If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Classic experiments: DNA as the genetic material

Experiments by Frederick Griffith, Oswald Avery and his colleagues, and Alfred Hershey and Martha Chase.


Our modern understanding of DNA's role in heredity has led to a variety of practical applications, including forensic analysis, paternity testing, and genetic screening. Thanks to these wide-ranging uses, today many people have at least a basic awareness of DNA.
It may be surprising, then, to realize that less than a century ago, even the best-educated members of the scientific community did not know that DNA was the hereditary material!
In this article, we'll look at some of the classic experiments that led to the identification of DNA as the carrier of genetic information.

Protein vs. DNA

The work of Gregor Mendel showed that traits (such as flower colors in pea plants) were not inherited directly, but rather, were specified by genes passed on from parents to offspring. The work of additional scientists around the turn of the 20th century, including Theodor Boveri, Walter Sutton, and Thomas Hunt Morgan, established that Mendel's heritable factors were most likely carried on chromosomes.
Scientists first thought that proteins, which are found in chromosomes along with DNA, would turn out to be the sought-after genetic material. Proteins were known to have diverse amino acid sequences, while DNA was thought to be a boring, repetitive polymer, due in part to an incorrect (but popular) model of its structure and composition1.
Today, we know that DNA is not actually repetitive and can carry large amounts of information, as discussed further in the article on discovery of DNA structure. But how did scientists first come to realize that "boring" DNA might actually be the genetic material?

Frederick Griffith: Bacterial transformation

In 1928, British bacteriologist Frederick Griffith conducted a series of experiments using Streptococcus pneumoniae bacteria and mice. Griffith wasn't trying to identify the genetic material, but rather, trying to develop a vaccine against pneumonia. In his experiments, Griffith used two related strains of bacteria, known as R and S.
  • R strain. When grown in a petri dish, the R bacteria formed colonies, or clumps of related bacteria, that had well-defined edges and a rough appearance (hence the abbreviation "R"). The R bacteria were nonvirulent, meaning that they did not cause sickness when injected into a mouse.
  • S strain. S bacteria formed colonies that were rounded and smooth (hence the abbreviation "S"). The smooth appearance was due to a polysaccharide, or sugar-based, coat produced by the bacteria. This coat protected the S bacteria from the mouse immune system, making them virulent (capable of causing disease). Mice injected with live S bacteria developed pneumonia and died.
As part of his experiments, Griffith tried injecting mice with heat-killed S bacteria (that is, S bacteria that had been heated to high temperatures, causing the cells to die). Unsurprisingly, the heat-killed S bacteria did not cause disease in mice.
The experiments took an unexpected turn, however, when harmless R bacteria were combined with harmless heat-killed S bacteria and injected into a mouse. Not only did the mouse develop pnenumonia and die, but when Griffith took a blood sample from the dead mouse, he found that it contained living S bacteria!
Diagram illustrating Frederick Griffith's experiment with S and R bacteria.
  1. Rough strain (nonpathogenic). When this strain is injected into a mouse, the mouse lives.
  2. Smooth strain (pathogenic). When this strain is injected into a mouse, the mouse gets pneumonia and dies.
  3. Heat-killed smooth strain. When heat-killed smooth cells are injected into a mouse, the mouse lives.
  4. Rough strain & heat-killed smooth strain. When these two types of cells are injected into a mouse as a mixture, the mouse gets pneumonia and dies.
_Image modified from "Griffith experiment," by Madeleine Price Ball (CC0/public domain)._
Griffith concluded that the R-strain bacteria must have taken up what he called a "transforming principle" from the heat-killed S bacteria, which allowed them to "transform" into smooth-coated bacteria and become virulent.

