 Okay. So thank you for the lovely introduction. This is where I'm from. So this is Sydney in Australia. I've been really enjoying getting to know Switzerland a little bit while I've been here. But they're both very beautiful countries. In Australia these days we do an acknowledgement of country, which is to say that we say the First Nations people where we are, what land we work on. And so I work on the land of the Gadigal people of the Urination. And I pay my respects to Aboriginal elders past, present and future. All right, so today we are talking about, well, I'm talking about cross-contaminated cell lines. And I think to make sense of the story, it makes sense to go back to where it started in order to try and understand. So I'm going to take us back to cell culture in the 1940s, which was an entirely different beast. At that time, highly specialised, a handful of people worldwide who could do cell culture. You know, this period would be well under 100 people. It was, well, non-standardised is a fair way of explaining it. Everything was handmade, reagents, glassware was hand-blown. The example I have here is of Virginia Evans from the National Cancer Institute. That lab is one that established the tea flask, the rectangular flask that we use today, hand-blown from glass tubing. And they did many other techniques, so that's a whole other talk. But the point being, of course, this is highly variable. This is the first efforts to standardise cell culture, date back to this time. So then in 1943, Breakthrough was made by this lab at NCI where Virginia Evans worked, and they established the first cell line. So that was established from the L-Strain that was taken from mice, a particular mouse that was treated with a carcinogen. And then the culture was cloned to give a few different names for it. NCTC clone-99, but I'm going to call it L-99 just for short in this talk. The photo we have here is actually of the establishment of L-99. So they would clone these cultures in glass capillary tubes. And then the clones that grew, they would snap the tube and then out would come the cells with further culture. And then, of course, the breakthrough that we all know about, 1951, when scientists at Johns Hopkins established the first human cell line, which was, of course, Hila. And anyone who hasn't read the book The Immortal Life of Henry Lacks, it's definitely well worth reading to give the full story of this saga. But those cells, as we know, were derived from Henry Lacks and were taken without her consent from a biopsy of her cervical cancer that she then went on to die from. So the photo shown here is of Hila's cells in culture at day three. So these two breakthroughs were indeed regarded as breakthroughs at the time. Those first two cell lines were used to standardize the cell culture media and reagents, plastic wear, glass wear that followed. Hila's cells were used very rapidly afterwards for the first clinical achievement, which was that they were grown at scale for field trials of vaccine efficacy for the production of polio vaccine. So that work was done by a gentleman there and a team of all Black scientists at the Tuskegee Institute who produced millions of vials in the 1950s for the purpose of this vaccine testing. So people discovered that cell lines were useful. Cell lines started to be deposited at the cell collections for future use. Cell culture began to be used for a discovery platform. There was a huge sense of excitement that all of a sudden we had these breakthroughs and now these new vistas unfolded where cell culture could be used for many different purposes. However, there was a hidden problem that was popping up, which is what we're looking at today. So in 1957, the scientists who established the L929 cell line noticed the first case of cross contamination. So they published it almost as an incidental finding. They were working with skin cells that they were culturing and they noticed a different cell type that was not what they were expecting. Being very experienced, they knew what they were looking at was not correct and very rapidly worked out that it was a cost species contamination from chicken cells that had been present in one of the reagents that was used to culture the cells. So they considered it a rare event. They published it because it was not worthy and then just kind of moved along. Then in 1962, Lewis Corriell described, well really it's a proof of concept study. He deliberately introduced healer cells into other cultures to see what would happen. And then the quote I've got there is from a book that described what he saw, merely pulling a stopper from a test tube or dispensing liquid from a dropper could launch tiny airborne droplets containing a few healer cells. When the drops landed on open petri dishes holding live cultures, the healer cells began growing so feverishly that in three weeks they overwhelmed the original cultures. And that's really the perfect description of what we've seen ever since that time, that proof of concept study. But the limitation at that time was that there was no test to detect whether human cells were cross contaminated by healer. There was cytogenetic analysis could be done for species. Some of those early cell collections did do that, but they didn't have the capacity to test for interest species contamination. And then a geneticist by the name of Stanley Gottler discovered such a test in the 1960s. So Stan Gottler was doing research on isoenzyme variations, originally purely for research purposes. And he obtained 20 of what he thought were well characterized human cell lines to look at isoenzyme variation between the set. And what he discovered was that all 20 had the same isoenzyme variant. So the odds of that were, shall we say, extremely low, considering they're all meant to come from