Led by Judy Lieberman and Tom Kirchhausen, researchers at the Immune Disease Institute and the Program in Cellular and Molecular Medicine at Children's Hospital Boston (PCMM/IDI) have discovered that two types of immune killer cells use the host cell's membrane repair pathway and the pore-forming protein perforin to deliver lethal granular enzymes, or granzymes. As reported in Nature Immunology, dramatic improvements in visualization technology allowed the team at IDI to be the first to watch perforin and granzymes as they leave the killer cell, enter the target cell, and cause apoptosis-all in real time.
Though cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells belong to different arms of the immune system (adaptive and innate, respectively), both use the same weapons to attack host cells infected with viruses or other pathogens and those showing dysfunction, including tumor cells. First the killer cell makes contact with the target cell, forming a tightly controlled interface called the immune synapse. Just as with the neural synapse, vesicles formed on one side (the killer cell) release molecules that have their effect on the other (the target cell).
It is well known that the vesicles released by the killer cells contain perforin and granzymes. However, exactly how granzymes reach the target cell's cytosol, or intracellular fluid, has been unclear. The simplest hypothesis was direct delivery: perforin makes holes in the plasma membrane through which the granzymes enter.
However, the new findings by the team at IDI fundamentally revise that initial model. Dr. Kirchhausen explains, "One of our new discoveries is that those holes are actually too small to let granzymes through. But they are large enough for calcium ions, and there's sufficient damage to induce a repair response, a form of wound healing."
Dr. Lieberman adds, "People previously thought that the membrane repair process mostly involved internal vesicles donating their membrane to patch holes. But as we reported in an earlier paper spearheaded by the same postdoctoral fellow in my lab, Jerome Thiery (Thiery et al., Blood 2011 115:1582-1593), it actually consists of very active removal of damaged membrane into endosomes inside the cell."
In the newly proposed model (illustrated below), after perforin makes small holes in the target plasma membrane, and the damaged membrane as well as perforin and granzymes are endocytosed, target cells rapidly form large vesicles positive for early endosomal antigen-1 (EEA-1) that contain both perforin and granzymes, which the team at IDI call 'gigantosomes.'
Dr. Lieberman and other scientists had previously shown that granzymes must reach the cytosol to find their targets and activate killing by apoptosis, or programmed cell death. Perforin, having already made tiny holes in the plasma membrane, also forms membrane pores in the gigantosome, gradually releasing granzymes. Then the final discovery by the IDI team: gigantosomes actually rupture, completing the one-two punch that finally kills the target cell.
Dr. Kirchhausen credits live-cell imaging microscopy with allowing the new findings. "Ten years ago, you might have taken pictures with an electronic camera and a standard microscope. Each image needed an exposure of a second or two. If you wanted three dimensions, like the optical sections of a CAT scan, recording 10 images could take perhaps 20 seconds. However, everything in the cell moves so fast, it would all be a blur and you would have seen nothing useful. Our new instrumentation provides sharp images of objects inside a cell 100 or more times faster than that."
Viewed sequentially, these images form an intracellular 3D movie. Dr. Kirchhausen continues, "We actually see the molecules first in the vesicles, then in the synapse, in the gigantosome in the receiving cells, and then in the cytosol. We have also been able to observe how the gigantosomes break."
Dr. Lieberman believes their findings may be applicable to other instances in which cells ingest something that eventually reaches the cytosol, such as viruses, bacteria, and RNA. "This is a great model with which to begin to think about how undesirable things get into the cytosol or how to manipulate this biology for therapeutic purposes, like getting RNA drugs into the cell."
Dr. Kirchhausen adds, "The normal biology is that anything taken into a cell is always contained within a vesicle; small, large, various shapes, but they're all vesicular or tubular endosomal carriers. They're used for transport to the lysosome for destruction; moving something across and dumping it out the other side of the cell; or as in the placenta, transporting antibodies from the mother to the baby. But what is interesting here is that the container loses its integrity."
Drs. Lieberman and Kirchhausen continue to collaborate. Dr. Lieberman is very involved in developing small RNAs for therapeutic purposes. "There are many methods for getting RNAs into endosomes, but very few get the RNAs out of the endosomes into the cytoplasm so they can do their work. If we can image the small RNAs as they are taken up by cells and see where they go, hopefully we can use the same technology to design new RNA drugs."
Dr. Kirchhausen's lab is applying the new visualization technology to study viral entry. "Some viruses like HIV have membranes; they fuse with the host cell to reach the cytosol. Other viruses without membranes must physically break into the cell. We color-labeled a virus and actually watched it in real time during translocation. That would have been impossible with the earlier imaging technology."
Dr. Lieberman sums it up: "I think the message is that a picture is worth a thousand words; when you see it happening before your eyes, it's just so much clearer than studying hundreds of immunoblots and immunoprecipitations. It's very powerful."
Thiery J, Keefe D, Boulant S, Boucrot E, Walch M, Martinvalet D, Goping IS, Bleackley RC, Kirchhausen T, Lieberman J. Perforin pores in the endosomal membrane trigger the release of endocytosed granzyme B into the cytosol of target cells. Nat Immunol. 2011 Jun 19. doi: 10.1038/ni.2050.
