�����ƽ��

Professor Carmen Gerlach: For me, the most problematic thing was that we had so many different parallel definitions for the same terms. So, if you were to read one paper and you read about effector cells, and then you read a second paper, and again, the term “effector cells” comes up—these two different studies may actually be referring to different cell populations. And I think the reason for that is something that they've explained in detail, which is that these terms often, in particular from the textbook definitions, are made up of different aspects of a cell, like the longevity or the function of the cell; the differentiation state—all of these were lumped together. And if then one research group decided to use function as a definition for a factor and another for longevity, or the absence of longevity, then we would end up actually looking at different populations. That has become very confusing with the old nomenclature. And now, with a modular nomenclature, we can actually pull these things apart, so we don't have to lump them together. We can just refer to the function, differentiation state, and so forth, separately, which will increase the clarity a lot.

Professor Rafi Ahmed: We emphasize this as version 1 because the field is rapidly evolving. A lot of new data is coming in. It has to not only be updated, but also integrated appropriately into the nomenclature article that Dave and all of us have put together. I think the key thing here will be at two levels. One, of course, is the new advances that happen; we need to incorporate those. But I think the other one is: how well is the model and nomenclature adopted? How well is it doing and how many people are using it? Also, in version 2, there will be recommendations for optimizing or improving the model or nomenclature. I think that a second version is essential. How are we going to do this? We haven't discussed the details yet, but our plan is to have a smaller group of people who are authors on this article, plus, perhaps, bring some additional people. But it'll be a smaller group that will meet on a regular basis, seeing how the recommendations are being implemented, how we can improve implementation if that's needed, and how we can also improve what we have suggested.

Professor David Masopust: Those are two really good questions, important questions. Is it going to be hard? I would say we're lucky. Immunologists are intelligent scientists, but it's also a field that I think is very comfortable with jargon. We have a number of CD molecules and transcription factors and so on. And I think people will recognize that this really is a movement towards simplification, not complexification. And so, with a limited vocabulary, one could describe a lot of different kinds of T-cell states. I think, in my experience, that there will be a subset, if you will (no pun intended), of people who have an initial aversion when they see the modular nomenclature. And for most people, that aversion goes away as they start to think about how it would influence what they do and how they would think about things. So I think it will be self-correcting and not that hard to learn if you give it a chance. Now, because people will have an aversion, you asked about the coexistence of two ways of doing it. It's inevitable. I remember back when I was a grad student and I had learned the names of some chemokines, and then that was really reorganized. MIG (monokine induced by interferon gamma) became CXCL9 (C-X-C motif chemokine ligand 9), and then all these chemokines and chemokine receptors had a lot of Cs and Xs in them, and I thought, this is madness. What are we doing here? But I survived the transition, but I think for a long time, people would say CXCL9 and then put MIG in parentheses or what have you. And then we've largely dispensed with that and the field has moved on and is probably a lot better off for having done so. My point is, we've made transitions. In some ways this transition will be easier, but to be fair, in some ways not. We're not just swapping names for molecules. We are changing the way we name T cells from one tidy label that incorporates a lot of different properties, into thinking about T cells the way we describe them, as a collection of discrete properties that we bundle together. And that may influence the flow of the science a little bit, but I imagine, in a transition period, that people may refer to the legacy name and the new name at the same time. The silver lining here is this, though: If you adopt solution number one, and in the methods section, you say, “I'm studying cell X, and how I define that is via these proxy assays,” there's no room for confusion. I think we can all deal with that. That's my two cents, but it's an important, weighty question, and maybe others have a different opinion.

Yvonne Bordon: Your point you make with the chemokine nomenclature, I think it was much better to move to a more agnostic naming approach, where you're not biased by the name. For example, monocyte chemoattractant protein; they attract other things. So, I think that's what you're trying to get with the T-cell nomenclature, too. The existing names make you think something about the biology of the cell that may not be true.

Professor Carmen Gerlach: I just want to second this point about the chemokine nomenclature and CD molecules. So, we've had a bunch of these changes, and indeed, you just need to get used to them, and most of the time, these changes have been good in the end.

