WEBVTT 00:00:00.000 --> 00:00:00.000 How's this? 00:00:00.000 --> 00:00:00.000 That's perfect. Um, that is perfect, thank you. We will go ahead and get started, since it's already noon. 00:00:00.000 --> 00:00:00.000 So, welcome everyone to this Reskill Learning webinar series from Sales to Solutions. 00:00:00.000 --> 00:00:00.000 Emerging tools for studying health and disease. This is the second session, 3D Models and Technologies to Illuminate Biological Effects of Contaminants. 00:00:00.000 --> 00:00:00.000 This webinar is sponsored by the National Institute of Environmental Health Sciences. 00:00:00.000 --> 00:00:00.000 Superfund Research Program. My name is Molly Velasco, and I am a support contractor with the NIEHS Superfund Research Program. 00:00:00.000 --> 00:00:00.000 We will get started in just a few moments, but while we wait for others to log on, I will cover a few housekeeping items. 00:00:00.000 --> 00:00:00.000 So, when you registered for today's webinar. You should have received a confirmation email with instructions to join us for the live event. 00:00:00.000 --> 00:00:00.000 We also send the same instructions. Um, in our reminder email yesterday. 00:00:00.000 --> 00:00:00.000 Those emails will both point you to the seminar homepage. 00:00:00.000 --> 00:00:00.000 I encourage you to check it out, as there are important pieces of information. 00:00:00.000 --> 00:00:00.000 including the session description, contact information for our presenters. 00:00:00.000 --> 00:00:00.000 links to download presentation materials, as well as options to browse through related websites. 00:00:00.000 --> 00:00:00.000 There is even an option to provide feedback after the webinar. 00:00:00.000 --> 00:00:00.000 So, this session is being hosted using Zoom Webinar. 00:00:00.000 --> 00:00:00.000 For those already familiar with Zoom meetings, you should find this a very similar interface. 00:00:00.000 --> 00:00:00.000 Audio will be available online with your device speakers or headphones. 00:00:00.000 --> 00:00:00.000 You may also listen via telephone. Regardless of how you listen, all participants in our webinar will automatically be muted. 00:00:00.000 --> 00:00:00.000 We will not be using the chat feature in today's webinar. 00:00:00.000 --> 00:00:00.000 Instead, if you have a question, comment, or need to report any technical issues. 00:00:00.000 --> 00:00:00.000 Please use the Q&A box. You can use that window to privately submit feedback at any time. 00:00:00.000 --> 00:00:00.000 Closed captioning is available for this webinar. Also, remember that our session today is being recorded. 00:00:00.000 --> 00:00:00.000 You will receive an email when the recording becomes available. 00:00:00.000 --> 00:00:00.000 Lastly, I encourage you to stay with us until the very end of the program. 00:00:00.000 --> 00:00:00.000 We will cover a number of important reminders, such as how to share your feedback on this webinar. 00:00:00.000 --> 00:00:00.000 This presentation has been provided as part of a U.S. Environmental Protection Agency webinar. 00:00:00.000 --> 00:00:00.000 The document does not constitute EPA policy. mention of trade names or commercial products does not constitute endorsement or recommendation for use. 00:00:00.000 --> 00:00:00.000 Links to non-EPA websites do not imply any official EPA endorsement of or. 00:00:00.000 --> 00:00:00.000 irresponsibility for their opinions, ideas, data, or products presented at those locations. 00:00:00.000 --> 00:00:00.000 or guarantee the validity of the information provided. Links to non-EPA servers are provided solely as a pointer to information that might be useful to EPA staff and the public. 00:00:00.000 --> 00:00:09.000 With that, I will now turn it over. to our moderator, Dr. Thad Shook from NIEHS, for opening remarks. 00:00:09.000 --> 00:00:12.000 that whenever you're ready. 00:00:12.000 --> 00:00:17.000 Good afternoon, everyone. I'm Pad Shugg. I'm a program officer at NIEHS. 00:00:17.000 --> 00:00:20.000 So, it's my pleasure to introduce the first speaker. 00:00:20.000 --> 00:00:30.000 We'll start with Steven Ferguson, an investigator with the mechanistic toxicology branch of the Division of Translational Toxicology, or DTT. 00:00:30.000 --> 00:00:37.000 Leading multiple research initiatives. to model and predict human responses to chemical exposures. 00:00:37.000 --> 00:00:44.000 He will discuss liver and kidney spheroid cultures. used to model chemical toxicity and disease. 00:00:44.000 --> 00:00:49.000 So, Dr. Ferguson, I'll turn it over to you. 00:00:49.000 --> 00:01:01.000 Thank you, Thad, and let me share my screen… 00:01:01.000 --> 00:01:05.000 Okay, so you able to see my screen and hear me? 00:01:05.000 --> 00:01:07.000 Yes. 00:01:07.000 --> 00:01:13.000 Yes. Perfect. Well, thank you so much for, uh, inviting me today, all the organizers, uh, particularly. 00:01:13.000 --> 00:01:24.000 Uh, given that Bev and Englewood, who I am not, was not able to be here today, but I am excited to share some of the work that we've been doing as we are. 00:01:24.000 --> 00:01:33.000 both kindred spirits and our aspirations for organotyping. in vitro models that help us better understand chemical toxicity and safety. 00:01:33.000 --> 00:01:38.000 Today, I'm going to share many of the learnings that we've had over recent years. 00:01:38.000 --> 00:01:51.000 towards advancing translational toxicology research with these types of models, and I must acknowledge, as a federal employee that I have no financial relationships to disclose, and the statements and opinions and conclusions that I make are. 00:01:51.000 --> 00:01:57.000 are my own. So… In the early models of human liver. 00:01:57.000 --> 00:02:04.000 Uh, such as these, these Hep G2 cells that you're seeing on the right, uh, these immortalized cancer cells that are proliferating on a plastic dish. 00:02:04.000 --> 00:02:12.000 They… they hold little resemblance to the functional elegance and complexity of human liver tissue and. 00:02:12.000 --> 00:02:18.000 This is, you know, while these models have been useful, this is a particular challenge. These cells are proliferating, they're. 00:02:18.000 --> 00:02:25.000 adhered to rigid plastic and really incapable of many of the differentiated hallmarks that we might observe in the. 00:02:25.000 --> 00:02:36.000 in the organ. But we've learned a great deal in recent decades about how to model liver epithelium using patient-derived cells. 00:02:36.000 --> 00:02:42.000 To better resemble the architectures. So you see here in this small caption, this cobblestone-like network of cells. 00:02:42.000 --> 00:02:49.000 And you're seeing that when those primary cells are isolated with sufficient viability and fidelity. 00:02:49.000 --> 00:02:59.000 And they're able to be reassembled into. tissue-like architectures that begin to behave much like the tissues, and so this is. 00:02:59.000 --> 00:03:07.000 An exciting opportunity. to begin to evaluate chemical safety with microphysiological systems. 00:03:07.000 --> 00:03:15.000 However, a major challenge with. these patient-derived cells is once they're ripped out of their native architecture. 00:03:15.000 --> 00:03:21.000 Uh, they rapidly de-differentiate, and without intervention, these cells will essentially become. 00:03:21.000 --> 00:03:30.000 uh, you know, dead. They just lose all their viability. They do have some short-term applications in drug metabolism, but outside of that. 00:03:30.000 --> 00:03:38.000 Uh, it's a real challenge to do this work. However, pioneering work from one of my mentors and many others in the field. 00:03:38.000 --> 00:03:48.000 Uh, in this case, uh, highlight Ed Leclus and his colleagues, uh, revealed critical factors for reestablishing epithelial-like canalicular networks. 00:03:48.000 --> 00:03:55.000 with these patient-derived cells to allow them to begin to manifest hallmarks of hepatocellular function. 00:03:55.000 --> 00:04:00.000 The key to this is culturing the cells with sufficient density. 00:04:00.000 --> 00:04:06.000 to enable them to be more three-dimensional, to be able to enable them to be more, uh. 00:04:06.000 --> 00:04:12.000 cuboidal in their structures that resembles better the in vivo architecture. 00:04:12.000 --> 00:04:22.000 And in doing so, we see rescue and recovery, maintenance of a number of hallmarks, such as drug metabolism activity. 00:04:22.000 --> 00:04:27.000 and signal transduction pathways that regulate the expression of. 00:04:27.000 --> 00:04:30.000 Uh, drug metabolizing enzymes and a myriad of other. 00:04:30.000 --> 00:04:43.000 transporters and related toxicological. Pathways. And so, these fundamental principles appear to be operative for many other epithelial tissues as well, not just the liver. 00:04:43.000 --> 00:04:51.000 Uh, but we're… we're learning from this over recent decades, and how we can begin to use these principles. 00:04:51.000 --> 00:04:57.000 to develop more sophisticated microphysiological systems that. are able to achieve tissue-like functionality, and. 00:04:57.000 --> 00:05:07.000 The good news is many of these systems have actually matriculated from concept and research into regulatory context. 00:05:07.000 --> 00:05:15.000 Presently, creating these types of culture systems and integrating assay platforms that allow us to make. 00:05:15.000 --> 00:05:24.000 interpretations and predictions for drug metabolism and drug-drug interactions can enable us to then make regulatory decisions and put. 00:05:24.000 --> 00:05:30.000 labels on new drugs that basically indicate whether there's a likelihood. 00:05:30.000 --> 00:05:40.000 of a response, whether clinical trials are needed. And so, the challenge we have is that for chemical safety research, this is not the case as we are. 00:05:40.000 --> 00:05:47.000 trying to much more extensively cover the myriad of effects that could happen in an animal or in a human. 00:05:47.000 --> 00:05:55.000 And in… along that route, we collect more than 120 tissues in a guideline animal study with clake biofluids. Many, many things can happen. 00:05:55.000 --> 00:06:05.000 And what's even more challenging is that these culture models, while they've been effective, are really limited. The traditional culture systems are in static. 00:06:05.000 --> 00:06:16.000 culture conditions in these sort of semi-2D environments, they exhibit poor nutrient exchange and are tethered to rigid plastic. 00:06:16.000 --> 00:06:22.000 in a survival mode that has super physiological concentrations of insulin and many other factors. 00:06:22.000 --> 00:06:29.000 And so these are, in essence, dying disease models that have a limited utility, but for chemical safety research. 00:06:29.000 --> 00:06:39.000 are really challenging, and so that's led to. sort of an explosion of different tissue engineering systems that have envisioned a way to sort of maintain. 00:06:39.000 --> 00:06:47.000 patient-derived cells with higher fidelity for weeks and months of exposure that include three-dimensional platforms. 00:06:47.000 --> 00:06:53.000 Microfluidic and barrier models, as well as multi-tissue axes, where we begin to. 00:06:53.000 --> 00:06:57.000 combine the liver and, say, the thyroid or other models. 