It was my great honor to be your President. Before I write on the topic of my President’s Address, I want to comment on a number of things that I considered to be highlights for the American Association of Immunologists (AAI) in the last year. The year 2018 was the 105th anniversary of our society—the largest professional society of immunologists in the world. We have over 7600 members, and we welcomed new members who were delighted to join our society.

Wayne M. Yokoyama

What does the AAI do? I want to emphasize several points in this article: we had over 3300 attendees from 40 countries at our annual meeting this past year (2018), which included 700 speakers and 197 sessions, 130 symposia, 9 plenary lectures, 72 poster abstract sessions, and 1780 abstracts. It was a very big meeting, and we owe a great deal of gratitude to Gene (Eugene M.) Oltz from Washington University in St. Louis, who was the program chair, and the rest of the Program Committee.

Those are the statistics, but I have heard a lot of people say, “The AAI meeting is too big; I’d rather go to my small [you name the conference] meeting.” I always loved the AAI meeting and have regularly attended for many years because over the course of my career, I have met many other people who are like me in the sense that we are all interested in immunology. But when I go to my specialized meeting, I don’t see most of these people. I see the people who are working in my specific area of interest, for sure, and that’s great for getting up to date on my field, but it’s the AAI meeting where I see everyone else. I see them at the sessions at spectacular talks, given by people I know! But I also see them in the hallways, in hotel lobbies, at the bars, and on the street. It’s wonderful to see again people I might have met for the first time 20 or more years ago. It’s really fun to catch up with them and to meet their friends, other immunologists, often for the first time. So, AAI meetings offer great opportunities for networking in immunology, and there’s no better chance for you to do that than at an AAI meeting.

I challenge all of you to come to our annual meeting to catch up with old friends and to get to know someone you’ve never met before, probably because they are not in your specific field of immunology. But it’s amazing what you will learn from them and vice versa. And you can benefit from that networking in other ways, perhaps even advice on experiments, or job opportunities! So I hope you will attend upcoming AAI meetings!

Another important aspect of what the AAI does is that we now publish two journals, including the new ImmunoHorizons. I thank the leadership of the journal, Leslie Berg and Michael Krangel, for helping to get this journal up and running. And of course, we have our flagship journal, The Journal of Immunology (The JI), the premier immunology journal in the world; we have always published the most immunology papers. Almost none of the manuscripts are triaged, and nearly all are reviewed. I really want to thank Pam Fink from the University of Washington, who was our editor in chief until she handed over the reins on July 1, 2018 to Gene Oltz from the other Washington University, the one in St. Louis. Again, thank you Pam and Gene for devoting a good portion of your time and effort to support the AAI and The JI. Finally, it is important to note that the journals are an important source of revenue for the AAI, allowing our society to sponsor other activities in support of our membership and immunology, so I encourage all of you to publish in our journals.

A key AAI activity is that we support the next generation of immunologists with many career development awards. A new one, Intersect Fellowship Program for Computational Scientists and Immunologists, was just approved by the Council this past year. It’s a novel fellowship program; the idea is to bring computational scientists into immunology and vice versa. Because the details of this new program are on our Web site, let me summarize the program by saying we hope to find applicants who are computational biologists who want to spend time learning immunology in an immunology laboratory. Alternatively, we hope that there are immunologists that want to spend time in a computational biology laboratory. For those of you who fit this picture, please apply, and if it’s not for you, then encourage your friends to apply.

We also have a new fellowship program that we call the Career Reentry Fellowship Program. I’m especially proud of this because it arose from a suggestion I made to the AAI Council. It’s based on the observation that in the course of my career, I’ve been very fortunate because I managed to stay on track, if you will. But there are many others who have fallen off the track. Sometimes it’s for fortunate things like the birth of a child, but then there are others who got off track because they were ill. Or they were the primary caregivers when illness struck someone else in their family. Or there was elderly care involved. Or they served the country in the military. Or anything else that life threw at them and got them off track. For those whose careers went off track, it’s clearly very hard to get back on track. With this innovative new program, the AAI is doing something that nobody else is doing, to our knowledge, to really help the next generation of immunologists and scientists by helping them get back on track. I encourage you to learn more about it and to encourage your friends and colleagues to apply.

Another really important aspect of what the AAI does is that we’re advocates for immunology and biomedical research on Capitol Hill. And I think this is really important to note, particularly in this age, that each one of us has one voice, but together we have a voice of over 7000 people. The AAI, through Beth Garvey who headed our Committee on Public Affairs, and through her committee and the rest of the AAI staff, especially Lauren Gross and her colleagues, is tirelessly addressing and advocating for all of us as immunologists. In particular, they pay attention to the debates on the Hill and at the National Institutes of Health (NIH) with respect to the budget allocations, NIH trends, and other relevant government issues.

One particularly concerning situation happened a few months ago, related to the potential tax on graduate student tuition waivers, when it was being discussed as becoming part of the tax bill on Capitol Hill. The AAI was up front on this topic, and I’m pleased to say that that issue has gone away. I know a lot of our graduate students, certainly at my institution at Washington University, were very relieved at this outcome. I think they couldn’t do this type of advocacy all by themselves; they needed organizations like the AAI to help in this discussion, not only to advocate but to help find timely information and advise the best ways to voice our opinions.

