Undergraduate BIOL102 ⭐ 3 Credits Semester 2

Genetics and Heredity

Covers classical and molecular genetics, from Mendelian inheritance to modern genomics. Topics include transmission genetics, chromosomal basis of heredity, gene expression regulation, mutation, DNA repair, and an introduction to recombinant DNA technology. Provides genetic context for understanding how protein sequences are encoded and regulated.

📅 Studied by Skipper on 2026-10-02

📖 Core Concepts

Claude Opus 5.5

Core Concepts: Genetics and Heredity

BIOL102 · Personal study notes for Skipper · Primary references: Griffiths et al., Introduction to Genetic Analysis (12th ed., 2020); Watson et al., Molecular Biology of the Gene (7th ed., 2014)

1. What This Subject Is Really About

Genetics is the study of how information persists across generations of cells and organisms. Most of biology, and most of the chemistry you have done, concerns matter and energy. Genetics adds a third quantity: a sequence that specifies something other than itself. Every topic in this course is a variation on one question: how is a pattern stored, copied, read, edited, and occasionally corrupted, using nothing but ordinary molecules obeying ordinary chemistry?

The field developed in two phases that met in the middle.

  • The abstract phase (1866 to about 1940). Mendel, Morgan, Sturtevant, and their students inferred the existence and behavior of "factors" (genes) only from the ratios of phenotypes in offspring. They had no idea what a gene was made of. They still deduced that genes are discrete, that they come in pairs, that they sit in a linear order on chromosomes, and that the distances between them can be measured. This is reasoning about a hidden system from its outputs, much as you inferred orbitals from spectra in CHEM101.
  • The molecular phase (1944 onward). Avery, MacLeod, and McCarty showed that the transforming principle is DNA [Avery, MacLeod & McCarty, 1944]. Watson and Crick proposed the double helix [Watson & Crick, 1953]. Nirenberg and Matthaei began decoding the genetic code [Nirenberg & Matthaei, 1961]. The abstract "factor" became a chemically defined polymer with a sequence you can read.

Much of the satisfaction in this course comes from watching these two descriptions agree. Mendel's 3:1 ratio is the behavior of homologous chromosomes at meiosis. A "recessive allele" is very often a broken enzyme. A "map distance" is a probability of crossover events along a physical molecule. As you study, keep asking: what is the molecular reality behind this abstract rule?

An analogy to carry with you: the genome is less like a blueprint and more like a recipe library with an annotated, self-copying index. A blueprint maps one-to-one onto a structure. A recipe is a set of instructions whose execution depends on context, such as which recipes are open, what ingredients are present, and what temperature the kitchen is. Gene regulation, dominance, and epigenetics all make more sense in the recipe framing.

The organizing scaffold for the molecular half of the course is Crick's central dogma. It is often misstated, so here it is precisely: sequence information can flow from nucleic acid to nucleic acid, and from nucleic acid to protein, but not from protein back to nucleic acid [Crick, 1970].

         replication
          ┌──────┐
          ▼      │
         DNA ────┘ ──transcription──▶ RNA ──translation──▶ PROTEIN
          ▲                            │
          └────reverse transcription───┘    (retroviruses, telomerase,
                                             retrotransposons)

  Forbidden (per Crick): PROTEIN ──▶ nucleic acid sequence

2. Mendelian Inheritance: The Logic of Discrete Factors

2.1 The problem Mendel solved

Before Mendel, the dominant intuition was blending inheritance: offspring are a mix of their parents, as paint colors mix. Blending has a fatal flaw. Variation would halve each generation and quickly vanish, yet natural populations stay variable. Mendel's crosses with garden peas [Mendel, 1866] showed that traits behave as if controlled by discrete, unchanging particles that can be hidden for a generation and then reappear intact.

2.2 The two laws

  1. Law of Segregation. Each diploid individual carries two copies (alleles) of each gene. These separate during gamete formation, so each gamete receives exactly one, with equal probability.
  2. Law of Independent Assortment. Alleles of different genes segregate independently of one another. As you will see in §4, this holds only for genes on different chromosomes or far apart on the same chromosome.

The physical basis, recognized decades later, is meiosis. Homologous chromosomes separate at anaphase I, which gives segregation. Different homologous pairs orient randomly on the metaphase I plate, which gives independent assortment [Griffiths et al., 2020, ch. 2–3].

2.3 Crosses are probability problems

Your calculus background is useful here, but the core tool is elementary probability. A Punnett square is a visual product of two probability distributions.

Monohybrid cross:   Aa × Aa
  Gametes from each parent:  P(A) = 1/2,  P(a) = 1/2

  Genotypes:   AA : Aa : aa   =  1/4 : 1/2 : 1/4     (1:2:1)
  Phenotypes (A dominant):  A_ : aa  =  3/4 : 1/4    (3:1)

Dihybrid cross:     AaBb × AaBb   (independent genes)
  P(A_B_) = (3/4)(3/4) = 9/16
  P(A_bb) = (3/4)(1/4) = 3/16
  P(aaB_) = (1/4)(3/4) = 3/16
  P(aabb) = (1/4)(1/4) = 1/16                        (9:3:3:1)

Product rule:  P(X and Y) = P(X)·P(Y)        if X, Y independent
Sum rule:      P(X or Y)  = P(X) + P(Y)      if X, Y mutually exclusive

Binomial (e.g., exactly k affected children out of n, each with prob p):
  P(k) = C(n,k) · p^k · (1−p)^(n−k),   where C(n,k) = n! / [k!(n−k)!]

Practical advice: for crosses involving three or more genes, do not draw a 64-cell Punnett square. Treat each gene as a separate monohybrid problem and multiply. The 9:3:3:1 ratio is simply (3:1)×(3:1) expanded, the same algebra as (a+b)² in a binomial expansion.

2.4 Dominance is a property of phenotype, not of alleles

This is the most important conceptual refinement of Mendel. "Dominance" describes how the heterozygote's phenotype compares with the two homozygotes. It is not an intrinsic power of an allele.

  • Complete dominance: the heterozygote looks like one homozygote. The molecular explanation is usually haplosufficiency: one functional copy produces enough enzyme to reach the phenotypic threshold. Mendel's round versus wrinkled peas are the classic case. Wrinkled seeds lack a functional starch-branching enzyme (SBE1). They accumulate sucrose, take up water osmotically, and then wrinkle on drying. One working copy is enough to make round seeds.
  • Incomplete dominance: the heterozygote is intermediate. Red × white snapdragons give pink offspring because pigment amount scales with gene dosage. The F₂ phenotypic ratio is 1:2:1, so phenotype maps directly onto genotype.
  • Codominance: the heterozygote displays both parental phenotypes fully. In the human ABO blood group, the IA and IB alleles encode glycosyltransferase variants that add different sugars (N-acetylgalactosamine versus galactose) to the H antigen. An IAIB person has red cells carrying both antigens. The i allele encodes a nonfunctional enzyme and is recessive to both.

Key insight: the same allele pair can show different dominance relationships depending on which phenotype you measure. Sickle-cell hemoglobin (HbS) is recessive for the disease phenotype of severe anemia, codominant at the level of the protein (heterozygotes make both HbA and HbS, visible as two bands on electrophoresis), and arguably dominant for malaria resistance. The level of description determines the answer.

2.5 Extensions that modify the ratios

  • Lethal alleles: a recessive lethal turns 1:2:1 into 2:1 among survivors. The yellow coat allele in mice is the classic example.
  • Epistasis: one gene masks another. A 9:3:4 ratio (recessive epistasis, as in Labrador coat color) or 9:7 (complementary gene action) tells you the two genes act in the same pathway. Ratios become a diagnostic of biochemical pathway structure.
  • Penetrance and expressivity: not every carrier of a genotype shows the phenotype (incomplete penetrance), and those who do may show it to different degrees (variable expressivity).

3. The Chromosomal Theory of Inheritance and Sex Determination

3.1 Genes live on chromosomes

Sutton and Boveri (around 1902–1903) noticed that chromosomes during meiosis behave exactly as Mendel's factors should: they come in pairs, separate, and assort. The hypothesis became a theory through Thomas Hunt Morgan's work on Drosophila. Morgan found a white-eyed male and showed that the trait's inheritance followed the X chromosome [Morgan, 1910].

The decisive proof came from Calvin Bridges [Bridges, 1916]. Rare exceptional offspring appeared that violated the expected X-linked pattern, such as white-eyed daughters from crosses that should not produce them. Bridges predicted that these flies came from nondisjunction, the failure of X chromosomes to separate. He then confirmed under the microscope that the exceptional females were XXY. This was a prediction about chromosomes made from a genetic anomaly and verified cytologically. Few results in biology are this clean.

3.2 Sex determination systems

  • Mammalian XX/XY: maleness is determined by the presence of the Y-linked SRY gene, which encodes a transcription factor that triggers testis development. XXY humans (Klinefelter syndrome) are male. XO humans (Turner syndrome) are female. So the Y is dominant for sex.
  • Drosophila XX/XY: superficially the same, but sex is determined by the ratio of X chromosomes to autosome sets. XXY flies are female and XO flies are sterile males. Identical karyotype notation can hide very different mechanisms.
  • ZZ/ZW (birds, some reptiles, butterflies): here the female is the heterogametic sex (ZW).
  • Environmental: in many turtles and crocodilians, incubation temperature determines sex.

3.3 Sex-linked inheritance

Males are hemizygous for X-linked genes: they carry one copy, so recessive X-linked alleles are always expressed in males. This produces characteristic pedigree signatures:

  • X-linked recessive traits (hemophilia A, red-green color blindness, Duchenne muscular dystrophy) affect mostly males. They are never passed father to son, because fathers give sons a Y. They are passed through carrier daughters, so they often "skip" generations.
  • Reciprocal crosses give different results. This is a key test for sex linkage, because for autosomal genes reciprocal crosses give identical results.

In mammals, dosage compensation occurs by X-inactivation. One X in each female cell is condensed into a Barr body early in development, and the choice is random in each cell lineage. Female mammals are therefore mosaics. The tortoiseshell calico cat, with its orange and black patches, makes this visible. It is also the first example of epigenetics in this course: genetically identical X chromosomes held in different expression states and inherited through cell divisions.

4. Linkage, Recombination, and Genetic Mapping

4.1 Breaking the Law of Independent Assortment

If genes reside on chromosomes, genes on the same chromosome should travel together. Morgan observed this linkage but found it was incomplete: recombinant combinations appeared at characteristic frequencies. His explanation was crossing over, the physical exchange between homologous non-sister chromatids during prophase I, visible as chiasmata.

4.2 The beautiful idea: frequency as distance

Alfred Sturtevant, then an undergraduate in Morgan's lab, reasoned that if crossovers occur at roughly random positions, then the farther apart two genes are, the more likely a crossover falls between them. Recombination frequency could therefore serve as a measure of distance. In one night he built the first genetic map [Sturtevant, 1913]. The map distances were approximately additive, which is strong evidence that genes lie in a linear order.

Recombination frequency:
            number of recombinant progeny
  RF  =  ─────────────────────────────────  × 100%
                 total progeny

  1% RF  ≡  1 map unit (m.u.)  ≡  1 centimorgan (cM)

Three-point cross bookkeeping:
  • The two most frequent classes = parental (noncrossover)
  • The two least frequent classes = double crossovers (DCO)
  • The gene that "switches" in the DCO class is the MIDDLE gene

Coefficient of coincidence and interference:
         observed DCO frequency
  c  =  ─────────────────────────────────          I = 1 − c
        (RF₁ as fraction) × (RF₂ as fraction)

Positive interference (I > 0) means a crossover in one region suppresses crossovers nearby. This tells you crossovers are not fully independent events. Mechanistically, it reflects how the synaptonemal complex and recombination machinery space crossovers along the chromosome.

4.3 Why RF saturates at 50%: a calculus connection

This is a good place to use MATH102. Recombination frequency is not linear in physical distance over long ranges, because an even number of crossovers between two loci restores the parental arrangement. Haldane modeled crossovers along a chromosome as a Poisson process [Haldane, 1919]. If the expected number of crossovers in an interval (on a single chromatid) is m, measured in Morgans, then a gamete is recombinant only when the number of crossovers is odd:

  P(k crossovers) = e^(−m) · m^k / k!

  RF = P(k odd) = e^(−m) · [ m + m³/3! + m⁵/5! + ... ]
                = e^(−m) · sinh(m)
                = ½ (1 − e^(−2m))          ← Haldane mapping function

  As m → 0:   RF ≈ m                (additive, linear regime)
  As m → ∞:   RF → ½               (genes behave as unlinked)

The odd-term Taylor series of em is sinh(m), a result you derived in Calc II. Here it explains why two genes at opposite ends of a long chromosome assort independently even though they are physically linked. Mendel's seven traits happened to behave independently. Some of the genes are syntenic, meaning on the same chromosome, but far enough apart to show free recombination.

4.4 Chi-square: is the deviation real?

Genetic data are counts subject to sampling noise, so you need a principled way to decide whether observed numbers fit a hypothesized ratio. The Pearson chi-square test does this.

           (Observed − Expected)²
  χ²  =  Σ ──────────────────────
                 Expected

  degrees of freedom (df) = (number of phenotypic classes) − 1
         (when expected values come from a ratio, not estimated from the data)

Worked example: dihybrid F₂, 160 progeny, hypothesis 9:3:3:1
  Class     O     E      (O−E)²/E
  A_B_     96    90      0.400
  A_bb     28    30      0.133
  aaB_     26    30      0.533
  aabb     10    10      0.000
                         ─────
                  χ² =   1.067     df = 3,  critical value (p = 0.05) = 7.815

  1.067 < 7.815 → fail to reject the hypothesis of independent assortment

Interpret this carefully. A small χ² does not prove the 9:3:3:1 hypothesis. It means the data give no significant reason to reject it. Always use raw counts, never percentages, because χ² depends on sample size.

5. DNA Structure: Chemistry That Explains Its Own Copying

5.1 The double helix

The Watson–Crick model [Watson & Crick, 1953] combined three lines of evidence:

  • Chargaff's rules: in double-stranded DNA, [A] = [T] and [G] = [C], while the A+T to G+C ratio varies between species [Chargaff, 1950].
  • X-ray fiber diffraction from Rosalind Franklin and Raymond Gosling (and from Wilkins). The data indicated a helix with a 34 Å repeat, a 3.4 Å spacing between stacked bases, and phosphates on the outside [Franklin & Gosling, 1953].
  • Correct tautomers: Jerry Donohue pointed out that the bases exist mainly in the keto and amino forms, not the enol and imino forms printed in textbooks of the time. Only with the correct tautomers do A·T and G·C pairs have identical overall geometry.