00:06:57.000 --> 00:07:05.000 That, uh, that these tissues sort of work together in vivo in an axis of pathophysiology in response to chemical exposures. And so, how can we begin to. 00:07:05.000 --> 00:07:15.000 Look at this landscape of things alongside, uh… useful, uh, enzyme and biomarker assays to then begin. 00:07:15.000 --> 00:07:24.000 begin to make decisions about chemical safety research. And these systems really do have a lot of potential, but. 00:07:24.000 --> 00:07:33.000 Uh, what's challenging right now is that we're not able to fully establish the context of use for all these different systems. There's a wide range of them. 00:07:33.000 --> 00:07:42.000 And then we ultimately have to do the heavy lifting to qualify these systems, to calibrate how to use data from these systems with human therapeutics. 00:07:42.000 --> 00:07:49.000 to understand xenobiotic exposure responses with environmental chemicals, drugs, botanicals. 00:07:49.000 --> 00:08:00.000 and other… other things. So, to address these needs, we've been advancing research and development efforts in multiple avenues, and one of the key partnerships that I'd like to highlight today. 00:08:00.000 --> 00:08:05.000 has been our work with the TIX VAL Consortium, led by Von Rusen at Texas A&M. 00:08:05.000 --> 00:08:12.000 Uh, to explore the landscape of these different systems in collaboration with industry and government researchers. 00:08:12.000 --> 00:08:24.000 Uh, led by laboratory technical experts like Courtney Sokolish in his lab. The TEXVAL team has comparatively evaluated a range of microphysiological systems. 00:08:24.000 --> 00:08:31.000 And provided methods to partner labs like ours that help us to extend. 00:08:31.000 --> 00:08:35.000 Uh, this kind of technology that's coming out of academic research. 00:08:35.000 --> 00:08:48.000 into, uh… additional research applications, and so we're really excited about continuing this partnership as we evaluate a number of device systems and. 00:08:48.000 --> 00:09:00.000 Look at clinical biomarker assays. In our laboratory here at NIHS, we've been working to extend emerging science and technology to understand and predict human responses to drug. 00:09:00.000 --> 00:09:07.000 natural product and environmental chemical exposures for many years, and our laboratory. 00:09:07.000 --> 00:09:14.000 looks to develop these microphysiological systems. alongside biomarker assays and high-dimensional assays that give us mechanistic insights. 00:09:14.000 --> 00:09:22.000 And then, ultimately, using computational models, interpret those assays into decision context. And we've been doing that work for quite a while. 00:09:22.000 --> 00:09:28.000 Uh, and really, our primary aims are to develop and qualify these physiologically relevant. 00:09:28.000 --> 00:09:33.000 microphysiological systems and assays into NAMS that are actually relevant. 00:09:33.000 --> 00:09:44.000 Uh, and useful. And to lead into innovative toxicological investigations of environmental substances, and I'll speak about some of our work today on PFAS and aqueous film-forming foams. 00:09:44.000 --> 00:09:51.000 And then, to address critical gaps in NAMM-based toxicology research. 00:09:51.000 --> 00:09:58.000 Our pioneering work in introducing free-floating spheroid cultures in human hepatocytes for chemical safety research. 00:09:58.000 --> 00:10:03.000 has revealed these little powerhouse liver microtissues can be. 00:10:03.000 --> 00:10:09.000 robust, uh, little engines that are able to emulate. 00:10:09.000 --> 00:10:17.000 hallmarks of hepatocellular function, and what's really great about these is they support not just days of exposure that we can see in our 2D cultures. 00:10:17.000 --> 00:10:23.000 but actually months of high-fidelity functionality. They're able to then, uh, as a result of that. 00:10:23.000 --> 00:10:29.000 Help us to predict the potency ranges for drug-induced liver injury and their highly efficient and cost. 00:10:29.000 --> 00:10:43.000 cost efficient, and as they use so many fewer cells compared to a standard two-dimensional culture model. And this is the type of model that Bevin and I have been working on for a number of years. 00:10:43.000 --> 00:10:52.000 And we're really excited to see how these can work. And in particular, I would just highlight the fact that they begin to polarize and form. 00:10:52.000 --> 00:10:59.000 Uh, liver-like architecture, uh, just spontaneously, that's much more elaborate than what we see in two-dimensional cultures. 00:10:59.000 --> 00:11:14.000 Now, one of the sort of key… hallmarks of hepatocellular function that's really distinguished the less evolved or differentiated systems like iPS-derived hepatocytes at HEPG2 cells. 00:11:14.000 --> 00:11:21.000 from the liver tissue and suspensions of hepatocytes that we've seen is their drug metabolism proficiency. 00:11:21.000 --> 00:11:28.000 And so, in this, this sort of, uh, this is your life photo, you see an example on the left where we're looking at the median. 00:11:28.000 --> 00:11:37.000 Drug metabolism activity for CYP1A2, 2B6, and 3A4 across hundreds of donor preparations of human hepatocytes. 00:11:37.000 --> 00:11:44.000 And so, in comparing that median to where we were in TOX 21, uh, in using immortalized cells. 00:11:44.000 --> 00:11:53.000 You can see we had just minimal metabolic activity for these major clinical enzymes that are important for drug clearance in humans. 00:11:53.000 --> 00:11:59.000 And then 2D models of human hepatocytes, you can see how they're about only 10% of what we see. 00:11:59.000 --> 00:12:05.000 Uh, in terms of their initial starting point. So even then, we see that they're de-differentiating to some extent. 00:12:05.000 --> 00:12:09.000 But what we were excited to discover is that when we took. 00:12:09.000 --> 00:12:15.000 Uh, the surrogate for primary human hepatocytes that we've been using, these HIPAA-RG cells, we found that. 00:12:15.000 --> 00:12:20.000 When you cultured them in increasingly sophisticated culture conditions. 00:12:20.000 --> 00:12:30.000 that we saw a dramatic increase in the enzymatic activities that began to rival what we see in the median suspension human hepatocytes. And so. 00:12:30.000 --> 00:12:46.000 This suggests not only that they're able to. to be responsive to… to… to the culture conditions, but they're also indicating that there's something really special going on in these free-floating microtissues, and I'll talk a little bit more about that later. 00:12:46.000 --> 00:12:57.000 Um, but what we're excited about is not only are these metabolically proficient, they're also able to retain liver-like signal transduction pathways. They're not just emulating zone 3 of the liver. 00:12:57.000 --> 00:13:06.000 But they're further inducible, analogous to liver tissue. to mimic these major hepatic receptor signaling pathways, many of which are involved in cancer. 00:13:06.000 --> 00:13:13.000 And to polarize, as I mentioned earlier, to form these biliary excretion networks that enable them to. 00:13:13.000 --> 00:13:20.000 to not be cholestatic, but to actually be a viable model that's actively, uh. 00:13:20.000 --> 00:13:27.000 effluxing bile acids into these bile pockets, and then ultimately out into the media. 00:13:27.000 --> 00:13:43.000 But with their small size, that dilution is quite good, and with the Brownian motion that can happen in moving around, we think that they're really exquisite in their ability to exchange nutrients and waste products and achieve this level of metabolic proficiency. 00:13:43.000 --> 00:13:50.000 And we found that when we looked at drug analogs, like trofloxacin versus levofloxacin, that had varied. 00:13:50.000 --> 00:13:59.000 clinical outcomes for liver injury, we're able to readily distinguish the hepatotoxic molecules from the non-hepatotoxic molecules. 00:13:59.000 --> 00:14:07.000 troglitazone and rosiglitazone was another example, and we've had many of these while we look at molecules like aspirin and gemfibrozil that have. 00:14:07.000 --> 00:14:17.000 Minimal reports of liver injury, and these models seem to be able to emulate that type of effect, and we've dug more deeply into these mechanisms and pathways. 00:14:17.000 --> 00:14:27.000 Uh, and we've also applied them in some ways to external things. So botanicals are particularly an area where we have intentional human exposures. 00:14:27.000 --> 00:14:37.000 And Adam Pearson in the lab was able to take these microtissue type models and apply them under our botanical safety consortium research. 00:14:37.000 --> 00:14:47.000 Uh, using these 384 wall plates. And a combination of cell painting and diagnostic-type assays for liver enzyme leakage and albumin production. 00:14:47.000 --> 00:14:57.000 And CYP3A4 inhibition. As well as transcriptomics have enabled us to have, uh, create data for a number of manuscripts in this space. 00:14:57.000 --> 00:15:02.000 And really demonstrated to us the power of these little microtissues. 00:15:02.000 --> 00:15:08.000 to inform a variety of applications, and I'll talk a bit more about some of those in the next few slides, but. 00:15:08.000 --> 00:15:11.000 One of the things I wanted to highlight to folks is that. 00:15:11.000 --> 00:15:16.000 when we look at, um, the longevity of 2D. 00:15:16.000 --> 00:15:22.000 primary rat hepatocytes. It's typically 3 to 5 days, and then the cells really start. 00:15:22.000 --> 00:15:30.000 Excuse me, starting to struggle and die. But those same cells in these 3D microtissues have months of longevity. 00:15:30.000 --> 00:15:35.000 And not only longevity, but functional longevity, and able to sort of. 00:15:35.000 --> 00:15:43.000 rescue metabolic proficiency, uh, respond to those CAR-type activators and other signaling pathways. 00:15:43.000 --> 00:15:53.000 and stain histologically in collaboration with Dr. Darlene Dixon, to reflect the same types of architecture that we see in liver tissue, in this case, glycogen staining and. 00:15:53.000 --> 00:15:59.000 CK19 staining for, um. cholangiocyte-like cells. 00:15:59.000 --> 00:16:17.000 And what this enables… is now an ability to bridge the gap between our animal studies and our human studies using this parallelogram-type approach of microtissues that we can compare between rat and human, and we know these systems do emulate. 00:16:17.000 --> 00:16:26.000 Um… interspecies comparisons, and the molecular potencies across species can oftentimes be very predictive. 00:16:26.000 --> 00:16:30.000 And so we're excited to see how this type of enabling technology. 00:16:30.000 --> 00:16:43.000 can expand to complement. our animal studies, and really, due to their small size, even give us an opportunity to reduce the number of animals that we're using in toxicology research while we're. 00:16:43.000 --> 00:16:55.000 building confidence in their context of use. So, we've also been integrating omics assays, and a few years back, we published this Power of Resolution Manuscript. 00:16:55.000 --> 00:17:01.000 that describes our efforts to develop high-throughput transcriptomics with these types of models. 00:17:01.000 --> 00:17:06.000 Uh, and we were able to successfully show that this platform defined a threshold. 00:17:06.000 --> 00:17:13.000 transcriptomically defined threshold. that enables you to define the relative potency. 00:17:13.000 --> 00:17:18.000 across a range of compounds based on where they cross this threshold. 