Avery, McCarty, and MacLeod: Identifying the transforming principle

In 1944, three Canadian and American researchers, Oswald Avery, Maclyn McCarty, and Colin MacLeod, set out to identify Griffith's "transforming principle."
To do so, they began with large cultures of heat-killed S cells and, through a long series of biochemical steps (determined by careful experimentation), progressively purified the transforming principle by washing away, separating out, or enzymatically destroying the other cellular components. By this method, they were able to obtain small amounts of highly purified transforming principle, which they could then analyze through other tests to determine its identity2.
Several lines of evidence suggested to Avery and his colleagues that the transforming principle might be DNA2:
  • The purified substance gave a negative result in chemical tests known to detect proteins, but a strongly positive result in a chemical test known to detect DNA.
  • The elemental composition of the purified transforming principle closely resembled DNA in its ratio of nitrogen and phosphorous.
  • Protein- and RNA-degrading enzymes had little effect on the transforming principle, but enzymes able to degrade DNA eliminated the transforming activity.
These results all pointed to DNA as the likely transforming principle. However, Avery was cautious in interpreting his results. He realized that it was still possible that some contaminating substance present in small amounts, not DNA, was the actual transforming principle3.
Because of this possibility, debate over DNA's role continued until 1952, when Alfred Hershey and Martha Chase used a different approach to conclusively identify DNA as the genetic material.

The Hershey-Chase experiments

In their now-legendary experiments, Hershey and Chase studied bacteriophage, or viruses that attack bacteria. The phages they used were simple particles composed of protein and DNA, with the outer structures made of protein and the inner core consisting of DNA.
Hershey and Chase knew that the phages attached to the surface of a host bacterial cell and injected some substance (either DNA or protein) into the host. This substance gave "instructions" that caused the host bacterium to start making lots and lots of phages—in other words, it was the phage's genetic material. Before the experiment, Hershey thought that the genetic material would prove to be protein4.
To establish whether the phage injected DNA or protein into host bacteria, Hershey and Chase prepared two different batches of phage. In each batch, the phage were produced in the presence of a specific radioactive element, which was incorporated into the macromolecules (DNA and protein) that made up the phage.
  • One sample was produced in the presence of 35S, a radioactive isotope of sulfur. Sulfur is found in many proteins and is absent from DNA, so only phage proteins were radioactively labeled by this treatment.
  • The other sample was produced in the presence of 32P, a radioactive isotope of phosphorous. Phosphorous is found in DNA and not in proteins, so only phage DNA (and not phage proteins) was radioactively labeled by this treatment.
Each batch of phage was used to infect a different culture of bacteria. After infection had taken place, each culture was whirled in a blender, removing any remaining phage and phage parts from the outside of the bacterial cells. Finally, the cultures were centrifuged, or spun at high speeds, to separate the bacteria from the phage debris.
Centrifugation causes heavier material, such as bacteria, to move to the bottom of the tube and form a lump called a pellet. Lighter material, such as the medium (broth) used to grow the cultures, along with phage and phage parts, remains near the top of the tube and forms a liquid layer called the supernatant.
  1. One batch of phage was labeled with 35S, which is incorporated into the protein coat. Another batch was labeled with 32P, which is incorporated into the DNA.
  2. Bacteria were infected with the phage.
  3. The cultures were blended and centrifuged to separate the phage from the bacteria.
  4. Radioactivity was measured in the pellet and liquid (supernatant) for each experiment. 32P was found in the pellet (inside the bacteria), while 35S was found in the supernatant (outside of the bacteria)
_Image modified from "Historical basis of modern understanding: Figure 3," by OpenStax College, Biology (CC BY 3.0)._
When Hershey and Chase measured radioactivity in the pellet and supernatant from both of their experiments, they found that a large amount of 32P appeared in the pellet, whereas almost all of the 35S appeared in the supernatant. Based on this and similar experiments, Hershey and Chase concluded that DNA, not protein, was injected into host cells and made up the genetic material of the phage.

Remaining questions

The work of the researchers above provided strong evidence for DNA as the genetic material. However, it still wasn't clear how such a seemingly simple molecule could encode the genetic information needed to build a complex organism. Additional research by many scientists, including Erwin Chargaff, James Watson, Francis Crick, and Rosalind Franklin, led to the discovery of DNA structure, clarifying how DNA can encode large amounts of information.

Want to join the conversation?