Professor Rafi Ahmed: I think I'm not fully aligned. I think Dave and I have slightly different opinions on this. So, the CD nomenclature is a good example, but there you're dealing with a single molecule, so it's very easy to make that transition. Here, you're dealing with complex biology. So I think it will take time because one cannot abandon the names that we are describing, even though they're imperfect. Because the challenge for the model and nomenclature will be: is it really capturing the biology? Are we losing some of the biology by just using the model and nomenclature? And I think time will tell how that evolves. So in the beginning, I think there's, at least in my opinion, nothing wrong with using existing books; describing the cells as you like to describe them using the terms that we're using now. Maybe sharpen those, but Dave pointed out that a key recommendation, which is what really led to the T cell nomenclature, and Dave articulated this very nicely in the first part of this lecture, was the confusion about terms being used without a clear definition of how the cells were defined; what was the context in which you were studying it? So, I think that part really has to be the one thing we would insist on that the editors do. If you're describing any cell type, describe in detail how it was analyzed and in what context. And Dave pointed out an important thing about the role of antigen. The T cell differentiation state and phenotype changes dramatically if there's antigen or if there isn't antigen. That needs to be documented. I think the biggest message is: please describe the context in which you're studying the cell and describe all the markers you're using to give it a name. That's okay. If the name is wrong, at least the people know exactly what was done, what markers were used, and what the context was. The modular nomenclature, I think, is a very creative idea, and I hope more and more people adopt it. And even if they don't like it, I would encourage them to at least give it a try; include it in your publication with the standard way that you're doing it. I think that will be the real test. So my plea to the people in the audience is, even if you don't like the modern nomenclature, give it a try, include it, but also include your description of the terminology that you're more used to.

Professor Carmen Gerlach: We're not forcing you to put a cell in a box where you don't think it belongs. So if you think that the cell has an important feature that isn't captured by an aspect of the modular nomenclature, just add this to the name.

Professor David Masopust: I would more or less just second what Carmen said. The point here is choice and combinatorial complexity. There are options within the modular nomenclature that have already been dictated, that you could adhere to and utilize for you and then add on any additional descriptors that really don't fit our current vocabulary. We don't have terms for that. And the temptation in the legacy nomenclature would be, perhaps, to invent a new subset name. This is part of the problem and where I feel, with the current way of doing things, we're going to run out of runway here at some point, and we're heading towards more and more confusion. I think what we are relying on is the creativity and ingenuity of our peers as they make new discoveries to figure out, in their particular example, how best to incorporate it. But I think it should be very achievable.

Professor David Masopust: It may be a question for the Nature team, really. I don't think there are any current plans within my lab or the lab of the 64 colleagues, as part of this paper, to pursue that.

Professor Rafi Ahmed: We did describe a lot of names in version 1. When we do version 2, I think we can revise that. Maybe some can be deleted and new ones can be added. Sometimes, it isn't actually a different cell subset; it's just the same cell dividing. Maybe the genes are different at that time because proliferation is going on. I think our ability to make this distinction between it truly being a different subset or the same cell at a different stage of the cell cycle will start getting better. This is the first step in the right direction. That's why it's called version 1.

Professor Carmen Gerlach: This version of the nomenclature guidelines we have now can serve as a translator for this. So, if in the future you're going to be confused about what, in the past, a certain cell would have been called, you could go back to this version and look that up, keeping in mind that in the old literature, sometimes the same name was used for different things. But that will also be explained here. We could have an archival of version 1, version 2. You could go back in history and search the pre-version 1 nomenclature.

Professor Rafi Ahmed: Let me add briefly to the question of searches in the literature. I think that's one of the problems. And Dave actually addressed this in the talk. When you look at it, we all do the same thing; we are all guilty of this. We have a cell, and then we look for the signature. We go into the literature and say, well, let's see the effect of the signature. The problem is that when we look up a given signature, we often don’t know exactly how that cell was originally defined. One of the main recommendations is to describe exactly what you did. I think that information will now help us compare it with the right signature.