00:17:18.000 --> 00:17:26.000 And using these types of data, we can now distinguish drug-induced liver injury compounds like trofofloxacin and levofloxacin. 00:17:26.000 --> 00:17:35.000 both qualitatively and in terms of their relative potencies. And so that begins to create an opportunity for predictive toxicology and chemical safety. 00:17:35.000 --> 00:17:45.000 that better defines margins of exposure, and the potency ordering of those firing sequences that occurs between. 00:17:45.000 --> 00:17:54.000 biological effects that might be off-target or, you know, reversible adaptive responses relative to those that are stressful and cytotoxic. 00:17:54.000 --> 00:17:59.000 is a key opportunity within these types of data to just sort of. 00:17:59.000 --> 00:18:07.000 resolve the relative potencies of biological responses, and then relating those with. 00:18:07.000 --> 00:18:15.000 biomarkers of clinical pathology and cellular imaging begins to build confidence, and one example that I highlight here is with aflatoxin. 00:18:15.000 --> 00:18:22.000 where we were actually able to observe scarring that occurred in these 2D cultures of Hepner RG. 00:18:22.000 --> 00:18:33.000 And we were able to observe clearly that at those same concentration ranges, we were getting molecular pathway responses related to injury and repair and DNA damage. 00:18:33.000 --> 00:18:40.000 And so this type of response with aflatoxin B1 just highlights the potential for these types of systems. 00:18:40.000 --> 00:18:53.000 to more comprehensively inform chemical safety research. And we're excited to see where these can go. And we've actually applied them recently in an ES&T publication. 00:18:53.000 --> 00:19:03.000 Uh, where we've been looking deeply in support of the Department of Defense and their needs to identify safer aqueous film-forming foams for military use. 00:19:03.000 --> 00:19:08.000 And in our coordination with them, published this manuscript. 00:19:08.000 --> 00:19:15.000 that basically is highlighting an evaluation of 30 test substances. 00:19:15.000 --> 00:19:23.000 that included 5 aqueous film-forming foams, and a number… a large number of PFAS perfluorinated alkyl substances. 00:19:23.000 --> 00:19:28.000 As well as human drugs that help to anchor the data set. 00:19:28.000 --> 00:19:35.000 And in this work, we were able to use a combination of this transcriptomic thresholding approach. 00:19:35.000 --> 00:19:45.000 Uh, to predict the potencies for liver injury. We were also able to use multiple data streams to build a weight of evidence for that in these systems. 00:19:45.000 --> 00:19:51.000 Uh, we were able to, as with the drugs, we were able to identify mechanistic pathway potencies. 00:19:51.000 --> 00:20:02.000 And then, to define biological response similarities, such as what you see here by comparing PFAS, which is a very extensively studied PFAS and is known to be toxic. 00:20:02.000 --> 00:20:13.000 with structurally similar but shorter chain PFH excess. And this poorly understood and data-poor molecule, 6-2 fluoroteomer sulfonate. 00:20:13.000 --> 00:20:21.000 And what we see is that they share a great deal of biological response similarity in these human hepatocyte microtissue models. 00:20:21.000 --> 00:20:38.000 And so the idea is to continue to. to help support our regulatory needs for read-across, and this type of data help us to decide whether these alternative or more modern molecules are toxic, and what's the relative potency. 00:20:38.000 --> 00:20:48.000 Uh, for those. We also have expanded that biological response similarity analysis to sort of more fully. 00:20:48.000 --> 00:20:54.000 capture which chemicals are similar to one another, but more importantly, to anchor their. 00:20:54.000 --> 00:20:59.000 Their aggregate biological effects at the very low concentrations. 00:20:59.000 --> 00:21:08.000 Uh, with… with model compounds like YF14643, which is a known activator of the peroxilone proliferator activator receptor alpha. 00:21:08.000 --> 00:21:15.000 Or, uh, omeprazole and phenobarbital, which can activate the pregnant X receptor and. 00:21:15.000 --> 00:21:27.000 the constituent of Androsane receptor, as well as a cluster here that was just containing cyclosporin A and a bunch of aqueous film-forming foams that might be more related to lipid disruption, as we know. 00:21:27.000 --> 00:21:40.000 cyclosporin A can cause fatty liver. And so, taking all of these sort of similarity relationships in mind, we can begin to also think about not only supporting our. 00:21:40.000 --> 00:21:45.000 our EPA colleagues and read across, but we also can begin to predict. 00:21:45.000 --> 00:21:50.000 Apical outcomes, and so in this case, we took the relative potencies for car. 00:21:50.000 --> 00:21:59.000 PPAR activation and liver injury. that we, uh, that are well-established modes of hepatomegaly, and then to use that. 00:21:59.000 --> 00:22:05.000 To predict whether we are, uh, what are the concentrations in vivo in the animal. 00:22:05.000 --> 00:22:09.000 And those are shown in the blue, or the purple triangles here. 00:22:09.000 --> 00:22:20.000 The concentrations at which we see. Um, in vivo effects to enlarge the liver, and what we're seeing is that when we look at these pathway potencies in human hepatocytes. 00:22:20.000 --> 00:22:26.000 relative to the in vivo liver weight lobe, we're getting a reasonable overlap, and so we're able to predict. 00:22:26.000 --> 00:22:40.000 The animal OL using internal doses. And we've taken all these types of data and begun to extend them both in terms of our in vivo investigation of PFAS bioaccumulation from aqueous film-forming foam exposures. 00:22:40.000 --> 00:22:45.000 Our in vivo biology from AFFF exposures and our in vitro mechanistic data. 00:22:45.000 --> 00:22:54.000 Uh, to begin to… to… to enable development of prioritization frameworks that can identify safer products. 00:22:54.000 --> 00:23:01.000 And ultimately, uh, help to improve our environment with safer options. 00:23:01.000 --> 00:23:10.000 So, we've also identified that PFAS accumulation is one of the fundamental concerns with, um. 00:23:10.000 --> 00:23:15.000 with PFAS, and we've been able to use tissue chip platforms, such as the Emulate platform. 00:23:15.000 --> 00:23:25.000 To demonstrate we can distinguish long and short half-life PFAS from one another, uh, based on their renal reuptake. So in this case, we established a kidney, uh. 00:23:25.000 --> 00:23:33.000 tissue chip system, and we're able to show PFAS extensively accumulated in epithelial cells from apical exposures compared to. 00:23:33.000 --> 00:23:43.000 PFBS. And we've done similar work with the liver and liver microtissues to demonstrate distinctions between the long and the short chain. 00:23:43.000 --> 00:23:51.000 PFAS, and we're also interested in addressing the more than 15,000 PFAS that have been identified in Commerce. 00:23:51.000 --> 00:24:01.000 Uh, by the US EPA in recent, recent, uh, recent publications, and to use computational models, uh, including the bioconcentration factor, to help stratify. 00:24:01.000 --> 00:24:07.000 PFAS, even if they don't cross historical thresholds, and using combinations of. 00:24:07.000 --> 00:24:16.000 of bioconcentration and hepatic clearance to be able to tier these molecules, these more than 15,000 molecules, into testable. 00:24:16.000 --> 00:24:29.000 units that help us fill the data gaps. I want to switch back to the three-dimensional microtissue models now, and just show you a couple examples where these micro-tissue models in 3D. 00:24:29.000 --> 00:24:42.000 convey superior, uh, ability to distinguish hepatotoxicant. responses with valproic acid and cyclophosphamide. And so here, uh, we're looking at two compounds that require metabolic activation. 00:24:42.000 --> 00:24:57.000 And what we see is robust responses in 3D compared to very small responses in 2D cultures. And so, these are two of the sort of less common examples, but clear examples, where metabolic activation. 00:24:57.000 --> 00:25:08.000 Uh, maybe a toxicity-determining step. Um, with benzoylpha-pyrin, we actually saw the opposite effect. With our proliferating HIPAA-RG, we saw they were more sensitive. 00:25:08.000 --> 00:25:12.000 than the 3D microtissues, and this may be related to. 00:25:12.000 --> 00:25:20.000 the fact that the benzoylfa pyroine is causing DNA damage, and the rate determining step for toxicity is different, and so. 00:25:20.000 --> 00:25:31.000 One of these may be more predictive than the other, and we're continuing to investigate this, but we… it's not always the case that three-dimensional cultures are more sensitive, and so we're digging into this more deeply. 00:25:31.000 --> 00:25:37.000 And we've done this type of analysis for many, many drugs, and one of the sort of interesting. 00:25:37.000 --> 00:25:42.000 trends we've seen. When we look at transcriptomic pathway enrichment. 00:25:42.000 --> 00:25:47.000 is that 3D hepatocyte cultures tend to be much more extensively enriched. 00:25:47.000 --> 00:25:54.000 for anticipated or known targets of these compounds. So you're looking at a range of compounds that have. 00:25:54.000 --> 00:26:00.000 known targets, fenofibric acid for PPAR signaling. phenobarbital for car signaling. 00:26:00.000 --> 00:26:05.000 rosiglitazone for PPAR signaling, and what we see across the board. 00:26:05.000 --> 00:26:16.000 Uh, is a better enrichment in 3D compared to 2D culture models, and so… In aggregate, this is telling us something in addition to the fact that many of these models are more predictive. 00:26:16.000 --> 00:26:22.000 Uh, across the board, and we're excited to see how this type of knowledge can then really. 00:26:22.000 --> 00:26:38.000 increase our ability to do predictive toxicology. Um, just want to highlight near the end here, just a couple of quick, uh, points. So, we've been working, uh, not only in liver, which is the vast majority of our work, but Adam Pearson also established in our lab. 00:26:38.000 --> 00:26:50.000 methods for looking at proximal tubule. Uh, we particularly know that the kidney is one of the target tissues of highest concern, and the proximal tubule being the most vulnerable. 00:26:50.000 --> 00:26:57.000 And he discovered that these small microtissues of renal proximal tubule cells actually remember. 00:26:57.000 --> 00:27:01.000 how large the proximal tubule was in vivo, and they just auto-assemble. 00:27:01.000 --> 00:27:11.000 into these small tubuloids is what we're calling them. And in comparing the 2D cultures with domes, so these would be considered a differentiated 2D culture. 00:27:11.000 --> 00:27:22.000 that's forming a fluid pocket under the epithelium. Uh, we're seeing sensitized response to compounds like cisplatin, and we've done a number of compounds. 00:27:22.000 --> 00:27:28.000 about 37 compounds we've screened with this platform, and those publications are in. 00:27:28.000 --> 00:27:37.000 in preparation. And finally, I just want to leave you with a couple of thoughts. When I joined the NIHS, one of the big things that we were concerned about is. 00:27:37.000 --> 00:27:46.000 Uh, you know, in an animal model, anything could happen. So we're trying to understand what it is that we can predict with NPS and NAM-based systems, and. 00:27:46.000 --> 00:27:54.000 To address this question, we first need to understand, well, what do we normally look for? And so, I turn to the NTP's histopathology glossary. 