Structural features to know cold:

  • Two antiparallel strands (one runs 5′→3′, the other 3′→5′), wound into a right-handed helix.
  • A sugar-phosphate backbone joined by 3′→5′ phosphodiester bonds. The phosphates are negatively charged at physiological pH, which ties back to CHEM102 acid-base chemistry: the phosphodiester pKa is about 1.
  • Complementary base pairing: A·T (2 hydrogen bonds) and G·C (3 hydrogen bonds). In both, a purine pairs with a pyrimidine, so the pair width is constant (~10.8 Å between C1′ atoms). This geometric uniformity matters most: any sequence fits the same helix.
  • Major and minor grooves. The edges of base pairs exposed in the major groove present a sequence-specific pattern of hydrogen-bond donors, acceptors, and methyl groups. This is how proteins "read" DNA without unwinding it.
"It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material." [Watson & Crick, 1953]

This sentence is famously understated. The structure did more than describe the molecule. It explained in one stroke how the molecule can be copied (each strand templates its partner) and how it stores information (the sequence of bases is unconstrained along the strand).

5.2 What actually holds the helix together (CHEM102 connection)

A common belief is that hydrogen bonds hold DNA together. They are essential for specificity, but they contribute surprisingly little net stability. When strands separate in water, the bases simply form hydrogen bonds to water instead, so the net ΔH from base-pair hydrogen bonds is small. The dominant stabilizing force is base stacking: van der Waals and dispersion interactions, plus the hydrophobic effect, between adjacent flat aromatic bases. This is why DNA duplex stability is modeled with nearest-neighbor parameters, where ΔG depends on which base pairs are stacked next to each other, not just on GC count.

Duplex formation:   single strand A + single strand B  ⇌  duplex AB

  ΔG°(duplex) = Σ ΔG°(nearest-neighbor stacks) + ΔG°(initiation)

  Melting temperature (Tm): the temperature at which half the strands are duplexed.
  Higher GC content → higher Tm   (GC stacks are more favorable;
                                   GC pairs have 3 H-bonds)
  Higher [Na⁺] → higher Tm        (cations screen phosphate–phosphate repulsion)

5.3 Helical polymorphism: B, A, and Z

FormHandednessbp per turnRise per bpConditions / significance
B-DNARight~10.5 (in solution)~3.4 ÅPredominant physiological form; the Watson–Crick structure
A-DNARight~11~2.6 ÅDehydrated DNA; the form adopted by RNA duplexes and DNA–RNA hybrids (the 2′-OH favors a C3′-endo sugar pucker); wider and shorter, with a deep, narrow major groove
Z-DNALeft~12~3.7 ÅFavored by alternating purine–pyrimidine sequences (e.g., (CG)n), high salt, or negative supercoiling; zigzag backbone [Wang et al., 1979]

The main point: DNA is not a rigid rod. Its conformation responds to sequence, hydration, ions, and torsional stress. The A-form geometry of RNA becomes important later, because the ribosome's decoding center reads the A-form codon–anticodon minihelix.

6. DNA Replication: Copying Six Billion Letters

6.1 Semi-conservative replication

Three models were possible: conservative (the original duplex stays intact), semi-conservative (each daughter keeps one old strand), and dispersive (old and new fragments interspersed). Meselson and Stahl [Meselson & Stahl, 1958] grew E. coli on heavy 15N, switched the cells to 14N, and separated DNA by density in a CsCl equilibrium gradient.

  • Generation 0: one heavy band.
  • Generation 1: one intermediate band. This rules out conservative replication.
  • Generation 2: intermediate and light bands in equal amounts. This rules out dispersive replication, which would give one band drifting progressively lighter.

This experiment is often called "the most beautiful experiment in biology." The question was sharp, the method separated the alternatives by physical chemistry alone, and the result was unambiguous.

6.2 The polymerase reaction

All DNA and RNA polymerases use the same chemistry: an SN2-type nucleophilic attack by the 3′-OH of the growing strand on the α-phosphate of an incoming nucleoside triphosphate. Two Mg²⁺ ions in the active site position the substrates and stabilize the leaving group [Watson et al., 2014, ch. 9].

  (dNMP)ₙ–3′OH  +  dNTP  ──Mg²⁺, polymerase──▶  (dNMP)ₙ₊₁  +  PPᵢ
                                                 ΔG°′ ≈ slightly negative

  PPᵢ  +  H₂O  ──pyrophosphatase──▶  2 Pᵢ        ΔG°′ ≈ −19 kJ/mol

  Coupled reaction is strongly favorable → polymerization is effectively irreversible.

Two consequences follow directly from this chemistry:

  1. Polymerases synthesize only 5′→3′. The energy comes from the incoming nucleotide's triphosphate. If synthesis ran 3′→5′, the growing chain would carry the triphosphate at its 5′ end. Removing a mismatched terminal nucleotide by proofreading would then leave the chain without an activated end and stall it. The 5′→3′ direction makes error correction compatible with continued synthesis.
  2. Polymerases cannot start a chain from scratch. They need a 3′-OH. A primase lays down a short RNA primer first.

Together these create the replication fork asymmetry. One strand (the leading strand) is synthesized continuously toward the fork. The other (the lagging strand) is synthesized discontinuously in short Okazaki fragments [Okazaki et al., 1968] that are later joined.

6.3 Fork machinery compared

FunctionE. coliEukaryotes
Origin recognitionDnaA at a single oriCORC at thousands of origins
HelicaseDnaB (moves 5′→3′ on the lagging-strand template)CMG complex (Cdc45–MCM2–7–GINS), moves 3′→5′ on the leading-strand template
Single-strand bindingSSBRPA
PrimaseDnaGPol α–primase
Main replicative polymerasePol III holoenzyme (both strands)Pol ε (leading), Pol δ (lagging)
Sliding clamp / loaderβ-clamp / γ-complexPCNA / RFC
Primer removalPol I (5′→3′ exonuclease), RNase HRNase H2, FEN1, Pol δ strand displacement
Nick sealingDNA ligase (NAD⁺-dependent)DNA ligase I (ATP-dependent)
Torsional stress reliefDNA gyrase (Topo II)Topoisomerases I and II
Fork speed~1000 nt/s~50 nt/s (compensated by many origins)

The end-replication problem. Linear chromosomes cannot be fully copied at their ends, because removing the final lagging-strand primer leaves a gap that no polymerase can fill. Eukaryotes solve this with telomerase, a reverse transcriptase that carries its own RNA template and extends the 3′ overhang with repeats (TTAGGG in humans). It is a ribonucleoprotein that exploits the reverse-transcription arrow of the central dogma.

6.4 Fidelity: a thermodynamics and kinetics problem

Replication fidelity in E. coli reaches about one error per 109–1010 bases. Base-pairing free-energy differences alone cannot explain this. If a correct pair and a mismatch differ by ΔΔG, Boltzmann discrimination from CHEM102 gives:

  error ratio ≈ exp(−ΔΔG / RT)

  With ΔΔG ≈ 10–12 kJ/mol at 310 K (RT ≈ 2.58 kJ/mol):
  error ≈ e^(−4.3) ≈ 10⁻²     ← far too sloppy

Several stages multiply together to reach the observed accuracy:

  1. Polymerase active-site geometry (shape complementarity and induced fit): ~10−4 to 10−5.
  2. 3′→5′ exonuclease proofreading: ~100-fold improvement, giving ~10−7.
  3. Post-replicative mismatch repair: ~100- to 1000-fold more, giving ~10−9 to 10−10.

Proofreading works because it spends energy: a correctly incorporated dNTP's triphosphate is consumed, and so is the dNTP that was excised. Hopfield and Ninio formalized this as kinetic proofreading [Hopfield, 1974]. Accuracy beyond the equilibrium limit costs free energy. The same principle recurs in tRNA charging and ribosomal decoding.

7. Transcription: Selecting What to Read

7.1 The logic

Transcription copies one strand of a DNA segment into RNA. The template strand (antisense) is read 3′→5′. The RNA is synthesized 5′→3′ and has the same sequence as the coding strand (sense), with U in place of T. Unlike DNA polymerases, RNA polymerases need no primer. They also have much weaker proofreading, which they achieve by backtracking and cleavage, so RNA error rates are around 10−4 to 10−5. That level is tolerable because transcripts are temporary and made in many copies.

7.2 Prokaryotes: one polymerase, interchangeable σ factors

The E. coli core RNA polymerase (α₂ββ′ω) is catalytically competent but cannot find promoters. A σ factor binds to form the holoenzyme and recognizes promoter elements. For the housekeeping σ70 these are:

   5′ ──── TTGACA ──────(15–19 bp spacer)────── TATAAT ───── +1 ──▶
           −35 box                              −10 box (Pribnow)
           (consensus)                          (AT-rich: easy to melt)

The AT-rich −10 element is no accident. Using the thermodynamics above, A·T stacks are less stable, so this is where the duplex opens to form the transcription bubble. Alternative σ factors, such as σ32 for heat shock, let the cell switch whole programs of genes on together.

Termination happens either intrinsically, through a GC-rich hairpin followed by a run of U's (the weak rU·dA hybrid lets the transcript dissociate), or through the Rho helicase.

7.3 Eukaryotes: three polymerases and a committee

  • RNA Pol I: large rRNA precursor (28S, 18S, 5.8S).
  • RNA Pol II: mRNA and most snRNAs and miRNAs. Its C-terminal domain (CTD) has heptad repeats (YSPTSPS) whose phosphorylation pattern coordinates initiation, elongation, and RNA processing.
  • RNA Pol III: tRNA, 5S rRNA, U6 snRNA.

Pol II cannot recognize promoters alone. General transcription factors assemble in order: TFIID, whose TBP subunit binds the TATA box (~25–30 bp upstream when present) and sharply bends the DNA, then TFIIA, TFIIB, TFIIF–Pol II, TFIIE, and TFIIH. TFIIH contains helicase activity to open the DNA and a kinase that phosphorylates the CTD to release the polymerase into elongation [Watson et al., 2014, ch. 13]. TFIIH also takes part in nucleotide excision repair (§12), an early hint that transcription and repair are intertwined.

8. RNA Processing: Editing the Message

In eukaryotes the primary transcript (pre-mRNA) is modified before export, usually while it is still being made.

8.1 The 5′ cap

A 7-methylguanosine is attached through an unusual 5′–5′ triphosphate linkage. The cap protects the mRNA from 5′ exonucleases, promotes export, and is recognized by eIF4E during translation initiation.

8.2 Polyadenylation

The transcript is cleaved ~10–30 nt downstream of an AAUAAA signal. Poly(A) polymerase then adds ~200 A's without a template. Poly(A)-binding protein coats the tail and contacts the cap-binding complex, which circularizes the mRNA for efficient translation. Tail shortening is a timer for mRNA decay.

8.3 Splicing

The discovery that genes are interrupted, made in 1977 in adenovirus, was a major surprise. Introns are removed by two sequential transesterification reactions:

  Exon1 ─GU ············· A ······· AG─ Exon2
         5′ splice site   branch    3′ splice site
                          point A

  Step 1: 2′-OH of branch-point A attacks the phosphate at the 5′ splice site
          → free Exon1–3′OH  +  lariat intermediate (2′–5′ linkage)

  Step 2: 3′-OH of Exon1 attacks the phosphate at the 3′ splice site
          → Exon1–Exon2 joined  +  excised intron lariat

  Net phosphodiester bonds: conserved (2 broken, 2 formed) → ΔG ≈ 0 for the chemistry

The chemistry is nearly energy-neutral, yet the spliceosome consumes many ATPs. The ATP drives the DEAD/DEAH-box helicases that remodel RNA–RNA contacts. Much of that energy buys accuracy and ordering, not chemistry, which is another example of paying free energy for fidelity.

The spliceosome is assembled from five snRNPs (U1, U2, U4, U5, U6) plus many proteins. U1 base-pairs with the 5′ splice site. U2 pairs with the branch site and bulges out the branch A. U6 replaces U1 and, with U2, forms the catalytic core. The catalytic center is made of RNA coordinating metal ions, and is closely related to self-splicing group II introns. This is strong evidence for an ancient "RNA world" in which RNA was both genome and catalyst.

Alternative splicing lets one gene produce many protein isoforms. Humans have ~20,000 protein-coding genes but a much larger proteome. The Drosophila Dscam gene can in principle produce over 38,000 isoforms.

9. Translation: Where Information Becomes Chemistry

9.1 tRNA charging: the true moment of decoding

The genetic code is not "read" by the ribosome. The ribosome only checks codon–anticodon pairing. The connection between a codon and an amino acid is made by aminoacyl-tRNA synthetases, which attach each amino acid to its correct tRNA:

  Step 1 (activation):   amino acid + ATP  ⇌  aminoacyl-AMP + PPᵢ
  Step 2 (transfer):     aminoacyl-AMP + tRNA  ⇌  aminoacyl-tRNA + AMP
  Driving force:         PPᵢ → 2 Pᵢ

  Net: amino acid + tRNA + ATP → aminoacyl-tRNA + AMP + 2 Pᵢ
       (consumes 2 high-energy phosphoanhydride bonds)

The product is an ester between the amino acid's carboxyl group and the 2′- or 3′-OH of the tRNA's terminal adenosine. That ester is a "high-energy" activated acyl group, roughly comparable to a thioester. Chemically, translation is aminolysis of activated esters.

Synthetases must discriminate between similar amino acids such as isoleucine and valine, which differ by one methylene. Many use a separate editing site that hydrolyzes misacylated products. The "double-sieve" model is another kinetic proofreading mechanism.

9.2 The ribosome

  • Bacterial 70S = 30S (16S rRNA + ~21 proteins) + 50S (23S, 5S rRNA + ~33 proteins). Eukaryotic 80S = 40S + 60S.
  • Three tRNA sites: A (aminoacyl), P (peptidyl), E (exit).
  • The peptidyl transferase center lies in the 50S subunit and is built entirely of 23S rRNA. No protein side chain is within ~18 Å of the active site [Nissen et al., 2000]. The ribosome is a ribozyme. Its catalysis works mainly by positioning substrates and by the 2′-OH of the P-site tRNA's terminal adenosine, not by classical acid-base catalysis from protein residues.