I encourage you all to learn more about the AAI. When you come to the annual meeting, we have a business meeting, and the Committee on Public Affairs has been sponsoring town hall meetings to invite more member input. You can visit the AAI booth. The AAI Web site is a good source to learn more about what your organization can do for you. There’s a lot that we can do, and if you’re not a member, I hope you will join to help advance your career and volunteer your efforts to help all immunologists.

Finally, I want to say a sincere thank you to everybody else I didn’t have a chance to mention for their support of the AAI. I thank Michele Hogan, who leads a dedicated professional staff at the AAI, but a large amount of the AAI’s work is from volunteers, those who review and edit papers for our journals, who serve on committees to advocate for immunologists on Capitol Hill, and who help the AAI with all of our activities in so many other ways. I really want to give them sincere thanks for their support of the AAI.

Now on to my presidential address, which typically allows the AAI President to write (and talk) about his or her career. The topic of my address is “Fifty Years (Well, Almost!) in Immunology.”

I got started in immunology in 1969; I don’t want to tell you how old I am, but I’m sure you’ll figure this out as I go on. The year 1969 was a tumultuous time in our country, just like it is now, but back then the consternation had a lot to do with the war in Vietnam. There were other issues, to be sure, but of course there were those who had a different idea of how to deal with the war: “Make love, not war.” So Woodstock was a big deal when it happened in the summer of 1969. Of course, one of the biggest deals that also happened that summer was when Neil Armstrong stepped on the moon on July 20, 1969. This obviously was historic; it was covered across the world, with newspapers everywhere having big headlines.

Meanwhile, I grew up in Hawaii, where it’s easy to be oblivious to many things happening in the rest of the world. A little bit more about my personal history because I think this might resonate with some of you, particularly in this time and age. Like many of you, I’m from an immigrant family; all four of my grandparents emigrated from Japan to Hawaii about 100 years ago. There was a time when they were welcomed; Hawaii was a territory of the United States at that time, having been annexed in 1898. But Congress later passed a bill to prevent further immigration from Japan (and other countries—a sad commentary on the current immigration debates).

In any event, my family had been in Hawaii for about 20 years when the Pearl Harbor attack happened. My parents were preteens at that time, and my paternal grandfather was taken from his home on Maui and interned in a camp in California, out of touch with his family for the duration of the war. This is despite the fact that two of his sons served in the U.S. Army, and one of them actually fought in the European theater and suffered from what now we know as posttraumatic stress disorder. A very unfortunate circumstance in our family (and many others), but, interestingly, nobody ever talked about it.

My parents met when they were students sitting next to each other in chemistry class at the University of Hawaii. They came from different islands, so they might never have met each other otherwise. They impressed upon me that education was important. I knew from when I was a little kid that I was going to go to college.

But this plan was almost derailed because my father died, suddenly, when I was 14 years old. I thought that there was just no way I was going to go to college when the breadwinner of our family died. My mother had not been working outside the home for many years—she was a homemaker at the time. She taught herself how to get into the education system as a teacher, ending up teaching in my high school. It was a little awkward growing up as a teenager, having my own mother in my high school, keeping an eye on me! But I was never a problem child.

Along the way, a really important mentor was Miles Muraoka, my high school biology teacher. Mr. Muraoka exposed me to the wonders of biology, and he helped me get a fellowship from the Hawaii Heart Association to work at the Blood Bank of Hawaii. So in 1969, as a rising high school senior, I had my start in immunology.

By the way, you can go to the AAI Web site, where we have a wonderful timeline set up by our history department. One of the highlights from 1969 is that the first flow cytometer was described in a paper from the Herzenberg laboratory in Science. At that point, it obviously wasn’t very widespread, and instead what was more well known—and this will resonate with some of the older immunologists—was the Ouchterlony method, also known as double immunodiffusion. The method used a series of wells punched out of an agar gel, layered on a glass slide, with the wells surrounding and equidistant from a central well. An antiserum could be pipetted into the central well, and then solutions containing different Ags were placed in the outside wells. If a precipitin line became clearly visible after a few hours between the central antiserum well and one of the surrounding wells, the line identified wells having an Ab and its corresponding Ag. This was fundamental.

The first chapter of every immunology text book for quite some time described the Ab–Ag precipitin curve that was the basis for what most immunologists did at that time. In fact, I recently did a little PubMed search and found that Ouchterlony or immunodiffusion accounted for one-third of all papers in The JI in 1969. But between 2010 and now, only one JI paper falls into this group. It’s not that the technique is useless, as immunologists learned a lot using this approach back then. (One notable example is the discovery of the “Australian antigen,” which led to the identification of the hepatitis B virus and the Nobel Prize!) (1, 2). But, of course, times have moved on, and there are other approaches that have replaced it.

For the rest of this article, I want to discuss studies that I participated in, describing some of the methods we used at that time, some of which were relatively new, and work my way to the present day. And, of course, I’m going to indulge you by using our work to illustrate how some of the approaches we used have become somewhat obsolete, even though we were able to make advances at the time.

