Science is a Process, not an Object.

Biotechnology and Genetic Engineering

Blog Entries

Pasted Graphic

Restriction Enzymes

Discovery of Restriction Enzymes

This was one of those serendipitous discoveries that characterize the history of biotechnology.
In the 1950s and 60s, several labs were working on a phenomenon in bacteria known as the "restriction/modification system." Basically, if you took the phage T4 (a virus that infects the bacterium E. coli). The phage looks like this, except that the colors are made up:

The image is again from Wikipedia. The phage infects some common lab strains of E. Coli. But, there were some strains that were "restrictive" to growing it. That is, if you tried to infect one strain of E. coli (known as RY13) with this T4, it was resistant to infection, while other strains were easily infected and killed by the same preparation of T4.
Rarely, one of the RY13 cells would become infected. What was more confusing was that phage purified from that rare infection could now infect other RY13 cells with ease. This became known as restriction/modification. RY13 could recognizes foreign DNA that got into it from the phage and somehow destroy it. But, if the infection escaped early elimination, the phage DNA was modified somehow so that it now was not recognized as foreign by other RY13 cells. This was a "temporary change." That is, all that mattered was what strain of bacteria the phage infected last.
This was sort of an innate immunity for bacteria. Genes were identified by standard mutational analysis. Mutant forms of RY13 were found that were not able to restrict growth of the phage (so, their innate immunity had been "broken" by the mutation). These strains were still able to modify the DNA of phage that infected it so that it was protected when infecting normal (wild-type, or "wt") colonies of RY13. That separate gene was also found. The genes were called EcoRIR (
E. coli Ry13 I Restriction) and EcoRIM (the "M" is for modification).

Some breakthroughs came in 1970 and 1971, when the protein encoded by EcoRIR gene was shown to be a nuclease—an enzyme that cuts DNA—and the modification gene was founds to encode a "methyl transferase" that added an additional methyl group to certain Adenine bases in DNA. The enzymes were very specific. The nuclease, which became known as EcoRI (E coli Restriction enzyme I…that's a Roman numeral one). The name is pronounced like "echo-R-one."
Both enzymes were really specific. EcoRI cut DNA at the sequence 5'-GAATTC-3'. You may notice that the "bottom" strand would be the same, 5'-GAATTC-3'. The cut in the phosphodiester backbone occurred on both strands between the G and the first A.
EcoRI site

The methyl-transferase adds a methyl group (CH3) to the second A on each strand. That doesn't change the hydrogen bonding with T, but it does prevent the EcoRI nuclease from cutting the DNA. Thus, we have the explanation for this innate immunity:
  • Phage grown in a modification-minus strain does not have the methyl group on the second A of the sequence GAATTC
  • When infecting strain RY13, the EcoRI nuclease cuts the phage anywhere that sequence occurs. There are 4 bases, A, T, G, C. If the sequence is random, the sequence GAATTC should occur (¼)6 or roughly every 4000 basepairs. The phage DNA is nearly 170,000 base pairs long.
  • In the rare case where the infection succeeded, Phage DNA replicated in RY13 would get modified with the methyl group at the A. After that, it would infect other RY13 bacteria easily.
  • Phage grown in one of the modification-minus strains would again be unable to infect RY13 (except rarely, as before).

Again, the site is a palindrome, read the same way backward and forward (if you read 5'→3', as all good molecular biologists do. This is important from an evolutionary point of view. The protein is actually two identical subunits, each of which reads ½ the site and cuts one strand. Together, they recognize the required 6 base sequence and cut both strands. The methyl group, if present, sticks out in the major groove and prevents the restriction endonuclease from cutting.

OK, here's the big "what can I build with this" question. Because of the staggered cut, any DNA cut with this enzyme would have a 4-base 5' overhang with the sequence AATT. So, two different pieces of DNA cut with this enzyme would have complimentary "sticky" ends. If you mixed them up together, the four bases would hold the two different pieces of DNA together. The backbone would still be broken. But, if the temperature were low enough, they would stay together. If you mixed that up with the enzyme DNA ligase and ATP for energy, ligase will re-join the backbone, splicing the two pieces together.
Herb Boyer's lab at UCSF was the first to show this. In 1977, along with Bob Swanson, Herb founded a company called Genentech, where I worked in the lat 80s and early 90s. Kathleen Dana and Jeremy Nathans, found that the mammalian virus SV40 had only one site in its genome that was GAATTC. They cut it there and were able to splice it into one of Boyer's plasmids. Biotechnology was born.

Rich Roberts, whom I knew well when I was at Cold Spring Harbor Lab in the early 1980s went another route. He realized that there would be similar systems in lots of other bacteria. He started out purifying restriction enzymes from lots of different strains. Each one would cut at a specific sequence. BamHI cuts at GGATCC, between the two Gs, HinDIII cuts at AAGCTT, between the two As. He started a company called New England Biolabs (NEB), where most of us get our enzymes to do this. Currently, NEB sells around 290 different enzymes with different sequence specificity.
Basically, we got to the point where just about any two sequences could be spliced together. Newer technologies have really removed all the boundaries. I can splice anything these days.

Here is a link to a chapter on this from the
perspective of Paul Berg. He shared the Nobel price in 1978 for the cloning SV40 DNA into a plasmid (the work was done by Kathleen Dana, who was a grad student. She was left off the prize, sadly).
Here's another
perspective from Rich Roberts. The Berg biography is less on point for us. But, you might want to look at Rich's paper. It's got some useful stuff in it.
Rich would go on to win the Nobel for another accomplishment…I'll talk about that later. The person doing the work on that paper was Louise Chow, who was my boss at Cold Spring Harbor. She was left off the Nobel Prize…sensing a pattern of who wins and who does not?
Call that the "Jim Watson Effect."

MathJax Font