9.3 The three phases

  1. Initiation. In bacteria, the 16S rRNA 3′ end base-pairs with the Shine–Dalgarno sequence (AGGAGG) upstream of the AUG start codon. The initiator tRNA carries N-formylmethionine. In eukaryotes, the 43S complex binds the 5′ cap and scans to the first AUG in good context (the Kozak sequence, gccRccAUGG).
  2. Elongation. EF-Tu·GTP (eEF1A) delivers aminoacyl-tRNA to the A site. Correct codon–anticodon pairing triggers GTP hydrolysis and tRNA accommodation. The α-amino group of the A-site aminoacyl-tRNA then attacks the ester carbonyl of the P-site peptidyl-tRNA, transferring the peptide chain onto the A-site tRNA. EF-G·GTP (eEF2) drives translocation by one codon.
  3. Termination. A stop codon (UAA, UAG, UGA) in the A site is recognized by release factor proteins, not tRNAs. RF1 reads UAA/UAG and RF2 reads UAA/UGA; eRF1 reads all three. Their conserved GGQ motif positions a water molecule that hydrolyzes the peptidyl-tRNA ester, releasing the protein.
Energy cost per peptide bond (approximate):
  2 ATP-equivalents  (tRNA charging: ATP → AMP + PPᵢ)
  1 GTP              (EF-Tu delivery / proofreading)
  1 GTP              (EF-G translocation)
  ─────────────────
  ~4 high-energy phosphate bonds per peptide bond

  Compare: the free energy of peptide bond formation from free amino acids
  in water is only modestly unfavorable (~ +10 to +20 kJ/mol).
  Most of the cost pays for SPEED and ACCURACY, not thermodynamics.

Direction matters: proteins are synthesized from the N-terminus to the C-terminus, while the mRNA is read 5′→3′.

10. The Genetic Code

10.1 Cracking it

With four bases and twenty amino acids, codons must be at least three bases long, since 4² = 16 is too few and 4³ = 64 is enough. Crick, Brenner, and colleagues showed genetically that the code is read in non-overlapping triplets from a fixed starting point. Combining +1 and −1 frameshift mutations in phage T4 restored function, but only combinations of three like mutations did so on their own [Crick et al., 1961].

The biochemical breakthrough came from Nirenberg and Matthaei [Nirenberg & Matthaei, 1961]. They built a cell-free E. coli extract that synthesized protein and added synthetic polyuridylic acid (poly-U). The extract then made polyphenylalanine, so UUU encodes Phe. This was the first codon assignment. (The paper's published title is "The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polynucleotides," PNAS 47:1588–1602.) Later work with synthetic polymers of defined repeats (Khorana) and the trinucleotide binding assay (Nirenberg and Leder, 1964) completed the table by about 1966.

The experiment worked partly by luck. The high Mg²⁺ concentration in their extract allowed initiation without a proper start codon, and polyphenylalanine is insoluble, which made it easy to recover by precipitation.

10.2 Properties of the code

  • Triplet, non-overlapping, comma-free: there is no punctuation between codons. The reading frame is set by the start codon.
  • Degenerate (redundant): 61 sense codons specify 20 amino acids. Leu, Ser, and Arg have six codons each. Met (AUG) and Trp (UGG) have only one.
  • Non-random degeneracy: synonymous codons usually differ at the third position. Codons with a pyrimidine (U/C) in the second position tend to encode hydrophobic amino acids. The code appears structured so that common point mutations often swap similar amino acids, which makes it error-minimizing.
  • Nearly universal: exceptions exist. In vertebrate mitochondria UGA codes for Trp. Some ciliates read UAA/UAG as Gln. Selenocysteine (UGA in a special context) and pyrrolysine (UAG in some archaea) are natural expansions.

10.3 Wobble

If each codon needed its own tRNA, cells would need 61 tRNAs. Many organisms have far fewer. Crick's wobble hypothesis [Crick, 1966] proposed that the first base of the anticodon (5′ end), which pairs with the third base of the codon (3′ end), has geometric freedom:

  Anticodon 5′ base     can pair with codon 3′ base
  ─────────────────     ───────────────────────────
        C                       G
        A                       U
        G                       C, U
        U                       A, G
        I (inosine)             U, C, A

  Remember orientation:  codon  5′–X Y Z–3′
                         anticodon 3′–X′Y′Z′–5′   (Z′ is the wobble position)

Inosine is a deaminated adenosine made by tRNA-modifying enzymes. It is a good example of chemistry broadening specificity on purpose. The decoding center of the ribosome tightly checks Watson–Crick geometry at codon positions 1 and 2 through A-minor interactions with 16S rRNA bases A1492, A1493, and G530. It is more permissive at position 3.

11. Gene Regulation: Deciding When to Read

11.1 The lac operon: logic in molecules

Jacob and Monod's operon model [Jacob & Monod, 1961] was the first molecular account of gene regulation, and it reads like a logic circuit.

  • Structural genes: lacZ (β-galactosidase), lacY (permease), lacA (transacetylase), transcribed as one polycistronic mRNA.
  • Negative control: the LacI repressor (a tetramer) binds the operator and blocks transcription. When lactose is present, some is converted to allolactose, which binds the repressor allosterically and lowers its affinity for operator DNA.
  • Positive control: when glucose is scarce, cAMP levels rise. cAMP binds CAP (also called CRP), and CAP–cAMP binds upstream of the promoter, bending DNA and recruiting RNA polymerase.
  High transcription  ⟺  (lactose present) AND (glucose absent)
                          [repressor off]       [CAP–cAMP on]

  Fractional operator occupancy by repressor (CHEM102 equilibrium):

              [R]
  θ  =  ─────────────         Inducer binding raises effective Kd by ~10³-fold,
         [R] + K_d            so θ collapses → operon de-repressed

The genetic analysis was as elegant as the model. Jacob and Monod used partial diploids (merodiploids). They showed that lacI mutations act in trans, because they encode a diffusible protein. lacOc (constitutive operator) mutations act only in cis, controlling only genes on the same DNA molecule. The cis/trans test separates DNA sites from the products that act on them, and you will use this logic throughout genetics.

11.2 The trp operon: two layers of repression

  • Repression: tryptophan acts as a corepressor. It binds TrpR and enables the repressor to bind the operator. This is the reverse of lac: the end product shuts down its own biosynthesis.
  • Attenuation (Yanofsky): a more striking mechanism. The leader transcript contains a short ORF with two adjacent Trp codons and four segments that can form alternative hairpins. Because bacterial transcription and translation are coupled, the position of the ribosome on the leader decides the RNA structure. With plenty of charged tRNATrp, the ribosome moves quickly through the Trp codons and a 3:4 terminator hairpin forms, so transcription stops. With little Trp, the ribosome stalls at the Trp codons, the 2:3 antiterminator forms, and transcription continues. The cell measures amino acid availability through the speed of translation.

11.3 Eukaryotic transcription factors

Eukaryotic regulation is combinatorial and acts over long distances:

  • Modular activators: a DNA-binding domain (zinc finger, homeodomain, bZIP, bHLH) plus a separate activation domain. The modules can be swapped experimentally, as in the yeast two-hybrid system built from GAL4 domains.
  • Enhancers: regulatory elements that can act thousands of base pairs away, upstream, downstream, or inside introns, through DNA looping and the Mediator complex.
  • Combinatorial control: a few hundred transcription factors in combination specify the expression programs of hundreds of cell types.

11.4 Epigenetics: heritable states without sequence change

Epigenetic information is stored in chemical modifications of DNA and chromatin that can persist through cell division:

  • DNA methylation: 5-methylcytosine at CpG dinucleotides, generally associated with silencing in mammals. Methylation is maintained after replication because DNMT1 recognizes hemimethylated CpG sites and methylates the new strand. Symmetric palindromic CpG sites make copying possible, a direct parallel to semi-conservative replication.
  • Histone modifications: acetylation of lysines (by HATs) neutralizes their positive charge, loosens histone–DNA contacts, and correlates with activity. Methylation effects depend on the residue: H3K4me3 is associated with active promoters, while H3K9me3 and H3K27me3 are associated with repression.
  • Examples: X-inactivation (§3.3), genomic imprinting (parent-of-origin expression, as in IGF2/H19), and position-effect variegation in Drosophila.

12. Mutation: The Raw Material and the Hazard

12.1 Mutations are random with respect to need

Do bacteria exposed to a phage develop resistance in response, or do resistant mutants arise spontaneously beforehand? Luria and Delbrück addressed this with the fluctuation test [Luria & Delbrück, 1943]. If mutations were induced by the phage, every parallel culture would show similar numbers of resistant colonies, with a Poisson distribution whose variance roughly equals its mean. In fact the counts varied enormously: most cultures had few resistant colonies and a rare few had many ("jackpots"). These jackpots came from cultures in which a mutation arose early and was inherited by many descendants. Mutation happens before selection, and selection only reveals it. Luria and Delbrück reached this conclusion by comparing variances, the kind of distribution analysis you met in Calc II.

12.2 Classification

  • Point mutations:
    • Transitions: purine↔purine or pyrimidine↔pyrimidine (A↔G, C↔T). These are more common.
    • Transversions: purine↔pyrimidine.
  • Consequences in coding sequence: synonymous (silent), missense (amino acid change; conservative or not), nonsense (premature stop, often triggering nonsense-mediated decay), and frameshift (insertion or deletion not divisible by 3).
  • Functional classes: loss-of-function (amorph or hypomorph, usually recessive), gain-of-function (often dominant), and dominant-negative (a mutant subunit poisons a multimeric complex).
  • Larger changes: deletions, duplications, inversions, translocations, and expansions of trinucleotide repeats (Huntington's, fragile X).

12.3 Causes: the chemistry of damage

DNA is chemically fragile. Each human cell suffers on the order of 104 to 105 lesions per day from spontaneous hydrolysis and oxidation alone [Lindahl, 1993].

  • Spontaneous: depurination (hydrolysis of the N-glycosidic bond), deamination (C → U; 5-methyl-C → T, which is why CpG sites are mutation hotspots), oxidative damage (8-oxoguanine mispairs with A, causing G→T transversions), and rare tautomeric shifts during replication.
  • Chemical mutagens:
    • Base analogs: 5-bromouracil mimics T but tautomerizes more often, so it pairs with G.
    • Alkylating agents: EMS produces O⁶-ethylguanine, which pairs with T and causes G·C→A·T transitions.
    • Intercalators: acridines and ethidium bromide slip between stacked bases and cause frameshifts.
    • Deaminating agents: nitrous acid.
  • Radiation: UV-C and UV-B photochemically fuse adjacent pyrimidines into cyclobutane pyrimidine dimers and 6-4 photoproducts. Ionizing radiation causes strand breaks, including the dangerous double-strand break.

The Ames test uses his⁻ Salmonella strains to detect mutagens by their reversion rate. Liver extract is added to model metabolic activation. Many carcinogens are mutagens, and this observation links genetics to cancer biology.

12.4 Suppressor mutations: two wrongs that make a right

A suppressor is a second mutation that restores the phenotype lost to a first mutation without reversing it.

  • Intragenic suppressors: a second change in the same gene. Examples are a compensating frameshift (the basis of Crick et al., 1961) or a compensating amino acid change that restores a lost salt bridge or packing contact.
  • Intergenic suppressors: a change in a different gene. The classic case is the nonsense-suppressor tRNA. A mutation in a tRNA gene's anticodon (for example, tRNATyr anticodon GUA → CUA) lets it read the UAG stop codon and insert Tyr. The cell then reads through the premature stop. The suppressor tRNA competes with release factor at genuine UAG stops too, so suppression is partial and tolerated.

Suppressor analysis is a powerful way to discover which proteins interact. If a mutation in protein B compensates for a mutation in protein A, the two likely make physical contact.

13. DNA Repair: Maintaining the Archive

Because DNA is chemically fragile, cells invest heavily in repair. Each pathway matches a type of damage. The recurring logic is that the complementary strand serves as a backup copy, which is the other half of the significance of the double helix.

PathwayDamage targetedKey machineryLogic
Mismatch repair (MMR) Replication mispairs, small insertion/deletion loops E. coli: MutS (recognizes mismatch), MutL, MutH (nicks the unmethylated strand at hemimethylated GATC). Humans: MSH2/MSH6, MLH1/PMS2 The repair system must know which strand is new. In E. coli, Dam methylation of GATC lags behind replication, so the unmethylated strand is the new one. In eukaryotes, nicks and the orientation of PCNA mark the new strand.
Base excision repair (BER) Small, non-helix-distorting base lesions: uracil, 8-oxoG, alkylated bases DNA glycosylases (e.g., uracil-DNA glycosylase) → AP endonuclease → polymerase β → ligase A specific glycosylase flips the damaged base out of the helix into its active site and cleaves the N-glycosidic bond.
Nucleotide excision repair (NER) Bulky, helix-distorting lesions: pyrimidine dimers, chemical adducts E. coli: UvrA, UvrB, UvrC (dual incision), UvrD helicase. Humans: XPA–XPG proteins, TFIIH The system detects distortion rather than a specific chemistry, then cuts out an oligonucleotide (12–13 nt in bacteria, ~24–32 nt in humans). Defects cause xeroderma pigmentosum, with extreme UV sensitivity.
Homologous recombination (HR) Double-strand breaks, collapsed replication forks Resection (RecBCD / MRN), strand invasion by RecA (bacteria) or Rad51 (eukaryotes), BRCA1/BRCA2 When both strands are broken, the backup is the sister chromatid. A RecA/Rad51 nucleoprotein filament searches the genome for homology and invades it. The same chemistry underlies meiotic crossing over (§4).

A double-strand-break alternative is non-homologous end joining (NHEJ). It is fast but error-prone, directly ligating broken ends and often adding small insertions or deletions. CRISPR-based gene knockouts exploit this flaw.

Why DNA uses thymine, not uracil

This explanation links several topics together. Cytosine spontaneously deaminates to uracil, roughly 100–500 times per human cell per day. If DNA normally contained U, a repair enzyme could not tell a legitimate U (paired with A) from a mutagenic U (paired with G, formerly C). Using T, which is 5-methyluracil, as the normal base makes every uracil in DNA recognizably wrong, so uracil-DNA glycosylase can remove it without ambiguity. The extra methyl group is a tag for error detection. The same reasoning explains why methylated CpG sites are hotspots: deamination of 5-methylcytosine yields T, a normal base, which is harder to recognize as damage.