  • marcimus pink style avatar for user Mamaian1230
    it was good
    but i wanted to ask that why do twins have different dna structure??
    (12 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Jonas Gelinių Nagų Priauginimas
      It happens, because normally there is only one egg at the time of ovulation in the ovarian duct. And if a sperm is present there is more than enough of spermatozoids. So one lucky sperm fertilizes one egg. This way an zygote is produced which later becomes embryo and eventually a baby.
      But in some cases, there is two or even more eggs in the ovarian duct, and as we know there are hundreds of sperms, both can get fertilized. In this case two different zygotes form, hence the name "dizygotic" twins. In more simple way, it is like getting pregnant twice, when female body gives two eggs that can be fertilized. And different eggs have different sets of DNA (because of myosis, crossover and other things). So dizygotic twins are related as much as simple brothers (or sisters).
      (18 votes)
  • leafers sapling style avatar for user Adi.price
    I don't understand how the mice died with the smooth strain but not the heat killed smooth strain.
    (5 votes)
    Default Khan Academy avatar avatar for user
  • mr pants teal style avatar for user Kimberly
    I'm kind of confused on Avery's experiment on transforming factors. I understand the process of it, but I don't know what exactly was grown in the petri dish, or how it proved that DNA is the transforming molecule.

    Thank you,
    (9 votes)
    Default Khan Academy avatar avatar for user
    • piceratops tree style avatar for user Jen
      There were three major components to their experiment: (1) separated, subcellular components of heat-kill virulent S. pneumoniae (2) non-virulent S. pneumoniae and (3) mice. To see which subcellular component (the "purified substance," as Khan refers to it) was the transforming principle (responsible for "passing on" virulence to non-virulent bacteria as shown in Griffith's experiment), they systematically injected one type of substance per trial into non-virulent bacteria. This initially non-virulent bacteria then was injected into mice; if the mice lived, then the substance injected could not have been the transforming principle. They then noticed that one specific substance (which we now know is DNA) transformed the initially non-virulent bacteria and killed the mice when injected.

      The results of three other tests then gave these scientists additional evidence that supported DNA as the transforming principle, which are listed in the article. The first two should be self explanatory: the substance tested positive for DNA detection and negative for protein detection. Its chemical composition also matched that of DNA. The last one may be a little confusing without knowing about the experimental setup above. They essentially treated the substance with either (1) protein/RNA degrading enzymes or (2) DNA degrading enzymes. Injecting the substance treated with (1) showed that non-virulent bacteria were still transformed (became virulent) and mice still died. Therefore, the transforming principle could not be protein/RNA, which had been degraded by enzymes in that sample. Injecting the substance treated with (2) showed that non-virulent bacteria were not transformed and mice did not die. This suggested that DNA was indeed the transforming principle because when it was degraded, virulent bacteria were unable to "pass on" their virulence.
      (11 votes)
  • female robot amelia style avatar for user janny liggyman
    You know how DNA makes us different from any other person, well could it be so complex that we could have an attached mutation that makes us different from the human race?
    (5 votes)
    Default Khan Academy avatar avatar for user
    • leaf green style avatar for user Nikki
      Mutations are present in every individual. It is part of life, the older we get the more mutations we acquire. Mutations occur in our DNA, its a structural change (a mutation is not necessarily attached to our chromosome) that alters our DNA, these alterations can be large or small. So you can think of a mutation as a genetic change in the sequence of our DNA. For example, individuals with the disease Sickle Cell Anemia are still human (and part of the human race), but they have a mutation in their genome.
      So to answer your question: mutations do not necessarily make us different from the human race-because you would still have the same human genetic makeup as everyone else. The only difference between you and other people is that your DNA is slightly altered (with a mutation).
      (7 votes)
  • blobby green style avatar for user raveshwadhwani
    What does the word transformation mean?
    (4 votes)
    Default Khan Academy avatar avatar for user
    • piceratops ultimate style avatar for user RowanH
      Transformation in normal, everyday language, means to change something.
      In the bacterial transformation experiment, Griffith saw that the R strain of bacteria could be changed, or 'transformed' into S-strain bacteria when they were mixed with dead S-strain bacteria.
      Nowadays, 'transformation' has a very specific meaning in biology. It describes the process of bacteria taking up DNA directly from their surroundings, like in Griffith's experiment. There are also other ways DNA can get into bacteria, and they have their own names. 'Transduction' is when DNA is injected into bacteria by phage (like mentioned in Hershey and Chase's experiment above), and under certain circumstances, DNA can also be passed from one bacterium to another by 'conjugation'.

