how we proved DNA carries genes

The Mechanism

In May 1952, the *Journal of General Physiology* received a manuscript titled *"Independent Functions of Viral Protein and Nucleic Acid in Growth of Bacteriophage."* The two authors were *Alfred Day Hershey* (born Owosso, Michigan, *4 December 1908*; died Syosset, Long Island, *22 May 1997*, aged 88), a 43-year-old bacteriophage geneticist working at the *Carnegie Institution of Washington's Department of Genetics* at *Cold Spring Harbor Laboratory* on Long Island, and *Martha Cowles Chase* (born Cleveland, Ohio, *30 November 1927*; died Lorain, Ohio, *8 August 2003*, aged 75), Hershey's 24-year-old research assistant, who had arrived at the laboratory in 1950 with a bachelor's degree in biology from the College of Wooster. The paper was published in the journal's issue of *September 1952* (volume 36, pp. 39-56). It described a single experiment — performed in a long sequence of refinements in late 1951 and the first months of 1952 in Hershey's *converted barn laboratory* at the Carnegie Institution's *Demerec laboratory complex* in the village of Cold Spring Harbor — that resolved one of the central open questions of twentieth-century biology: whether the *genetic material* — the molecule that carries hereditary information from parent organism to offspring — was *protein* or *nucleic acid (DNA)*. The question had been argued since the 1920s. The chromosomes inside cell nuclei were known to be the carriers of heredity; the chromosomes were known to be composed of *roughly equal parts protein and DNA*; the question was *which* of the two molecules carried the actual instructions. The dominant view through the 1930s and 1940s, championed by the influential biochemists *Phoebus Levene* of the Rockefeller Institute and *Max Bergmann* of Harvard, was that DNA could *not* possibly be the genetic material: DNA was a simple, repetitive molecule (Levene's *"tetranucleotide hypothesis"* held that DNA was a monotonous repeating unit of the four bases adenine-cytosine-guanine-thymine, with no capacity for variation or information storage); proteins, with their twenty distinct amino acids and combinatorial sequence complexity, were the only molecules with the right kind of information-carrying capacity. The contrary view had been growing since the 1944 publication by *Oswald Avery, Colin MacLeod, and Maclyn McCarty* at the Rockefeller Institute of their *"transforming principle"* paper — a careful demonstration that the chemical extracted from one strain of *Pneumococcus* bacteria that could transmit hereditary properties (specifically, the smooth-versus-rough surface phenotype) to a different strain was, by every assay they could devise, *deoxyribonucleic acid*. The Avery-MacLeod-McCarty paper convinced many but not all of the field; the dissenters held that the transforming principle might contain traces of contaminating protein that were the actual carriers. The question needed a *cleaner experiment*. Hershey was a member of the so-called *"phage group"* — the small international collection of biologists (Hershey, Max Delbrück, Salvador Luria, and their students) who had agreed, in the late 1940s, to focus the field's foundational genetic work on a single experimental system: the *bacteriophage T2*, a virus that infects the bacterium *Escherichia coli*. T2 was a small, elegant, geometrically precise object — about *200 nanometres long*, with an *icosahedral head* (a 20-faced polyhedron about *95 nm across*), a *contractile tail sheath*, *six long thin tail fibres*, and a baseplate that the tail fibres attached to. The virus's life cycle was well-described: a phage particle would attach to the outer surface of an *E. coli* cell via its tail fibres, the tail sheath would contract like a syringe, and *something* — some component of the virus — would be injected into the bacterial cell. About 25 minutes later, the bacterium would burst open and release a *hundred or more* new phage particles. The question was: *what gets injected?* In *late 1951*, Hershey and Chase devised an experiment built on a single chemical accident of biology. *Proteins contain sulfur* (in the amino acids cysteine and methionine) but *do not contain phosphorus*. *DNA contains phosphorus* (in the backbone of the molecule, one phosphate group per nucleotide) but *does not contain sulfur*. The two molecules can therefore be selectively radioactively labelled: phage grown in a medium containing *radioactive sulfur-35* incorporates the ³⁵S into its protein coat but not its DNA; phage grown in a medium containing *radioactive phosphorus-32* incorporates the ³²P into its DNA but not its protein. Hershey and Chase grew two separate stocks of T2 phage: one batch in ³⁵S, one batch in ³²P. They then performed two parallel infections of *E. coli*: in the first, the bacteria were exposed to ³⁵S-protein-labelled phages; in the second, to ³²P-DNA-labelled phages. After allowing five minutes for the phages to attach to the bacterial surfaces and inject their contents, they faced a technical problem: the empty phage *protein coats* (the *"ghosts"*) remained stuck to the bacterial cell walls after injection, mixing the radioactive label on the outside of the bacteria with whatever radioactive label was inside. They needed to strip the empty coats off without lysing the bacteria. The solution they tried, after several other shearing methods had failed, was a *Waring brand commercial kitchen blender* — the same model used in cocktail bars and home kitchens — running at high speed for *two minutes*. The mechanical shearing was strong enough to break the tail-fibre attachments and tear the empty phage coats off the bacterial surfaces, but gentle enough that the bacteria themselves remained intact. The blender contents were then *centrifuged at low speed*: the heavier *bacteria* (with whatever phage component had been injected) sedimented into a pellet at the bottom of the tube, while the lighter *empty phage coats* remained suspended in the supernatant. The two fractions were then counted in a scintillation counter to determine where the ³⁵S and ³²P had ended up. The result was unambiguous. In the ³⁵S infection — the protein-labelled phage — *80 % of the radioactive sulfur* ended up in the *supernatant* (the empty coats torn off by the blender), and only *20 %* in the bacterial pellet. In the ³²P infection — the DNA-labelled phage — *more than 70 % of the radioactive phosphorus* ended up in the *bacterial pellet* (inside the bacteria), and only the residual fraction in the supernatant. The interpretation followed directly. The phage *protein coat* stayed on the *outside* of the bacterium: it was not what entered. The phage *DNA* went *inside* the bacterium: it was what entered. The DNA was what was *required* for the production of progeny phage; the protein coat was a delivery vehicle. The genetic material of T2 was its DNA. The result was not a knock-down proof — there were technical critiques, most prominently that about 20 % of the sulfur remained in the bacterial fraction and might or might not represent injected protein, and that some traces of protein could still have been transferred with the DNA — and the paper itself was characteristically *cautious* about its conclusions. But the experiment crystallised the field. Within twelve months, the *Watson-Crick double helix* (April 1953) had given a structural account of how DNA could store information; within a few years, the *Hershey-Chase blender experiment* had become the canonical undergraduate-textbook entry-point demonstration that DNA is the genetic material. *Alfred Hershey* shared the *1969 Nobel Prize in Physiology or Medicine* with *Salvador Luria* and *Max Delbrück* "for their discoveries concerning the replication mechanism and the genetic structure of viruses." *Martha Chase*, whose contribution to the experiment was technical and substantial, *was not included in the Nobel*; she completed her PhD at the University of Southern California in 1964, struggled with what is now understood to be undiagnosed dementia from the 1960s onward, and lived in obscurity until her death in 2003. The Waring blender that Hershey and Chase used is on display, alongside Hershey's lab bench, in the *James D. Watson Genome Center* at *Cold Spring Harbor Laboratory*.