Recently my heart was filled with joy to see a paper published by the laboratory where I did my PhD.
David Schatz’s lab at Yale has obtained a crystal structure of part of the protein RAG-1 bound to DNA. My congratulations to first author Fang Fang Yin.
RAG-1 and its sister RAG-2 form the scissors that cut the DNA in immature white blood cells during VDJ recombination, a process where the developing immune system slices and reassembles antibody genes in a unique, partly random way.
Because VDJ recombination can happen in a different way in each cell, the immune system has a library of millions of antibodies that can react with many different antigens out in the environment.
The signals on the DNA that mark off the gene segments that get cut and pasted together again come in two flavors. They look like this:
Antibody gene segment — CACAGTG – 12 letters – ACAAAAACC
Antibody gene segment — CACAGTG – 23 letters – ACAAAAACC
The puzzle that attracted me to this area of research was this: the RAGs need one of each kind, 12 and 23, to cut the DNA. That’s weird.
Another weird aspect is that the part of the DNA opposite the CACAGTG is turned into a hairpin, with one strand of the DNA double-helix joined to the other.
We can imagine a process like this: the RAG proteins grab one of each signal and pull them together, forming a loop. What do the proteins look like when this loop forms? What are the geometric requirements for assembling this complex?
These are the questions I was trying to answer, or at least move closer to answering, during my PhD research.
Besides being a fascinating puzzle and the centerpiece of the adaptive immune system, VDJ recombination has connections to:
- DNA repair, because all sorts of monkey wrenches can interfere with the immune system’s ability to smush the DNA back together again
- Evolution, because the transposon-like arrangement suggests how antibodies suddenly appeared in vertebrate biology millions of years ago
When I started it hadn’t rigorously been proven that the RAGs actually cut DNA directly. With the help of Thomas Leu, a postdoc from Switzerland, I put together a system for examining how they cut DNA in the test tube. I also did some experiments where I crosslinked the RAGs to the DNA with UV light.
But actually getting a crystal structure, that’s really impressive and difficult. Now David’s group (and other scientists) can start to build a model of the whole complex.
X-ray crystallography might not allow visualization of an entire ready-to-cut structure with both 12 and 23 signals because the whole shaky bridge might be too flexible for crystals to form.
The structure published in Nature SMB doesn’t include the CACAGTG part, which is where the DNA gets cut, or RAG2, which is necessary for cleavage to occur.
But it does show that two RAG proteins intertwine and that unit binds to two separate ACAAAACC elements at once. That’s a great step forward.