Back at the Blood Bank of Hawaii, my mentor was Mitsuo Yokoyama. He’s not a relative. For those of you who don’t know otherwise, Yokoyama in Hawaii is like Smith on the mainland. There are a lot of Yokoyamas, and as far as I know, we are not relatives. Perhaps all of us should do a 23andMe, and we could learn otherwise! In any event, I was exposed to words like “immunoglobulin” that I couldn’t even pronounce, never mind spell. And we knew relatively little, of course, about how the immune system works.

What I did that summer was to determine the reactivity of lymphocytes from individual normal blood donors to alloantisera. I isolated lymphocytes (this was before the advent of Ficoll-Hypaque, which to me was revolutionary because of how difficult it was to do it the old way; I don’t have time to explain that to you) and incubated them with individual sera from patients with multiple pregnancies or transfusions. I then added a source of complement to cause Ab-dependent lysis. The tools I used in the laboratory included a repeating Hamilton syringe. You can still buy this instrument today, but it is not disposable, it is hard to sterilize, and you can’t change the tips. We pipetted lymphocytes and antisera into Terasaki plates that were the forerunner of the 96- and 384-well plates with which you are familiar. Development of the Terasaki plate itself was also a revolution because I used to make my own plates with a wax pencil and a glass coverslip!

Then what? Well, you look at the wells under the microscope. There’s evidence that I actually did this: a picture from the local newspaper in 1970. What was I looking at? I looked at the cells using the trypan blue exclusion technique for viability that we still use today. Some antisera lysed nearly all cells in most donors, whereas others had minimal reactivity to most donors. It was really amazing that a given antiserum would lyse all the white cells from a given donor. Then in other cases, it would do nothing, and in other donors, it would lyse a few of the cells, maybe just 20%. The patterns of reactivity clearly were very different between donors, and it was thought that these reactivities reflected “tissue types.” By the way, this cytotoxicity assay is one approach that led to the discovery of the HLA system, but we didn't know the details of HLA back then.

It was mind-boggling how complicated the patterns were, but nevertheless, we did make some findings that correlated with the grouping of the donors into different ethnicities. And as a result of that, I actually won the Hawaii State Science Fair and got to go for the first time to the mainland in the spring of 1970, to the International Science Fair in Baltimore, Maryland (Fig. 1). I even published a paper (3). But much more importantly, in terms of my career, was what happened that summer of 1969: the first organ (kidney) transplant was done in Hawaii.

Figure 1.

The author presenting his poster at the International Science Fair in Baltimore, MD, spring 1970.

Figure 1.

The author presenting his poster at the International Science Fair in Baltimore, MD, spring 1970.

Close modal

Kidney transplants, of course, now are commonplace, having started back in the 1950s and 1960s. But in Hawaii, the first kidney transplant occurred at a local hospital in the summer of 1969 while I was working at the Blood Bank. It was a very big deal because, prior to this, patients from Hawaii had to go to the mainland to get a kidney transplant. It was covered extensively by the local news media. I remember watching a black and white television and a news reporter asking the transplant surgeon something like, “Now why was it that you used the healthy kidney from the brother of the recipient?” I knew the answer: they were more likely to have identical tissue types, the same tissue types that I was studying in the laboratory! That was my epiphany, if you will, that started the course of my career.

I became convinced I wanted to have a career in “medical research.” Now, all of my early mentors were physicians doing research, so their advice was, “Go to medical school.” And I entered college with that idea. I went to the University of Rochester, and there I worked with Parker Staples, who was a physician-scientist studying autoimmune disease in NZB/W mice. What I didn’t realize is how much snow there was in Rochester. I had never seen snow before I got there, and when it was still snowing my freshman year in mid-May, I said, “That’s it.” I wasn’t a particularly good student at Rochester, in part because I was spending a lot of time in the research laboratory, trying to learn as much immunology as I could by being an experimentalist. But I did finish my bachelor’s degree and returned to Hawaii to attend my state medical school.

By the way, I never took a formal course in immunology. I tested out of the immunology course in medical school and informally taught my classmates immunology at night, all based on knowledge I gained by doing research in immunology laboratories!

At the University of Hawaii, I was fortunate to meet another physician-scientist, Eugene Lance, and was his first employee when I worked in his laboratory during the summer between my first and second years of medical school. Gene received his Ph.D. with Sir Peter Medawar, the father of the classic laws of tissue transplantation. Through a lot of encouragement from Gene, and actually getting a chance to meet Sir Peter when he came to visit Gene, I was inspired that I could have a productive career in medical research because a person who worked with a Nobel laureate said I was pretty good.

So off I went to the University of Iowa for clinical training. I did an internal medicine internship and residency and then a clinical rheumatology fellowship. I then joined the laboratory of Bob (Robert F.) Ashman, the new chief of rheumatology at the time. When I got back in the laboratory after all that medical and clinical training (about 7 years), I felt that I was home again, even though I liked clinical work. But that’s when I realized that the field had passed me by in the time I was doing clinical training, and I really needed to expand on the knowledge I gained in Bob’s laboratory (it was almost like getting a Ph.D. in Bob’s laboratory), so then I went off to the Laboratory of Immunology at the NIH in Bethesda, MD.