      Also, another separate meaning of transformation in biology is when normal cells become 'transformed' to cancer cells.
      (4 votes)
  • marcimus red style avatar for user metra azimi
    what evidence was there to lead hershey and chase to conclude that only DNA from the bacteriophage entered the bacterium leaving protein sheath behind
    (4 votes)
    Default Khan Academy avatar avatar for user
  • orange juice squid orange style avatar for user dezaraestone
    What is the function of the enzymes protease, RNase, and DNase in relation to the mice experimentation?
    (3 votes)
    Default Khan Academy avatar avatar for user
    • winston baby style avatar for user Ivana - Science trainee
      RNase and DNases are not designed/used or connected specifically to the mice.

      Mice are used as model organisms, but it means that the same DNase would have the same effect on Caenorhabditis Elegans (nematode model organism).

      RNase is a type of enzymes that cut RNA molecules, and DNase is enzymes (endonucleases or exonucleases) that specifically cut DNA molecules. But it does not matter which organism is experimented on.

      Proteases on the other hand, are used to cut proteins.
      (3 votes)
  • aqualine ultimate style avatar for user Jorge R. Martinez Perez-Tejada
    I still do not fully understand how the rough strain bacteria kill the mice when there's heat killed smooth strain bacteria. Do they incorporate the dead bacteria's DNA into theirs and start replicating it and producing proteins based on that newly acquired sequence? how does that happen or why? Is that similar to what viruses do?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • leaf green style avatar for user Cooper Collier
      When bacteria incorporate DNA outside of themselves into their genome, it's called transformation. This is what happened in the experiment, as the rough bacteria sucked up the still-intact DNA from the dead smooth bacteria. Yes, they did incorporate the dead bacteria's DNA into theirs and started replicating it to produce proteins from the newly acquired sequence. By the way, what viruses do is slightly different--they inject their own DNA into the bacteria which is called transduction.
      (4 votes)
  • duskpin seedling style avatar for user Hritwik Haldar
    In Griffith's experiment, why he is not considering different immunity power of the mice?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • winston baby style avatar for user Ivana - Science trainee
      I got it what you ask - so you allude to those different types of engineered mice - with S strain, R train heat killed R and helat killed S, right?

      He has done nothing. He made the experiment with premise that mice upon changes he has done have fairly uniform immune system (meaning not compromisd mice).

      I think, just like in human trials, that animals undergo screening as well. The always take healthy, young, male animals. Or if they take female, they must not be pregnant.

      So the differences in their immune system he created caould not be altered by mice immune system by default.
      (2 votes)
  • primosaur tree style avatar for user Jellybean😆
    How was DNA found?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • piceratops ultimate style avatar for user Jonathan
      I was interested in finding this out as well, so I searched on wikipedia. I see this is 2 years old, but I still want to answer!

      Apparently DNA was discovered by a Swiss scientist named Friedrich Miescher in 1868, so almost contemperaneous to Mendel. Doesn't roll off the English speaking tongue as easily as Watson and Crick (or Franklin). Also, part of the experiment might be even grosser than fruit flies. He was interested in studying the chemistry of the nucleus in neutrophils, a type of white blood cell. Unfortunately, these are not nearly as easy to come by as fruit flies or pea plants — and this is the 1860s — so he went to the hospital to obtain used surgical bandages with pus on them from which to extract the neutrophils. Yep, the 19th century was pretty spooky sometimes!

      A chemistry expert can probably explain this part better. He used sodium sulfate to filter the cells, then just let them settle in a beaker since the centrifuge was not quite finished being invented yet. Wikipedia doesn't explain how he removed the nucleus from the cytoplasm, but then he subjected the nuclei to fancy chemistry techniques, and he discovered a new substance he decided to call "nuclein" because he found it in the nucleus. I couldn't find where the name got changed to DNA, but scientists must have settled on it decades later once they had a good understanding of its chemical makeup.
      (2 votes)