The molecular basis of heredity

Genetics, as BIOL-201 showed, can be done without knowing what genes are made of. Mendel worked with abstract “factors”, and the first genetic maps were drawn decades before anyone knew the chemistry of a chromosome. Molecular biology asks the next question: what is the physical substance of heredity, and how does it store, copy and use information? This course follows that question from the experiments that identified DNA as the genetic material, through the machinery that copies DNA and reads it into RNA and protein, to the regulation that decides which genes are used when, and finally to the techniques that let us cut, copy, read and edit DNA in the laboratory.

This first lesson sets the scene. It reviews the experiments that led from “something in the chromosome” to DNA, introduces the central dogma and its exceptions, and practises the quantitative habits that run through the course: thinking in base pairs, kilobases and nucleotides per second, and checking claims against numbers. BIOL-107 introduced nucleic acids as one of the four classes of biological macromolecules, and BIOL-201 lesson 1 told the history of genetics in outline. Here we look more closely at how the key experiments were designed and why their conclusions were convincing.

Protein or nucleic acid?

By the 1920s it was clear that genes sat on chromosomes, and chemical analysis showed that chromosomes contain both protein and DNA. Most biologists assumed proteins were the genetic material. Proteins are built from 20 different amino acids and come in an enormous variety of shapes, while DNA, with only four kinds of nucleotide, looked too simple. A popular idea, the tetranucleotide hypothesis of Phoebus Levene, held that DNA was a monotonous repeat of the four bases in equal amounts, a structural scaffold rather than an information store. That idea turned out to be wrong, but it shaped thinking for two decades.

Griffith and transformation

In 1928 Frederick Griffith, a British medical officer studying pneumonia, worked with two strains of Streptococcus pneumoniae. The S strain made a polysaccharide capsule, formed smooth colonies and killed mice. The R strain lacked the capsule, formed rough colonies and was harmless. Heat-killed S bacteria were also harmless. But when Griffith injected mice with a mixture of live R bacteria and heat-killed S bacteria, many mice died, and he recovered live S bacteria from their blood. Something from the dead S cells had permanently changed the R cells into the virulent form, and the change was inherited by their descendants. Griffith called this transformation, without identifying the transforming substance.

Avery, MacLeod and McCarty

Oswald Avery, Colin MacLeod and Maclyn McCarty, working at the Rockefeller Institute in New York, spent more than a decade purifying the transforming principle. MacLeod was Canadian, trained at McGill. In 1944 they reported that the active material behaved chemically like DNA: it had the elemental composition of DNA, it was unaffected by enzymes that destroy protein (proteases) or RNA (ribonuclease), and it was destroyed by an enzyme that degrades DNA (deoxyribonuclease). The logic was elimination: if destroying a component abolishes activity, that component is needed; if destroying it leaves activity intact, it is not.

Many scientists were not fully persuaded. Critics argued that tiny amounts of contaminating protein might be responsible, and that bacterial transformation might be a special case. The evidence was strong, but the conclusion was revolutionary, and revolutions in science usually need more than one line of evidence.

Hershey and Chase

The second line came in 1952 from Alfred Hershey and Martha Chase at Cold Spring Harbor, using bacteriophage T2, a virus that infects Escherichia coli. A phage is little more than DNA packed inside a protein coat. When it infects a cell, new phages appear inside after about 20 minutes. Hershey and Chase asked which part of the phage enters the cell and directs the production of new phages. Their insight was to label the two components separately with radioactive isotopes: protein contains sulfur (in the amino acids cysteine and methionine) but almost no phosphorus, while DNA contains phosphorus in its backbone but no sulfur.

They grew one batch of phages with ³⁵S, labelling the protein coats, and another with ³²P, labelling the DNA. Each batch infected unlabelled bacteria. After a few minutes they agitated the mixtures in a kitchen blender to shear the phage coats off the bacterial surfaces, then centrifuged. The heavier bacterial cells formed a pellet; the lighter phage coats stayed in the liquid above (the supernatant). Most of the ³²P ended up in the pellet with the cells, and that label appeared in the next generation of phages. Most of the ³⁵S stayed in the supernatant with the empty coats.

