Molecular Basis of Inheritance

The chapter builds outward from molecule to organism. It opens with the physical structure of DNA — the polynucleotide chain, the double helix, and how a two-metre-long molecule folds into a micrometre-scale nucleus via histones, nucleosomes and chromatin. It retraces the historical proof that DNA is the genetic material, from Griffith through Avery-MacLeod-McCarty to Hershey-Chase, and asks why DNA displaced RNA as the dominant hereditary molecule even though RNA came first. The engine room is the central dogma: semiconservative replication (Meselson-Stahl), transcription within a defined transcription unit, the triplet genetic code, and translation via tRNA adapters and the ribosome. It closes with regulation (the lac operon) and two landmark applications — the Human Genome Project and DNA fingerprinting. Three worked derivations trace the Meselson-Stahl logic, the gene-to-polypeptide path in a eukaryotic cell, and the lac operon switch.

The DNA molecule: from nucleotide to double helix

A nucleotide has three components: a nitrogenous base, a pentose sugar (deoxyribose in DNA, ribose in RNA), and a phosphate group. Nitrogenous bases are purines (adenine, guanine) or pyrimidines (cytosine, thymine, uracil); cytosine occurs in both DNA and RNA, thymine only in DNA, and uracil replaces it in RNA. A base attached to the sugar's $1'$ carbon gives a nucleoside; adding a phosphate at the $5'$ carbon completes the nucleotide. Successive nucleotides join by a $3'$-$5'$ phosphodiester linkage into a polynucleotide chain, with a free phosphate at the $5'$ end and a free hydroxyl at the $3'$ end, bases projecting sideways. RNA differs from DNA by an extra $2'$-OH on its sugar, and by uracil in place of thymine. The double-stranded structure was proposed in 1953 by James Watson and Francis Crick, building on the X-ray diffraction images of Maurice Wilkins and Rosalind Franklin and on Erwin Chargaff's observation that in double-stranded DNA the ratio of adenine to thymine, and of guanine to cytosine, is always $1$ (Chargaff's rule). The salient features of the double helix: (i) two polynucleotide chains, sugar-phosphate backbone outside, bases projecting inward; (ii) antiparallel polarity — where one strand runs $5' \to 3'$, its partner runs $3' \to 5'$; (iii) bases pair across the strands by hydrogen bonds — adenine with thymine via two H-bonds, guanine with cytosine via three — so a purine always faces a pyrimidine, keeping the helix diameter uniform; (iv) right-handed coiling with a pitch of $3.4\ \text{nm}$ and roughly $10$ bp per turn, a rise of about $0.34\ \text{nm}$ per bp; (v) stacked base pairs, with the H-bonds, stabilise the helix. Because the strands are complementary rather than identical, each can template a new partner — the basis of both replication and Francis Crick's central dogma: information flows DNA to RNA to protein (reverse transcription, RNA to DNA, occurs in some viruses).

Packaging the genome: nucleosomes, chromatin and chromosomes

A haploid human cell holds about $3.3 \times 10^9$ base pairs of DNA, which, laid end to end, stretches roughly $2.2$ metres — far longer than the nucleus containing it. In prokaryotes such as E. coli, which have no defined nucleus, the negatively charged DNA is held by proteins in a region called the nucleoid, organised into large loops. Eukaryotes use histones, small, positively charged proteins rich in lysine and arginine: eight histone molecules form a histone octamer, around which DNA wraps to form a nucleosome (about $200$ bp each), the repeating unit of chromatin, seen under the electron microscope as beads on a string. With additional non-histone chromosomal (NHC) proteins, this chromatin coils and condenses at metaphase into visible chromosomes. In the interphase nucleus, loosely packed, transcriptionally active chromatin is euchromatin; densely packed, inactive chromatin is heterochromatin.

