting step of fatty acid synthesis; the rate-limiting step of cholesterol synthesis is instead HMG-CoA reductase, the target of statins. Note that aspirin does not affect cholesterol synthesis; it acts on COX. Acetyl-CoA carboxylase is regulated on two tracks: allosteric activation by citrate and allosteric inhibition by long-chain acyl-CoA; on the covalent side, AMPK phosphorylates and inactivates it when energy is low, and insulin dephosphorylates and switches it on — so during fasting or exercise, AMPK comes online, acetyl-CoA carboxylase shuts off, fat synthesis stops, carnitine palmitoyltransferase I opens wide, and β-oxidation burns freely. The final round of β-oxidation of an odd-chain fatty acid produces propionyl-CoA, which is carboxylated by biotin into methylmalonyl-CoA and then converted by the B12-dependent mutase into succinyl-CoA, entering the TCA cycle — one of the few pathways by which a fatty acid can become glucogenic, and also the source of methylmalonic acid accumulation in B12 deficiency. The eicosanoid substrate must be arachidonic acid, C20:4 ω-6; the saturated palmitate, stearate, and acetyl-CoA are never COX substrates. COX produces prostaglandins and thromboxane A2, which promotes platelet aggregation, and prostacyclin, which inhibits it; 5-lipoxygenase produces the leukotrienes responsible for inflammation and bronchoconstriction. Aspirin irreversibly acetylates COX. Among membrane lipids, cardiolipin is almost exclusively confined to the inner mitochondrial membrane, not the plasma membrane; integral membrane proteins require a detergent for extraction; platelet-activating factor has a C1 alkyl-ether, and plasmalogen has a C1 vinyl-ether — do not mix up the two. High-density lipoprotein carries out reverse transport, esterified by LCAT.

The signaling axis first splits into lipid-soluble and water-soluble. Lipid-soluble hormones enter the cell, with receptors inside acting as transcription factors, slow and long-lasting; water-soluble hormones stay outside the membrane, relying on second messengers, fast and brief. Heat shock protein 90 binds the nuclear receptor's ligand-binding domain, not its DNA-binding domain. Among G-protein-coupled receptors, Gs raises cAMP, Gi lowers cAMP, and Gq proceeds through IP3 plus DAG. Epinephrine's glycogenolysis runs through the β receptor plus Gs plus adenylate cyclase plus cAMP plus protein kinase A plus glycogen phosphorylase kinase plus glycogen phosphorylase; the β receptor does not directly activate Ras. Cholera toxin ADP-ribosylates Gsα, inhibiting its GTPase and locking it on; pertussis toxin modifies Giα so it cannot bind GTP, locking it off; both ultimately raise cAMP, though locked in opposite positions. Insulin's MAPK sequence runs IRS-1 to Grb2 plus Sos to Ras to Raf to MEK to ERK — MEK is upstream of ERK, do not reverse it; atrial natriuretic peptide acts through membrane-bound guanylate cyclase to produce cGMP; JAK-STAT handles cytokines, growth hormone, EPO, leptin, and interferon. In the cell cycle, cyclin is the regulator and CDK is the catalyst — cyclin cannot phosphorylate anything on its own.

The death cap mushroom patient's liver enzymes begin to skyrocket at forty-eight hours; α-amanitin has an extremely high affinity for RNA Pol II, and completely inhibits Pol II at vanishingly low concentrations, so mRNA synthesis grinds to a halt, protein synthesis stops, and hepatocytes die. Pol I is completely unaffected, and Pol III is only inhibited at high concentrations — this specificity is itself an exam question. The division of labor among the three eukaryotic RNA polymerases is remembered in one line as "one big, two messenger, three small": Pol I works in the nucleolus making the large rRNAs, Pol II makes pre-mRNA and most small nuclear RNA, and Pol III makes 5S rRNA, transfer RNA, and small RNAs. Mammalian nuclei have no Pol IV responsible for mRNA — that belongs to plant RNA interference; prokaryotes have only one RNA polymerase, which recognizes promoters via a σ factor.

The three major modifications of mRNA are completed entirely within the nucleus, co-transcriptionally, not in the endoplasmic reticulum or the Golgi apparatus. The 5′ cap — a 7-methylguanosine joined by a 5′-5′ triphosphate bond — is added as soon as just over twenty nucleotides have been transcribed; the 3′ end receives a poly(A) tail, not phosphorylation, with the core signal AAUAAA lying upstream of the cleavage site; splicing is carried out by the spliceosome's U1, U2, U4 plus U6, and U5. The key to the three types of splicing is who holds the blade. Group I introns use the 3′-OH of an exogenous free guanosine as the nucleophile; group II introns and the eukaryotic spliceosome both attack with the 2′-OH of the branch-point adenosine inside the intron, coiling into a lariat. Remember "external G, internal A" and you will never get it wrong. The 3′ end of every tRNA is CCA-OH, and the amino acid is esterified onto this 3′-OH by aminoacyl-tRNA synthetase, regardless of species or amino acid.