Because blending was imperfect, neither separation was clean, and Hershey and Chase themselves were cautious. But the experiment was simple, the result was easy to repeat, and coming after Avery’s work it settled the question for most biologists.

Chargaff’s rules and the double helix

Meanwhile Erwin Chargaff, at Columbia University, measured the base composition of DNA from many species. Two observations demolished the tetranucleotide hypothesis. First, base composition varies between species, which is what an information-carrying molecule should do. Second, within any double-stranded DNA the amount of adenine equals the amount of thymine, and the amount of guanine equals the amount of cytosine. These became known as Chargaff’s rules.

The structure that explained these rules was published in April 1953 by James Watson and Francis Crick at Cambridge. It drew critically on X-ray diffraction images made at King’s College London by Rosalind Franklin and her student Raymond Gosling, especially the image known as Photograph 51, and on measurements from Franklin’s unpublished report. Maurice Wilkins, also at King’s, shared the image with Watson without Franklin’s knowledge. The same issue of Nature carried papers by Wilkins and colleagues and by Franklin and Gosling supporting the model. Watson, Crick and Wilkins received the 1962 Nobel Prize; Franklin had died of ovarian cancer in 1958 at 37, and the prize is not awarded posthumously. Historians now agree that her data and analysis were essential to the discovery, and the story is often used to discuss credit and recognition in science.

Lesson 2 examines the structure in detail. The point for now is that the double helix immediately suggested how DNA could be copied: each strand carries the information to specify its partner. Watson and Crick ended their paper with a famously understated sentence noting that the specific pairing “immediately suggests a possible copying mechanism”. Lesson 3 shows how that mechanism was demonstrated by Meselson and Stahl in 1958.

The code below shows Chargaff’s rules emerging from base pairing. It generates a random single strand with a deliberately uneven composition, builds the complementary strand, and counts bases in the single strand and in the duplex. The single strand need not obey the rules; the double-stranded molecule always does.

import random
random.seed(207)

pair = {"A": "T", "T": "A", "G": "C", "C": "G"}
weights = {"A": 0.35, "C": 0.15, "G": 0.20, "T": 0.30}   # deliberately uneven
strand = "".join(random.choices(list(weights), weights=list(weights.values()), k=10_000))
partner = "".join(pair[b] for b in strand)

def pct(seq):
    return {b: round(100 * seq.count(b) / len(seq), 1) for b in "ACGT"}

print("one strand:      ", pct(strand))
print("double-stranded: ", pct(strand + partner))
duplex = strand + partner
print("A = T?", duplex.count("A") == duplex.count("T"), "  G = C?", duplex.count("G") == duplex.count("C"))

one strand:       {'A': 35.3, 'C': 15.0, 'G': 19.8, 'T': 29.9}
double-stranded:  {'A': 32.6, 'C': 17.4, 'G': 17.4, 'T': 32.6}
A = T? True   G = C? True

The central dogma

In 1958 Francis Crick stated what he called the central dogma of molecular biology: information, once it has passed into protein, cannot get out again. The familiar summary is that DNA is copied into DNA (replication), DNA is transcribed into RNA (transcription) and RNA is translated into protein (translation). Crick’s original formulation was about which transfers of sequence information are possible, not about the importance of each molecule, and it allowed for transfers he thought might occur occasionally.

Several of the special transfers were later found. Retroviruses such as HIV carry RNA genomes and use the enzyme reverse transcriptase to copy RNA into DNA, discovered independently by Howard Temin and David Baltimore in 1970. Many RNA viruses, including influenza and the coronaviruses, copy RNA directly into RNA. Telomerase, which maintains the ends of our chromosomes (lesson 4), is a reverse transcriptase that carries its own RNA template. And a large fraction of the human genome consists of retrotransposons, sequences that spread by being transcribed and then reverse transcribed back into DNA (BIOL-201 lesson 10). What has never been found is a mechanism that copies the sequence of a protein back into nucleic acid. Prions, the infectious misfolded proteins behind diseases such as chronic wasting disease in deer and elk, transmit a shape, not a sequence, and so do not break the rule.