The search for the genetic material: from Griffith to Hershey-Chase

Even though Meischer's nuclein and Mendel's factors surfaced around the same time, decades passed before anyone connected them. The decisive clue came from Frederick Griffith's transformation experiment (1928) with Streptococcus pneumoniae. Smooth (S) strain bacteria, coated in polysaccharide, are virulent and kill mice; rough (R) strain, lacking the coat, are harmless. Heat-killed S alone did not kill mice, but heat-killed S mixed with live R did — and living, virulent S bacteria could be recovered from the dead mice. Griffith concluded some transforming principle from the dead S cells had permanently converted R cells into S cells, though he could not identify its chemical nature.

Oswald Avery, Colin MacLeod and Maclyn McCarty (1933-44) supplied the biochemical answer, purifying proteins, DNA and RNA from heat-killed S cells and testing each for transforming ability. Proteases and RNases did not stop transformation, but DNase did — proving DNA, not protein or RNA, was the transforming principle, though many biologists still doubted so "simple" a molecule could carry such information. The unequivocal proof came from Alfred Hershey and Martha Chase (1952), with bacteriophages infecting E. coli. One batch of phages was grown in radioactive phosphorus ($^{32}\text{P}$, labels DNA), another in radioactive sulfur ($^{35}\text{S}$, labels protein). After infection, a blender sheared off phage coats from the bacterial surface, and centrifugation separated phage from bacteria. Bacteria infected with $^{32}\text{P}$-phages were radioactive; those with $^{35}\text{S}$-phages were not — proving DNA, not protein, physically enters the bacterium and directs new phage synthesis.

DNA versus RNA, and the RNA world

A genetic material must (i) generate its own replica, (ii) be chemically and structurally stable, (iii) allow slow mutation for evolution, and (iv) express itself as Mendelian characters. Both DNA and RNA can direct their own duplication by complementary base pairing, so both pass the first test. But RNA's extra $2'$-OH group makes it chemically reactive and structurally labile, and it is also catalytic in some contexts; DNA, lacking that group and carrying the less reactive thymine instead of uracil, is more stable. Both can mutate, but the less stable RNA mutates — and evolves — faster, which is why RNA viruses evolve rapidly. RNA translates directly into protein, while DNA depends on RNA as an intermediate. Taken together, DNA's stability suits it to long-term storage, while RNA's reactivity and direct translatability suit it to the dynamic roles of messenger and adapter. This division of labour is thought to echo evolutionary history: essential life processes — metabolism, translation, splicing — likely evolved around RNA first, acting as both genetic material and catalyst, before DNA evolved from RNA through modifications that made it more stable and repairable. This period is called the RNA world.

DNA replication: semiconservative and enzyme-driven

Watson and Crick's own 1953 paper noted, famously, that the specific base pairing they had proposed "immediately suggests a possible copying mechanism for the genetic material." Their scheme — the two strands separating, each acting as a template for a new complementary strand, so that every daughter duplex carries one parental strand and one newly made strand — is called semiconservative replication. The direct experimental proof came from Matthew Meselson and Franklin Stahl (1958), growing E. coli for many generations in the heavy nitrogen isotope $^{15}\text{N}$, transferring to normal $^{14}\text{N}$ medium, and measuring DNA density by caesium chloride (CsCl) gradient centrifugation at successive generations — walked through in full in the first derivation below. Similar logic with radioactive thymidine, applied by Taylor and colleagues (1958) to the faba bean (Vicia faba), confirmed that chromosomal DNA also replicates semiconservatively. Replication needs a battery of enzymes. The principal one, DNA-dependent DNA polymerase, uses a DNA template to polymerise deoxynucleotides, working at roughly $2000$ bp per second in E. coli with high fidelity. Deoxyribonucleoside triphosphates (dNTPs) serve a dual role, as substrates and, via their high-energy terminal phosphates, as the energy source. Since DNA polymerase synthesises only $5' \to 3'$, and the parental strands are antiparallel, replication at the replication fork is asymmetric: the $3' \to 5'$ template is copied continuously, the $5' \to 3'$ template discontinuously, as short Okazaki fragments later joined by DNA ligase. Replication starts only at a defined origin of replication — which is why any DNA propagated via recombinant DNA needs a vector supplying one. In eukaryotes, replication is confined to the S phase; a failure of division after replication causes polyploidy.