00:27:54.000 --> 00:28:00.000 Um, where there are a thousand lines of things that a pathologist could find in an animal. 00:28:00.000 --> 00:28:07.000 And I basically started to map. those morphologies and those locations, and trying to then say, well. 00:28:07.000 --> 00:28:14.000 in a liver microphysiological system, what could we potentially see with the existing or emerging models? 00:28:14.000 --> 00:28:20.000 And then work to develop a Rosetta Stone that translates liver MPS data into. 00:28:20.000 --> 00:28:33.000 recognizable pathology findings that we can then meet the regulators and other decision makers in a more recognizable place, and then ultimately calibrating those with drugs so that we can quantitatively use them. 00:28:33.000 --> 00:28:39.000 And just to show you a high-level summary of what I was able to determine, but essentially. 00:28:39.000 --> 00:28:44.000 even though we're only looking at one tissue out of the 120 that you might collect. 00:28:44.000 --> 00:28:50.000 Uh, and we're not really able to identify all the locators across all of pathology. 00:28:50.000 --> 00:29:01.000 We are starting to make an impact in the number of morphologies and modifiers that we could observe there, and I'm going to be going more into this and partnering with folks in pathology to. 00:29:01.000 --> 00:29:08.000 to sort of roll these out and vet them as we go further, but I think the initial point is to say we're trying to understand. 00:29:08.000 --> 00:29:16.000 what it is we want to predict, and then to tune microphysiological systems towards these endpoints that can then be used in decision making. 00:29:16.000 --> 00:29:27.000 And finally, I'm going to leave you with feature directions. We're already working to, uh, you know, manifest toxicological phenotypes, as I mentioned, and this pathology sort of mindset. 00:29:27.000 --> 00:29:36.000 Uh, we also want to begin to do a much better job in addressing factors of inter-individual susceptibility to chemical exposures. 00:29:36.000 --> 00:29:41.000 Uh, we have, oftentimes, adult models, because that's where clinical trials are run, and. 00:29:41.000 --> 00:29:47.000 But we don't have models of fetal, although iPS-derived cells. 00:29:47.000 --> 00:29:52.000 could be used in that way. Uh, geriatric childhood neonatal, we don't have those. 00:29:52.000 --> 00:29:59.000 Uh, it's sad to say we don't have. sexually dimorphic culture conditions for hepatocytes. 00:29:59.000 --> 00:30:07.000 We put them in a monosex medium, um, that… and we may use donors from a male and female, but we need to do a much better job of this. 00:30:07.000 --> 00:30:26.000 Um, genetic predisposition and disease haplotypes, and then. Probably most importantly, from a susceptibility standpoint, is modeling pre-existing disease, like steatosis. We're starting to do that now with high-fat media, and then looking to see how these might differentially sensitize. 00:30:26.000 --> 00:30:36.000 Uh, to chemical toxicity. And then finally, just expanding tissue coverage and axes of pathophysiology, as I mentioned earlier, the liver, thyroid, axis. 00:30:36.000 --> 00:30:42.000 And the concerns with PFAS. So, with that, I'll be happy to take any questions and just really. 00:30:42.000 --> 00:30:51.000 It's an honor to work at the NIHS, and it was such an amazing group of talented scientists, and thank you for your time. 00:30:51.000 --> 00:30:58.000 Thank you, Steve. Um, so, listeners, you can enter questions through the, um, question-answer pod. 00:30:58.000 --> 00:31:13.000 And looks like we have a few already on deck here, so… How do you deal with floating HEPA RG cells that do not readily aggregate into spheroids? 00:31:13.000 --> 00:31:32.000 Yeah. So if the HEPARG cells are not aggregating into spheroids, there's probably something, uh, not… ideal in the cultural conditions. They will self-aggregate if the cell numbers and the cell configurations are correct, and we often find a more conical. 00:31:32.000 --> 00:31:40.000 uh, shape will make that happen faster, but it typically takes between 3 and 7 days for those cells to coalesce into a microtissue. 00:31:40.000 --> 00:31:51.000 Now, once they're formed into that microtissue, we do use special plates, and there are a few different types. Some plates are more or less sensitive to the floating. 00:31:51.000 --> 00:32:03.000 But we use standard liquid handler, in our case, the Viaflo, that we stamp a 384-wheel plate, we remove half the media, and in some cases, we remove 80% or 90% of the media. 00:32:03.000 --> 00:32:14.000 And then process the cells that way, but it's, it's pretty straightforward. The primary hepatocytes tend to be more dense than the HEPARGs after about 10 days in culture. 00:32:14.000 --> 00:32:28.000 Um, so that's… that's one distinction that we've observed, but… but that's how we handle floating HepRGs, is that they… they settle to the bottom, and with liquid handling, where you're only exchanging about half the medium every other day or every day. 00:32:28.000 --> 00:32:31.000 It's no problem. 00:32:31.000 --> 00:32:38.000 Okay, uh, next question. Can you comment on how your assays, such as HEPARG, stack up? 00:32:38.000 --> 00:32:43.000 Oh, it's all good. Got it. Fix this here, wait. 00:32:43.000 --> 00:32:45.000 Just one second… Yes, I guess most models, do you see the same responses? 00:32:45.000 --> 00:32:50.000 Against mouse models, I think that's it, yeah. 00:32:50.000 --> 00:32:58.000 Yeah, we haven't made mouse, uh, spheroids yet. Uh, we've made rat, we feel very comfortable that we're seeing, um. 00:32:58.000 --> 00:33:02.000 things that make sense there. We know that 2D models of mouse. 00:33:02.000 --> 00:33:08.000 are good, but very short-term. They're even shorter-lived than the rat. 00:33:08.000 --> 00:33:15.000 Um, and so… but there have been reports of mouse spheroids being formed, and they should be a good model. 00:33:15.000 --> 00:33:31.000 And again, you would use far fewer mice, because they don't eat well, would require a very small amount of cells, and so, um… I'm very encouraged by what I've seen so far with the rat, and… but we haven't done the mousework yet. 00:33:31.000 --> 00:33:40.000 Okay, thank you. So, are you comparing biological pathway readouts using MPS, etc, with responses seen in human population studies? 00:33:40.000 --> 00:33:43.000 If so, how do these compare? 00:33:43.000 --> 00:33:50.000 Yeah, we haven't… we haven't, in our lab, gotten to that point. I think we… we would be looking to partner with researchers who. 00:33:50.000 --> 00:34:01.000 have existing, you know, epidemiological observations and hypotheses, and what we're thinking about is more developing those states of susceptibility. 00:34:01.000 --> 00:34:05.000 And we don't know what our end's gonna be yet for any given state. 00:34:05.000 --> 00:34:19.000 But the idea is to capture those states, and then we'll do power calculations to understand how many we would need to see what effect. And for acute effects, the numbers probably won't be that large, but when we start to get to something that's slower developing. 00:34:19.000 --> 00:34:24.000 And it takes much longer exposure. The end's gonna likely increase, and so I think. 00:34:24.000 --> 00:34:29.000 That's where the scientific community has to decide what's the value of. 00:34:29.000 --> 00:34:45.000 you know, digging deeper compared to the decisions you could make shorter term. But, um, you know, that… we're a long ways from being able to do that type of work, but the technology at scale is there to do it. 00:34:45.000 --> 00:34:51.000 So the next question, are we missing the inflammatory contributors to toxicology? 00:34:51.000 --> 00:34:53.000 our toxicity in these models, and any comments? 00:34:53.000 --> 00:35:07.000 That's a… that's a great comment, and we're… we are and we're not, so… Some of the models that I showed you have resident macrophages, the Kupfer cells that are already in these microtissues, and. 00:35:07.000 --> 00:35:16.000 When they're in a happy 3D microtissue, they become quiescent if they're not already, typically they're already isolated to be quiescent. 00:35:16.000 --> 00:35:27.000 Um, and then they will become activated when toxicity occurs. What we don't have is circulating macrophages in your typical microtissue model. 00:35:27.000 --> 00:35:51.000 And so, being able to see infiltration and a lot of things you're going to see in pathology is something that could be done, might be more amenable to some of the tissue chip systems where there's been some progress with that, but it could also work with microtissues with the right tuning. We just haven't gotten far enough to do that, but there is an inflammatory component there, but it's not… the amplifier is probably not there to really. 00:35:51.000 --> 00:35:57.000 uh, you know, manifested, you know, an organ-level toxicity. 00:35:57.000 --> 00:36:03.000 Okay, next question. What might be correlated more by microtissia 3D systems, the toxokinetico. 00:36:03.000 --> 00:36:07.000 Technetic or toxico… dynamic aspects. 00:36:07.000 --> 00:36:15.000 Well, if we've done it properly, it should be both, but um… That's a great point. These microtissues. 00:36:15.000 --> 00:36:24.000 are very small, and so they don't… They don't reflect the allometric scaling of blood to liver. 00:36:24.000 --> 00:36:28.000 And so we have to tune and calibrate these systems. 00:36:28.000 --> 00:36:37.000 for the decision context that they're useful for, but what I can say as a general principle is if we've modeled the toxicokinetics. 00:36:37.000 --> 00:36:46.000 Real… in a relevant way. the Tosico dynamics and the translation of an individual effect, whether it's benign or toxic. 00:36:46.000 --> 00:36:55.000 should be more accurate. And what we're seeing in these microtissue models and other MPS systems, that they have a better capacity to. 00:36:55.000 --> 00:37:05.000 partition their own internal exposure, uh, better than we would think, but still, um, they're going to be limitations unless we have. 00:37:05.000 --> 00:37:13.000 a full accounting for that, and that's why I'm… really interested in pooled spheroid models, where we can have a much bigger liver than we do a thyroid, or. 00:37:13.000 --> 00:37:18.000 other tissue so that we can try to scale those axes to have metabolomic. 00:37:18.000 --> 00:37:29.000 endocrine… paracrine signaling that's more equivalent to what we would see in VO, and how those tissues would adapt and interact with one another. 00:37:29.000 --> 00:37:37.000 Okay, um, thanks a lot for your presentation, Steven, it was… Terrific, and also perfect timing. 00:37:37.000 --> 00:37:45.000 So, let's move on to our, um, second presentation today. Next, we have Susan Tilton. 00:37:45.000 --> 00:37:50.000 Associate Professor in the Department of Environmental and Molecular Toxicology. 00:37:50.000 --> 00:37:55.000 At Oregon State University and a project leader at Oregon State University. 00:37:55.000 --> 00:38:05.000 Superfund Research Center. Her lab uses molecular and computational tools to predict toxic… toxicity of chemical contaminants in the environment. 00:38:05.000 --> 00:38:11.000 Dr. Tilmutter introduced a 3D human lung respiratory model of asthma. 