By Abby Van Voorhees, MD, July 02, 2012
In this month’s Acta Eruditorum column, Physician Editor Abby S. Van Voorhees, MD, talks with Thomas S. Kupper, MD, about his recent Nature article, “Skin infection generates non-migratory memory CD8+ T(RM) cells providing global skin immunity.”
Dr. Van Voorhees: Immunity to infections has been thought to be based in T cells found in the lymph nodes and blood. Can you tell us what made you think about the local immune memory found in skin?
Dr. Kupper: Almost seven years ago, in work that was led by Rachel Clark, MD, PhD, who was in my lab at the time and now has her own lab, we found quite by accident that if we looked in normal, un-inflamed human skin and tried to extract T cells from it, surprisingly large numbers of T cells would come out of that skin. It turns out that there are about a million T cells per square centimeter of skin; if you do the math and extrapolate to the surface area of a person that’s about 20 billion T cells. That’s twice as many as you have in your blood. Virtually all of the T cells in your skin are memory cells; they have skin-homing markers, meaning they express things like CLA and CCR4 on their surface. There are no nave cells in there, so it’s a much less heterogeneous population than what circulates in your blood.
Having made this observation and having looked in various skin diseases and finding that there were more T cells in T-cell inflammatory skin diseases but not that many more — it wasn’t a 50:1 ratio, it was more like 3:1 — we had to really try to understand how these cells got there. We moved into some mouse models, because while human studies can give you snapshots of what’s happening at a point in time, they can’t really tell you how those things got there. We made some mouse models, including a model in which we infected the skin of mice with vaccinia virus, which happens to be the way the smallpox vaccine is given. It’s an infection of skin with a live virus that grows well in skin. What we were able to do is actually track the T cells that were made in response to that skin infection. It turns out that the first time an animal sees the virus there are no virus-specific T cells. However, there are cells circulating in blood and lymph nodes. But as a result of the infection, you make a very large population of virus-specific T cells, initially in the draining lymph node, and then these cells are distributed all over the body. They are distributed most effectively to the infected skin site; they’re also seeded to all different areas of skin. It’s a kind of carpet-bombing of your skin with these cells. These cells stayed there for a long time — a significant portion were still there a week, a month, two months, six months, and even a year later.[pagebreak]
Dr. Van Voorhees: What kinds of cells carry this information? Do these cells communicate with other parts of the skin or do they provide a strictly local immune protection?
Dr. Kupper: Virus-specific T cells, both CD4 and CD8 cells, carry the information. The ones we decided to take a very close look at are the CD8 cells. These are killer cells designed to kill virus-infected cells. The CD8 killer cells for vaccinia virus were the ones we did our more extensive experiments on.
What we found is that they’re remarkably effective at getting into skin and living in skin for long periods of time. The remarkable thing we found is that we could get rid of other parts of the immune system — B cells and even CD4 cells, with genetically engineered models — and it turned out these CD8 cells still went to skin.
The next question was: What do these cells do there? Are they just artifacts or are they actually disease-protecting T cells? The way to find that out was to challenge the skin with vaccinia and ask how rapidly virus disappears from the skin, which is a function of virus-infected cells being killed by T cells. It turned out that these T cells were very effective at rapidly eliminating virus. And not just the T cells at the spot where the virus had infected before, but really all throughout the skin. We went on to show that this was not due to antibodies but was all due to T cells. We finally asked, how do we know these T cells in skin are more important that the T cells we know circulate in blood? A significant number of T cells also remain in blood and in lymph nodes. We wanted to find out which ones were more important and whether the T cells in the skin were really doing much at all.
So we devised experiments that showed us that the cells in skin actually could protect without the participation of the cells circulating in the blood. The skin-resident memory T cells were effective all by themselves at eliminating virus from skin. That was encouraging because it told us all those T cells we found in skin are there for a reason: They’re protecting us from things that we’ve encountered previously through skin.[pagebreak]
Dr. Van Voorhees: Are they specific for a particular virus or are they non-specific?
Dr. Kupper: The ones that were generated in response to a viral infection are specific for that virus. But as you encounter many things over your lifetime — a virus one day, staph another day, strep another day — you accumulate these cells in skin and they live quite harmoniously together. So your skin is basically kind of a library of T cells that have encountered things through your skin in the past and are ready to respond in the future.
What works for host defense can also be subverted against the host. In contact hypersensitivity and with contact allergens, for instance, T cells don’t know whether what they’re responding to is a viral peptide or a hapten-modified peptide from a contact sensitizer. The same way you have virus-specific T cells and bacteria-specific T cells in your skin, you also have poison ivy-specific T cells and nickel sulfate-specific T cells. Those remain in your skin long-term as memory cells. We believe that’s what mediates allergic contact dermatitis. We believe the same is true in psoriasis. Whatever the autoantigen is in psoriasis, the T cells that are specific for that are in your skin and they remain there as resident populations of T cells. In between flares of psoriasis we think what you’re doing is quieting these cells down but not removing them from the skin.