Transcription: copying DNA into RNA

Transcription is the copying of genetic information from one strand of DNA into RNA, governed by the same complementarity rule as replication except that adenine now pairs with uracil rather than thymine. Unlike replication, transcription copies only a defined segment of DNA, and only one of its two strands: copying both would give two different, complementary-but-not-identical RNAs (potentially two different proteins from one DNA segment), and those two RNAs would themselves pair into an untranslatable double-stranded molecule. A transcription unit in DNA is defined by three regions: a promoter, the structural gene, and a terminator. Because RNA polymerase synthesises only $5' \to 3'$, the strand running $3' \to 5'$ acts as the template strand and is actually copied; the other strand, sharing the RNA's sequence (T instead of U), is displaced and is called, counter-intuitively, the coding strand, since the transcription unit's boundaries are described relative to it. The promoter lies upstream (towards the coding strand's $5'$ end) and binds RNA polymerase; the terminator lies downstream ($3'$ end) and usually ends transcription. A gene is the functional unit of inheritance; genes for tRNA or rRNA exist too, so a gene resists a purely DNA-sequence definition. Defining a cistron as a DNA segment coding for one polypeptide lets a structural gene be called monocistronic (mostly eukaryotes) or polycistronic (mostly bacteria). Eukaryotic genes are typically split: coding, expressed exons are interrupted by non-coding introns absent from the mature RNA. Bacteria have three RNA types — mRNA, tRNA, rRNA — made by one DNA-dependent RNA polymerase, which catalyses only elongation, aided transiently by an initiation factor ($\sigma$) and a termination factor ($\rho$). With no nuclear membrane, bacterial transcription and translation can be physically coupled. Eukaryotes differ in two ways: three nuclear RNA polymerases divide the labour (Pol I makes rRNAs $28S/18S/5.8S$; Pol III makes tRNA, $5S$ rRNA, snRNAs; Pol II makes the mRNA precursor, hnRNA); and that primary transcript, carrying both exons and introns, is non-functional until capping (methyl guanosine triphosphate at the $5'$ end), tailing ($200$-$300$ adenylate residues at the $3'$ end, template-independently) and splicing (introns removed, exons joined) convert it to mature mRNA, the only form exported for translation.

The genetic code and mutations

Replication and transcription copy nucleic acid into nucleic acid by simple complementarity; translation is different, since it must convert a nucleotide sequence into an amino acid sequence, between which no chemical complementarity exists. Physicist George Gamow reasoned that $4$ bases specifying $20$ amino acids demanded a combinatorial code; a doublet gives only $4^2 = 16$ combinations, too few, so he proposed a triplet code, giving $4^3 = 64$ codons. Proving and decoding it took Har Gobind Khorana (chemical synthesis of defined RNA sequences), Marshall Nirenberg (a cell-free translation system that revealed each codon's amino acid), and Severo Ochoa (the enzyme polynucleotide phosphorylase, for template-independent RNA synthesis) — together yielding the complete codon table. The salient features of the genetic code: (i) the codon is a triplet, $61$ of $64$ coding for amino acids, the remaining $3$ (UAA, UAG, UGA) as stop codons; (ii) the code is degenerate — most amino acids have more than one codon; (iii) codons are read contiguously, with no punctuation; (iv) the code is nearly universal, with known exceptions only in mitochondrial and some protozoan codons; (v) AUG both codes for methionine and acts as the start codon. Mutation studies illuminate gene-DNA relationships. Large deletions or rearrangements simply cause loss or gain of a gene's function; a point mutation at a single base pair is subtler, best shown by sickle-cell anaemia, where one base-pair change in the beta-globin gene substitutes valine for glutamate. Because the code is read in contiguous, punctuation-free triplets, inserting or deleting one or two bases is far more disruptive: it shifts the reading frame from that point onward, scrambling every downstream codon — a frameshift mutation. Inserting or deleting three bases, or an exact multiple of three, adds or removes whole codons but leaves the downstream reading frame intact.

tRNA and translation: from code to protein

Francis Crick predicted, before the code was even deciphered, that an adapter molecule must link code to amino acid. That adapter is tRNA (once sRNA, soluble RNA): its anticodon loop pairs with a codon on mRNA, and its amino-acid acceptor end carries the matching amino acid; tRNAs are specific to each amino acid, a distinct initiator tRNA reads the start codon, and no tRNA exists for a stop codon. Its secondary structure is a cloverleaf; in three dimensions it folds into a compact inverted L.