00:38:11.000 --> 00:38:24.000 So, Dr. Tilton? Please feel free to begin when you're ready. 00:38:24.000 --> 00:38:29.000 All right, thank you. 00:38:29.000 --> 00:38:34.000 So let me see, um… 00:38:34.000 --> 00:38:38.000 just make sure that you can hear me okay, and you can see my slides. 00:38:38.000 --> 00:38:39.000 Yes, we can. 00:38:39.000 --> 00:38:46.000 Okay. Um, well, thank you for that introduction, um, and thank you for the invitation to speak today. 00:38:46.000 --> 00:38:52.000 Um, as that said, I'm an associate professor and principal, um. 00:38:52.000 --> 00:39:00.000 Project principal investigator, uh, for the Oregon State University and Pacific Northwest National Laboratory Superfund Research Center. 00:39:00.000 --> 00:39:04.000 Uh, so today, I plan to share with you, um. 00:39:04.000 --> 00:39:10.000 Some of our work using primary human bronchial epithelial cells cultured at the air-liquid interface. 00:39:10.000 --> 00:39:15.000 As a 3D model to investigate health effects of polycyclic aromatic hydrocarbons. 00:39:15.000 --> 00:39:23.000 associated with, um, both contaminated sites and wildfire smoke. 00:39:23.000 --> 00:39:32.000 So I'd like to begin, um, just by briefly highlighting our center and, um, our integrated projects and cores. 00:39:32.000 --> 00:39:39.000 So our center is titled, uh, the Center for Science, Technology, and Emerging Health Risk of PAHs. 00:39:39.000 --> 00:39:45.000 It's, uh, driven by a mission to identify pHs in the environment, characterize their toxicity. 00:39:45.000 --> 00:39:51.000 determine environmental concentrations below which they pose minimal risk to human health. 00:39:51.000 --> 00:39:58.000 And also to develop safe and sustainable approaches for remediating pH-contaminated sites. So. 00:39:58.000 --> 00:40:05.000 Um, overall, our goals are to understand the composition of these complex mixtures containing PAHs. 00:40:05.000 --> 00:40:15.000 how they can change in the environment, and the implications for both these pHs and mixtures for human health. 00:40:15.000 --> 00:40:23.000 Uh, research program relies heavily on the use of integrated, um, new approach methodologies, or NAMs. 00:40:23.000 --> 00:40:30.000 So this includes, um, cell-based systems, non-mammalian whole animal models, like superfish. 00:40:30.000 --> 00:40:39.000 And in silico approaches that are used, um. Throughout the program, and today I'll talk about some work in collaboration, in particular with. 00:40:39.000 --> 00:40:43.000 Jordan Smith in the predictive dosimetry and Metabolism core. 00:40:43.000 --> 00:40:49.000 Um, also coupled to passive sampling techniques to quantify environmental fate. 00:40:49.000 --> 00:40:58.000 and human exposure to chemicals. So together, these tools allow us to efficiently link environmental measurements with biologically relevant. 00:40:58.000 --> 00:41:04.000 human outcomes. Some of the key goals, uh, for my project in particular. 00:41:04.000 --> 00:41:14.000 Um, include using 3D respiratory models to establish quantitative relationships between chemical exposure and toxicity. 00:41:14.000 --> 00:41:21.000 advancing these systems… Um… to predict pH toxicity in susceptible populations. 00:41:21.000 --> 00:41:32.000 And then understanding mechanisms, uh, driving lung toxicity from novel substituted pHs, and those in complex mixtures. So this is particularly relevant. 00:41:32.000 --> 00:41:41.000 For, um, multi-source exposures near Superfund sites, which in the Pacific Northwest often include, um, concurrent exposures. 00:41:41.000 --> 00:41:49.000 Uh, to wildfire smoke, for example, among others. So I'll focus on some of these goals today. 00:41:49.000 --> 00:42:04.000 In addition, within our program, an essential component is just the engagement with state and federal agencies, um, industry partners, and local communities, so through these collaborations, uh, we communicate our science to support informed decision making. 00:42:04.000 --> 00:42:14.000 related to the assessment, management, and mediation, remediation of contaminated sites throughout our region. 00:42:14.000 --> 00:42:19.000 So I've mentioned that our program focuses on polycyclic aromatic hydrocarbons. Um, these are. 00:42:19.000 --> 00:42:26.000 Uh, ubiquitous contaminants in the environment that can occur from both naturally occurring and anthropogenic sources. 00:42:26.000 --> 00:42:32.000 pHs account for 3 of the top 10 chemicals of concern at priority pollutant sites. 00:42:32.000 --> 00:42:48.000 And are associated with multiple adverse health effects. Our Superfund program in particular, um… Over a number of years has identified both parent PHs and transformation products that are formed in the environment. 00:42:48.000 --> 00:42:59.000 Um, across, uh, multiple phases in soil, water, and air, um, associated with Superfund sites and other hazardous waste sites. 00:42:59.000 --> 00:43:03.000 In addition, I wanted to highlight, um, through some of the work. 00:43:03.000 --> 00:43:09.000 Uh, and Kim Anderson's, uh, project, um, so she's a project lead in our program. 00:43:09.000 --> 00:43:14.000 Uh, working towards identifying, um, that alkylated pHs in particular. 00:43:14.000 --> 00:43:21.000 rank among the most abundant pHs measured in environmental samples, um, from multiple sources. So this includes. 00:43:21.000 --> 00:43:27.000 Petrochemicals associated with Superfund sites, um, and thermal decomposition of organic matter. 00:43:27.000 --> 00:43:35.000 So, for example, in this case, we're passive samplers were deployed, um, in air and water at a browns-filled creosote site. 00:43:35.000 --> 00:43:52.000 Um, at all locations, um, and sampling dates. Uh, we can see that alkylated pHs that are summed here by the blue dots were more abundant, um, than the unsubstituted or parent pHs that are quantified by the orange dots. 00:43:52.000 --> 00:43:56.000 Um, so this is in both air and surface water. 00:43:56.000 --> 00:44:02.000 Um, and again, across all sites, um, and time points. 00:44:02.000 --> 00:44:09.000 Her group is also observed in passive sampling with wristbands, so these are used to measure personal exposure to PAHs. 00:44:09.000 --> 00:44:19.000 Uh, that seasonally, many of the more highly abundant pHs that are measured are 2- to 3-ring alkylated PAHs. 00:44:19.000 --> 00:44:26.000 Um, so some of the primary sources during summer are associated with wildfire smoke exposure. 00:44:26.000 --> 00:44:33.000 Or in the winter with, um, indoor sources, like, uh, indoor heating or cooking. 00:44:33.000 --> 00:44:44.000 So while priority PAHs. And those shown here, which include these 16 unsubstituted parent PAHs, are the ones that we continue to know the most about. 00:44:44.000 --> 00:44:49.000 Um, in terms of their toxicity. Their potential for human exposure. 00:44:49.000 --> 00:45:00.000 And frequency of occurrence at hazardous waste sites. Um, there are more than 1,500 chemicals total in the broader class of polycyclic aromatic compounds. 00:45:00.000 --> 00:45:07.000 That include pHs which with. diverse structural features, um, and substitutions. 00:45:07.000 --> 00:45:23.000 that we currently have very little data on in terms of sources, exposure, toxicity, and mechanisms. So… Overall, our program focuses on these novel pHs, and particularly these highly abundant alkylated PAHs, which are also found. 00:45:23.000 --> 00:45:34.000 As part of environmentally relevant mixtures. We have tested a number of these 3-5 ring alkylated, um. 00:45:34.000 --> 00:45:40.000 pH families for toxicity and primary human bronchioepithelial cells. 00:45:40.000 --> 00:45:49.000 And found that many of the alkylated pHs. Cause toxicity at lower concentrations than their unsubstituted parents. 00:45:49.000 --> 00:46:01.000 Um, or unsubstituted forms across multiple endpoints, uh, so including cytotoxicity, oxidative stress, mitochondrial membrane potential, um, and also including DNA damage. 00:46:01.000 --> 00:46:09.000 So, in this table, the asterisks, um, indicate that the alkylated pH, um, listed. 00:46:09.000 --> 00:46:17.000 cause a statistically significant response at a concentration lower than its corresponding parent. So all of the parents are highlighted in boxes, and they're. 00:46:17.000 --> 00:46:34.000 Responding alkylated forms are listed below. So while we're still in the process of… of analyzing this data, we can preliminarily say that we are observing that alkylated pHs may cause toxicity through unique mechanisms compared to their parent. 00:46:34.000 --> 00:46:41.000 Potentially also causing genotoxicity, um, uniquely compared to parent forms. 00:46:41.000 --> 00:46:49.000 And collectively, this data really suggests that current risk assessments may underestimate pH risk from inhalation exposure. 00:46:49.000 --> 00:46:58.000 due to lack of data and lack of our knowledge about contribution to toxicity from these alkylated forms. 00:46:58.000 --> 00:47:04.000 So, in order to study toxicity from inhaled pHs, our lab primarily works with. 00:47:04.000 --> 00:47:15.000 3D airway organotypic culture model. So, this is a cell system in which primary human cells are differentiated at the air-liquid interface, or ALI. 00:47:15.000 --> 00:47:22.000 Um, primary human bronchial epithelial cells are collected from donors and then cultured on a transmembrane insert. 00:47:22.000 --> 00:47:34.000 And, um, in this case, under conditions in which they're allowed to differentiate for approximately 25 days, uh, to induce a three-dimensional structure with these multiple cell types. 00:47:34.000 --> 00:47:38.000 Um, that mimic the lining of the bronchial epithelium. 00:47:38.000 --> 00:47:44.000 So in our model, the cells differentiate into the pseudostratified structure, um, with ciliated cells. 00:47:44.000 --> 00:47:51.000 Um, mucus producing goblet cells, producing stem-like basal cells, among others, um, to. 00:47:51.000 --> 00:47:58.000 Allow us to more accurately reflect. Um, exposure at tissue interfaces and better recapitulate. 00:47:58.000 --> 00:48:22.000 In vivo response. Towards the end of my talk today, I will spend some time highlighting how we are continuing to work, develop, and modify our ALI models to study disease phenotypes, um, also generate co-culture models to better reflect in vivo systems. 00:48:22.000 --> 00:48:32.000 Overall, and Steven referenced this very well in his talk, too, these types of advanced cell culture models and organotypic models. 00:48:32.000 --> 00:48:37.000 You know, which can range from these differentiated cell models, um, all the way to tissue models. 00:48:37.000 --> 00:48:42.000 are really an important part of this translational pipeline in toxicology. 00:48:42.000 --> 00:48:50.000 Helping us translate from Encylico and simple in vitro testing to what we know about adverse outcomes in humans. 00:48:50.000 --> 00:48:56.000 And while they can potentially, um… serve as a replacement for animal models. 00:48:56.000 --> 00:49:02.000 Um, and also serve to supplement our knowledge. Uh, from animals and human relevant systems. 00:49:02.000 --> 00:49:16.000 These complex models also allowed for translation to humans and improved understanding of key events leading to adverse outcomes for better predictions of toxicity. 00:49:16.000 --> 00:49:23.000 Um, as I mentioned at the beginning, early on, one of our goals is to use our 3D respiratory model to establish. 