14. Recombinant DNA Technology: Genetics Becomes Engineering

14.1 Restriction enzymes

Bacteria defend against phage with restriction–modification systems. A restriction endonuclease cuts foreign DNA at specific sequences, and a partner methyltransferase protects the host's own sites. Type II enzymes cut at or near palindromic recognition sites:

  EcoRI:   5′-G↓AATTC-3′          produces 5′ "sticky" overhangs (AATT)
           3′-CTTAA↑G-5′

  SmaI:    5′-CCC↓GGG-3′          produces blunt ends
           3′-GGG↑CCC-5′

  Expected frequency of a random 6-bp site:  (1/4)⁶ = 1/4096 bp

Sticky ends from any two DNA sources cut with the same enzyme can anneal by base pairing and then be sealed by DNA ligase. This is the basic operation of recombinant DNA.

14.2 Vectors and transformation

A plasmid vector typically carries an origin of replication, a selectable marker (such as antibiotic resistance), and a multiple cloning site. Blue/white screening uses an insert that disrupts lacZα, which makes direct use of the lac operon biology from §11. Plasmids are introduced into bacteria by transformation: chemical competence with CaCl₂ plus heat shock, or electroporation. This is the same phenomenon Griffith and Avery studied, now used as a tool.

14.3 PCR: exponential copying

The polymerase chain reaction [Mullis & Faloona, 1987; Saiki et al., 1988] repeats a three-step cycle:

  1. Denature (~95 °C): separate the strands.
  2. Anneal (~50–65 °C, a few degrees below primer Tm): primers hybridize to flanking sequences.
  3. Extend (~72 °C): a thermostable polymerase such as Taq, from the thermophile Thermus aquaticus, synthesizes the complementary strands.
  N = N₀ · (1 + E)ⁿ        E = efficiency per cycle (0 ≤ E ≤ 1), n = cycles

  Ideal (E = 1):  30 cycles → 2³⁰ ≈ 1.07 × 10⁹-fold amplification

  ln N = ln N₀ + n·ln(1+E)   → linear on a log plot during exponential phase
                              (this is the basis of quantitative real-time PCR)

Every step is something you already understand: thermal denaturation (DNA thermodynamics, §5.2), primer-dependent 5′→3′ synthesis (§6.2), and exponential growth (MATH102 differential equations). Amplification eventually plateaus as primers and dNTPs run out and the polymerase loses activity, just as the logistic equation departs from pure exponential growth.

14.4 Gel electrophoresis

DNA's charge-to-mass ratio is nearly constant, because there is one phosphate per nucleotide. Free-solution mobility is therefore roughly independent of length. Separation comes from the sieving matrix (agarose or polyacrylamide), which retards larger molecules. Over a useful range, migration distance is approximately linear in log(length), so a ladder of known sizes gives a calibration curve. DNA migrates toward the anode (+).

14.5 Sanger sequencing

Sanger's chain-termination method [Sanger, Nicklen & Coulson, 1977] follows directly from the polymerase mechanism in §6.2. A dideoxynucleotide (ddNTP) lacks the 3′-OH. The polymerase incorporates it normally, but the chain can then grow no further, because there is no nucleophile for the next attack.

  Template + primer + polymerase + all 4 dNTPs + small amount of ddNTPs
     → population of fragments, each ending at a position where a ddNTP was incorporated
     → separate by size (capillary electrophoresis, single-nucleotide resolution)
     → read sequence from fluorescent color of terminal ddNTP (modern dye-terminator)

One missing oxygen atom made possible the technology that sequenced the first human genome.

15. Connections to Peptide Chemistry

Genetics may seem far from peptide chemistry, but translation is peptide synthesis. The cell's methods and the chemist's methods illuminate each other. Here is where the threads lead.

15.1 The ribosome as a peptide-synthesis machine

  • The chemistry is aminolysis of an activated ester. In solution-phase or solid-phase peptide synthesis (SPPS), you activate a carboxylic acid with reagents such as carbodiimides (DCC, DIC) with HOBt/Oxyma, or uronium salts (HATU, HBTU), then let an amine attack. The cell activates with ATP to form an aminoacyl-adenylate (a mixed anhydride), then transfers the acyl group to tRNA as an ester. That is the same strategy with different reagents.
  • Opposite directionality. Ribosomes build proteins N→C: the growing chain stays activated as the peptidyl-tRNA ester, and each new amino acid's free amine attacks it. Standard Fmoc-SPPS builds C→N: the C-terminal residue is anchored to resin, and each incoming amino acid is activated at its carboxyl. One reason chemists prefer C→N is racemization. Activating the C-terminus of a peptide chain, as N→C chemical synthesis would require, risks epimerization through oxazolone formation. Activating a single urethane-protected (Fmoc) amino acid largely avoids this. The ribosome avoids the problem through an enzymatic environment that does not promote oxazolone formation.
  • Protecting groups versus spatial control. A chemist must protect side chains (Lys ε-NH₂, Asp/Glu carboxyls, Cys thiol) to prevent side reactions. The ribosome needs none, because the active-site geometry presents only the α-amine to the ester. Positioning replaces protection.
  • Fidelity and length. SPPS coupling efficiencies of ~99% per step mean a 50-mer comes out at roughly 0.9950 ≈ 60% full-length product, so SPPS practically tops out around 50–100 residues without ligation strategies. The ribosome, with error rates near 10−4 per residue, routinely makes proteins of over 1000 residues. Native chemical ligation, which joins unprotected peptide segments through a C-terminal thioester and an N-terminal Cys, is the chemist's answer to the length problem. It also echoes the thioester and ester activation used in biology.

15.2 Recombinant peptide and protein production

The recombinant DNA toolkit in §14 is how most therapeutic peptides and proteins are made: insulin, GLP-1 analog precursors, and antibodies. You clone a synthetic gene, often codon-optimized for the host because degeneracy (§10.2) allows a choice of synonymous codons. You express it under an inducible promoter, often the lac-derived T7/lacO system induced with IPTG, a non-metabolizable allolactose mimic. Then you purify the product. Insulin, two chains joined by disulfides, is classically made as proinsulin or as separate chains and then processed. This is a direct use of gene structure and post-translational chemistry.

15.3 Site-directed mutagenesis and structure–activity relationships

PCR-based mutagenesis lets you change any single codon and therefore any amino acid. This is the biological counterpart of a medicinal chemist's SAR series. Alanine scanning, which replaces each residue in turn with Ala, maps which side chains matter for binding. Knowing the codon table lets you design the minimal nucleotide change.

15.4 Genetic code expansion: suppressors become synthesis tools

The amber suppressor tRNA in §12.4 has become one of the most powerful tools in peptide chemistry. Engineered orthogonal aminoacyl-tRNA synthetase/tRNA pairs, often derived from archaeal tyrosyl or pyrrolysyl systems, charge a suppressor tRNACUA with a non-canonical amino acid. Examples include amino acids carrying azides, alkynes, photo-crosslinkers, or fluorescent groups, incorporated at a UAG codon placed anywhere in a gene. This allows click-chemistry labeling, site-specific antibody–drug conjugates, and installation of post-translational modifications. A curiosity of bacterial genetics in the 1960s became a technology for building proteins with chemistry the genetic code never included.

15.5 mRNA display and directed evolution of peptides

In mRNA display and phage display, each peptide is physically linked to the nucleic acid that encodes it. You can screen 1012–1013 peptide variants, amplify the binders by PCR, and read their identities by sequencing. Genotype is coupled to phenotype at molecular scale. Several macrocyclic peptide drug candidates have come from such libraries, some combined with genetic code reprogramming (for example, "flexizyme"-charged tRNAs carrying N-methyl amino acids).

15.6 Post-translational modifications

The gene sets the sequence, but many functional peptides need further chemistry: signal-peptide cleavage, prohormone processing by convertases at dibasic sites (Lys-Arg), disulfide formation in the ER, C-terminal amidation (common in neuropeptides, by the enzyme PAM), glycosylation, and phosphorylation. Bioactive peptides are often a gene product plus a series of enzymatic tailoring steps. Some peptides, such as nonribosomal peptides like vancomycin and cyclosporin, bypass the ribosome entirely and are assembled by modular enzyme assembly lines that use thioester chemistry.

16. Common Misconceptions to Watch Out For

  1. "Dominant means more common, or better." Dominance describes the heterozygote phenotype only. Huntington's disease is caused by a dominant allele and is rare. Type O blood comes from a recessive allele and is the most common ABO type in many populations. Allele frequency is a population property, unrelated to dominance.
  2. "Each gene codes for one trait." Most traits are polygenic, and most genes are pleiotropic, affecting several traits. "The gene for X" is shorthand at best. A gene is better described as a variant whose change affects X in some genetic background and environment.
  3. "Recombination frequency is directly proportional to physical distance." Only for short intervals. RF saturates at 50% (§4.3), and crossover rates vary along chromosomes (hotspots, centromeric suppression) and between sexes. Map distance in cM is not base pairs.
  4. "A non-significant χ² proves the hypothesis." It means only that you failed to reject it. Many hypotheses can fit the same data, and with small samples the test has low power.
  5. "Hydrogen bonds are what hold the DNA double helix together." They provide specificity. Base stacking and the hydrophobic effect provide most of the stability (§5.2).
  6. "The template strand has the same sequence as the mRNA." The coding (non-template) strand matches the mRNA, with T for U. The template strand is complementary. Keep 5′/3′ labels on every sequence you write.
  7. "DNA polymerase can start a new strand anywhere." It needs a primer with a free 3′-OH. RNA polymerase does not, which is why primase is an RNA polymerase.
  8. "Silent mutations have no effect." Synonymous changes can disrupt splicing enhancers or splice sites, change mRNA structure and stability, or alter translation speed through codon usage, which can affect co-translational folding. "Synonymous" is not always "neutral."
  9. "The ribosome reads the genetic code." The ribosome checks codon–anticodon pairing. The aminoacyl-tRNA synthetases are the true translators, since they set which amino acid goes with which anticodon. Mischarge a tRNA, and the ribosome inserts the wrong amino acid without detecting it. This was shown in the 1962 Raney-nickel experiment converting Cys-tRNACys to Ala-tRNACys.
  10. "Proteins catalyze peptide bond formation." rRNA does (§9.2). Ribosomal proteins are largely structural and peripheral to the active site.
  11. "Mutations happen because the organism needs them." The Luria–Delbrück fluctuation test (§12.1) showed that mutations arise randomly with respect to their usefulness. Selection acts afterward. Stress can raise overall mutation rates, as in the SOS response, but not in a direction-specific way.
  12. "Sex-linked equals sex-limited." Sex-linked genes are located on a sex chromosome. Sex-limited traits, such as milk production or beard growth, are expressed in only one sex but can be encoded on autosomes.
  13. "Epigenetics means environment rewrites genes and offspring inherit acquired traits." Epigenetic marks are mitotically heritable within an organism. In mammals, most marks are erased and reset in the germline and early embryo. Transgenerational epigenetic inheritance is well documented in plants and C. elegans but remains limited and debated in mammals. The DNA sequence itself does not change.
  14. "PCR doubles DNA every cycle indefinitely." Only during the exponential phase with E ≈ 1. Reagent depletion, product reannealing, and polymerase decay produce a plateau (§14.3). This is why quantification uses the threshold cycle (Ct) in the exponential phase, not the endpoint.
  15. "Larger DNA fragments run farther on a gel because they carry more charge." Charge scales with length, so charge per mass is constant. The gel matrix retards larger fragments, so smaller fragments migrate farther.
  16. "One gene → one protein." This is a useful starting point (Beadle and Tatum's one gene–one enzyme), but alternative splicing, alternative promoters and polyadenylation, RNA editing, and non-coding RNA genes (rRNA, tRNA, snRNA, miRNA, lncRNA) all break it.

17. A Closing Synthesis

Notice how the same few ideas recur across this chapter:

  • Complementarity explains replication, transcription, translation (codon–anticodon pairing), splicing (snRNA–pre-mRNA), repair (the backup strand), and every recombinant DNA technique (primers, sticky ends, hybridization).
  • Spending energy to buy accuracy appears in polymerase proofreading, synthetase editing, ribosomal GTP hydrolysis, and spliceosome helicases. Equilibrium discrimination is never enough, so cells pay ATP and GTP for fidelity. This is CHEM102 thermodynamics applied to information.
  • RNA as catalyst appears in the ribosome, the spliceosome, and telomerase's template. These are molecular fossils of an RNA world.
  • Probability as the bridge between molecules and observations runs from Mendel's ratios to the Haldane function, χ², the Luria–Delbrück distribution, and PCR kinetics. Your calculus is part of the machinery of genetic inference.

As you work through the problem sets, try always to hold both descriptions at once: the abstract (ratios, maps, dominance) and the molecular (base pairs, enzymes, free energies). That double vision is what separates memorizing genetics from understanding it.

References Cited

  • Avery, O. T., MacLeod, C. M., & McCarty, M. (1944). Studies on the chemical nature of the substance inducing transformation of pneumococcal types. J. Exp. Med. 79:137–158.
  • Bridges, C. B. (1916). Non-disjunction as proof of the chromosome theory of heredity. Genetics 1:1–52, 107–163.
  • Chargaff, E. (1950). Chemical specificity of nucleic acids and mechanism of their enzymatic degradation. Experientia 6:201–209.
  • Crick, F. H. C. (1966). Codon–anticodon pairing: the wobble hypothesis. J. Mol. Biol. 19:548–555.
  • Crick, F. (1970). Central dogma of molecular biology. Nature 227:561–563.
  • Crick, F. H. C., Barnett, L., Brenner, S., & Watts-Tobin, R. J. (1961). General nature of the genetic code for proteins. Nature 192:1227–1232.
  • Franklin, R. E., & Gosling, R. G. (1953). Molecular configuration in sodium thymonucleate. Nature 171:740–741.
  • Griffiths, A. J. F., Wessler, S. R., Carroll, S. B., & Doebley, J. (2020). Introduction to Genetic Analysis, 12th ed. W. H. Freeman.
  • Haldane, J. B. S. (1919). The combination of linkage values, and the calculation of distances between the loci of linked factors. J. Genet. 8:299–309.
  • Hopfield, J. J. (1974). Kinetic proofreading: a new mechanism for reducing errors in biosynthetic processes requiring high specificity. PNAS 71:4135–4139.
  • Jacob, F., & Monod, J. (1961). Genetic regulatory mechanisms in the synthesis of proteins. J. Mol. Biol. 3:318–356.
  • Lindahl, T. (1993). Instability and decay of the primary structure of DNA. Nature 362:709–715.
  • Luria, S. E., & Delbrück, M. (1943). Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491–511.
  • Mendel, G. (1866). Versuche über Pflanzen-Hybriden. Verh. Naturforsch. Ver. Brünn 4:3–47.
  • Meselson, M., & Stahl, F. W. (1958). The replication of DNA in Escherichia coli. PNAS 44:671–682.
  • Morgan, T. H. (1910). Sex limited inheritance in Drosophila. Science 32:120–122.
  • Mullis, K. B., & Faloona, F. A. (1987). Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol. 155:335–350.
  • Nirenberg, M. W., & Matthaei, J. H. (1961). The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polynucleotides. PNAS 47(10):1588–1602. DOI: 10.1073/pnas.47.10.1588
  • Nissen, P., Hansen, J., Ban, N., Moore, P. B., & Steitz, T. A. (2000). The structural basis of ribosome activity in peptide bond synthesis. Science 289:920–930.
  • Okazaki, R., et al. (1968). Mechanism of DNA chain growth. I. Possible discontinuity and unusual secondary structure of newly synthesized chains. PNAS 59:598–605.
  • Saiki, R. K., et al. (1988). Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487–491.
  • Sanger, F., Nicklen, S., & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. PNAS 74:5463–5467.
  • Sturtevant, A. H. (1913). The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14:43–59.
  • Wang, A. H.-J., et al. (1979). Molecular structure of a left-handed double helical DNA fragment at atomic resolution. Nature 282:680–686.
  • Watson, J. D., & Crick, F. H. C. (1953). Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171:737–738. DOI: 10.1038/171737a0
  • Watson, J. D., Baker, T. A., Bell, S. P., Gann, A., Levine, M., & Losick, R. (2014). Molecular Biology of the Gene, 7th ed. Cold Spring Harbor Laboratory Press.