Translation polymerises amino acids into a polypeptide in the codon order of the mRNA, joined by peptide bonds. Since bond formation needs energy, amino acids are first activated with ATP and attached to their cognate tRNA — aminoacylation, or charging. The ribosome — structural RNA plus roughly $80$ proteins, in large and small subunits — provides two sites for successive charged tRNAs, and rRNA itself ($23S$ in bacteria) catalyses peptide-bond formation, making the ribosome a ribozyme. A translational unit runs start codon (AUG) to stop codon, flanked by untranslated regions (UTRs) at both ends that aid translation without being translated. Translation proceeds in three stages: initiation (small subunit binds mRNA at AUG, read by initiator tRNA), elongation (charged tRNAs pair codon by codon as the ribosome moves along, adding residues one at a time), and termination (a release factor at the stop codon releases the finished polypeptide).

Regulation of gene expression: the lac operon

Gene expression can, in principle, be regulated at several points — transcriptional, processing (splicing), transport (of mRNA out of the nucleus), or translational — but in prokaryotes, control of the rate of transcriptional initiation dominates. Regulatory proteins bind sequences called operators, usually adjacent to the promoter, acting positively (activators) or negatively (repressors) on RNA polymerase. A cluster of structural genes sharing a common promoter and regulatory sequences is an operon — lac, trp, ara, his and val operons are common bacterial examples. The lac operon, elucidated by Francois Jacob and Jacques Monod, is the prototype of a transcriptionally regulated system. It has one regulatory gene, the i gene (named for "inhibitor," not "inducer"), coding for the lac repressor, and three structural genes: z, for $\beta$-galactosidase, which hydrolyses lactose into glucose and galactose; y, for permease, which increases the cell's permeability to $\beta$-galactosides; and a, for a transacetylase. Lactose, as the substrate that switches the operon on, is called the inducer. In its absence, the constitutively made repressor binds the operator and blocks RNA polymerase, keeping the operon off (bar a low basal leak needed to let some lactose in at all); once lactose enters and its derivative allolactose inactivates the repressor, RNA polymerase transcribes the polycistronic z-y-a mRNA. Because a repressor is what does the switching, this is negative regulation — the third derivation below walks the full mechanism step by step.

The Human Genome Project and DNA fingerprinting

Launched in 1990 and completed in 2003, the Human Genome Project (HGP) was a $13$-year international "mega project" coordinated by the U.S. Department of Energy and the National Institutes of Health, with the Wellcome Trust and partners in Japan, France, Germany and China. Its goals: identify all $20{,}000$-$25{,}000$ human genes, sequence all $3 \times 10^9$ base pairs, store the data in databases, improve analysis tools, transfer the technology outward, and address the ethical, legal and social issues (ELSI) it raised. Sequencing combined two strategies — Expressed Sequence Tags (ESTs) to spot expressed genes, and blind whole-genome sequencing followed by sequence annotation to assign function afterward. DNA was fragmented, cloned in bacterial or yeast vectors (BACs, YACs), sequenced by Frederick Sanger's method on automated sequencers, and the overlaps aligned by purpose-built software — the origin of large-scale bioinformatics. Salient findings: the genome holds $3164.7$ million bp; the average gene is about $3000$ bases, the largest (dystrophin) $2.4$ million bases; the gene count is only about $30{,}000$, far below earlier estimates of $80{,}000$-$140{,}000$; over $50\%$ of genes have unknown function; less than $2\%$ of the genome codes for protein; repeats make up a large fraction of it; chromosome $1$ has the most genes ($2968$), the Y chromosome the fewest ($231$); and about $1.4$ million single nucleotide polymorphisms (SNPs) have been catalogued. That $99.9\%$ sequence identity across humans is also the premise of DNA fingerprinting. The differing fraction lies disproportionately in repetitive DNA, which forms distinct minor peaks (satellite DNA, further split into micro- and mini-satellites) when genomic DNA is spun in a density gradient. Satellite DNA is highly polymorphic, largely non-coding, identical in polymorphism across all tissues of one individual, and inherited — the basis of forensic and paternity testing. Polymorphism formally means a variant allele occurring at a locus with population frequency above $0.01$, and it persists more readily in non-coding DNA, where mutations rarely affect reproductive fitness.