Yvonne Bordon: That's a good point. If that's actually stated in all the papers now, it should be easier going forward to at least understand what the scientist was doing at the time.

Professor David Masopust: We've gone through this process in the past. I got a very nice email from someone involved in the description of Lyt-2 being associated with different T cell properties. And I had read that paper. It was published 50 years ago, so I read it long after it came out. What's Lyt-2? I didn't have AI to help me out. I still figure things out and it's nice to know the history. I think it's a little easier today. As Carmen pointed out, there's a Rosetta Stone with this document and there are the search tools that are available to us today that will likely handle that problem.

Professor David Masopust: Clearly, I think It will help, and I think most of us think it will help. The status quo is not particularly great, first of all. When you talk to those that are a little bit outside of living and breathing T cell immunology every day, there really is a lot of confusion. What is an effector memory T cell exactly, and how do I know when I should call it that versus something else? There is this proliferation of names. And I want to emphasize that it is not getting better. It is getting worse because as we learn more things, there's more ad hoc names. And these names serve in a way of a single name to capture multiple properties that are expressed by a cell or measured by a cell within a single moment in time. And often these processes are dynamic. And so by stepping away and really changing how we describe these cells as a collection of distinct properties that aren't co-regulated with a very small vocabulary, we can say so much more. And so I think it's future-proof. I think new terms can be added when we want to start addressing additional properties. And I think it is a move towards simplification. Yes, there's a learning curve, but there was already a learning curve that people have invested in anyway. And I don't think this learning curve will be any steeper.

Professor Carmen Gerlach: I think if you're speaking to the general public, you would also, with the current nomenclature, not talk about central memory or effector memory cells. Instead, you would use broader, less specific terms to communicate clearly. That type of communication doesn’t change with this new framework. I think between people within the field, it's only going to add clarity and therefore be a good thing. And then, for people that are somewhat familiar with the field, you could have had the previous issues with the older nomenclature where people think, "I actually know what ‘effector’ means. I know that word”. But they don’t have the awareness of how complicated that term actually has been. And therefore, for that level of communication also, I think this additional clarity will only be helpful.

Professor Carmen Gerlach: One can just spell out the abbreviations and add the word mouse or human to it. And if in upcoming revisions, there is a strong need to clarify that with a letter, that can be discussed and can be done.

Alexandra Flemming: You can add descriptors in front of the modular nomenclature as before.

Professor David Masopust: If there's context and important information that exists outside what is currently being clearly delineated, one would just put that into words. And then if there becomes a major need for this in version 2, we can add something.

Professor Carmen Gerlach: The antigen presence or absence could actually be a helpful indicator in this case.

Professor David Masopust: The first thing I would emphasize is that if that is how you are inferring functional properties, and I'm not saying that's not a good thing to do; the intention of this is not to criticize how you do things. Again, it's between you and your science and what you have available to you. But, if you are very upfront in the methods section by briefly stating that this is how you arrived at these conclusions, so there's no ambiguity, then it's easy for the reader to evaluate what they're learning and incorporate it into what they already know. So that's where I think we're simplifying and clarifying. And, in every certain case, one is going to have to evaluate how good a proxy is for the reality that we're inferring. It's going to be true or not true in a case-by-case scenario. That's really a job for the 10,000 immunologists out there to work their way through.

Professor Rafi Ahmed: For using transcription factors to define subsets, I always felt that that's a good way of defining them because there are some canonical transcription factors that define the current subsets that we're using. I think that might be included in version 2. But I think in that modular nomenclature, indicating the transcription factor would be the value.

Professor Rafi Ahmed: Yes. It might make the modular nomenclature more acceptable.

Professor Carmen Gerlach: The concept can definitely be translated, or copy and pasted into other cell types.

Professor David Masopust: I agree. It will be up to those scientists. But there is a subtle paradigm shift here in how one describes things, not as a single cell type, but more as a description of that cell. And I think that may be helpful in some other fields, and they could use the same vocabulary a lot of the time and coin their own.