There I was mentored by Ethan Shevach, another physician-scientist who gave me a lot of guidance yet independence; he showed me how to pursue independent research by following the findings of our experiments. I also trained with David Cohen, who taught me the fundamentals of molecular biology. I worked on many things in Ethan’s laboratory, but two papers became the foundation of the work that I pursued when I took an independent faculty position at the University of California, San Francisco (UCSF) (4, 5). They are both JI papers that are summarized in this article.

In these papers, we discovered what we know now as Ly49A. I used the cDNA cloning system that was newly described at the time developed by Brian Seed, using an mAb called A1, produced by Osami Kanagawa against the T cell tumor EL4. The Ag was peculiar in that it was expressed on some T cell tumors, but when the Ab was used to stain splenocytes or thymocytes by flow cytometry, there was no signal. However, when splenocytes or thymocytes were radiolabeled, the Ab immunoprecipitated a characteristic disulfide-linked dimer, similar to findings with EL4, suggesting that the Ag was rarely expressed on splenocytes and thymocytes. I was able to expression clone the cDNA for this Ag from a cDNA expression library I produced from EL4, revealing a type II integral membrane protein with an external domain that resembles C-type lectins. We later renamed it Ly49, and later when we realized it belongs to a whole family of molecules, we called the original Ag Ly49A. (My wife, Lynn Yokoyama, believes that I named Ly49 after her, except there were 48 other molecules designated Ly before then!)

Ironically, I cloned Ly49A from a T cell tumor, but it turns out that Ly49A is not constitutively expressed on T cells. In the second paper, we used Southern blots (I think most people in my laboratory don't know how to do a Southern blot anymore) to demonstrate restriction fragment length polymorphic variants (RFLPs) for Ly49a. The paper has several blots showing a series of lanes, each having DNA from a different strain of mouse, digested with a restriction enzyme, run on a gel, blotted to nitrocellulose, and then probed with the cDNA for Ly49A. What we noticed relatively quickly was that there were a lot of different RFLP variants, suggesting that Ly49A itself likely displays allelic polymorphism. There were a lot of bands in each mouse as well, suggesting that there was a whole family of these molecules. We then used RFLP analysis of recombinant inbred mice to map the gene for Ly49A to the distal portion of mouse chromosome 6. Because all RFLP variants were inherited together as a group, it was likely that the entire putative Ly49 family was genetically clustered. [Later on, we cloned other members of the Ly49 family—another JI paper! (6)].

Simultaneously, a group in Dallas, Texas, directed by Vinay Kumar and Mike Bennet, was mapping the gene for the NK1.1 Ag. Their prior work strongly suggested that NK1.1 is the best marker of NK cells on CD3 splenocytes in the C57BL/6 mouse spleen. They had been using the same recombinant inbred mouse panel that we had employed for our RFLP analysis of Ly49a. So the gene for Ly49A must map next to the gene for NK1.1. Later on, we actually cloned the cDNA for the Nkrp1 family that includes Nkrp1c, which encodes NK1.1, and mapped Nkrp1 genes to mouse chromosome 6, next to the Ly49s (another JI paper!) (7). Thus, these data established the NK gene complex, a genomic complex containing families of genes for NK cell receptors.

These studies prompted us to look at Ly49A expression on NK cells, and we determined that Ly49A is constitutively expressed on a subset of NK cells (5). When I left Ethan’s laboratory and started my own laboratory, we had the observations that Ly49A is expressed by a subset of NK cells and appeared to belong to a polymorphic family of molecules, leading to the question of whether it is involved in “immune recognition” by NK cells. (I should note that many of my colleagues wondered why I gave up on other potential projects that Ethan would have let me continue, but encouragement from Ira Goldstein, who recruited me to UCSF, and Bill Seaman, who was also working on NK cells at UCSF, was instrumental in instilling confidence that we should continue our studies of NK cells.)

Franz Karlhofer, my first postdoctoral fellow, expanded NK cells in vitro from the mouse spleen by using high doses of IL-2 and then isolated Ly49A+ NK cells by panning and Ly49A NK cells by Ab and complement lysis, two techniques we don't use today. People in my laboratory today want to use flow cytometry or use fancy expensive beads to isolate these cells. Our old approaches are much cheaper and very efficient, but I can’t convince anybody to do it the old way anymore. In any event, we learned that Ly49A+ and Ly49A NK cells were otherwise fairly similar with respect to other surface molecules, so we asked the question, “Are there functional differences between Ly49A+ and Ly49A NK cells?”

The assay that Franz used was a standard radioactive chromium-51 release assay. Today, we have the institutional approval to use chromium-51, but nobody seems to want to use radioactivity in my laboratory. In fact, we get more people from around our university, knowing that we have the approval, to come and use our instruments than our own laboratory members. Anyway, Franz used a gamma counter to determine radioactivity in the cell-free supernatant from chromium-51–labeled targets incubated with Ly49A+ or Ly49A NK cells. When pictures of this instrument are viewed, what is dramatically evident is that there is no computer! Instead, the cpm were printed out for each sample on rolls of paper, just like a cash register, although I suppose many are not familiar with this either! Franz had to collect all those rolls of paper and use a standard formula to calculate specific lysis. He went on to examine a large number of different targets. Let me summarize those data.