Dr. Van Voorhees: Does repeat exposure lead to higher levels of these protective cells?
Dr. Kupper: Exactly. That’s also an experiment we did in the mouse model. We thought that would make sense based on what we know about human disease and, indeed, the more times the mouse is exposed to the virus, even at completely different skin sites, the greater the absolute number of cells that accumulate in the skin. There are quantitative differences, too.
Imagine for things you’re allergic to, it may be that the first five times you encounter that thing you generate X number of cells in the skin and it’s only once you reach a certain threshold that you have enough allergen-specific cells in your skin, or certain parts of your skin, to mediate an allergic contact dermatitis response.[pagebreak]
Dr. Van Voorhees: Is there a mechanism to diminish levels of these protective cells?
Dr. Kupper: This provides a great explanation for why skin-directed therapies work in immunologically mediated skin diseases. I would always be puzzled by how a topical or a skin-directed ultraviolet light therapy could work in a disease like psoriasis or atopic dermatitis or even eczematous dermatitis if this was truly a T-cell mediated disease and the important T cells were circulating through blood and lymph nodes and just getting into skin episodically. But if you now consider that most of the disease-causing T cells in these diseases are actually skin-resident T cells, it makes sense that you’d be able to quiet these T cells down and in some cases eliminate them altogether — although I think most of the time we’re simply quieting them down.
Dr. Van Voorhees: Maybe that explains why, in psoriasis, people’s skin goes into remission with ultraviolet therapy; their skin stays clear until they repopulate. Many of my patients would be glad to understand why that occurs!
Dr. Kupper: Exactly. And that would explain why you have the lesions coming back in the same areas.
We used to do a lot of hand-waving to explain that before we made this kind of stunningly simple observation. It’s kind of shocking we didn’t know this 10 years ago, but it just shows that if you don’t look for something you won’t find it.
Dr. Van Voorhees: You showed that the skin-resident T cells don’t need the circulating T cells to protect the host. What if you get rid of the skin-resident T cells? Can the circulating T cells still do the job?
Dr. Kupper: We asked that question through a complicated series of experiments involving parabiotic mice; suffice to say we were able to make mice that had only circulating central memory cells versus mice that had only skin-resident cells. And then we looked at them side by side to see if they could get rid of a second viral infection equally. It turned out that the mice that had the skin-resident cells, as we predicted, got rid of the viral infection right away. The mice that had the circulating cells but not the skin-resident cells, however, had a really rough time getting rid of the infection. They ultimately did, and they were better than mice that had never seen the virus before. That memory was good for something. But it was a couple of orders of magnitude less effective than the skin-resident cells. These circulating cells are really a backup reserve force; they’re not the cells really in the front lines.[pagebreak]
Dr. Van Voorhees: Does this inform us about the mechanism of immune protection at the skin surface? Is this the basis for the way vaccines allow for protection as well? Have these same populations of cells been seen in these other sites?
Dr. Kupper: I think it’s the way our immune system protects us from infection. Things we see through the skin help us generate large populations of skin-resident cells. Things we see through the lung like influenza generate large populations of memory cells that hang around the lung — and we’ve actually demonstrated this in humans. The same, we believe, is true with rotovirus in the GI tract.
Based on this, our strategies for vaccination probably aren’t as effective as they could be. We’re delivering an antigen into the muscle, which is crazy: Muscle has never evolved any sort of method of protecting the host against infection, and it’s a crazy place to put something you’re trying to make an immune response to. The reasons things are delivered to muscle have to do with blood supply and ease of delivery. If you have a relatively untrained person it’s easy to find a gluteus or deltoid muscle. But it’s not the best way to immunize. When we look for protection after immunization, we look at antibodies in the blood, but our studies suggest that at least for viral infections these are, not unimportant, but less important than the resident T cells.
What we’d like to see are vaccines designed to generate robust populations of tissue-resident T cells to protect against infection. Imagine, for HIV, if you could generate populations of tissue-resident T cells that went to and remained in reproductive and anogenital mucosa — that’s really what you need for a vaccine. You don’t need T cells circulating in the blood.
Dr. Kupper is Thomas B. Fitzpatrick Professor of Dermatology at Harvard Medical School and chairman of dermatology at Brigham and Women’s Hospital’s Dana Farber Cancer Institute, as well as director of the cutaneous oncology program at the Dana Farber Brigham and Women’s Cancer Center. His article was published in Nature, Vol. 483, p. 227231. doi:10.1038/nature10851.