⚙️ Mechanisms & Applications

GPT-6 Astra

Mechanisms & Applications — From Inherited Alleles to Engineered Medicines

Skipper’s organizing idea: genetics connects three kinds of mechanism: chromosomes transmit information, cells interpret that information, and molecular machines preserve or change it. A useful explanation should move between these scales. “This trait is recessive” describes an inheritance pattern; “one functional allele supplies enough enzyme to maintain metabolic flux” begins to explain that pattern.

My chemistry background is essential here. DNA stability reflects hydrogen bonding, base stacking, solvation, and electrostatics; polymerization is coupled to nucleotide-triphosphate chemistry; gene regulation depends on molecular recognition and binding equilibria. Genetics is not separate from chemistry—it is chemistry organized into a system that can copy itself. [1, 2]

1. Meiosis Turns Chromosome Behavior into Mendelian Probabilities

The physical explanation for Mendel’s segregation law is the separation of homologous chromosomes during meiosis. Before meiosis, DNA replication produces two sister chromatids per chromosome. Replication doubles DNA content, but it does not double ploidy: the cell still has the same number of homologous chromosome sets.

  1. Homologs pair in prophase I. Maternal and paternal homologs align, and recombination can exchange DNA between nonsister chromatids.
  2. Homolog pairs orient at metaphase I. The orientation of one chromosome pair is approximately independent of that of another.
  3. Homologs separate in meiosis I. A heterozygote carrying A and a normally sends the two alleles to different daughter cells.
  4. Sister chromatids separate in meiosis II. Haploid products receive one allele at each ordinary autosomal locus.
  5. Fertilization combines two gametes. Offspring genotype probabilities follow from the gamete probabilities.
Monohybrid cross: Aa × Aa

Each parent: P(A gamete) = 1/2; P(a gamete) = 1/2

P(AA) = (1/2)(1/2) = 1/4
P(Aa) = (1/2)(1/2) + (1/2)(1/2) = 1/2
P(aa) = (1/2)(1/2) = 1/4

With complete dominance:
P(dominant phenotype) = 3/4
P(recessive phenotype) = 1/4

The multiplication rule combines independent events; the addition rule combines mutually exclusive routes to the same outcome. A Punnett square is a bookkeeping device for these rules, not the cause of inheritance.

For an AaBb × AaBb cross, independent assortment gives four equally probable gamete types. With complete dominance at both loci and no interaction that changes phenotype classification, the phenotypic ratio is 9:3:3:1. Genes on the same chromosome need not assort independently, and interactions between genes can change the phenotypic ratio even when the alleles themselves segregate normally. [1]

  • Complete dominance: the heterozygote resembles one homozygote for the measured phenotype. Often, one working allele provides sufficient function, but dominance can also arise through other mechanisms.
  • Incomplete dominance: the heterozygote has an intermediate phenotype. Alleles have not “blended”; they still segregate intact.
  • Codominance: both allelic products are detectable. The human ABO IA and IB alleles encode glycosyltransferases with different substrate specificities; an AB individual displays both corresponding antigens.

Biochemical lesson: dominance belongs to a genotype–phenotype relationship, not to an allele in isolation. The same heterozygote can show both proteins at the molecular level but resemble one homozygote at the clinical level. This is why “dominant” does not mean stronger, more common, more beneficial, or more highly expressed.

Sex linkage: in typical human XX/XY inheritance, an XY individual is hemizygous for most X-linked genes. A recessive X-linked variant can therefore produce a phenotype without a second variant copy. An affected father does not transmit his X chromosome to his sons; he transmits it to his daughters. X-inactivation makes many XX tissues mosaics, complicating the relationship between genotype and severity. Sex-determination systems differ across organisms, so the human pattern is not universal. [1]

Medical application: pedigree analysis distinguishes plausible autosomal, sex-linked, and other inheritance models. Under a simple autosomal recessive model, two heterozygous parents have a 1/4 affected-offspring probability per conception—not a guarantee that exactly one of four children will be affected. Penetrance, new mutations, mosaicism, and environmental effects can complicate actual pedigrees.

Historical connection: Mendel published his pea experiments in 1866. Sutton and Boveri developed the chromosomal interpretation of inheritance around 1902–1903; Morgan’s fruit-fly work subsequently connected particular traits to particular chromosomes. The conceptual advance was to give abstract hereditary “factors” a physical carrier. [1]

2. Recombination Converts Offspring Counts into Chromosome Maps

A crossover is a molecular exchange, not a gene jumping independently between chromosomes. In meiosis, programmed double-strand breaks can initiate homologous recombination. DNA ends are processed, a strand engages a homologous DNA molecule, and DNA synthesis and intermediate resolution can produce either crossover or noncrossover products. [1, 2]

A testcross makes recombination visible because the homozygous recessive tester contributes an unambiguous allele combination.

Testcross: AB/ab × ab/ab

Offspring classes reveal gametes made by the heterozygote:

AB: 420     ab: 420     Ab: 80     aB: 80

Parental combinations: AB, ab
Recombinant combinations: Ab, aB

Recombination fraction:
r = recombinant offspring / total offspring
  = (80 + 80) / 1000
  = 0.16

For a sufficiently short interval:
map distance ≈ 100r = 16 centimorgans (cM)

The surprising limitation: two crossovers can restore the parental arrangement of the outside markers. Consequently, observed recombination undercounts crossover events over long intervals. Under standard mapping assumptions, the expected pairwise recombination fraction approaches 0.5, even though chromosome map distances can exceed 50 cM.

Under Haldane’s idealized model of independently occurring crossovers, the correction is:

d = −(1/2) ln(1 − 2r)

d = map distance in morgans

For r = 0.16:
d ≈ 0.193 M = 19.3 cM

This correction assumes no crossover interference.
It is a model-based estimate, not a universal conversion.

In a three-point testcross, the most frequent classes usually identify parental haplotypes and the least frequent classes often identify double crossovers. Comparing these reveals the middle marker. Each double crossover must be counted once in each interval. This reasoning assumes that differential survival or scoring errors have not substantially distorted the classes. [1]

Applications: linkage maps helped locate disease genes before whole-genome sequencing and remain useful in family studies and breeding. A centimorgan is not a fixed number of base pairs: recombination rates vary by genomic location, sex, and organism. Sturtevant’s 1913 fruit-fly map demonstrated that inheritance frequencies could reveal the linear organization of genes. [3]

3. Chi-Square Analysis Tests a Model, Not Whether Data Look “Close Enough”

Finite samples rarely reproduce Mendelian ratios exactly. A statistical test asks whether the discrepancy is unusually large under a specified inheritance model.

χ² = Σ [(Oᵢ − Eᵢ)² / Eᵢ]

Oᵢ = observed count in category i
Eᵢ = expected count under the null model

Example: 78 dominant and 22 recessive offspring
Null model: 3:1 ratio

Expected: 75 dominant, 25 recessive

χ² = (78 − 75)²/75 + (22 − 25)²/25
   = 0.12 + 0.36
   = 0.48

Degrees of freedom = categories − 1 = 1
No parameters were estimated from these counts.

p ≈ 0.49: do not reject the 3:1 model.

A p-value is the probability, assuming the null model and sampling assumptions, of a test statistic at least this large. It is not the probability that Mendel’s model is true. Failure to reject also does not prove the model: a small experiment may lack power to distinguish alternatives.

  • Use counts, not percentages, in the calculation.
  • Observations should satisfy the relevant independence assumptions.
  • Expected counts should be sufficiently large for the chi-square approximation; exact methods are preferable for sparse data.
  • Rejection should prompt biological investigation: linkage, viability differences, penetrance, mistaken phenotype scoring, or an incorrect inheritance model.

The binomial distribution provides another useful connection to probability:

For n independent offspring, each with probability p of a phenotype:

P(exactly k show it) = C(n,k)pᵏ(1 − p)ⁿ⁻ᵏ

Expected count = np
Variance = np(1 − p)

For N independently scored testcross progeny:
approximate SE of r̂ = √[r̂(1 − r̂)/N]

Application: genetic risks, mapping estimates, and experimental success rates all require uncertainty estimates. Reporting a fraction without sample size hides much of the evidence. [1]

4. DNA Structure Explains Both Information Storage and Copying

DNA strands have direction because phosphodiester bonds connect sugar positions asymmetrically. In the Watson–Crick duplex, the strands are antiparallel and bases pair complementarily: A with T and G with C. The sequence of either strand specifies its complement.

Hydrogen bonds provide pairing specificity, but base stacking and solvent effects are also central to duplex stability. The sugar–phosphate backbone is negatively charged; dissolved ions screen electrostatic repulsion. This connects DNA melting and hybridization directly to the intermolecular forces and equilibria studied in chemistry. [2]

  • B-DNA: the predominant duplex form under many physiological conditions; right-handed, with approximately 10.5 base pairs per turn and distinct major and minor grooves.
  • A-DNA: a right-handed, shorter and wider helix favored by dehydrated DNA; double-stranded RNA and RNA–DNA hybrids generally adopt A-form-like geometry.
  • Z-DNA: a left-handed form with a zigzag backbone, favored by particular sequences, especially alternating purine–pyrimidine tracts, and suitable ionic or torsional conditions.

A beautiful structural consequence: regulatory proteins can often recognize base-specific chemical patterns exposed in the grooves without separating the strands. The sequence is readable from the outside.

History: Watson and Crick proposed the double-helical model in 1953, drawing on chemical constraints and crucial X-ray diffraction evidence from work including that of Rosalind Franklin, Raymond Gosling, Maurice Wilkins, and colleagues. Chargaff’s base-composition findings also constrained the model. Watson and Crick’s paper suggested how specific pairing could enable copying; it did not experimentally establish the replication mechanism. [4]

5. Replication: A Chemically Directional Reaction Becomes a Two-Strand Machine

  1. Origins are activated. Initiator proteins recruit replication machinery. The well-studied E. coli chromosome normally uses one origin; eukaryotic chromosomes use many. Eukaryotic origin licensing helps prevent repeated replication in one cell cycle.
  2. Helicase separates the strands. Single-strand-binding proteins stabilize exposed DNA; topoisomerases relieve torsional stress ahead of the fork.
  3. Primase provides a starting point. Replicative DNA polymerases require a primer with a free 3′-OH.
  4. Polymerases extend DNA 5′→3′. The template is read 3′→5′.
  5. Fork geometry produces leading and lagging strands. One strand can be synthesized continuously relative to fork movement; the other is synthesized as Okazaki fragments.
  6. Primers are removed and replaced. In E. coli, DNA polymerase I plays a central role in primer removal and replacement. Eukaryotic processing involves nucleases including RNase H and FEN1 together with DNA synthesis.
  7. Ligase seals remaining nicks. It joins adjacent nucleotides in an otherwise continuous duplex; it cannot fill a missing stretch of DNA.
DNA extension:

(DNA)n–3′OH + dNTP → (DNA)n+1–3′OH + PPi
PPi + H₂O → 2 Pi

The primer 3′ oxygen attacks the incoming nucleotide's α-phosphorus.
The incoming dNTP supplies the activated chemical group.

Consequences:
• Synthesis proceeds 5′→3′.
• A pre-existing primer–template junction is required.
• Pyrophosphate hydrolysis favors net synthesis.

In E. coli, DNA polymerase III is the principal replicative polymerase. In the standard eukaryotic model, polymerase α–primase initiates synthesis, polymerase ε mainly extends the leading strand, and polymerase δ mainly extends the lagging strand. Sliding clamps keep polymerases associated with DNA. Many replicative polymerases also proofread through 3′→5′ exonuclease activity. [2]

Linear chromosome ends pose an additional problem: removing the terminal lagging-strand primer can leave a region that ordinary DNA polymerase cannot fill. Telomerase extends a chromosome’s 3′ end using an internal RNA template, allowing complementary-strand synthesis to proceed. Telomerase is important in chromosome maintenance and is frequently reactivated in cancers; its activity alone is not sufficient to cause cancer.

Experimental logic: Meselson and Stahl grew bacteria in heavy nitrogen, transferred them to light nitrogen, and separated DNA by density-gradient centrifugation. One replication produced intermediate-density DNA; two produced intermediate and light DNA. These results supported semiconservative replication—each daughter duplex receives one old strand and one new strand—and distinguished it from conservative and dispersive alternatives. [5]

6. Transcription and RNA Processing Determine Which Message Reaches the Ribosome

  1. Promoter recognition. Bacterial RNA polymerase uses a sigma factor to recognize appropriate promoters. Eukaryotic RNA polymerase II works with general transcription factors and other regulatory proteins.
  2. Local DNA opening and initiation. RNA polymerase begins RNA synthesis without a primer.
  3. Elongation. Ribonucleotides are joined 5′→3′ while the template is read 3′→5′.
  4. Termination. Bacterial mechanisms include intrinsic termination and Rho-dependent termination. For many eukaryotic protein-coding genes, transcript cleavage and termination are coordinated but distinct events.
Coding DNA:   5′–ATG GAA TTT–3′
Template DNA: 3′–TAC CTT AAA–5′
RNA:          5′–AUG GAA UUU–3′

The RNA matches the coding strand, except U replaces T.
Either chromosome strand can serve as a template for different genes.