We started with a small panel of different target cell lines and determined whether there was differential susceptibility to killing by Ly49A+ versus Ly49A NK cells. A number of lines were equally killed, clearly showing that Ly49A+ and Ly49A NK cells were functionally similar in that they had the same killing capacity. There were a couple of targets in which there was no killing observed by either NK cell subpopulation, so for the purposes of what we were trying to do, these targets were uninformative, and we stopped working on them. But there were several targets that were differentially killed by Ly49A+ versus Ly49A NK cells. Always when we observed differential killing of a target, the Ly49A NK cells could kill it, but the Ly49A+ cells could not. We puzzled about this for a while, wondering what the basis was for this apparent specificity.

We kept on ordering cell lines from the American Type Culture Collection, thinking maybe it had something to do with tissue origin. But it turned out that this was not the case, and instead, differential killing correlated with the H-2 (MHC in mice) haplotype of the targets. Interestingly, targets killed by Ly49A but not by Ly49A+ NK cells were either of H-2d or H-2k haplotype, regardless of tissue origin, suggesting target susceptibility had something to do with Ly49A recognition of MHC alleles, either MHC class I or class II. But interestingly, there were two targets originating from the H-2k haplotype mice that were equally killed by Ly49A+ and Ly49A NK cells, providing exceptions to the apparent rule.

When we looked at these targets in greater detail, we found that MHC class I was not expressed by either of these cells, strongly suggesting that the H-2 correlation with differential killing had to be MHC class I related. It was at that time that we became interested in the work of Klas Kärre, who had observed that NK cells seem to ignore targets that normally express MHC class I (8). But when MHC class I is downregulated, NK cells appear to attack. Because MHC class I is ubiquitously expressed as “self,” he postulated that NK cells attack in the absence of self (i.e., NK cells detect the absence of “self-MHC” or “missing-self,” according to the missing-self hypothesis).

There were two major hypotheses to describe how the missing-self hypothesis would work. One was the “masking” hypothesis: MHC somehow covered up the ligand for a putative activation receptor on the NK cell. In the absence of MHC class I, the ligand is “uncovered” or “unmasked,” and that is when the NK cell attacks. The competing “inhibitory receptor” hypothesis suggested an NK cell inhibitory receptor. When it engages MHC class I, it sends an inhibitory signal that turns off the NK cell. When MHC class I is absent on a target, that inhibitory signal is lost, allowing the NK cell to kill. Most people favored the masking hypothesis because at the time, thinking in immunology was dominated by the TCR, which, of course, is an “activation receptor.”

Although we didn’t start off our studies of Ly49A with the missing-self hypothesis in mind, Franz’s findings suggested that we might be able to contribute to the understanding of missing-self. He found a couple of targets that were derived from F1 hybrid mice that were both H-2b and H-2d. Interestingly, these targets showed the phenotype of the homozygous H-2d haplotype targets that were not killed by Ly49A+ NK cells, rather being killed like H-2b targets. These findings strongly suggested that we could take an H-2b target that was equally killed by both Ly49A+ and Ly49A NK cells, transfect it with the cDNAs from the MHC class I alleles of the H-2d locus, and figure out what MHC allele was involved. That is when I got on the phone and called David Margulies at the NIH, who worked across the hall from Ethan’s laboratory, and he sent us the cDNAs that Franz used to express in our targets.

When Franz derived transfected targets expressing H-2Dd, the Ly49A+ NK cells could not kill them, whereas Ly49A NK cells could kill the targets. Meanwhile, other targets derived from transfection of H-2Kd or H-2Ld remained equally susceptible to killing by Ly49A+ and Ly49A NK cells, just like the original parental target. We posited that Ly49A on the NK cell must be recognizing H-2Dd.

Franz showed that the Ab against Ly49A allowed the Ly49A+ NK cells to kill the H-2Dd transfectants; it’s like reversing inhibition, if you will. An Ab against the α3 domain of H-2Dd had no effect, but an Ab against the α1/α2 domains of H-2Dd reversed inhibition and allowed target lysis. Taken together, these data, indicating that Ly49A must specifically recognize H-2Dd on the target and inhibit NK cell killing, provided strong support for the inhibitory receptor hypothesis for missing-self recognition, although we did not directly refute the masking hypothesis (9).

Here, I would note that Francisco Borrego graciously nominated our Nature paper for the Pillars of Immunology series in The JI (10). I should also point out that I offered David Margulies coauthorship on our paper (9), but he graciously declined, a very nice gesture from a senior investigator to a beginning junior faculty member, one that I try to emulate today.

Meanwhile, extensive analysis was being undertaken on the human NK cell receptors. Studies led by Marco Colonna, Eric Long, Alessandro Moretta, Lorenzo Moretta, Lewis Lanier, and others led to the molecular identification a few years later of what we now know as the killer Ig-like receptors (KIRs). As the name implies, these molecules have Ig-like domains; they’re actually type I integral membrane proteins, whereas the Ly49s are type II and C-type lectin-like, as already mentioned. This led to a lot of consternation in the field for a few years. I knew I was right, I knew Marco and the others were right, but everybody else in the world thought somebody was wrong—but this is where persistence helps.