00:49:23.000 --> 00:49:36.000 Quantitative relationship between chemical exposure. and toxicity and make comparisons between our cell model and tissues from an in vivo system. So, in collaboration with Jordan Smith. 00:49:36.000 --> 00:49:41.000 In the predictive metabolism and dosimetry core at the Pacific Northwest National Lab. 00:49:41.000 --> 00:49:47.000 We've developed a dosimetry, or physiologically based pharmacokinetic model. 00:49:47.000 --> 00:49:52.000 to improve our ability to translate between endpoints that we measure in vitro. 00:49:52.000 --> 00:49:56.000 Um, to toxicity or health outcomes that can occur in humans. 00:49:56.000 --> 00:50:04.000 So, historically, PBPK models have been utilized in what's called a forward dosimetry approach. 00:50:04.000 --> 00:50:08.000 Uh, to predict tissue concentrations from human exposure doses. 00:50:08.000 --> 00:50:15.000 for use in extrapolating to animal studies. But in our studies, we want to use the PVPK model. 00:50:15.000 --> 00:50:22.000 Uh, to develop a reverse dosimetry. Model for predicting human exposure concentrations. 00:50:22.000 --> 00:50:29.000 from in vitro data, with the idea that concentrations that cause toxicity in cells. 00:50:29.000 --> 00:50:35.000 would then mimic concentrations at the tissue level. 00:50:35.000 --> 00:50:46.000 So to do this, uh, we need to collect data in our model on the disposition and metabolism of pHs within our culture system. 00:50:46.000 --> 00:50:51.000 And so I was going to spend some time walking through, um, our efforts to do this. 00:50:51.000 --> 00:50:59.000 Uh, we had broken our model down into compartments, um, with the goal to quantify metabolism and distribution. 00:50:59.000 --> 00:51:04.000 Um, across dose and time, and develop parameters for the dosimetry model. 00:51:04.000 --> 00:51:11.000 So our fractions reflect the mucous apical compartment, the cells, the media or plasma compartment. 00:51:11.000 --> 00:51:20.000 And then we also wanted to evaluate potential loss of pHs that may bind to plastic material. 00:51:20.000 --> 00:51:30.000 We developed the model, um, using the pH benzoate, um, just as a reference chemical, also because we do have a lot of information about benzoepyrine and. 00:51:30.000 --> 00:51:40.000 and knowledge about its metabolites and standards for those metabolites, so… Um, we use benzoepyrene to develop. 00:51:40.000 --> 00:51:45.000 parameters, um… To collect parameters, actually, for developing essentially two models. 00:51:45.000 --> 00:51:51.000 A model to describe the parent chemical, um, benzoid pyrine itself, and how it moves within. 00:51:51.000 --> 00:52:00.000 our system, and then also develop a model describing formation of metabolites, particularly those that contribute to the formation. 00:52:00.000 --> 00:52:14.000 of the dial epoxide leading to DNA damage. So before, uh, looking at the metabolite data, we first tracked how benzoypirine moved in our culture system between compartments. 00:52:14.000 --> 00:52:22.000 Um, across multiple time points, up to 48 hours after exposure. Uh, we were able to recover. 00:52:22.000 --> 00:52:28.000 Um, approximately 100% of benzoapyrine that was put into the system. 00:52:28.000 --> 00:52:36.000 And we see the amount of benzoipyrene then decrease over time, as we would expect metabolites to form. 00:52:36.000 --> 00:52:43.000 So, uh, when we looked at. benzoyapyrine in each compartment, we can see that benzoipyrene, um. 00:52:43.000 --> 00:52:55.000 We'll leave the mucous apical compartment over time. Um, we can see it moving into the cells, um, up to 4 hours before, um, also decreasing. 00:52:55.000 --> 00:53:12.000 And we see some benzoe pyrene moving into the media, but if you look at the y-axis, um, you can tell that this is much lower… at much lower levels than where, um, measuring benzoid pyrine in the other compartments. So actually very little of the parent benzoepyrine itself is moving into the. 00:53:12.000 --> 00:53:30.000 media fraction. We also really, um, were not able to, um… Identify benzoipyrine bound to plastic and so, um, this did not serve as a significant source of loss for benzoypine from our system. 00:53:30.000 --> 00:53:35.000 So to measure metabolite formation, we also collected samples. 00:53:35.000 --> 00:53:43.000 Uh, multiple time points up to 48 hours for extraction and analysis by UPLC with fluorescence detection. 00:53:43.000 --> 00:53:54.000 So this is a schematic for benzoypirine metabolism. We see benzoipyrene here in the center, and different metabolites that can form, um, through. 00:53:54.000 --> 00:54:01.000 Uh, various pathways. So while we aren't able to quantify all of the metabolites. 00:54:01.000 --> 00:54:09.000 We can quantify many of them that we, um, have standards for, and so those are listed here. 00:54:09.000 --> 00:54:17.000 And out of those metabolites, we were primarily able to detect these 6 metabolites that are, um. 00:54:17.000 --> 00:54:22.000 highlighted in red, and we can see from the UPLC trace, um. 00:54:22.000 --> 00:54:35.000 the peaks for these metabolites. Uh, so this includes a number of metabolites, but particularly, as I mentioned, we were interested in those associated with the formation of. 00:54:35.000 --> 00:54:45.000 Um, the dial epoxide, benzoepyrine 7-8, diol 910 epoxide, um, that can interact with DNA and cause DNA damage. 00:54:45.000 --> 00:54:55.000 So, uh, we were able… while we can't… Uh, measure the dial epoxide itself. Due to stability, we can measure, um, metabolites upstream. 00:54:55.000 --> 00:55:03.000 Um, so we can measure the 7-8. Diol, and also the tetraols downstream. 00:55:03.000 --> 00:55:17.000 So after we look at formation of metabolites. Over time, we can see, um… formation of a number of metabolites, and I'm just going to highlight an increase in the 7-8 diol in blue. 00:55:17.000 --> 00:55:23.000 Um, and an increase in the tetrals in red, um, that are formed over time. 00:55:23.000 --> 00:55:29.000 Um, and again, these are upstream and downstream of the dial epoxide. 00:55:29.000 --> 00:55:34.000 We also see that, um, metabolites account for the majority. 00:55:34.000 --> 00:55:41.000 Of benzo a pyrene that's recovered at later time points. So, we start to, um. 00:55:41.000 --> 00:55:50.000 Really see measurable… uh, metabolites, um, contributing to total recovery at 8 hours, and then this contributes to a majority. 00:55:50.000 --> 00:55:55.000 of the metabolites that are formed. Um, at the later time points. 00:55:55.000 --> 00:56:00.000 So while, uh… detection of these metabolites. 00:56:00.000 --> 00:56:05.000 indicate that we are likely generating the dial epoxide itself, leading to DNA damage. 00:56:05.000 --> 00:56:10.000 Uh, we did also evaluate DNA damage in cells. 00:56:10.000 --> 00:56:16.000 After treatment, um, using a comment-based assay, um, in collaboration. 00:56:16.000 --> 00:56:24.000 with Bevin Englewart's lab at MIT. Uh, this assay detects single-strand breaks mediated by genotoxic intermediates. 00:56:24.000 --> 00:56:37.000 generated through metabolism of pHs. Um, this method uses trapping agents to block repair mechanisms, so damaged DNA can be viewed as comets, and you can see examples here of. 00:56:37.000 --> 00:56:45.000 undamaged DNA and DNA after treatment. So, using this assay, we were able to detect. 00:56:45.000 --> 00:56:55.000 Um, a dose-dependent increase in the DNA damage with benzoipyrene in the presence of trapping agents, which are here on the right-hand side. 00:56:55.000 --> 00:57:04.000 But not when, uh. trapping agents were not included in our negative controls, um, where we aren't able to. 00:57:04.000 --> 00:57:15.000 Um, observe significant DNA damage after benzoypirine treatment. So this exit, um, suggests that the damage that we were seeing, uh, was mediated by generation of these reactive. 00:57:15.000 --> 00:57:29.000 intermediates, um, of pH metabolism. Um, so we were… we were observing DNA damage at 24 hours. Um, this correlates, uh, with formation of the Phase I metabolites. 00:57:29.000 --> 00:57:33.000 Um, which are quantified in bulk here in blue. 00:57:33.000 --> 00:57:42.000 Um, and we start to see, again, significant, um, formation of metabolites in 8 hours that peak at 24 hours. 00:57:42.000 --> 00:57:49.000 This also correlates with induction of CYP1A1 enzyme activity in our cells, which is the black line. 00:57:49.000 --> 00:57:53.000 Um, which we see peak at approximately 6 to 8 hours. 00:57:53.000 --> 00:58:03.000 Um, after treatment. And CYP11 is one of the primary enzymes responsible for pH metabolism in lung cells. 00:58:03.000 --> 00:58:10.000 I will also mention that, um, we did see evidence of Phase II metabolism by glucuronides in our cells. 00:58:10.000 --> 00:58:16.000 Uh, so if we incubate our samples with glucuronidase to inhibit metabolism. 00:58:16.000 --> 00:58:23.000 We saw approximately 50% increase in measurable phase 1 metabolites at the 24 to 48. 00:58:23.000 --> 00:58:32.000 our time points, um, improving our overall recovery. Of, um, total BAP added to the system at these time points. 00:58:32.000 --> 00:58:42.000 We are continuing to, uh, still investigate, uh, where the rest of the total mass may be at these time points. 00:58:42.000 --> 00:58:53.000 just based on total recovery of what's put into the system, but it is likely… Uh, much of it is likely due to a combination of, um, Phase I metabolite peaks. 00:58:53.000 --> 00:59:04.000 Um, that we haven't identified yet. Uh, or other Phase 2 metabolite processes besides glucuronidation. 00:59:04.000 --> 00:59:10.000 So we use this data to, um, first build a diff symmetry model for the parent chemical benzoapyrine. 00:59:10.000 --> 00:59:17.000 Um, by calculating parameters for absorption, permeation through cells, partitioning into media. 00:59:17.000 --> 00:59:33.000 Elimination processes of metabolism and volatilization. Um, when we overlay our predicted or simulated model, which are the, um, represented by the lines onto our actual measurements, um, that were collected from ourselves. 00:59:33.000 --> 00:59:38.000 Which are the dots. We do see that we have a very good fit. 00:59:38.000 --> 00:59:45.000 of the model to our data. So, in red, we see prediction of, um, benzoe pyrene in the media. 00:59:45.000 --> 00:59:52.000 although is, um… we presented very little of the parent itself is, um, moving into the media. 00:59:52.000 --> 00:59:58.000 In Black, uh, we see, um, prediction of benzoipyrine in cells. 00:59:58.000 --> 01:00:02.000 And in blue, we see prediction of metabolite formation. 01:00:02.000 --> 01:00:08.000 And so, as I mentioned, while we can't account for all of the potential metabolism that's occurring. 01:00:08.000 --> 01:00:14.000 Um, our model accounts for at least the minimum amount of metabolism achieved. 01:00:14.000 --> 01:00:26.000 And so, um, and I'll note that. Um, well, we start to see significant metabolites at 8 hours, metabolites do begin to appear at approximately 4 hours. 01:00:26.000 --> 01:00:40.000 So we can incorporate, um, metabolite formation into the model using a very similar, um, using similar parameters that we did for the parent itself. So, calculating parameters for formation of the metabolite, permeation through cells. 