Killing Cancer From the Inside

There are 10,000 natural killer (NK) cells in every drop of blood (1) – and yet most people haven’t heard about them. That’s a shame, because these white blood cells are particularly good at killing cancer cells as well as virus-infected cells. They have the unique ability to detect cells under stress and can respond to abnormal states much faster than other components of the immune system. They can even detect cells that, due to a lack of MHC protein expression, are effectively hidden from surveillance by other lymphocytes.

Because these cells are so fascinating, much of our work has focused on trying to see what happens at the point of contact between NK cells and other cells. How do NK cells decide whether or not another  cell is diseased and should be killed? What occurs at that point of contact? Some years ago, we (along with other researchers like Avi Kupfer at Johns Hopkins and New York University’s Mike Dustin) learnt to our surprise that the contact immune cells make with other cells are reminiscent of what happens between neurons – and coined the term “immunological synapse” to describe it (2).

It has been difficult to find answers to a lot of questions that have arisen about the immunological synapse, though, because we simply haven’t had the technology to investigate them. But recently, super-resolution microscopy has come to the fore, especially with the 2014 Nobel Prize in Chemistry awarded “for the development of super-resolved fluorescence microscopy,” (3). This technology gives us the power to break through the diffraction limit and see structures as small as individual molecules. The techniques that our team use include photoactivated localization microscopy (PALM), stochastic optical reconstruction microscopy (STORM), structured illumination microscopy (SIM) and stimulated emission depletion (STED). Most of these methods rely on allowing only a few molecules in a sample to fluoresce and applying a mathematical model to find the positions of those molecules. We keep repeating the process, a few molecules at a time, until we’ve located every molecule in the sample. By using these techniques to study NK cells in multiple myeloma, we have been able to understand more about immune responses in disease states (Figure 1).

The two aspects of activation

Multiple myeloma is a cancer of the plasma cells, a specific type of B cell that produces antibodies. NK cells, which specialize in killing abnormal cells, are initially very good at dealing with multiple myeloma – but over time, the NK cell response decreases. One of the “blockbuster” drugs used to treat myeloma patients is lenalidomide, an immunomodulatory compound derived from thalidomide. We already knew that the drug worked to inhibit cancer cell growth directly, but whether or not it could also help NK cells attack the cancer cells was less clear.

We started by asking whether or not lenalidomide directly impacts on the NK cells’ ability to deal with myeloma. We looked at the release of interferon gamma (IFNγ), a small compound produced by NK cells that is involved in activating the immune system, in relation to the lenalidomide dose. We found that when we increased the dose of lenalidomide, we saw an increased IFNγ response from the NK cells. After verifying that it was a direct effect on the NK cells, we noticed something interesting – that lenalidomide actually seems to have two effects: not only does it make more cells produce IFNγ, but the amount of IFNγ secreted by each cell also increases (4).

The first effect – an increase in the number of cells responding – was easy to explain. At the point of contact, ligands on a cancer cell bind to receptors on the NK cell’s surface. Once enough ligands have bound, the NK cell is activated and attacks the cancer cell. We titrated the number of activating ligands versus the NK cells’ responses and found that when we added lenalidomide, the threshold shifted downward. And because fewer ligands were needed to activate each NK cell, more of them could respond to the same number of cancer cells. Lenalidomide is very effective at lowering this threshold. But that only explained one of the drug’s effects. How is it that each NK cell also secretes more IFNγ after lenalidomide treatment? To try and understand this, we turned to super-resolution microscopy to watch what happened.

Imaging inside immune cells

Underneath the surface membrane, NK cells have a dense cortical meshwork of actin. That raises a big question: how do vesicles containing cytokines, or proteins that kill diseased cells, move through this meshwork to exit the cell? We thought we had the answer in 2009, when we saw that the actin moves out to the periphery of the cell. It looked like the center of the synapse was entirely cleared of actin. But we were wrong, because our microscopes weren’t good enough. In 2011, we revisited the problem with super-resolution microscopy and saw that, in fact, the actin doesn’t entirely clear from the center of the cell. We wrote some software to analyze the periodicity of the meshwork and discovered that when the NK cell is activated, the periodicity changes and the holes in the mesh enlarge (5) (Figure 2).

We then wanted to know where in the cortical meshwork there were holes large enough to allow vesicles (which average about 250 nm in diameter) to pass through. Using a super-resolution microscope, we determined that, even after activation, the granules can’t pass through most of the meshwork. Only in small penetrable regions (about 4 percent of the total surface area), where the holes are largest, can they fit through (6). Lenalidomide increases the degree to which the meshwork opens up in activated NK cells, so that there are more regions from which vesicles and granules can be secreted. But it doesn’t affect cells that have not been activated; it just augments their actin remodeling response, which may be important in increasing the overall cytokine release. In general, lenalidomide may be able to enhance the immune response to allow the patient’s own immune system to fight the cancer.