Alec Jeffreys developed DNA fingerprinting using a highly polymorphic mini-satellite probe, VNTR (Variable Number of Tandem Repeats), whose length ($0.1$-$20$ kb) varies between individuals. The classical method uses Southern blot hybridisation: isolate DNA, digest with restriction endonucleases, separate fragments by electrophoresis, blot onto a membrane, hybridise with a labelled VNTR probe, and detect by autoradiography — giving a banding pattern unique to every individual except monozygotic twins. Coupling this with PCR made it sensitive enough for a single cell, extending its use to paternity testing and population-diversity studies.

Board level - Applying Chargaff's rule to find base composition

A double-stranded DNA molecule is found to contain $20\%$ cytosine. Calculate the percentage of adenine, guanine and thymine in this DNA.

Step 1. Recall Chargaff's rule for any double-stranded DNA: because adenine always pairs with thymine and guanine always pairs with cytosine across the two strands, the amount of adenine equals the amount of thymine, and the amount of guanine equals the amount of cytosine, so $\%A = \%T$ and $\%G = \%C$.

Step 2. Apply the given data: since $\%C = 20\%$, it follows immediately that $\%G = 20\%$ as well, so together guanine and cytosine account for $\%G + \%C = 40\%$ of the total bases.

Step 3. The remaining bases must be adenine and thymine, and together they make up the rest of the total: $100\% - 40\% = 60\%$. Since $\%A = \%T$, this $60\%$ splits evenly, giving $\%A = \%T = 30\%$.

Answer: $\%A = 30\%$, $\%T = 30\%$, $\%G = 20\%$, $\%C = 20\%$.

Board level - Writing the template strand and mRNA from a given coding strand

The coding strand of a transcription unit reads $5'\text{-ATGCATGCATGCATGCATGCATGCATGC-}3'$. Write (a) the template strand in its correct $3'$ to $5'$ orientation, and (b) the mRNA transcribed from this transcription unit.

Step 1. The template strand is complementary and antiparallel to the coding strand. Pairing each base of the coding strand (A with T, T with A, G with C, C with G) directly beneath it, and remembering the template strand runs in the opposite ($3' \to 5'$) direction, gives: $3'\text{-TACGTACGTACGTACGTACGTACGTACG-}5'$.

Step 2. The mRNA has the same sequence as the coding strand except that uracil replaces thymine, and it is written $5' \to 3'$ just like the coding strand itself: $5'\text{-AUGCAUGCAUGCAUGCAUGCAUGCAUGC-}3'$.

Step 3. Cross-check by building the mRNA the "long way," directly from the template strand by complementary base pairing (A pairs with U, T with A, G with C, C with G) — it reproduces exactly the Step 2 answer, confirming both routes agree.

Answer: template strand $= 3'\text{-TACGTACGTACGTACGTACGTACGTACG-}5'$; mRNA $= 5'\text{-AUGCAUGCAUGCAUGCAUGCAUGCAUGC-}3'$.

Competitive level (NEET) - The sickle-cell point mutation, GAG to GUG

The mRNA codon for the sixth amino acid of normal human $\beta$-globin is $5'\text{-GAG-}3'$, coding for glutamic acid. In individuals with sickle-cell anaemia, this codon is changed by a single base substitution to $5'\text{-GUG-}3'$. (a) Identify the amino acid now coded, (b) classify the type of mutation, and (c) state why this single substitution, and not a comparable single insertion, still leaves the rest of the polypeptide chain unaffected.