01:00:40.000 --> 01:00:48.000 partitioning into mucus, as well as active transport into other, um, compartments, and elimination of. 01:00:48.000 --> 01:00:53.000 Um, metabolites. my metabolism and volatilization. 01:00:53.000 --> 01:01:01.000 So, combined, uh, this provides, uh. pharmacokinetic model for benzoipyrene in our cell system. 01:01:01.000 --> 01:01:13.000 Accounting for how cells move throughout the system and how they are quantified in a… how they are metabolized in a quantitative manner. 01:01:13.000 --> 01:01:23.000 In addition to developing a dosimetry model, um. This effort also helped us to establish a better understanding of the metabolic. 01:01:23.000 --> 01:01:28.000 competency of, um… our ALI model. 01:01:28.000 --> 01:01:37.000 And… This can help us improve our ability to predict toxicity down the line. 01:01:37.000 --> 01:01:42.000 So we're currently working to expand our assessment of, um. 01:01:42.000 --> 01:01:48.000 dosimetry to other PAHs besides benzoapyrine, so that we can start to make. 01:01:48.000 --> 01:01:57.000 predictions for movement and metabolism. of pHs with different chemistries, um… in our 3D model, and see if we can. 01:01:57.000 --> 01:02:05.000 predict based on chemistry, um… what might be happening to chemicals within our culture system. 01:02:05.000 --> 01:02:15.000 Ultimately, uh, what we would like to do with this data is apply it for in vitro to in vivo extrapolation so that we can translate toxicity, um. 01:02:15.000 --> 01:02:29.000 that we're observing in our in vitro systems. to in vivo systems, and then… associate concentrations that cause toxicity in vitro with those that are being measured as part of our broader program through passive sampling. 01:02:29.000 --> 01:02:38.000 Um, and concentrations that we know individuals are exposed to in the environment. 01:02:38.000 --> 01:02:45.000 Um, as I mentioned earlier, we're also modifying our, um, ALI model to just. 01:02:45.000 --> 01:02:51.000 to study disease phenotypes and generate co-culture, so I just wanted to spend a few minutes, um. 01:02:51.000 --> 01:03:05.000 talking about some of these efforts fairly briefly, so… Um, we can induce disease phenotypes during differentiation, and we're currently working with an asthmatic phenotype where cells are differentiated in the. 01:03:05.000 --> 01:03:10.000 The presence of interleukin-13 over 14 days to generate a phenotype. 01:03:10.000 --> 01:03:18.000 We're the pathophysiology is similar to what's reported in individuals with type 2 allergic asthma, and so this is. 01:03:18.000 --> 01:03:30.000 been evaluated by us and by others in the literature, um… with the phenotype exhibiting mucociliary dysfunction, airway remodeling, mucous hypersecretion, and. 01:03:30.000 --> 01:03:38.000 And also loss of barrier integrity. So we're interested to understand these cumulative risks to public health from combined stressors. 01:03:38.000 --> 01:03:46.000 Um, recent animal models have identified pulmonary inflammation from respiratory disease, such as asthma as a possible modifier, and. 01:03:46.000 --> 01:03:54.000 And, um, risk factor for chemical toxicity in the lung after exposure to inhaled pollutants. 01:03:54.000 --> 01:04:04.000 Um, overall, asthma affects. 8% of the US population and more than 260 million people worldwide. Um, in our state, Oregon ranks 5th. 01:04:04.000 --> 01:04:08.000 Um, out of all states in asthma incidents overall. 01:04:08.000 --> 01:04:21.000 So rates of asthma in our state, um. do exceed the national average, um, range between 9 and 13%, and it's particularly relevant for communities adjacent to the Portland Harbor Superfund site. 01:04:21.000 --> 01:04:35.000 Which is highlighted here in gray. Um, so it makes them a unique and vulnerable population for subsequent chemical insults after respiratory exposure to inhaled chemicals from the Superfund site. 01:04:35.000 --> 01:04:41.000 So we have characterized our model in the presence, um, we've characterized the phenotype itself. 01:04:41.000 --> 01:04:50.000 Um, and can measure significant reduction in barrier integrity using transepithelial electrical resistance. 01:04:50.000 --> 01:04:58.000 We can also see increased detection of mucus and mucus-producing goblet cells in the phenotype itself. 01:04:58.000 --> 01:05:03.000 So, barrier integrity, uh. of the airway epithelium remains. 01:05:03.000 --> 01:05:23.000 Um, maintains homeostasis, uh, protects against environmental factors. Barrier dysfunction itself can increase risk of chemical uptake and also enhancement of chronic inflammation, so… Um, we have evaluated our, um, disease phenotype model in the presence of PAHs and. 01:05:23.000 --> 01:05:34.000 particular here, I'm showing some data with a mixture that contains primarily alkylated PAHs that was collected from air sampling at a creosote site during a wildfire event. 01:05:34.000 --> 01:05:44.000 Um, in our region. And, um, when we expose both the asthmatic phenotype and hash spars in normal cells and solid, um. 01:05:44.000 --> 01:05:50.000 to this mixture of pHs. We really only see, um, further reduction of barrier integrity. 01:05:50.000 --> 01:05:57.000 In the asthmatic phenotype, um. compared to normal cells, suggesting that. 01:05:57.000 --> 01:06:03.000 These individuals really may be more susceptible to chemical insult when compared to healthy individuals. 01:06:03.000 --> 01:06:15.000 Um, so we have, um, evaluated this through a number of ways, and using some global approaches, we… Um, know that as the asthmatic phenotype and lung cells in the asthmatic phenotype. 01:06:15.000 --> 01:06:25.000 Um, respond to PAHs, um. suggesting that they are… Causing toxicity through unique mechanisms in the disease model. 01:06:25.000 --> 01:06:33.000 Uh, we're continuing to explore how individuals with inflammation-based disease may respond differently to pH toxicity. 01:06:33.000 --> 01:06:42.000 And we're also doing this by utilizing a macrophage epithelial co-culture model to effectively evaluate immune contribution. 01:06:42.000 --> 01:06:53.000 Um, so… So one thing that we are exploring is, um, increasing the complexity of our culture system to better mimic in vivo studies. 01:06:53.000 --> 01:06:58.000 And, um, use this to help us evaluate some of these, um. 01:06:58.000 --> 01:07:02.000 highly abundant chemicals that we're able to measure, um. 01:07:02.000 --> 01:07:06.000 Associated with Superfund site, but also in wildfire smoke. 01:07:06.000 --> 01:07:10.000 Uh, we know that induction of airway inflammation is orchestrated. 01:07:10.000 --> 01:07:19.000 Uh, through signaling between airway epithelial cells. And macrophages, and so, um, the co-culture model incorporates, um. 01:07:19.000 --> 01:07:27.000 Macrophage cells into our 3D model, um, just as another way to expand the use of these 3D models to evaluate. 01:07:27.000 --> 01:07:32.000 the combined contribution of airway epithelial and macrophage cell communication. 01:07:32.000 --> 01:07:42.000 In terms of determining toxicity. So, in conclusion, I just, uh, want to emphasize the importance of utilizing. 01:07:42.000 --> 01:07:56.000 integrated approaches for understanding chemical toxicity. Uh, this has been important across our entire program and allows us to prioritize environmentally relevant chemicals and exposures for testing and filling critical data gaps for these. 01:07:56.000 --> 01:08:06.000 Understudied and undermeasured chemicals. Um, even within a class of what we would typically consider to be very well-studied polycyclic aromatic hydrocarbons. 01:08:06.000 --> 01:08:09.000 Um, there exist a lot of… a lot of data gaps. 01:08:09.000 --> 01:08:15.000 Uh, there are benefits to using cell-based NAMs combined within silico approaches to improve. 01:08:15.000 --> 01:08:24.000 Our ability to translate studies from in vitro to in vivo outcomes based on quantitative understanding of dosing and pharmacokinetics on our system. 01:08:24.000 --> 01:08:32.000 We can also use these tools to improve our understanding of susceptibility due to combined stressors like pre-existing disease. 01:08:32.000 --> 01:08:39.000 Um, that are common in target populations near relevant exposure sources, like the Portland Harbor Superfund site. 01:08:39.000 --> 01:08:47.000 Um, future studies in our, um, we are really interested in continuing to understand mechanisms of toxicity for alkyl pHs. 01:08:47.000 --> 01:09:03.000 Um, including self-specific responses, um. in our lung culture model and in our co-culture model. Um, so… As Steven mentioned, we also see less sensitivity in our 3D model. 01:09:03.000 --> 01:09:25.000 in response to pHs, like benzoypirine and others. You know, for broad endpoints like cytotoxicity, but for also specific endpoints, like DNA damage and SYP induction. And, um… We have some evidence that, um, this is likely due to differential response across the different cell types in the 3D model, and so we're continuing to explore that more. 01:09:25.000 --> 01:09:30.000 Um, as well as look at sources of variants that could contribute to inter-individual response. 01:09:30.000 --> 01:09:38.000 Um, to toxicity. So with that, I just want to thank, um, all of the. 01:09:38.000 --> 01:09:48.000 Um, past and current contributors. Uh, to all of this work, uh, within my group, and also our collaborators within the Superfund program. 01:09:48.000 --> 01:09:54.000 And outside of the Superfund program, as well as our funding sources. 01:09:54.000 --> 01:10:00.000 And, uh, with that, I thank you and can take, uh, questions. 01:10:00.000 --> 01:10:05.000 Okay, thank you for the presentation, Susan. It was excellent. 01:10:05.000 --> 01:10:10.000 Um, so we have several questions here in the question and answer box. 01:10:10.000 --> 01:10:13.000 We'll probably only be able to get through a few here. 01:10:13.000 --> 01:10:19.000 Um, versus, can you please comment on the confidence in separating the mucus fraction versus media? 01:10:19.000 --> 01:10:26.000 Or cells in the ALI model. 01:10:26.000 --> 01:10:36.000 Um… so… When we are collecting, um, the mucus, we are using… we are essentially. 01:10:36.000 --> 01:10:45.000 We're regularly washing the cells, um. generally weekly during differentiation, and so we use a similar process of. 01:10:45.000 --> 01:10:51.000 Um, washing the cells in terms of collecting the mucus fraction. 01:10:51.000 --> 01:11:00.000 Um, compared to collecting our cellular fraction. So there is possibly, you know, some overlap in that. 01:11:00.000 --> 01:11:04.000 in that collection process. 01:11:04.000 --> 01:11:10.000 Okay, so next question for benzoepyrine. Well, you didn't recover much from plastic. 01:11:10.000 --> 01:11:18.000 Did the amount measured in other compartments represent the most of the initial exposure added to the system? 01:11:18.000 --> 01:11:30.000 Yes, we were able to, um… we did go through and look at a mass balance, and we were able to account for essentially 100% of the mass that was put into the system, and. 01:11:30.000 --> 01:11:41.000 It's when we get to the later time points where we know we're having metabolism, and we're accounting for some of those metabolites with, um, the standards that we have. 01:11:41.000 --> 01:11:58.000 Um, and there are, um… those later time points where we still are working to identify additional metabolites that might be contributing to that total mass, but certainly for the majority of the time course. 