different donors. But what he really observed was a particular variant there that was only found in individuals of African descent. And he knew that that was the case with healer. Since he knew that the origin of healer was was in Henry Alex. So basically all 20 of those well characterized cell lines were cross contaminated with healer. So Gottler turned up at the 1966 cell tissue and organ culture meeting where the founders of those cell lines were presenting their work. Many of them were famous for these cell lines. So you can imagine someone turning up at a conference like this to say, I'm sorry, all your work is totally wrong. It didn't go down well. It's safe to say. However, Gottler was right. Those cell lines have been tested over and over again with many different methods and they have all shown the same result. Those cell lines are all derived from him. So at that point we have 20 cell lines that were known to be cross contaminated. So then let's look a little bit further forward to today's research and say, well, what impact is it having now. But before we get to that, I'm just going to do a brief spot of terminology. So it's true that the terms we use for these things do change over time and different people do have different versions. So for this talk, a cross contaminated cell line is one where cells have been introduced from another culture, usually leading to replacement of the original authentic material with the contaminant. So when I talk about a misidentified cell line, I'm mostly thinking about one that either fails to correspond to the reported donor. Or unexpectedly corresponds to another cell line. So healer is the example that we've just looked at. Now the field has moved away from cross contamination. I think mostly because we often don't know the mechanism. We tend to assume it's due to this event. But as Amos commented, it could be mislabeling. It could be various other things. And when I mean authentication testing, I'm talking about testing of a cell line to confirm its identity as coming from a particular donor, ideally, or at least verifying that it's from the correct species. And as best we know, comes from the correct donor doesn't correspond to other things unexpectedly. Okay, so today's literature. How much do people use cell culture? So this is a spot of literature review that in Freshnee and I did a few years ago, where we're looking at publications over the years appearing in PubMed. And we ran a search for four different phrases. So it was tissue culture, cell culture, culture cells or cell line. And then you see this interesting mountain image. So from the 1970s and explosion of literature using these terms using cell culture cell lines in various permutations and combinations. Now, if we think back to the history and say, well, how does that compare to the literature? This is what we see. And you'll see there 1943 is the first cell line. So L929 1951 when Hila was established 1957 that incidental report of cross contamination. And then 1967 when Stan Gatler turned up at the meeting with his findings about cross contamination. And you'll see all of those are before that explosion. So ever since then people have been using cell culture and cell lines. And the question is, well, how much of that is affected by this problem? Stan Gatler found 20 cell lines, but how many are there these days that are affected by that? And how many publications use them? So we can get a partial answer from the ICLAC register of misidentified cell lines. It's essentially a very simple list with apologies to band informaticians everywhere. It was developed by Ian Freshney and myself during on several older lists, and is now curated by the International Cell Line Authentication Committee. The ICLAC group is basically a voluntary group of experts in this area that review reports that come out of problematic cell lines and try and decide should they be registered as having this specific problem. So the latest version of the ICLAC register has 535 cell lines that are known to be misidentified with no known authentic material. So looking at those numbers about a quarter are actually Hila. So those are the same sort of issue as Stan Gatler found all that time ago, but the problem is now much wider. In addition to Hila, there are 145 other contaminating cell lines. About 12% come from a different species that we can find. So the classic reported to be human, but actually mouse. And about the same amount have an unknown cell line. So we don't know exactly what they are, but there is some form of data to indicate that there is a problem. So the classic is a cell line reported to come from a female donor that carries an XY chromosome. An additional 47 cell lines are separate to that. They were thought at one time to be misidentified, but thanks to investigational efforts from many different groups, authentic stocks have been found. And those are excluded from the numbers. Okay, so I mentioned that it's a very simple list, the ICLAC register, and the presence, the arrival of cellosaurus has really been quite transformational I think. I know when Ian and I were originally working on this list, we wanted to make it searchable and accessible, but we just don't have the skills to do it. So the move to have it included as part of cellosaurus has really been a radical move to me anyway. So for cell lines that are listed on ICLAC register, cellosaurus provides a problematic cell line warning so you'll see that in red on the entry that says very prominently there is a problem here to be aware of. Usually it lists the source references because it can be based on data that's often published. It includes the ICLAC registration ID which hopefully brings in more data, not all of that's publicly available. It includes authentication testing data as we've talked about