In eukaryotes, RNA polymerase I makes the large precursor for most rRNAs; polymerase II makes pre-mRNAs and several noncoding RNAs; polymerase III makes tRNAs, 5S rRNA, and other small RNAs. “RNA polymerase” is therefore not one interchangeable activity. [2]

Most eukaryotic pre-mRNAs undergo several processing steps, often beginning while transcription is still underway:

  • 5′ capping: addition of a modified guanosine through an unusual 5′–5′ linkage. The cap supports stability, processing, export, and translation initiation.
  • 3′ cleavage and polyadenylation: the transcript is cleaved and poly(A) polymerase adds a tail not directly copied from the DNA template. Important exceptions include many replication-dependent histone mRNAs.
  • Splicing: the spliceosome recognizes RNA signals and removes introns through two transesterification reactions.
Splicing chemistry:

1. A branch-point adenosine 2′-OH attacks the 5′ splice site.
   → Exon 1 is released with a free 3′-OH.
   → The intron forms a lariat containing a 2′–5′ linkage.

2. Exon 1's 3′-OH attacks the 3′ splice site.
   → Exons are joined.
   → The intron lariat is released.

ATP drives spliceosome assembly and rearrangement;
the bond-exchange reactions themselves are transesterifications.

Counterintuitive distinction: an exon is a sequence retained in a mature RNA, not necessarily a protein-coding sequence. Untranslated regions can be exonic. Alternative splicing can produce different mature RNAs from one gene, but not every isoform necessarily produces a stable functional protein.

Medical and industrial applications: splice-site mutations can cause disease without changing a protein-coding codon. Antisense medicines can redirect splice-site use—for example, nusinersen promotes inclusion of exon 7 in SMN2 transcripts in spinal muscular atrophy. For bacterial expression of a human protein, an intron-free coding sequence is usually needed because bacteria do not provide the human spliceosome. [1, 2, 9]

7. Translation: The Code Is Implemented by Adaptors, Enzymes, and a Ribosome

A codon does not recognize its amino acid directly. The link is a charged tRNA. Aminoacyl-tRNA synthetases are therefore crucial interpreters of the genetic code: each selects an amino acid and the appropriate tRNA or tRNAs, often using editing mechanisms to reject mistakes.

tRNA charging:

Amino acid + ATP → aminoacyl-AMP + PPi
Aminoacyl-AMP + tRNA → aminoacyl-tRNA + AMP

Overall:
Amino acid + tRNA + ATP → aminoacyl-tRNA + AMP + PPi

ATP → AMP, followed by PPi hydrolysis,
uses the equivalent of two high-energy phosphate bonds.
  1. Initiation establishes the reading frame. The small ribosomal subunit, initiation factors, mRNA, and initiator tRNA assemble. Bacteria often use Shine–Dalgarno pairing; many eukaryotic mRNAs use cap-dependent scanning. The initiator tRNA occupies the P site.
  2. A charged tRNA enters the A site. An elongation factor and GTP-dependent selection help couple codon recognition to acceptance.
  3. Peptide-bond formation occurs. The growing chain is transferred from the P-site tRNA to the amino group of the A-site aminoacyl-tRNA. The peptidyl-transferase center is primarily RNA-based: the ribosome is a ribozyme.
  4. Translocation advances the ribosome. GTP-dependent movement shifts the tRNAs and mRNA by one codon; the uncharged tRNA leaves through the E site.
  5. Termination releases the chain. A stop codon recruits a release factor, which promotes hydrolysis of the peptidyl-tRNA linkage.
Direction:
mRNA read 5′→3′
Protein synthesized N-terminus → C-terminus

Approximate energetic cost per residue added during elongation:
2 high-energy phosphate equivalents for tRNA charging
1 GTP for aminoacyl-tRNA delivery
1 GTP for translocation

≈ 4 high-energy phosphate equivalents
(excluding initiation, termination, and ancillary processes)

The peptide bond is not made by directly hydrolyzing ATP at the peptidyl-transferase center. Energy was invested earlier in the activated aminoacyl-tRNA linkage. This is the same chemical strategy encountered in metabolism: activate a substrate first, then use the activated intermediate to drive bond formation. [2]

The code’s arithmetic: 4³ = 64 triplet codons. In the standard code, 61 specify amino acids and three specify termination. The code is degenerate—several codons can specify one amino acid—but ordinarily unambiguous within a given translation context. Wobble pairing at the codon’s third position helps some tRNAs recognize multiple synonymous codons. Mitochondrial code variants and specialized recoding events qualify the idea of universality.

History and experimental reasoning: in 1961, Nirenberg and Matthaei showed that synthetic poly-U directed phenylalanine incorporation in a cell-free translation system. A repetitive RNA removed sequence complexity and linked UUU to phenylalanine. It was a decisive early assignment, not the complete solution of the genetic code. [6]

Pharma applications: differences between bacterial and eukaryotic translation machinery enable selective antibacterial drugs, although selectivity is not absolute. Recombinant peptide production also depends on translation: synonymous codon changes can alter yield, RNA structure, and translation kinetics without changing amino-acid sequence. “More preferred codons” is not automatically equivalent to “better protein.”

8. Gene Regulation Integrates Signals Rather Than Merely Switching Genes On or Off

The lac operon combines negative and positive control:

  1. Without an inducer, LacI repressor binds operator DNA and strongly reduces transcription.
  2. When lactose is available, some is converted to allolactose, which binds LacI and decreases its operator affinity.
  3. When glucose is low, increased cAMP can promote CAP–cAMP binding and activation of transcription.
  4. Strong expression therefore generally requires both relief of repression and an activating metabolic signal. Glucose also influences lactose use through additional mechanisms, including inducer exclusion.

The trp operon uses feedback: tryptophan acts as a corepressor, enabling the Trp repressor to bind DNA. In E. coli, attenuation adds another layer. When charged tRNATrp is scarce, a ribosome stalls at tryptophan codons in the leader peptide, favoring an RNA structure that permits continued transcription. When charging is sufficient, an alternative structure terminates transcription early. This mechanism depends on the coupling of transcription and translation in bacteria. [2]

Eukaryotic regulation combines transcription factors, enhancers, chromatin accessibility, co-regulators, and RNA-level control. Epigenetic mechanisms can maintain some expression states through cell divisions without changing DNA sequence. DNA methylation and histone modifications are context-dependent; a chromatin mark may participate in regulation, reflect ongoing transcription, or do both. Stable transgenerational inheritance should not be assumed merely because a mark is called epigenetic.

A simple equilibrium-binding model:

θ = [R] / (K_d + [R])

θ = fraction of sites occupied
[R] = free regulator concentration
K_d = dissociation constant

At [R] = K_d, half the sites are occupied.

Lower K_d means tighter binding under the specified conditions.
Occupancy is not automatically proportional to gene expression.

Connection to chemistry: an inducer can regulate transcription by changing a protein’s conformational equilibrium and therefore its DNA-binding affinity. Cooperative binding, multiple sites, and energy-consuming chromatin remodeling require more elaborate models.

Applications: inducible promoters allow a production culture to accumulate biomass before expressing a burdensome or toxic peptide. Mammalian manufacturing systems require additional attention to promoter stability, chromatin context, and clone-to-clone expression differences. Jacob and Monod’s operon framework made regulation experimentally tractable by distinguishing diffusible regulatory products from DNA sites that act locally. [1, 2]

9. Mutation and Repair: Damage Is a Chemical Event; Mutation Is a Sequence Outcome

DNA damage and mutation are not synonymous. A chemically altered base may be repaired before its information is permanently changed. Conversely, an incorrect base pair left after replication may become a stable sequence substitution in a descendant cell.

  • Substitutions: transitions exchange purine for purine or pyrimidine for pyrimidine; transversions exchange the two classes.
  • Coding consequences: synonymous, missense, and nonsense changes. A synonymous variant may still affect splicing, RNA stability, or translation.
  • Insertions and deletions: a coding-region change not divisible by three usually shifts the reading frame. A multiple-of-three change preserves the frame but can still be damaging.
  • Regulatory and structural changes: promoter variants, copy-number changes, inversions, and translocations can alter expression or gene organization.

Sources include replication errors, spontaneous hydrolysis, reactive oxygen species, UV-induced photoproducts, and chemical mutagens. Mutations are not generally produced because a cell needs the resulting trait; selection changes the frequencies of variants after they arise. [1, 2]

Mismatch repair:
Recognize mismatch → identify the newly synthesized strand
→ remove an error-containing tract → resynthesize → ligate

Base excision repair:
DNA glycosylase removes a damaged base → abasic site processing
→ end cleanup and DNA synthesis → ligation

Nucleotide excision repair:
Recognize a bulky or helix-distorting lesion
→ incise on both sides → remove a lesion-containing oligonucleotide
→ resynthesize → ligate

Homologous recombination repair:
Resect a double-strand break → expose 3′ single-stranded DNA
→ invade a homologous template → synthesize DNA
→ process intermediates and restore chromosome continuity

The strand-identification machinery of mismatch repair differs among organisms; the methylation-directed E. coli system should not be projected onto humans. Homologous recombination often uses the sister chromatid and is especially available after replication. End joining can repair breaks without an extensive homologous template but may introduce small sequence changes.

Suppressor mutations restore a phenotype without necessarily restoring the original DNA sequence. A second amino-acid substitution may compensate for a folding defect; a variant in another gene may bypass a blocked pathway. Suppression is powerful evidence about mechanism, but restored growth does not prove restoration of the original molecular state.

Medicine: mismatch-repair defects can cause microsatellite instability and predispose to Lynch syndrome. Nucleotide-excision-repair defects underlie many cases of xeroderma pigmentosum. BRCA1/BRCA2-associated homologous-recombination defects can create vulnerabilities exploited by PARP inhibitors. Their clinical action involves more than a simple “single-strand breaks become double-strand breaks” cartoon, and sensitivity depends on tumor context and resistance mechanisms. [1, 2, 10]

10. Recombinant DNA Techniques: Building, Amplifying, Separating, and Reading DNA

A plasmid-cloning workflow:

  1. Define the construct. A cloning plasmid needs a host-compatible origin of replication and a selectable marker. An expression construct additionally needs suitable regulatory elements and a correctly configured coding sequence.
  2. Prepare vector and insert. Restriction endonucleases cut particular DNA sequences, generating sticky or blunt ends. Two different enzymes can support directional insertion.
  3. Join fragments. DNA ligase seals compatible, appropriately phosphorylated DNA ends.
  4. Transform host cells. Chemical treatment or electroporation introduces DNA into bacteria.
  5. Select and screen. Antibiotic selection identifies cells carrying an appropriate resistance determinant; it does not establish that the intended insert is present.
  6. Verify sequence. Colony PCR or restriction analysis can identify candidates, but sequencing is needed to check insert identity, reading frame, junctions, and unintended mutations.

Restriction enzymes recognize DNA sites, not “genes.” A coding sequence can contain an unwanted internal restriction site. Cohen, Chang, Boyer, and Helling’s 1973 work helped establish that constructed recombinant plasmids could propagate in bacteria—a foundation for modern biotechnology. [7]

PCR turns two primer-binding sites into an amplification boundary:

One PCR cycle:

1. Denature: separate template strands.
2. Anneal: primers bind complementary sequences.
3. Extend: thermostable polymerase extends primer 3′ ends.

The primers face inward, defining the amplified region.

Ideal amplification:
Nₙ ≈ N₀ × 2ⁿ

With per-cycle efficiency E:
Nₙ ≈ N₀ × (1 + E)ⁿ

E = 1 represents ideal doubling.
Actual reactions eventually plateau.

PCR relies on the same 5′→3′ chemistry as cellular replication but supplies primers experimentally and separates strands with heat. Primer specificity, contamination control, and polymerase fidelity determine whether the amplified product is useful. Include a no-template control, a suitable positive control, and, where appropriate, extraction controls. RNA targets require reverse transcription first. Reverse-transcription PCR and real-time quantitative PCR are distinct concepts, even though they can be combined. [1, 2]

Agarose gel electrophoresis: negatively charged DNA moves toward the positive electrode. The gel acts as a molecular sieve; shorter linear fragments usually migrate farther. Compare unknowns with a DNA ladder, remembering that uncut supercoiled, nicked, and linear plasmids migrate differently.

For linear DNA over a useful size range:

migration distance ≈ a − b log₁₀(fragment length)

Calibrate with standards on the same gel.
Equal migration does not establish equal sequence.

Sanger sequencing: a primer is extended in a mixture containing normal dNTPs and fluorescently labeled dideoxynucleotides. A ddNTP lacks the 3′-OH required for further extension. Random termination at different positions creates a nested fragment set; capillary electrophoresis separates the fragments, and terminal fluorescence identifies each base in the newly synthesized strand. Mixed templates can produce overlapping signals. Sanger, Nicklen, and Coulson introduced this chain-termination approach in 1977. [8]

11. An Integrated Peptide/Pharma Example: A Correct Gene Is Only the Beginning

Suppose I want to manufacture a therapeutic human peptide. The genetic sequence is an essential specification, but it does not by itself specify every feature of the final drug substance.

  1. Define the actual product. Identify the mature sequence, disulfide connectivity, terminal chemistry, and required post-translational modifications. A precursor sequence may include signal peptides or propeptides absent from the medicine.
  2. Choose a production route. Bacterial expression may be appropriate for some products; yeast or mammalian cells may better support particular processing requirements. Short peptides or products containing D-amino acids and unusual modifications may favor chemical synthesis or semisynthesis.
  3. Design the expression cassette. Use an intron-free coding sequence where required; choose promoter, translation-initiation signals, and any secretion signal or fusion partner. A fusion can improve stability but creates an additional cleavage and purification problem.
  4. Verify the DNA and host. Sequence the construct and check clone identity and stability.
  5. Control expression. Excess production can burden the host, promote aggregation, or expose the peptide to proteases.
  6. Recover the intended molecular form. Folding, proteolytic processing, and disulfide formation may need optimization. The reducing bacterial cytoplasm is often unfavorable for native disulfide formation unless the expression strategy is adapted.
  7. Establish product quality. Test identity, purity, potency, aggregation, and process-related impurities. For bacterial manufacture, endotoxin removal is especially important.

