Matthew Krummel: Visions enumerated

Neurobiologists and embryologists have been doing 4D live imaging studies for some time. Why was there a lag in immunology?

I think, in part, there was a lag because there is still so much being found out about these molecules. I mean, the T cell receptors and the B cell receptors are still a subject of intense scrutiny. Because this is so complex, it might have taken longer to realize that spatial dimensions play into decisions that the immune system makes, that there's a bunch of different cell types, chemokines, and cytokines interacting in one microenvironment. Now we're getting to the phase where the question is how the immune system works in real space and time.

Have you faced any resistance from other immunologists when presenting an image rather than a quantitative chart or table?

Yes, I think people do have problems with that. It takes a re-appreciation of what an image is. Human eyes and brains are actually incredibly good at noticing quantitative information. We can see when something is a tad bit brighter or moving a little faster. At the back of the retina, individual pixels record how bright something is. A camera is nothing more than a spatial array of detectors. And so from that standpoint, there's an infinitely large number of things that you can quantify in an image.

Still, there is a pushback in the field, which has been using certain techniques for a very long time to do things in vitro. Yet how cells behave in their native environment is so important. This is where the biochemistry takes place, so that's where you have to meet it.

How can 4D imaging help us tackle diseases like cancer?

Historically, we've been working in a flat, two-dimensional space to look at cancer cell interactions. The way people study tumors and the T cell interaction in the tumors has just been to cut out a slice and see where cells are. But when you look over time, you realize that cells are behaving in groups. Sometimes those groups don't spend much time together, but they consistently come back to each other. And you start to understand that it's not two cell types talking to each other, but 50 different cell types, each of which has its own set of favorite cells to send information to.

This behavior can suggest why an immune response isn't clearing a tumor. At some point, a tumor might subvert an interaction. So, if you can pick cell types that might be subverted, track them over time, and see who've they've been talking to, you can learn a lot.

How do you think tumors subvert T cells?

It may not be the tumor cells themselves. It could be other cell types in the tumor environment such as monocytes, B cells, NK cells, or even nontransformed epithelial cells. We think that something must be happening to interfere with the T cells because if T cells are given antigen and are happily primed—especially CD8⁺ T cells—they will kill the target. And yet the reality of human cancer is that the T cells, even if primed, don't kill the tumor. And they do get primed, only somewhere along the way, they get turned off. So the logical place to start looking for how that happens is in local tissues.

What has the two-photon microscope revealed about regulatory T (T reg) cells?

Visualization helps to answer questions about what T reg cells are doing. There exist 10 or 12 different models on how these cells work. And the reality is that T reg cells are probably a collection of different kinds of cell types that work in different ways depending on the setting. So the key is to find out how they work in a defined setting. In our case, to understand autoimmune diabetes we looked at the lymph nodes of a diabetic mouse and asked how the T reg cells were preventing effector cells from getting activated (3).

In those studies, two things became clear. The first thing was that T reg cells weren't physically capturing or sequestering T cells through some kind of stable interaction—there just wasn't evidence for this. And the second was that while T reg cells efficiently interacted with antigen-bearing dendritic cells, their presence dramatically prevented interactions of helper cells with those same putative antigen-presenting cells. So, there has emerged strong support for a model in which T reg cells physically alter either the dendritic cells or the microenvironment in which they live.

What are you doing on sabbatical now?

I'm working at the Institut Curie in Paris with a friend of mine, Sebastian Amigorena. Although a lot of things in his group are fundamentally the same, the environment is totally different. At UCSF, my neighbors are almost all immunologists; at the Curie, there are a lot of biophysicists working on the role of force, rigidity, and structure. These aspects come into play when we start to look at the immune responses in different organs and tissues because the design of the organ matters. That's one of the things that I didn't necessarily think I would be thinking about, but now I am. In order to look at cell motility, I'm using photomasks from microchip fabrication to pour molds of structures that cells must navigate. I'm also trying to develop some stromal cell line cultures to examine the role of the stroma in T cell activation.

If a leprechaun granted you the microscope of your wildest dreams, what could it do?

I think the ultimate thing would be to have down-to-the-nanometer scale resolution of everything in three dimensions over time. And, I mean, if I've got a leprechaun, then I'd want to complement the microscope with a technology that allows you to perturb things while observing them.

When you were in graduate school at Berkeley, did you live in the infamously wild co-ops?

No, actually I had a funny gig at Berkeley where I was a houseboy for two of the first psychotherapists to set up a clinic in Berkeley. It was like the first taco stand in Mexico… It was a giant mansion right at the base of the Claremont Hotel. I trimmed the hedges, walked the dog, and did other chores in exchange for rent. The cool thing about doing it was that it forced me to have a balanced life; I had to go home as part of my job.

Has your technique as a P.I. changed since you became a 33-year-old assistant professor?

Yes, a little bit. A strange thing happens when you become a P.I. The day before you become faculty, you can walk around your old laboratory, and people take your opinions as just another opinion. But the minute you become a P.I., when you spout an opinion, everyone acts on it. So there's this complete change purely because of the title you now have. And so I learned that, my relationships with people had to evolve.

One of the best formulations I've come up with in the last two years is really to give people a sense of their own. I'm not watching over them quite as much as I did in the first five years. If people think their career is in their own hands, then they act more like their career is in their own hands.

Besides the science, are you enticed by simply creating beautiful images?

I very much like the aesthetic aspect of it. It's a nice field to be in because it humanizes the sciences. Other scientists might trivialize an image and say, “Yeah, nice photo,” and not think that it's quantitative. But at the other end of the spectrum, if you have a nonscientist look at it, they say, “Beautiful, I can't believe that's going on in my lymph node!” You tend to blow people's minds with this kind of stuff. If you show a layperson a gel, they have no idea what it means. Compare that to showing them videos and saying, “These are the cells trying to clear a tumor.”