and enables people to do their own STR similarity searches. So those are all transformational additions. So in terms of the number of papers that use these cell lines are just the ones on the ICLAC register. We're fortunate that researchers did an analysis on this a few years back, and they took the cell lines, the cell line names and said how many papers use them. So their findings was that at that time 32,755 articles use these known misidentified cell lines. Most of them were cited at least once and they estimated about half a million citations to those publications across the entire publication course. However, it's really important to emphasize the cell lines that are listed in this list are only part of the problem. And there's a few reasons for that. One is that there's often not enough data to really reach a sound conclusion. We often don't know enough about the origins of the cell line to even know if it's a problem. If there's no publication, then if it matches something else, perhaps it was legitimately derived from that cell line originally. If we're not sure, then we don't include it. And often there's just not enough testing data available. So if the original lab didn't test the cell line, if they've not deposited it at a place, a cell line repository where that testing is done, then no one may know. So it's like the tip of what could be a much larger iceberg. A further factor to consider is that where labs get their cell lines from is in itself a risk factor. So Willie Dirks may or may not talk about that later today, but you can kind of break this down into primary sources and secondary sources. Your primary source is either the original lab or the place where they have deposited it that's done the testing to look at its quality. The secondary source is essentially the lab down the corridor. You see the emails, I want a vial of whatever cell line it is, please can you give it to me. The sources typically handle the cell lines more so they're high passage and they test them less so there are much more high risk factor. So it's important to think about not just the cell lines that are listed but other sources of data and the editors at the International Journal of Cancer have made a recent publication which is a fantastic one and well worth reading in full. There's a figure here from one of their papers but just to quickly summarize the journal requires authentication testing and they have also tracked the metrics connected to that. So what they've reported is that at least 5% of cell lines in the manuscripts that are submitted to this journal for peer review are misidentified. Now the editors will work with those authors but even so about 4% of manuscripts are subsequently rejected for severe cell line problems but the particularly worrying aspect is that most of those papers then go on to be published in other journals. Okay, well time going well. Alright so how can we reduce this? So there is evidence here we have a problem. We may not have a full sense of the size of it but I hope you'll see it's a major problem for the life sciences research community. So what can we do about it at this point in time? And so this is what I want to then spend the rest of the time to think about and to me it comes down to four things. We need to test, we need to train, we need to report and we need to require. So I'm going to go through each of these. So first off we need to test. If we don't test cell lines for this problem we will never ever know about it. And as Andreas pointed out we need to use a shared test method for comparison. So historically and continuing the problems usually found when lab A tests their cell lines and compares their results to what's held by lab B. So they don't realise it until they bring them together and work out oops we've got what we thought was the same cell line but we've got different results, different genotypes. So it's essential to do testing with a shared method if you're going to work out that kind of problem. So any consensus method also must be based on genotype. There's been this constant to and fro we should be able to test based on expression of tissue specific markers things like that. No. The reason it's no is that cell lines adapt and it's possible that they will express markers that are not typical of what you would expect from the original tissue type. Now they're the universal adapters of the research reagent world. So it has to be genotype based. So as Amos mentioned, there's been an effort to standardise this and it was eventually published as an NC standard back in 2012. And that working group looked at different test methods and said no we think the best method for this comparison purpose is short tandem repeat profiling. And that was revised back in 2021-22 republished with the same finding. So there are other methods and very important and useful. So complimentary to the STR profiling. The more methods is good having more information. But for this sort of comparison, we have to have something we agree on. So as we talked about several times already, Salasaurus has a fantastic similarity search tool. So you can take your results and compare it to everybody else's. So it's fantastic and user friendly. You can put in your own profiles, not just one but multiple. There's would be good to have more STR profiles and hopefully that will happen. But it's the largest data set that I know of for this purpose and one with the most user options. Okay, so that's test. The second is train. So anybody who trains new users to their tissue culture lab. I salute you and you're doing a fantastic job. Because this is essential. We've always got new people coming into this field and people don't know unless they're taught. So the teaching of course teaching aseptic technique is really important. But it's also important to