Step 1. Consult the codon table for both triplets: GAG codes for glutamic acid (Glu), and GUG codes for valine (Val). The single base change from A to U at the second position of the codon has therefore converted glutamic acid to valine at that one position in the beta-globin chain.

Step 2. Classify the mutation. Exactly one base pair has changed, converting one codon into another codon that still specifies an amino acid (rather than a stop codon) — this is a point mutation, and more specifically a missense mutation, since it changes the amino acid encoded rather than leaving it silent or converting it to a stop signal.

Step 3. Explain why the rest of the chain is untouched: a substitution swaps one base for another without changing the total number of bases, so the reading frame downstream is completely undisturbed — every codon after the mutated one is read exactly as before. This is the key distinction from an insertion or deletion of one or two bases, which would instead be a frameshift mutation, scrambling every codon downstream of the change, not just the one affected codon.

Answer: The GAG-to-GUG substitution replaces glutamic acid with valine at the sixth position of the beta-globin chain — a single point (missense) mutation, not a frameshift, because a base substitution leaves the reading frame intact. This is the molecular basis of sickle-cell anaemia. Timing note: allow about $60$ seconds. The fastest route is to go straight to the codon table for both triplets, note that only one of the three letters differs, and conclude "substitution, not insertion or deletion" without needing to re-derive the whole reading frame — reserve the reading-frame argument only for justifying why the rest of the chain is unaffected, which is usually the actual question being tested.

The single most common error is describing replication, transcription or translation with the wrong direction or wrong strand: DNA and RNA polymerase both synthesise only $5' \to 3'$, so the template strand always runs $3' \to 5'$, and it is the template strand that is copied even though a transcription unit's landmarks are described relative to the coding strand. A second confusion is treating the "coding strand" as the one transcribed — it is not; it is the strand displaced, named only because its sequence matches the RNA product. Students also routinely confuse a frameshift mutation (insertion or deletion not a multiple of three, scrambling every downstream codon) with a point substitution like sickle-cell (which changes only one codon and leaves the frame intact). Another trap is assuming Meselson and Stahl's first-generation result alone proved semiconservative replication — a single hybrid band is also consistent with the dispersive model; only the second-generation result (two bands, equal amounts) rules dispersive out. Finally, DNA fingerprinting is often wrongly imagined as requiring the entire genome to be sequenced; its power comes from comparing only the small, highly polymorphic satellite/VNTR fraction.

• Draw and label a polynucleotide chain and the DNA double helix, and state the base-pairing rule (A=T with two H-bonds, G≡C with three) and Chargaff's rule.
• Describe how DNA is packaged into nucleosomes and chromatin, and distinguish euchromatin from heterochromatin.
• Recount the Griffith, Avery-MacLeod-McCarty and Hershey-Chase experiments and state precisely what each one did and did not prove.
• Explain, with reference to the $2'$-OH group and thymine versus uracil, why DNA is chemically more stable than RNA, and what the RNA world hypothesis proposes.
• Describe the Meselson-Stahl experiment and explain why its result rules out both the conservative and dispersive models of replication.
• Define a transcription unit (promoter, structural gene, terminator) and distinguish the template strand from the coding strand.
• Explain hnRNA processing in eukaryotes — capping, tailing and splicing — and why it has no equivalent in bacterial transcription.
• State the salient features of the genetic code, and distinguish a point (substitution) mutation from a frameshift mutation using the sickle-cell and reading-frame examples.
• Trace translation from initiation through elongation to termination, naming the role of tRNA, the ribosome, and the start and stop codons.
• Explain the lac operon's negative regulation mechanism, and summarise the goals, methods and salient findings of the Human Genome Project and the basis of DNA fingerprinting.

Ask the tutor: Want to work through more Chargaff's-rule or codon-table numericals, or trace how a mutation in this chapter's replication or transcription machinery could show up as a variation studied in the next chapter on evolution?