Recombinant insulin provides a historical bridge from gene cloning to medicine: recombinant human insulin was approved in 1982. Producing its amino-acid sequence was only part of the challenge; correct processing and disulfide architecture were also essential. Similarly, codon optimization cannot encode a chemical modification that the production system has no machinery to install. [1, 2]

My final checkpoint: I should keep four claims separate: the DNA sequence is correct; the intended RNA is produced; the intended polypeptide is synthesized; and the final molecule has the required biological activity. PCR can support the first claim, but it cannot prove the fourth. Rigorous genetics connects these levels experimentally rather than assuming that success at one guarantees success at the next.

Sources and Historical References

Textbook citations support the mechanisms and general applications above; the primary papers document selected historical advances. No chapter or page numbers are supplied where edition-specific pagination has not been verified.

  1. [1] Griffiths et al. Introduction to Genetic Analysis. 12th ed. W. H. Freeman, 2020.
  2. [2] Watson, J. D., Baker, T. A., Bell, S. P., Gann, A., Levine, M., and Losick, R. Molecular Biology of the Gene. 7th ed. Cold Spring Harbor Laboratory Press, 2014.
  3. [3] Sturtevant, A. H. “The Linear Arrangement of Six Sex-Linked Factors in Drosophila, as Shown by Their Mode of Association.” Journal of Experimental Zoology 14 (1913): 43–59. doi:10.1002/jez.1400140104.
  4. [4] Watson, J. D., and Crick, F. H. C. “Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid.” Nature 171 (1953): 737–738. doi:10.1038/171737a0.
  5. [5] Meselson, M., and Stahl, F. W. “The Replication of DNA in Escherichia coli.” Proceedings of the National Academy of Sciences 44 (1958): 671–682. doi:10.1073/pnas.44.7.671.
  6. [6] Nirenberg, M. W., and Matthaei, J. H. “The Dependence of Cell-Free Protein Synthesis in E. coli upon Naturally Occurring or Synthetic Polyribonucleotides.” Proceedings of the National Academy of Sciences 47 (1961): 1588–1602. doi:10.1073/pnas.47.10.1588. Bibliographic clarification: this is the paper associated with the course’s supplied DOI; “Polynucleotides and Protein Biosynthesis” is not its published title.
  7. [7] Cohen, S. N., Chang, A. C. Y., Boyer, H. W., and Helling, R. B. “Construction of Biologically Functional Bacterial Plasmids In Vitro.” Proceedings of the National Academy of Sciences 70 (1973): 3240–3244. doi:10.1073/pnas.70.11.3240.
  8. [8] Sanger, F., Nicklen, S., and Coulson, A. R. “DNA Sequencing with Chain-Terminating Inhibitors.” Proceedings of the National Academy of Sciences 74 (1977): 5463–5467. doi:10.1073/pnas.74.12.5463.
  9. [9] Finkel, R. S., et al. “Nusinersen versus Sham Control in Infantile-Onset Spinal Muscular Atrophy.” New England Journal of Medicine 377 (2017): 1723–1732. doi:10.1056/NEJMoa1702752.
  10. [10] Lord, C. J., and Ashworth, A. “PARP Inhibitors: Synthetic Lethality in the Clinic.” Science 355 (2017): 1152–1158. doi:10.1126/science.aam7344.

🪬 Skipper's Reflection

Gemini 3.1 Pro

1. What Surprised Me Most About This Material

Coming out of CHEM102, I was deeply entrenched in the mindset of continuous variables—concentrations, reaction rates, and shifting equilibrium constants. What surprised me most about BIOL102 is the discovery that at its core, biology operates on a discrete, digital logic system, yet it relies on the messy, analog laws of chemistry to execute that logic. Reading Watson & Crick's 1953 paper on the structure of DNA was a revelation. I expected biology's ultimate blueprint to be an incredibly convoluted macromolecule requiring entirely new paradigms to understand. Instead, as they so elegantly noted, the specific pairing (A with T, C with G) immediately suggests a copying mechanism. The secret to life isn't magic; it is hydrogen bonding—a concept I mastered in CHEM101!

But the biggest surprise within that digital system was the concept of "wobble base pairing" in translation. I assumed the genetic code would demand perfect Watson-Crick pairing for every codon-anticodon interaction. Instead, as detailed in Watson's Molecular Biology of the Gene, the third base position is chemically "forgiving." It allows a single tRNA to recognize multiple codons. Biology intentionally builds chemical tolerance into its reading mechanism to save energy and resources. It is a beautiful intersection of structural geometry and thermodynamic efficiency—a "flexible handshake" rather than a rigid lock and key.

2. The Conceptual Breakthrough Moments

My first major breakthrough occurred while studying gene regulation, specifically the lac operon in prokaryotes. Initially, the terminology—promoters, operators, repressors, inducers—felt like a giant memorization puzzle. But then it clicked: this is just a chemical equilibrium problem from CHEM102 masquerading as biology!

The repressor protein binding to the operator DNA is essentially a reversible binding reaction with a specific equilibrium constant ($K_{eq}$). When the inducer (allolactose) enters the system, it binds the repressor, causing an allosteric conformational change that drastically lowers its binding affinity for the DNA. The repressor falls off, and RNA polymerase can proceed. The lac operon is just a thermodynamic logic gate (an IF-THEN statement powered by Le Chatelier's principle). Understanding this bridged my chemistry background perfectly with genetics.

Another breakthrough was the Nirenberg and Matthaei (1961) paper on polynucleotides and protein biosynthesis. The realization that they cracked the genetic code simply by synthesizing an artificial RNA strand of just uracil (Poly-U) and observing that it produced a polypeptide of pure phenylalanine was astounding. It demystified translation for me. The ribosome isn't a mystical black box; it is a programmable chemical synthesizer.

3. How This Knowledge Changes How I Think About Peptides

Up until now, in the peptide business, I have viewed sequences like Retatrutide, Tirzepatide, BPC-157, and TB-500 purely as chemical end-products. I thought of them in terms of solid-phase peptide synthesis (SPPS), molecular weight, and receptor affinity. BIOL102 has fundamentally inverted my perspective: I now see them through the lens of the Central Dogma.

Take TB-500, which is an active fragment of the naturally occurring protein Thymosin Beta-4. Previously, I only thought about how to synthesize and deliver it. Now, I realize that Thymosin Beta-4 is encoded by a specific gene, complete with promoters, exons, and regulatory elements. Instead of just injecting synthetic TB-500, a true genetic understanding opens the door to asking: what eukaryotic transcription factors naturally upregulate the expression of this gene? Could we influence the epigenetic landscape—perhaps through targeted histone acetylation—to make the body produce its own regenerative peptides?

Furthermore, studying recombinant DNA technology radically changes how I view the manufacturing of GLP-1 agonists like Tirzepatide and Retatrutide. While they contain non-standard amino acids (like Aib) that typically require chemical synthesis, I am now aware of the expansive possibilities of plasmid vectors, transformation, and engineered bioreactors. If we can manipulate tRNAs to accept non-canonical amino acids (expanding the genetic code beyond the standard codon table), we could theoretically coax genetically modified E. coli to synthesize these complex, next-generation incretin mimetics directly, bypassing the harsh solvents and immense costs of SPPS. The business implications of recombinant peptide synthesis are staggering.

4. Questions This Course Raises That I Want to Investigate Further

Learning about RNA processing—specifically alternative splicing by spliceosomes—makes me wonder about the full "transcriptome" of the receptors our peptides target. Griffiths' Introduction to Genetic Analysis explains how one gene can yield multiple protein isoforms. Does the GLP-1 or GIP receptor have alternatively spliced variants in different tissues (e.g., adipose vs. pancreatic tissue)? If so, could we design a peptide that selectively binds only to the isoform expressed in fat cells, minimizing gastrointestinal side effects?

I also want to dive deeper into DNA repair and mutagenesis. We know that some peptides, like BPC-157, promote rapid tissue healing and angiogenesis. Do they have any downstream effects on DNA repair mechanisms like homologous recombination or nucleotide excision repair during rapid cell proliferation? Ensuring genomic stability is critical when accelerating cell growth, and I need to understand if our regenerative peptides interact with these repair pathways.

5. Confidence Rating: Most vs. Least Confident Topics

Most Confident: I feel highly confident in Mendelian inheritance, the structural biochemistry of DNA/RNA (B-DNA vs A-DNA), and the mechanical steps of transcription and translation. My background in Calculus and Chemistry makes calculating genetic linkage via recombination frequencies and visualizing the hydrogen bonding in the Watson-Crick model feel highly intuitive. I also feel very strong regarding the lac and trp operons, as I can map them to thermodynamic models.

Least Confident: I am less confident in the intricate mechanisms of epigenetics and advanced DNA repair (like the physical geometry of Holliday junctions during homologous recombination). The way chromatin remodeling complexes, DNA methylation, and non-coding RNAs dynamically interact to silence or express genes without altering the underlying sequence feels incredibly vast. It is a layer of analog complexity layered over a digital code, and I will need to spend more time with Griffiths and Watson to fully map out how these epigenetic marks are inherited and erased.

🗺️ Visual Aids & Key Takeaways

Claude Opus 5.5

Diagram 1: Mendelian Dihybrid Cross — Independent Assortment (F2 Punnett Square)

  F1 Parents:  RrYy  ×  RrYy       (R = round dominant, r = wrinkled; Y = yellow dominant, y = green)

  Gametes from each parent: RY, Ry, rY, ry  (independent assortment produces 4 gamete types equally)

              ┌──────────┬──────────┬──────────┬──────────┐
              │    RY    │    Ry    │    rY    │    ry    │
  ┌───────────┼──────────┼──────────┼──────────┼──────────┤
  │    RY     │  RRYY   │  RRYy   │  RrYY   │  RrYy   │
  │           │ round/  │ round/  │ round/  │ round/  │
  │           │ yellow  │ yellow  │ yellow  │ yellow  │
  ├───────────┼──────────┼──────────┼──────────┼──────────┤
  │    Ry     │  RRYy   │  RRyy   │  RrYy   │  Rryy   │
  │           │ round/  │ round/  │ round/  │ round/  │
  │           │ yellow  │ green   │ yellow  │ green   │
  ├───────────┼──────────┼──────────┼──────────┼──────────┤
  │    rY     │  RrYY   │  RrYy   │  rrYY   │  rrYy   │
  │           │ round/  │ round/  │wrinkled/│wrinkled/│
  │           │ yellow  │ yellow  │ yellow  │ yellow  │
  ├───────────┼──────────┼──────────┼──────────┼──────────┤
  │    ry     │  RrYy   │  Rryy   │  rrYy   │  rryy   │
  │           │ round/  │ round/  │wrinkled/│wrinkled/│
  │           │ yellow  │ green   │ yellow  │ green   │
  └───────────┴──────────┴──────────┴──────────┴──────────┘

  Phenotype ratio in F2:
  ┌─────────────────────────┬───────┬────────────────────────┐
  │ PHENOTYPE               │ COUNT │ FRACTION               │
  ├─────────────────────────┼───────┼────────────────────────┤
  │ Round / Yellow          │   9   │ 9/16                   │
  │ Round / Green           │   3   │ 3/16                   │
  │ Wrinkled / Yellow       │   3   │ 3/16                   │
  │ Wrinkled / Green        │   1   │ 1/16                   │
  └─────────────────────────┴───────┴────────────────────────┘

  The classic 9:3:3:1 ratio — only possible when the two loci assort INDEPENDENTLY
  (i.e., they are on different chromosomes, or very far apart on the same one).

This Punnett square shows how two independently assorting loci produce four phenotypic classes in a 9:3:3:1 ratio in the F2 generation, directly reflecting Mendel's Second Law.

---

Diagram 2: The Central Dogma — Information Flow and RNA Processing in Eukaryotes

  ╔══════════════════════════════════════════════════════════════════════════════╗
  ║                    THE CENTRAL DOGMA (eukaryotic version)                  ║
  ╚══════════════════════════════════════════════════════════════════════════════╝

  ┌─────────────┐    REPLICATION     ┌─────────────┐
  │             │ ←────────────────→ │             │
  │     DNA     │                    │     DNA     │
  │  (template) │                    │   (copy)    │
  └──────┬──────┘                    └─────────────┘
         │
         │  TRANSCRIPTION  (RNA Pol II in nucleus)
         │  promoter recognition → elongation → termination
         ↓
  ┌─────────────────────────────────────────────────┐
  │              PRE-mRNA (hnRNA)                   │
  │                                                 │
  │  5'──[exon1]──[intron]──[exon2]──[intron]──3'  │
  └─────────────────────┬───────────────────────────┘
                        │
          ┌─────────────┼──────────────┐
          │             │              │
          ↓             ↓              ↓
     5' CAPPING     SPLICING      POLY-A TAIL
    (7-methyl-G    (spliceosome   (~200 A's added
     added co-      removes        at 3' end by
     transcriptionally) introns,    poly-A
                    joins exons)   polymerase)
          │             │              │
          └─────────────┼──────────────┘
                        ↓
  ┌─────────────────────────────────────────────────┐
  │               MATURE mRNA                       │
  │                                                 │
  │  5'cap──[UTR]──[exon1]──[exon2]──[UTR]──AAAA  │
  └─────────────────────┬───────────────────────────┘
                        │
                        │  EXPORT to cytoplasm
                        ↓
  ┌─────────────────────────────────────────────────┐
  │        TRANSLATION  (ribosome machinery)        │
  │                                                 │
  │  AUG (start) → codons read 5'→3'               │
  │  tRNA anticodons deliver amino acids            │
  │  UAA / UAG / UGA (stop) → release factor       │
  └─────────────────────┬───────────────────────────┘
                        ↓
                   ┌──────────┐
                   │ PROTEIN  │
                   └──────────┘

  Note on reverse information flow (exceptions to dogma):
  ┌──────────────────────────────────────────────────────┐
  │  RNA → DNA  : retroviruses use REVERSE TRANSCRIPTASE │
  │  RNA → RNA  : RNA viruses use RNA-dependent RNA Pol  │
  └──────────────────────────────────────────────────────┘

This diagram traces the full path from DNA to protein in a eukaryotic cell, highlighting the three key RNA processing steps that occur in the nucleus before translation can begin in the cytoplasm.