teach the theory that this sort of problem. And it's important to be aware that there are more educational resources in this area compared to a few years ago. So a couple of screen captures there. One is to the Salisaurus educational material. Another is to an eye-clack animated video, which was designed to be kind of student level and above school student. I mean, which talks about these are cell line problems. And Salisaurus. I don't need to belabor this one because we're probably all in this room using it, but it's a great resource to use. Okay, so number three is report. And what I mean by that is when you publish what information do you include that can help reduce this problem? And the most important thing I can think of is to include the RID. So I'm not going to steal Anita's thunder who's going to talk about this this afternoon. But it has an impact on problematic cell lines. And there's a manuscript that I'm sure she will go back and talk about later. That says that even just including the RID results in a reduced usage of misidentified cell lines. A little mysterious in terms of why that would be so effective, but reasons and I'm sure she'll talk about that. Report is also what other information to should people put in their materials and methods. And this is the little data set that I think is a good idea. So apart from the RID, is it known to be misidentified? Has the testing been performed for either authentication or microplasma? Is the source of the cell line reported? And the basic is sufficient information given to replicate experiments. That's always a variable question, but there are some obvious basics. And then we've got require. So require I think is at two levels. One is for labs. So everyone who's who's running a lab, managing a lab with cell lines. Can you get your users, your people to do testing? So it's a question of how does that get done? So there are some very simple things that are helpful. So setting up some kind of shared folder, shared database to record results over time. Starting off with what are the most important cell lines this lab uses? What do we really rely on? What has to be right? And then rolling that out to testing everything. So before you start work on it, testing before publication is of course important, but the earlier you test and realize there's a problem, the earlier you can either jettison that sample is not appropriate or even pick up problems. So the screen capture here is of an early cross-contamination event that was picked up through testing. And in that case, the lab was able to go back to earlier passage sources and retrieve something that had not been affected by this problem. And lo and behold, problem fixed. But in addition to the lab involvement, and that can be fantastic in some labs, there's a need unfortunately for everybody else to mandate, you know, to require. So journals and funders, having them involved has really been a transformational thing. Requiring testing and reporting of testing as a condition of either publication or funding. So I've seen, and we won't have time to talk about it today, but a decline I think in usage certainly of some misidentified cell lines since 2014. And I don't know exactly which of these factors is most effective. Perhaps it's altogether, but I do think the policy requirements are particularly important. So when NIH released our policy on authentication of key resources, that was one where anecdotally testing went up. People realised, oh, we've got to do it in order to get funding. And again, returning to the International Journal of Cancer, probably too small to see, but this is again a little snippet from their recent paper. And I think we're all singing from the same songbook. Our recommendations are broadly similar. Requiring RIDs, going to cellosaurus, testing cell lines, it all fits together. And so that's my four areas. And at that point, as I'm in time, I shall stop and ask for questions. Thank you very much for your talk. I have a question which is not so much about the identification of a cell line, but of using a certain name when a cell line has been cultivated for decades under unknown conditions, and yet the name of the 1950s is used today in commercial manufacturing of proteins for therapy. What is your comment? I would comment that is extremely common, in fact, more typical than not. And yes, extremely problematic. I think raises the broader question of when a derivative is developed. So you handle a cell line, you treat it in a certain way, and it does evolve, grown under certain circumstances. And you get to a point where you say, this is different. This behaves differently. The genotype is demonstrably different. There is no line in the sand, which is exceptionally difficult. I think to balance that, the need that labs need to have is for the best culture practice approach of cell banking, where instead of that long process of handling, you say, OK, I have just established this cell line. And the first step that I do is to store it. And then to do the approach where you have the master bank and the working bank. So you have continuing resources to low passage material. And that's part of training and an important step to try and minimize what is most definitely a huge problem. Thank you for this very interesting talk. You mentioned something named Severe Cell Line Problems that would lead to rejection of publication, right? I was wondering what does this actually means? Because I guess it's a bit arbitrary. So that particular example, I mean, Constantine may wish to comment further, is that there is a very clear process and that's made aware to authors. So authors within the author guidelines are told that the journal requires authentication testing. The testing will be provided to the journal editors and reviewers. If that testing