On this page
Restriction enzymes are the "molecular scissors" in your biotechnology chapter. NCERT tells you that Hind II cuts at a six base pair sequence and EcoRI recognises GAATTC. This post tells how those scissors were found, who found them, and how to calculate how often a given enzyme should cut a piece of DNA.
Who shared the 1978 Nobel Prize?
The 1978 Nobel Prize in Physiology or Medicine went jointly to Werner Arber, Daniel Nathans and Hamilton O. Smith "for the discovery of restriction enzymes and their application to problems of molecular genetics". The Nobel press release splits the credit clearly:
Arber, working in Geneva in the 1960s, showed that "host-controlled restriction" of bacterial viruses (phages) involved changes in the viral DNA. He divided it into restriction (breakdown of DNA) and modification (methylation that protects DNA), and predicted enzymes that act at specific sequences.
Smith, in Baltimore, verified the idea with a purified enzyme and showed it cuts in the middle of a specific symmetrical sequence.
Nathans, also at Johns Hopkins, used Smith's enzyme to cut the DNA of the virus SV40 into 11 fragments and went on to build the first genetic map made by a chemical method.
Smith was born in New York in 1931 and died in October 2025, according to his Nobel facts page. Arber was born in Switzerland in 1929.
How did a lost DNA sample lead to a discovery?
In his Nobel lecture, Smith describes working in 1968 with graduate student Kent Wilcox on *Haemophilus influenzae*, a bacterium that readily takes up DNA. In one experiment they fed the cells labelled DNA from phage P22 and could not recover it. They suspected the cells were destroying the foreign DNA.
The next day they set up two viscometers, one with P22 DNA and one with *Haemophilus* DNA, and added cell extract to both. Long DNA makes a solution thick; cut DNA makes it runny. As Smith tells it, the viscosity of the *Haemophilus* DNA "held steady while the P22 DNA viscosity fell". The new enzyme needed only magnesium ions, which suggested it was simpler than the *E. coli* restriction enzymes known at the time.
Smith and Wilcox purified it and published in the Journal of Molecular Biology in 197090149-X). A second paper with Thomas Kelly worked out the base sequence of the recognition site90150-6).
Why did this enzyme matter more than earlier ones?
Smith's lecture explains that the enzymes from *E. coli* K and B recognise a site but cut at random points far from it. That makes them useless for cutting DNA into predictable pieces. Smith's enzyme cut at the site itself every time. Enzymes of this kind, which need only Mg²⁺, became the tools of genetic engineering.
The sequence Kelly and Smith found reads the same on both strands in the 5′ to 3′ direction, the kind of DNA palindrome described in NCERT Class 12 Biology, Chapter 9. Smith's lecture describes it as a six-nucleotide sequence ending Pu-A-C, cut evenly in the middle, which is written GTPy↓PuAC (Py is a pyrimidine, C or T; Pu is a purine, A or G).
Worked example: how often should an enzyme cut?
Assume, as Smith did, a random DNA sequence in which each of the four bases is equally likely.
EcoRI (GAATTC). Each of the six positions must be one specific base, with probability 1/4.
Probability = (1/4)⁶ = 1/4096, so EcoRI should cut about once every 4096 base pairs.
Hind II (GTPy PuAC). Four positions are fixed (G, T, A, C), each 1/4. Two positions only need "a pyrimidine" or "a purine", each 1/2.
Probability = (1/4)⁴ × (1/2)² = 1/256 × 1/4 = 1/1024.
Smith's lecture gives exactly this: the enzyme "would be expected to cut once every 1024 base pairs", which agreed with what they had seen.
Common mistakes:
Using (1/4)⁶ for every six-base site. Degenerate positions (Py, Pu) raise the probability.
Forgetting that real DNA is not random, so actual fragment sizes differ from the estimate.
What happened to Smith's bacterium later?
In 1995, a team including Smith and J. Craig Venter published the complete genome of *Haemophilus influenzae* Rd, 1,830,137 base pairs, then the only complete genome of a free-living organism. The bacterium that gave the first site-specific scissors also gave one of the first full genome sequences.
To see how these enzymes were used to build the first recombinant plasmids, read our post on Cohen and Boyer. For the exam view of the tools, see Biotechnology principles for NEET.
What can a student take from this story?
Notice anomalies. A missing sample was the clue.
Pick a fast, sensitive test. Two viscometers answered the question in a day.
Turn biology into numbers. The 1024 estimate let Smith check his sequence against his fragment sizes.
If you want to practise these probability problems with a solution that shows each step, the Fermi Problem Library has a step-by-step solver that explains the reasoning rather than printing an answer.
Next step: work out how often a four-base cutter such as one recognising GATC should cut random DNA, then check your answer is 4⁴.