Genes, genomes and scale

Molecular biology is full of numbers, and getting a feel for them is part of understanding the field. DNA length is measured in base pairs (bp), kilobase pairs (kb, 1,000 bp) and megabase pairs (Mb, one million bp). The E. coli genome is about 4.6 Mb and encodes roughly 4,300 proteins; the human haploid genome is about 3.1 billion bp and contains roughly 20,000 protein-coding genes, which together occupy only about 1 to 2 per cent of the sequence. One base pair adds about 0.34 nanometres of length to the double helix, so the DNA in a single human cell, stretched out, would be about two metres long, packed into a nucleus a few micrometres across. Lesson 2 shows how chromatin achieves that packing.

# Lengths of DNA molecules from base pair counts (0.34 nm per base pair)
genomes = {
    "phage lambda": 48_502,
    "E. coli": 4_641_652,
    "baker's yeast (haploid)": 12_100_000,
    "human (haploid)": 3_100_000_000,
}
NM_PER_BP = 0.34
for name, bp in genomes.items():
    length_m = bp * NM_PER_BP * 1e-9
    if length_m < 1e-3:
        print(f"{name:24s} {bp:>15,d} bp  ->  {length_m * 1e6:8.1f} micrometres")
    else:
        print(f"{name:24s} {bp:>15,d} bp  ->  {length_m * 1e3:8.1f} millimetres")
diploid_m = 2 * genomes["human (haploid)"] * NM_PER_BP * 1e-9
print(f"\nDNA in one diploid human cell: about {diploid_m:.1f} m")

phage lambda                      48,502 bp  ->      16.5 micrometres
E. coli                        4,641,652 bp  ->       1.6 millimetres
baker's yeast (haploid)       12,100,000 bp  ->       4.1 millimetres
human (haploid)            3,100,000,000 bp  ->    1054.0 millimetres

DNA in one diploid human cell: about 2.1 m

Note the units in the output: the code converts to micrometres and millimetres because those are easier to picture than raw metres. A useful habit is to estimate before calculating. The human genome is about 3 × 10⁹ bp; multiplying by 3.4 × 10⁻¹⁰ m gives about 1 m per haploid set, so roughly 2 m per diploid cell, which matches the output.

Model organisms and tools

The founding experiments of molecular biology used bacteria and their viruses because they grow fast, in enormous numbers, on simple media. E. coli divides every 20 minutes or so under good conditions, so a single cell can become billions overnight. Phages made it possible to study gene structure at the level of individual nucleotides decades before DNA could be sequenced. Later, baker’s yeast (Saccharomyces cerevisiae) became the workhorse for eukaryotic molecular biology, joined by the fruit fly, the nematode Caenorhabditis elegans, the plant Arabidopsis thaliana, the zebrafish, the mouse and cultured human cells. Throughout this course, examples come from whichever organism revealed a mechanism most clearly, with the understanding that the core machinery of replication, transcription and translation is shared across life.

Canadian laboratories contributed to several stories in this course. Oswald Avery was himself born in Halifax, Nova Scotia, and MacLeod grew up in Prince Edward Island and Saskatchewan. Later, in the 1970s and 1980s, groups in Canada helped identify genes for cystic fibrosis and Duchenne muscular dystrophy (BIOL-201), and Canadian scientists contributed to the lipid nanoparticle technology used in mRNA vaccines, which lesson 12 discusses.

How to read this course

Each lesson moves from mechanism to quantity. You will meet the molecules involved, see how experiments revealed their roles, and then calculate: replication times, error rates, melting temperatures, binding fractions, amplification factors and sequencing coverage. Short Python programs, written with only the standard library or numpy, model these processes and handle DNA sequences as strings of letters. You do not need to be a programmer to follow them; each one is explained in the surrounding text, and the output is shown. If you want to experiment, change a number and predict the result before running the code again.