Now that we have whole genome sequencing available, it’s very clear if you look at human chromosome 12p13, the region syntenic to mouse chromosome 6, where the mouse NK gene complex resides, there are no functional human Ly49 genes. Meanwhile, if you look at mouse chromosome 7, syntenic to the human chromosome 19q, where the leukocyte receptor complex resides and which contains the human KIR genes, there are no mouse KIR genes. That’s despite the many other similarities between the human and mouse NK receptors for MHC class I, including the basis for inhibitory function residing in their shared cytoplasmic ITIMs (11). Parham and Gumperz were the first to point out that this is an outstanding example of convergent evolution, in which each species came up with a different genetic solution to a very important issue related to its survival (12).

I should emphasize to the students that this shows that you can’t always just do a simple BLAST search to find genes for molecules with homologous function. Sometimes you have to look deeper and think about what principles you are trying to reveal. We’ve been convinced over the years that insights from studies of Ly49s in mice will inform our understanding of the human KIRs and NK cells in general. Of course, it’s worthwhile doing this because it is much harder to study human immune responses. I think this notion is generally well accepted, although there are always some doubters, despite the history that most advances in NK cell receptor biology happened first with insights from studying the Ly49s.

Back to our Ly49 story. Interestingly, in many situations, even though MHC class I is downregulated, the NK cell doesn’t attack. This brings up our studies on murine CMV (MCMV), which can very effectively downregulate MHC class I on infected cells. We began a longstanding collaboration when Tony (Anthony A.) Scalzo at the University of Western Australia discovered that a single Mendelian locus, termed Cmv1, controls resistance to MCMV infections in certain strains of mice, like B6 mice, for example (13). Other mouse strains, such as BALB/c, are susceptible and die of infection, and resistant mice can be made susceptible if NK cells are depleted. Tony showed that Cmv1 was located on mouse chromosome 6, raising the possibility that it was somehow related to the NK gene complex. In collaboration with the Scalzo laboratory, we positionally cloned the resistance allele of Cmv1 and discovered it encodes an activation receptor in the Ly49 family, called Ly49H, that is required for resistance of B6 mice to infection and that was missing in susceptible mice (14). It took us about 10 years to positionally clone Ly49h; today, with whole-genome sequences available, I suspect it would be a lot faster.

In studies done by both our laboratory and Lewis Lanier’s laboratory, it was determined that the ligand for Ly49H is actually coded by the virus itself; it’s called m157 and is expressed on the surface of MCMV-infected cells (15, 16). In other words, the Ly49H activation receptor specifically recognizes a ligand expressed by infected cells, and this is required for NK cell activation. These findings and others helped show that NK cells express both activation and inhibitory receptors, and it’s the integration of their signaling events that results in whether the NK cell is activated and a target is killed.

Let me move on to a different Ly49 study led by Sungjin Kim in my laboratory (17). David Raulet’s laboratory had published many years previously that YAC-1 targets, the prototypic NK-sensitive target in mice, were poorly killed when exposed to NK cells from MHC class I–deficient mice (18). This phenomenon was not well understood at the time and could have been due to the effects of MHC class I on the NK cell receptor repertoire, for example. Sungjin took NK cells from MHC-sufficient B6 mice as well as MHC-deficient mice, exposed them to YAC-1, and then looked for intracellular IFN-γ production by use of a relatively new flow cytometry–based, intracellular cytokine staining assay (17). NK cells from MHC-sufficient mice readily made IFN-γ, whereas they did so very poorly when they were derived from MHC-deficient mice, in many respects recapitulating the Raulet results with bulk killing assays. Sungjin then stimulated the NK cells with anti-NK1.1 Ab immobilized on plates, allowing cross-linking of an activation receptor, NK1.1, which is expressed on all NK cells at normal levels, regardless of MHC class I expression. Just as with the YAC-1 targets, NK cells from the wild-type mice were much better at making IFN-γ with NK1.1 cross-linking than NK cells derived from MHC class I–deficient mice.

I want to emphasize the use of intracellular cytokine assays to determine individual NK cell responses compared with bulk populations in killing assays that Franz used a few years earlier. To distinguish what the Ly49A+ cells were doing versus Ly49A cells, Franz had to isolate and purify the cells first, before putting in the killing assay. But in the intracellular IFN-γ assay, Sungjin could just stimulate unfractionated NK cells and use flow cytometry to determine what happened to individual NK cells, allowing relatively easy phenotyping of the responding cells. And that turned out to be incredibly informative.

Sungjin then used, as a source of NK cells, different MHC-congenic mice, all on the C57BL/10 genetic background (17). As far as we know, they all had the same NK gene complex and the same Ly49 alleles but differed in their MHC alleles. Sungjin found the Ly49A+ NK cells from several mice did not work particularly well in the immobilized anti-NK1.1 assay. However, when they were derived from B10.A, B10.D2, or B10.D2 (R107) mice, they readily made IFN-γ. Because these mice expressed H-2Dd whereas nonresponding strains did not, this was a perfect correlation of IFN-γ production by Ly49A+ NK cells with H-2Dd, the ligand for Ly49A, as self–MHC class I, suggesting that Ly49A engagement of self-MHC was required for optimal activation receptor function. We termed this education effect “licensing.”