The three stages of translation use prokaryotes as the framework. In initiation, IF1 occupies the A site, IF2 carries fMet-tRNA into the P site, and IF3 prevents the large subunit from joining prematurely; in elongation, EF-Tu delivers aminoacyl-tRNA into the A site, a peptide bond forms, and EF-G carries out translocation; in termination, a release factor recognizes one of the three stop codons. Peptide bond formation is catalyzed by the peptidyl transferase of the 23S rRNA, so the ribosome is a ribozyme — RNA, not a protein enzyme, is the catalytic core. The polypeptide runs from N to C, and the mRNA from 5′ to 3′. Prokaryotic initiation relies on complementary pairing between the purine-rich Shine-Dalgarno sequence and the 3′ end of 16S rRNA to position the ribosome on the initiator AUG; eukaryotes have no SD sequence and instead start from the 5′ cap, scanning for the first AUG — the Kozak sequence. The eukaryotic initiator tRNA carries methionine, unformylated, while only prokaryotes use formylmethionine — assigning the 80S ribosome to prokaryotes is a common reversal. The small subunit binds mRNA first, and the large subunit joins only at the end; mistakenly believing that the complete ribosome assembles before mRNA binds is a frequently tested trap.

Antibiotics first split into those that inhibit the cell wall and those that inhibit protein synthesis. Ampicillin and the other β-lactams hit the transpeptidase and penicillin-binding proteins of the cell wall, unrelated to the ribosome — a frequent distractor option. The latter group further splits by ribosomal subunit: the 30S subunit is targeted by aminoglycosides and tetracycline, with aminoglycosides causing misreading and blocking initiation, and tetracycline blocking aminoacyl-tRNA from entering the A site; the 50S subunit is targeted by chloramphenicol, which inhibits peptidyl transferase, and macrolides, which block translocation. Global regulation runs through eIF2α: when iron is deficient, heme synthesis falls, free heme falls, HRI — the heme-regulated inhibitor kinase — is activated, and it phosphorylates serine 51 of eIF2α; phosphorylated eIF2 becomes a competitive inhibitor of eIF2B, eIF2-GTP cannot be recycled, the 43S initiation complex cannot assemble, and global translation — especially of hemoglobin — is shut down. This is also the common final pathway of the integrated stress response, one that PKR, PERK, and GCN2 all converge on; as long as eIF2α is phosphorylated, initiation stops. Formula for calculation questions: the number of amino acids is roughly the DNA base pairs divided by three, minus one stop codon; the molecular weight is roughly the number of amino acids multiplied by one hundred ten daltons. For example, nine hundred base pairs is three hundred codons, minus one is two hundred ninety-nine amino acids, multiplied by one hundred ten is roughly thirty-two thousand nine hundred daltons, roughly thirty-three kilodaltons. A few details of protein structure need to be clear: a glycine-proline sequence tends toward a β-turn, because proline's rigid ring bend plus glycine's minimal side chain lend flexibility — not an α-helix; the disulfide bond is the only covalent bond in tertiary structure, not a force that maintains the α-helix, since the α-helix relies purely on backbone hydrogen bonds, formed between the i-th carbon-oxygen double bond and the (i+4)-th nitrogen-hydrogen bond; codon and anticodon pairing is by hydrogen bonds, not covalent, and the third position allows wobble.

The core of enzyme kinetics lies in a handful of parameters from the Michaelis-Menten rate equation. Km is the substrate concentration at which the reaction reaches half of Vmax — a small Km means high affinity; Vmax is proportional to the total amount of enzyme; kcat is how many substrate molecules each enzyme converts per second; kcat divided by Km is the catalytic efficiency, with an upper limit of roughly the diffusion rate, ten to the eighth to ninth power. When the substrate concentration is far below Km, the reaction simplifies to v approximately equal to Vmax divided by Km, multiplied by [S], and the proportionality constant is kcat divided by Km — that is, Vmax divided by Km — not kcat alone, and not one divided by Km; this is an easy point to drop in calculation questions. The three types of reversible inhibition can be told apart at a glance with a double-reciprocal plot: the x-intercept is negative one divided by Km, and the y-intercept is one divided by Vmax. Competitive inhibition binds only the free enzyme, competing for the active site; Vmax is unchanged, Km rises, the lines share a y-intercept with a steeper slope, and a large excess of substrate can rescue Vmax. Noncompetitive inhibition binds an allosteric site, binding both the enzyme and the enzyme-substrate complex; Km is unchanged, Vmax falls, and the lines share an x-intercept. Uncompetitive inhibition binds only the enzyme-substrate complex; Km and Vmax fall together, giving parallel lines. Irreversible inhibition — such as aspirin acetylating COX or organophosphates inhibiting acetylcholinesterase — causes permanent covalent inactivation, kinetically resembling a reduction in the amount of enzyme. Protein quantification relies on absorbance at 280 nm: the absorbance comes strongest from tryptophan, next from tyrosine, weaker from phenylalanine; the non-aromatic asparagine, glycine, and alanine do not absorb; 280 nm absorbance can only quantify, never identify a protein's identity or molecular weight. Choosing the right separation technique depends on what you are separating by: SDS-PAGE confers a uniform negative charge, destroys activity, and separates by molecular weight; native gel electrophoresis preserves activity; isoelectric focusing separates by isoelectric point and measures the isoelectric point; ion exchange separates by charge; molecular sieving separates by size, with large molecules eluting first, and cannot identify identity; affinity chromatography separates by specific binding. Do not reach for isoelectric focusing to measure molecular weight — that measures the isoelectric point instead; this too is a high-frequency, fatal trap. Think through where the molecules flow across the whole chapter, work out exactly which parameter has shifted, and every question will fall out on its own.