is not provided, then the editor will come back to the author, I believe, to kind of flag that that is an issue. And so that's kind of a requirement for publication. So that is identified. I know that this particular journal will review the data, will use Celesaurus Sancluster and come back based on data. So they'll come back and say a problem has been identified relating to the STR profile. It matches X. It's not human, whatever the problem may be. It's in order to meet the quality requirements, the rigor and reproducibility requirements. It's necessary to address those issues. Of course, the question you always come back to is how much does it matter for the purpose of the paper? But it's not just the paper. It's also the usage of the paper by others. So if the cell line is not authentic, but it's not clearly shown as such within the paper, then others will use it without a sense of what the problem is. So there's that sense where it's for the quality of what appears in the journal, but also the broader quality of the citation corpus and other usage of the cell line in the future. Amanda, at earlier times when I worked with neuroblastoma cells, we didn't have possibilities to genetically identify cells. So I was looking, is it developing action potentials and to what neurotransmitters is it sensitive? Would that nowadays also be an acceptable way to identify cells? I would say it's a complementary method. If you do come back to genotype, is it based on something genotypic? And I think that's for experience because it's constantly surprising the behaviors that cell lines can exhibit. So it can still trip you up even if it's a characteristic that should be quite specific of the cell type or the tissue type. But all of these methods add up to, I guess, a body of data. You need all of that data in order to reach a conclusion about whether a cell line is what you think it is. So, I mean, also it's important to recognize that some genotype based methods like SNP analysis are more convenient for researchers. So there are other reasons why they may look at other tests and that's all important. But it's essential that you need at least one STR profile in order to do this comparison with the broader communities holdings of the cell line. And at that point you say, OK, you've got that confidence that it's not Hila, not any of these other possibilities. So, related to the question before, this would also mean that the journals ask for the original data of the STR typing. So it's not sufficient just to write the sentence, cell lines were authenticated. That's it. Journals vary quite considerably in what they require or what they state in their author guidelines. It is the very rare journal that is as stringent as this one. So the journal I know made a decision some years ago that for cancer research, this is important. It's one where cancer researchers probably are the field that know the problem best. A number of these cell lines are cancer cell lines. So therefore they have more likely to have the background of knowledge to understand the need for it. Other journals, it may be phrased in a sense of you should include any methods where the cell line has been tested and what the result was. Some they don't require, but they will encourage. So they will say that you should test your cell lines. That is still useful. It's raising awareness. It's like the RIDs in that sense that it's it's flagging that there is this process that needs to be done and hopefully combined with RIDs and cellosaurus to bring you quickly to the information when there is a known problem. But yeah, there's considerable variation in terms of what journals require. When you're using RNA sequencing, do you think you can also use this data to identify your cell? Is it like complementary, as you said just before? There's a lot of sequence based testing, yes that can be used. I think a lot of those have advantage for in-house testing. So if you're a large, if you're a facility or a research team that has some kind of NGS based technique, something that's based on sequence and sequence variance, then it can have utility for this sort of testing alongside the consensus for the comparison. Yeah, so RNA is not one that I have much experience with, but I do know there has been some very helpful work in looking at a viral integration, which again is complementary, so that you're able to look at that in terms of the comparison to a likely parental cell line. So there's all kinds of utility for these different methods that are all telling you something that's a unique attribute that adds to your knowledge. Very interesting talk. Coming back to the requirement for the journals, sometimes I work in a services core facility, so we do provide STR authentication for the researchers. Sometimes what happens is that maybe they have authenticated the cell line two years ago and they are publishing now. So any misidentified cell line at least once in his life was authentic and then it becomes misidentified. So maybe it would be better to have from the journals a kind of an STR done in the last month with a date where the date is taken accountable. Otherwise, the researchers can come back to us as a provider and say, oh, you say they were authentic, but they are not authentic anymore. So it creates a lot of false information in terms of the sense is authenticated. Yeah, but three years ago. Yeah, that's a very good point. And I suspect a lot of labs will not, yeah, not test nearly as much as any of us would like to. I mean, I do have a sense that needs to be worked at how to do it in an efficient way though. I think, again, it varies what journals will require. Some do have time limits. So again, it would be nice to have that more consistent. Yeah, but that can be challenging as we know, getting that standardized approach.