We then took advantage of a single-chain trimer MHC class I molecule produced by my colleagues, Ted Hansen and Daved Fremont, at Washington University. They produced a single DNA construct encoding a single polypeptide, the single-chain trimer, which includes H-2Kb H chain, β2-microglobulin, and SIINFEKL from OVA in the leader sequence, all connected by linkers. This molecule folds like a conventional H-2Kb molecule loaded with SIINFEKL and binds only Ly49C on primary NK cells (17). They produced a transgenic mouse expressing only one MHC class I molecule (i.e., the single-chain trimer-Kb transgene) on a triple knockout (β2m, and both MHC class I H chain alleles [H-2Kb and H-2Db]) background. In this mouse, only Ly49C+ NK cells were licensed, strongly supporting the hypothesis that Ly49 engagement with self-MHC is required for licensing.

In summary, the “inhibitory” MHC class I–specific Ly49 receptors have two functions: inhibiting NK cell activation in effector responses and licensing, or educating, NK cells via host MHC class I in vivo. These properties allow the functional coupling of the polymorphic Ly49 alleles with their MHC class I ligands, also encoded by polymorphic genes (11). These principles have been generally accepted for the KIRs as well (19).

Now, obviously immunology has advanced significantly in almost 50 years. I went from studying lymphocytes that I isolated at the Blood Bank of Hawaii using polyclonal antisera, microcytotoxicity, and trypan blue exclusion assays. Now we can do amazing things with much more advanced approaches, as I have discussed. And we are continuing the Ly49 story with CRISPR-Cas9–based approaches, studies too premature to put in writing in this article.

For the trainees, I want to emphasize an important theme in my story, and that is that techniques become obsolete. We can still use these techniques, but they have become outdated, so no one uses them anymore. I direct the M.D.–Ph.D. program at Washington University, and we have many talented students, many of them brilliant. Occasionally, though, they’ll come to my office and say, “I want to go into [name a famous professor]’s lab so I can learn a new technique.” And I have to remind them that’s not what it’s all about because techniques become obsolete. To the trainees, it’s really about learning how to do science, how to follow up on interesting findings, and how to keep up with advances in science.

So what’s ahead in immunology in the next 50 years? It’s hard to imagine that I peered into a microscope, first being exposed to immunology, almost 50 years ago. And it’s really hard to predict what the future is going to be. I don’t think anybody has that power despite what some people say. But I can tell you one thing that disturbs me a lot, and that is the current environment that’s anti-science. We have to do something about this, and I encourage all scientists to get involved to counter this sentiment. Join the AAI and help us convince policy makers and the general public to strongly support science!

I want again to acknowledge all my mentors that I mentioned along the way. I want to acknowledge Ira Goldstein, John Atkinson, and Emil Unanue. Ira recruited me to UCSF and then to Mount Sinai, and gave me strong support when I was just getting started as an independent investigator. Unfortunately, he died at too young an age. John and Emil recruited me to Washington University, and they’ve been wonderful mentors beyond the research laboratory. I’ve had many talented students, fellows, and staff working in my laboratory, and I want to acknowledge their contributions, too numerous to mention, although I did highlight a few of them.

I’ve had many great colleagues at Washington University and elsewhere, beyond the few I mentioned; we’ve published papers together and shared many fruitful discussions. I want to acknowledge the funding support that we’ve had. We’ve been privileged to be part of the Howard Hughes Medical Institute, and we’ve had wonderful support from the NIH (the National Institute of Allergy and Infectious Diseases, in particular); my prior institutions, UCSF and Mount Sinai in New York; and now my current institution, Washington University in St. Louis, including the new Andrew M. and Jane M. Bursky Center for Human Immunology and Immunotherapy Programs, and the Barnes-Jewish Hospital Foundation.

I want to acknowledge my family for their understanding and support: my daughter Christine and her husband Anthony, my son Reid and his wife Kolina, and the three new delights in our life: Thomas, Kayla, and Remy Lynn. And words are insufficient to describe my appreciation for my wife Lynn’s contributions to all of this.

Finally, I want to thank all of you for indulging me in this chance to relive some of my own scientific explorations. I want to thank you again for your support of the AAI and all that we’re doing. It was my privilege to serve as your President.

I thank John Emrich in the History Department at the AAI for a transcript of my oral presentation.

This work was supported by National Institutes of Health Grants U19-AI109948, R01-AI129545, R01-AI128845, R01-AI131680, and R01-AI140397 and the Barnes-Jewish Hospital Foundation.

Abbreviations used in this article:

AAI

American Association of Immunologists

JI

The JI, The Journal of Immunology

KIR

killer Ig-like receptor

MCMV

murine CMV

NIH

National Institutes of Health

RFLP

restriction fragment length polymorphic variant

UCSF

University of California, San Francisco.