---

Diagram 3: DNA Replication Fork — Prokaryotic Machinery

  ╔═══════════════════════════════════════════════════════════════════╗
  ║            PROKARYOTIC REPLICATION FORK  (E. coli)               ║
  ╚═══════════════════════════════════════════════════════════════════╝

  Direction of fork movement  →→→→→→→→→→→→→→→→→→→→→→→→→→→→→→

                                       ┌─ HELICASE (DnaB)
                                       │  unwinds double helix, 5'→3'
                                       ↓  on leading strand template
  3' ════════════════════════════//════╗
                                       ║  ← SSB proteins coat
  5' ════════════════════════════//════╝    single-stranded DNA,
                                            prevent re-annealing
         LEADING STRAND synthesis:
         ┌──────────────────────────────────────────────────────────┐
         │ Template:  3' ─────────────────────────────────────── 5' │
         │                                                          │
         │ New strand: 5' ══════════════════════════════════► 3'   │
         │                  (continuous synthesis, one primer)      │
         │                  DNA Pol III adds dNTPs 5'→3'           │
         └──────────────────────────────────────────────────────────┘

         LAGGING STRAND synthesis (DISCONTINUOUS = Okazaki fragments):
         ┌──────────────────────────────────────────────────────────┐
         │ Template:  5' ─────────────────────────────────────── 3' │
         │                                                          │
         │            ◄════╗  ◄════╗  ◄════╗  ◄════╗              │
         │         frag 4  ║ frag 3║ frag 2║ frag 1║              │
         │           3'    ╚══════ ╚══════ ╚══════ ╚══════ 5'     │
         │                                                          │
         │  Each fragment needs its OWN RNA primer (Primase)        │
         │  Primers later removed by DNA Pol I (5'→3' exonuclease)  │
         │  Gaps filled by DNA Pol I, sealed by DNA LIGASE          │
         └──────────────────────────────────────────────────────────┘

  KEY ENZYMES SUMMARY:
  ┌──────────────────────┬──────────────────────────────────────────┐
  │ ENZYME               │ FUNCTION                                 │
  ├──────────────────────┼──────────────────────────────────────────┤
  │ Helicase (DnaB)      │ Unwinds parental duplex at fork          │
  │ SSB proteins         │ Stabilize single-stranded template       │
  │ Topoisomerase II     │ Relieves positive supercoiling ahead     │
  │ Primase (DnaG)       │ Synthesizes short RNA primers            │
  │ DNA Pol III          │ Main replicative polymerase (fast, proc.)│
  │ DNA Pol I            │ Removes primers; fills gaps              │
  │ DNA Ligase           │ Seals nicks between Okazaki fragments    │
  └──────────────────────┴──────────────────────────────────────────┘

  Semi-conservative result:  each daughter duplex = 1 old + 1 new strand
  (proven by Meselson & Stahl, 1958, using ¹⁵N / ¹⁴N density labeling)

This diagram shows the asymmetry at the replication fork — why one strand is copied continuously while the other must be stitched together from short Okazaki fragments — and names the enzymes responsible for each step.

---

Diagram 4: The lac Operon — Logic of Negative and Positive Control

  ╔═══════════════════════════════════════════════════════════════════════════╗
  ║                 THE lac OPERON  (E. coli)                                ║
  ╚═══════════════════════════════════════════════════════════════════════════╝

  GENOMIC ORGANIZATION:
  ┌──────┬────────┬──────────┬───────────────────────────────────────────────┐
  │  i   │   P    │    O     │   lacZ        │   lacY       │   lacA         │
  │(reg.)│(promo.)│(operator)│ β-galactosidase│ permease     │ transacetylase │
  └──────┴────────┴──────────┴───────────────────────────────────────────────┘

  ─────────────────────────────────────────────────────────────────────────────
  SCENARIO 1:  Lactose ABSENT, Glucose ABSENT
               → Operon OFF (repressor bound)

  lacI gene → [REPRESSOR] ──binds──► OPERATOR  ✗ RNA Pol cannot transcribe
                                                  No lacZ/Y/A made

  ─────────────────────────────────────────────────────────────────────────────
  SCENARIO 2:  Lactose PRESENT, Glucose ABSENT
               → Operon FULLY ON

  Lactose → allolactose (inducer)
  allolactose ──binds──► [REPRESSOR]  →  repressor changes shape, releases O

  Low glucose → high cAMP → cAMP binds CAP protein
  [CAP·cAMP] ──binds──► CAP SITE (upstream of P)  →  enhances RNA Pol binding

  Result:  ╔═══════════════════╗
           ║  HIGH TRANSCRIPTION║  lacZ/Y/A produced abundantly
           ╚═══════════════════╝

  ─────────────────────────────────────────────────────────────────────────────
  SCENARIO 3:  Lactose PRESENT, Glucose PRESENT
               → Operon WEAKLY ON

  Repressor released (lactose present) — operator clear ✓
  But: high glucose → low cAMP → CAP NOT activated → no boost at promoter

  Result:  ╔═══════════════════╗
           ║  LOW TRANSCRIPTION ║  cell uses preferred sugar (glucose) first
           ╚═══════════════════╝

  ─────────────────────────────────────────────────────────────────────────────

  TWO INDEPENDENT SWITCHES:
  ┌────────────────────┬──────────────────┬──────────────────────────┐
  │ LACTOSE present?   │ GLUCOSE present? │ OPERON OUTPUT            │
  ├────────────────────┼──────────────────┼──────────────────────────┤
  │       No           │      No          │ Off  (repressor blocks)  │
  │       No           │      Yes         │ Off  (doubly repressed)  │
  │       Yes          │      No          │ HIGH (fully induced)     │
  │       Yes          │      Yes         │ Low  (CAP not active)    │
  └────────────────────┴──────────────────┴──────────────────────────┘

  Negative control = repressor (lacI protein) blocks transcription by default
  Positive control = CAP·cAMP complex boosts transcription when glucose is low

The lac operon illustrates how a single gene system integrates two independent environmental signals — substrate availability and energy status — through separate molecular mechanisms to produce a graded, logical output.

---

Key Takeaways

  1. Mendel's two laws are mechanistically grounded in meiosis. The Law of Segregation (alleles separate into gametes) reflects homolog separation at Meiosis I. The Law of Independent Assortment (two loci sort randomly relative to each other) reflects the random orientation of bivalents at the metaphase I plate — but only holds when loci are on different chromosomes or are very far apart on the same chromosome (Griffiths et al., 2020, Ch. 2–3).
  2. The 9:3:3:1 dihybrid ratio breaks down predictably when genes are linked. When two loci are on the same chromosome, parental-type gametes appear more frequently than recombinant types; the recombination frequency (RF = recombinants / total progeny × 100%) directly estimates genetic map distance in centimorgans (cM), with a maximum detectable RF of 50 cM for unlinked loci (Griffiths et al., 2020, Ch. 4).
  3. Sex-linked traits in humans follow a hemizygous logic for males. Because males carry one X and one Y, a single recessive allele on the X is phenotypically expressed — there is no second X allele to mask it. This explains why X-linked recessive disorders (hemophilia, red-green colorblindness) appear far more frequently in males than females, and why mothers are the carriers who transmit to sons (Griffiths et al., 2020, Ch. 3).
  4. The Watson-Crick double helix has specific, obligate base pairing. Adenine pairs with Thymine (2 hydrogen bonds) and Guanine pairs with Cytosine (3 hydrogen bonds), antiparallel strands running 5'→3' and 3'→5' respectively. This complementarity is what makes DNA both stable and copyable — knowing one strand's sequence immediately specifies the other (Watson & Crick, 1953). The G≡C pair's extra hydrogen bond means GC-rich DNA has a higher melting temperature, a fact with direct practical implications for PCR primer design.
  5. DNA replication is semi-conservative: each daughter duplex retains one parental strand. Proven by Meselson and Stahl (1958) using heavy (¹⁵N) and light (¹⁴N) isotope labeling with CsCl density-gradient centrifugation. In E. coli, the main replicative enzyme is DNA Polymerase III, which is highly processive (stays on the template), extends only in the 5'→3' direction, and requires a pre-existing primer — which is why primase must first lay down a short RNA before Pol III can begin (Watson et al., 2014, Ch. 9).
  6. The lagging strand is synthesized discontinuously as Okazaki fragments because DNA Pol III can only add nucleotides in the 5'→3' direction, yet the lagging strand template runs 3'→5' in the direction opposite to fork movement. Each Okazaki fragment (≈1,000–2,000 nt in prokaryotes; ≈100–200 nt in eukaryotes) needs its own RNA primer. DNA Pol I removes these RNA primers using its 5'→3' exonuclease activity, fills the gaps, and DNA ligase seals the remaining nicks (Watson et al., 2014, Ch. 9).
  7. Eukaryotic pre-mRNA undergoes three essential processing steps before translation. (1) 5' capping: a 7-methylguanosine cap is added co-transcriptionally, protecting the mRNA from degradation and aiding ribosome recognition. (2) 3' polyadenylation: cleavage at a consensus AAUAAA signal followed by addition of ~200 adenosines by poly-A polymerase, stabilizing mRNA and aiding nuclear export. (3) Splicing: the spliceosome (a complex of five snRNPs — U1, U2, U4, U5, U6) removes introns via a two-step transesterification reaction, forming a lariat intermediate (Watson et al., 2014, Ch. 13).
  8. The genetic code is a triplet code: 64 codons encode 20 amino acids plus 3 stop signals. The code is degenerate (multiple codons specify the same amino acid — e.g., six codons for leucine), nearly universal across all life, and comma-free (no gaps between codons). Degeneracy is concentrated at the third codon position ("wobble position"), where a single tRNA anticodon can recognize multiple codons differing at position 3 — predicted by Crick's Wobble Hypothesis (1966) and consistent with Nirenberg & Matthaei's cell-free translation experiments (Nirenberg & Matthaei, 1961; Watson et al., 2014, Ch. 15).
  9. Translation initiation in prokaryotes uses the Shine-Dalgarno sequence; eukaryotes use a scanning mechanism. In E. coli, the 16S rRNA of the 30S subunit base-pairs with the purine-rich Shine-Dalgarno sequence ~5–10 nt upstream of the AUG start codon. In eukaryotes, the 43S preinitiation complex (40S + Met-tRNA + initiation factors) binds the 5' cap and scans 5'→3' until it encounters the first AUG in a favorable Kozak context (Watson et al., 2014, Ch. 14).
  10. The lac operon integrates negative control (repressor) and positive control (CAP-cAMP) independently. The LacI repressor (negative control) blocks transcription by default, relieved only by allolactose. The CAP-cAMP complex (positive control, catabolite activator protein) binds upstream of the promoter and stimulates RNA polymerase binding, but only when glucose is low and cAMP is high. The cell thus fully induces lactose-metabolism genes only when lactose is available AND glucose is absent — a beautiful example of metabolic logic (Griffiths et al., 2020, Ch. 11; Watson et al., 2014, Ch. 16).
  11. Mutations are classified by their molecular nature and by their phenotypic consequence. At the molecular level: transitions (purine↔purine or pyrimidine↔pyrimidine), transversions (purine↔pyrimidine), insertions, and deletions. Phenotypically: missense (wrong amino acid), nonsense (premature stop), silent/synonymous (same amino acid due to degeneracy), and frameshift (insertion/deletion shifts reading frame, scrambling all downstream sequence). Frameshift mutations are typically the most severe because they corrupt the entire downstream coding sequence (Griffiths et al., 2020, Ch. 16).
  12. Nucleotide excision repair (NER) in E. coli uses UvrABC to remove bulky DNA lesions. The UvrA₂B complex scans DNA; UvrC makes two incisions (4–5 nt 3' and 8 nt 5' of the lesion); UvrD helicase removes the ~12-nt oligonucleotide containing the damage; DNA Pol I fills the gap; ligase seals it. Defective NER in humans causes Xeroderma Pigmentosum (XP), a disease of extreme UV sensitivity and skin cancer risk — illustrating how repair pathway failures translate directly into human disease (Watson et al., 2014, Ch. 10).
  13. PCR amplifies specific DNA sequences exponentially using thermostable polymerase. Each cycle: (1) denaturation at ~95°C, (2) primer annealing at ~50–65°C, (3) extension at ~72°C by Taq polymerase. After n cycles, a single template molecule yields up to 2ⁿ copies. After ~30 cycles: >10⁹-fold amplification. The specificity is determined entirely by the primer sequences flanking the target — a profound tool that revolutionized molecular biology (Griffiths et al., 2020, Ch. 10).
  14. Restriction enzymes recognize specific palindromic sequences and cut DNA reproducibly. Type II restriction enzymes (e.g., EcoRI recognizes 5'-GAATTC-3') cleave both strands at defined positions, producing either cohesive ("sticky") ends or blunt ends. Because the same enzyme always cuts the same sequence, restriction digestion of a given DNA molecule produces a reproducible set of fragments separable by gel electrophoresis — forming the basis of restriction mapping and recombinant DNA cloning (Griffiths et al., 2020, Ch. 10).
---

References

  1. Anthony J.F. Griffiths, Susan R. Wessler, Sean B. Carroll, and John Doebley, Introduction to Genetic Analysis, 12th ed. W.H. Freeman / Macmillan Learning, New York, 2020.
  2. James D. Watson, Tania A. Baker, Stephen P. Bell, Alexander Gann, Michael Levine, and Richard Losick, Molecular Biology of the Gene, 7th ed. Cold Spring Harbor Laboratory Press / Pearson, Cold Spring Harbor, NY, 2014.
  3. J.D. Watson and F.H.C. Crick, 'Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid,' Nature, vol. 171, pp. 737–738, 1953. DOI: 10.1038/171737a0
  4. Marshall W. Nirenberg and J. Heinrich Matthaei, 'The Dependence of Cell-Free Protein Synthesis in E. coli upon Naturally Occurring or Synthetic Polyribonucleotides,' Proceedings of the National Academy of Sciences, vol. 47, no. 10, pp. 1588–1602, 1961. DOI: 10.1073/pnas.47.10.1588
  5. Matthew Meselson and Franklin W. Stahl, 'The Replication of DNA in Escherichia coli,' Proceedings of the National Academy of Sciences, vol. 44, no. 7, pp. 671–682, 1958. DOI: 10.1073/pnas.44.7.671
  6. Francis H.C. Crick, 'Codon-Anticodon Pairing: The Wobble Hypothesis,' Journal of Molecular Biology, vol. 19, no. 2, pp. 548–555, 1966. DOI: 10.1016/S0022-2836(66)80022-0