1
Blumberg
,
B. S.
,
H. J.
Alter
,
S.
Visnich
.
1965
.
A “New” Antigen in Leukemia Sera
.
JAMA
191
:
541
546
.
2
Blumberg
,
B. S.
1977
.
Australia antigen and the biology of hepatitis B
.
Science
197
:
17
25
.
3
Yokoyama
,
W. M.
,
M. M.
Yokoyama
.
1973
.
Incidence of HL-A antigens in the Japanese population of Hawaii
.
Jpn. J. Exp. Med.
43
:
81
86
.
4
Yokoyama
,
W. M.
,
L. B.
Jacobs
,
O.
Kanagawa
,
E. M.
Shevach
,
D. I.
Cohen
.
1989
.
A murine T lymphocyte antigen belongs to a supergene family of type II integral membrane proteins
.
J. Immunol.
143
:
1379
1386
.
5
Yokoyama
,
W. M.
,
P. J.
Kehn
,
D. I.
Cohen
,
E. M.
Shevach
.
1990
.
Chromosomal location of the Ly-49 (A1, YE1/48) multigene family. Genetic association with the NK 1.1 antigen
.
J. Immunol.
145
:
2353
2358
.
6
Smith
,
H. R. C.
,
F. M.
Karlhofer
,
W. M.
Yokoyama
.
1994
.
Ly-49 multigene family expressed by IL-2-activated NK cells
.
J. Immunol.
153
:
1068
1079
.
7
Yokoyama
,
W. M.
,
J. C.
Ryan
,
J. J.
Hunter
,
H. R.
Smith
,
M.
Stark
,
W. E.
Seaman
.
1991
.
cDNA cloning of mouse NKR-P1 and genetic linkage with LY-49. Identification of a natural killer cell gene complex on mouse chromosome 6
.
J. Immunol.
147
:
3229
3236
.
8
Ljunggren
,
H. G.
,
K.
Kärre
.
1990
.
In search of the ‘missing self’: MHC molecules and NK cell recognition
.
Immunol. Today
11
:
237
244
.
9
Karlhofer
,
F. M.
,
R. K.
Ribaudo
,
W. M.
Yokoyama
.
1992
.
MHC class I alloantigen specificity of Ly-49+ IL-2-activated natural killer cells
.
Nature
358
:
66
70
.
10
Borrego
,
F.
2006
.
The first molecular basis of the “missing self” hypothesis
.
J. Immunol.
177
:
5759
5760
.
11
Yokoyama
,
W. M.
2013
.
Natural killer cells
. In
Fundamental immunology
, 7th Ed
W. E.
Paul
, ed.
Lippincott Williams & Wilkins
,
Philadelphia
, p.
395
431
.
12
Gumperz
,
J. E.
,
P.
Parham
.
1995
.
The enigma of the natural killer cell
.
Nature
378
:
245
248
.
13
Scalzo
,
A. A.
,
N. A.
Fitzgerald
,
A.
Simmons
,
A. B.
La Vista
,
G. R.
Shellam
.
1990
.
Cmv-1, a genetic locus that controls murine cytomegalovirus replication in the spleen
.
J. Exp. Med.
171
:
1469
1483
.
14
Brown
,
M. G.
,
A. O.
Dokun
,
J. W.
Heusel
,
H. R.
Smith
,
D. L.
Beckman
,
E. A.
Blattenberger
,
C. E.
Dubbelde
,
L. R.
Stone
,
A. A.
Scalzo
,
W. M.
Yokoyama
.
2001
.
Vital involvement of a natural killer cell activation receptor in resistance to viral infection
.
Science
292
:
934
937
.
15
Arase
,
H.
,
E. S.
Mocarski
,
A. E.
Campbell
,
A. B.
Hill
,
L. L.
Lanier
.
2002
.
Direct recognition of cytomegalovirus by activating and inhibitory NK cell receptors
.
Science
296
:
1323
1326
.
16
Smith
,
H. R.
,
J. W.
Heusel
,
I. K.
Mehta
,
S.
Kim
,
B. G.
Dorner
,
O. V.
Naidenko
,
K.
Iizuka
,
H.
Furukawa
,
D. L.
Beckman
,
J. T.
Pingel
, et al
.
2002
.
Recognition of a virus-encoded ligand by a natural killer cell activation receptor
.
Proc. Natl. Acad. Sci. USA
99
:
8826
8831
.
17
Kim
,
S.
,
J.
Poursine-Laurent
,
S. M.
Truscott
,
L.
Lybarger
,
Y. J.
Song
,
L.
Yang
,
A. R.
French
,
J. B.
Sunwoo
,
S.
Lemieux
,
T. H.
Hansen
,
W. M.
Yokoyama
.
2005
.
Licensing of natural killer cells by host major histocompatibility complex class I molecules
.
Nature
436
:
709
713
.
18
Liao
,
N. S.
,
M.
Bix
,
M.
Zijlstra
,
R.
Jaenisch
,
D.
Raulet
.
1991
.
MHC class I deficiency: susceptibility to natural killer (NK) cells and impaired NK activity
.
Science
253
:
199
202
.
19
Vivier
,
E.
,
D. H.
Raulet
,
A.
Moretta
,
M. A.
Caligiuri
,
L.
Zitvogel
,
L. L.
Lanier
,
W. M.
Yokoyama
,
S.
Ugolini
.
2011
.
Innate or adaptive immunity? The example of natural killer cells
.
Science
331
:
44
49
.

The author has no financial conflicts of interest.