01:11:58.000 --> 01:12:09.000 We can account for, uh, definitely benzoate pyrine, and even benzo pyrine in combination with metabolites, um, is a majority of what's been added to the system. 01:12:09.000 --> 01:12:22.000 Okay, so we have time for one more question. The PK modeling in vitro is a major advance. Thank you. Can you comment on how this model and data is helping to bridge in vitro and in vivo situations? 01:12:22.000 --> 01:12:29.000 In other words, what is the benefit of this very involved work in human risk assessment? 01:12:29.000 --> 01:12:37.000 So, um, one thing that we really want to be able to do is start to make predictions in our system about. 01:12:37.000 --> 01:12:48.000 Um, chemicals with different chemistries. how we can predict how they move within our system, um, how they're metabolized, and at what rates, um. 01:12:48.000 --> 01:12:57.000 at what rates they're metabolized, how that is linked to, um, specific endpoints in our lung cell system. 01:12:57.000 --> 01:13:06.000 So that we can start to identify threshold concentrations that cause toxicity, and then make very specific, more quantitative comparisons to. 01:13:06.000 --> 01:13:16.000 how the concentration in vitro, um, is then. Um, link to potential exposure concentrations where. 01:13:16.000 --> 01:13:28.000 We know we have, um… Our collaborators are in the field with passive samplers, um, measuring chemicals in the environment. We know what these exposure concentrations are, and so we'd like to be able. 01:13:28.000 --> 01:13:34.000 To start to prioritize some of our studies, um, based on both what we're seeing in terms of toxicity. 01:13:34.000 --> 01:13:47.000 But also, combined with, um, the exposure concentrations for what we know are, um, most abundant and most important in these mixtures. 01:13:47.000 --> 01:13:50.000 Well, thank you very much, Susan. We appreciate it. 01:13:50.000 --> 01:14:02.000 So, our final speaker today is Aram Hahn, a professor in the Department of Electrical and Computer Engineering and Biomedical Engineering at Texas A&M University. 01:14:02.000 --> 01:14:06.000 He is also faculty with the Texas A&M Health Science Center. 01:14:06.000 --> 01:14:13.000 and the Texas A&M Institute for Neuroscience. His research interests are in solving grand challenge problems. 01:14:13.000 --> 01:14:19.000 and the broad areas of health and energy through the use of micro and nanosystem technologies. 01:14:19.000 --> 01:14:24.000 Dr. Han will discuss the fetal-maternal interface tissue chip. 01:14:24.000 --> 01:14:28.000 to study the effect of environmental substances on preterm birth. 01:14:28.000 --> 01:14:32.000 So, Dr. Hahn, I'll turn it over to you. 01:14:32.000 --> 01:14:41.000 Thank you. Um… I guess for some reason, I don't see the, uh, share screen button, so I guess somebody from the NIHSI. 01:14:41.000 --> 01:14:49.000 Could maybe just a share. the presentation. 01:14:49.000 --> 01:14:54.000 Thank you, appreciate it. So, uh, my name is Arum Khan, I'm a faculty at Texas A&M University. 01:14:54.000 --> 01:15:06.000 I'm going to talk about organoid chip models of the fetal motor interface, enabling rapid hazard analysis of environmental contaminants impacting pregnancy. Next slide. 01:15:06.000 --> 01:15:10.000 So… I've been collaborating with colleagues. 01:15:10.000 --> 01:15:16.000 Uh, we're very much interest in looking into preterm birth and the adverse pregnancy outcome. 01:15:16.000 --> 01:15:25.000 Um, this just shows the global preterm birth rate, where unfortunately, we have 150 million babies born prematurely in the last decade. 01:15:25.000 --> 01:15:31.000 And our worldwide preterm birth rate, about 11% of all of the pregnancy. 01:15:31.000 --> 01:15:44.000 Next. And, uh, so there's a strong need to understand the mechanism, identify high-risk pregnant subjects, and provide proper intervention, which is very much lacking. Next. 01:15:44.000 --> 01:15:53.000 So, uh, if you look at the preterm birth, it has a fairly complex etiology. Some of the preterm birth is caused by infection-associated inflammation. 01:15:53.000 --> 01:15:59.000 relatively easily identifiable, and that's about 40% of the preterm birth, but the rest of them. 01:15:59.000 --> 01:16:04.000 is due to variety different risk factor induced by non-infectious inflammation. 01:16:04.000 --> 01:16:08.000 The cause of this non-infectious inflammation could vary a lot. 01:16:08.000 --> 01:16:12.000 And that really, uh, makes it very difficult to study this one. 01:16:12.000 --> 01:16:17.000 Where this inflammation compromises… once it compromises the fetal motor interface. 01:16:17.000 --> 01:16:30.000 that could trigger the pretembrance pathway. Next. So, we… our team, in collaboration, uh, all of this work has been in collaboration with Dr. Ramku Menon. 01:16:30.000 --> 01:16:34.000 A professor in obstetric and gynecology, University of Texas Medical Branch in Galveston. 01:16:34.000 --> 01:16:44.000 And together in the past 6-7 years, together with Dr. Menon, we have been working on developing a variety of different NPS models of the female reproductive tract. 01:16:44.000 --> 01:16:49.000 We initially, our initial work was started out with the M9 membrane organ chip. 01:16:49.000 --> 01:16:55.000 In advance, it had some technical advancement, built into fetal material fetal membrane over a chair. 01:16:55.000 --> 01:17:03.000 And then, uh, we developed a variety of different ownership with the goal of really understanding the entirety of the female reproductive tech. 01:17:03.000 --> 01:17:11.000 track, uh, both from mechanistic understanding as well as a variety of different factors that may impact female reproduction. 01:17:11.000 --> 01:17:24.000 Next. So, um, amongst all variety of different organ systems, uh, today I would like to really focus on the fetal maternal interface and the complex nature of this interface. 01:17:24.000 --> 01:17:30.000 If you look at the fetal motor interface, in other words, the entire interface that surrounds the fetus. 01:17:30.000 --> 01:17:37.000 There are two different interfaces. One is the placental interface, which is probably most well studied. 01:17:37.000 --> 01:17:41.000 And the other one is the fetal fetal membrane interface. 01:17:41.000 --> 01:17:47.000 And each of these interfaces are composed of multiple complex layers. Next. 01:17:47.000 --> 01:17:57.000 So, if you look at the… challenge of current models to study the fetal-maternal interface, there are many challenges that are very similar to any other organ system. 01:17:57.000 --> 01:18:02.000 Where 2D culture explains human cell cultures have a variety of different limitations. 01:18:02.000 --> 01:18:08.000 animal models such as mouse models also have a variety of different limitations. 01:18:08.000 --> 01:18:18.000 However, for pregnancy-specific animal model poses additional challenge because the pregnancy, for example, the pregnancy mouse is very different from pregnancy in human. 01:18:18.000 --> 01:18:22.000 So, a much bigger difference to some degree compared to other organ systems. 01:18:22.000 --> 01:18:33.000 And, uh, non-human primate, probably the most, uh, mimicking in terms of human pregnancy, but doing studies using non-human primate really, uh, in terms of cost, uh, in terms of cost. 01:18:33.000 --> 01:18:42.000 is really not a routinely usable tool. Explained perfusion studies at term placenta have been utilized, but there are also somewhat limited. 01:18:42.000 --> 01:18:52.000 Next. So, we've been working on, uh, developing a chip system for the fetal membrane, specifically. 01:18:52.000 --> 01:18:59.000 A lot of initial work was enabled by two, uh, funding, one from, from an NICHD. 01:18:59.000 --> 01:19:06.000 are one that studies intercellular injection that defines cell migration and transition, maintaining field membrane homeostasis. 01:19:06.000 --> 01:19:14.000 was really using, developing a, uh. and PS system of a fetal membrane and utilize this one to study a variety of different mechanisms. 01:19:14.000 --> 01:19:19.000 The other one is the clinical trial part of the NCAS clinical trial and CHIP. 01:19:19.000 --> 01:19:24.000 Where we are one of the 10 teams that, uh, received funding. 01:19:24.000 --> 01:19:34.000 It was a joint NCIS NICHD funding to look to develop a maternal-fetal interphase on a chip and use it to generate preclinical data. 01:19:34.000 --> 01:19:45.000 Next. So, if I'm focusing on the fetal, uh, fetal maternal interface, as I mentioned, we are looking into two interfaces. 01:19:45.000 --> 01:19:52.000 One is the fetal membrane, the other one is the placental interface, and each of them are composed of multiple different cellular layer. 01:19:52.000 --> 01:20:04.000 Next. And so, uh, we have developed a field member on a chip. We focus on 4 different cell types, uh, one maternal cell type, three different fetal cell type. 01:20:04.000 --> 01:20:10.000 I'm gonna go into this particular design configuration a little bit more in detail in a later slide. 01:20:10.000 --> 01:20:15.000 But basically, it's a full concentric ring structure of silica compartment. 01:20:15.000 --> 01:20:21.000 connected by Aria for Microfluid channel, and so basically a 4-cell type cochlear system. 01:20:21.000 --> 01:20:35.000 Next, the placental, uh, placenta interface organ chair. focusing on 3 different cell types, and it is a rectangular-shaped co-culture chamber, organically connected by OSF of a microfluid channel. 01:20:35.000 --> 01:20:45.000 So these two have been our main ownership of the fetal maternal interface that we have developed and have been utilizing for a variety of different studies. 01:20:45.000 --> 01:20:51.000 Next. So if you look at the basic key features of our particular MPS model. 01:20:51.000 --> 01:20:58.000 And there are typically a vertical co-culture model, and there are horizontal co-culture model. 01:20:58.000 --> 01:21:03.000 In our case, the basic configuration is basically a horizontal co-code model. 01:21:03.000 --> 01:21:08.000 We have two, in this case, in this example, we have two different cell culture compartments. 01:21:08.000 --> 01:21:14.000 connected by RAF microfluid channel. So these smoker fluke channels are small enough that prevents. 01:21:14.000 --> 01:21:19.000 Passive, uh, movement of cells from one compartment to another compartment. 01:21:19.000 --> 01:21:27.000 In other words, when you load the cell into cell chamber A, they don't simply flow into cell chamber B, so that we can keep them separate. 01:21:27.000 --> 01:21:35.000 However, they are large enough to allow any biochemicals to transition or diffuse from one chamber to the other one. 01:21:35.000 --> 01:21:41.000 Uh, so these are a variety of different cell-produced factors as well as the toxicant that we would like to apply. 01:21:41.000 --> 01:21:49.000 Um, so this is a two-chamber example, and these channels, depending on what particular part of the fetal membrane we would like to mimic. 01:21:49.000 --> 01:21:59.000 These channels can be actually filled with extracellular matrix, so on the lower right side, you will see a scanning electron microscope image of a microfluid channel. 01:21:59.000 --> 01:22:03.000 filled with collagen in this case, so we have already of this channel. 01:22:03.000 --> 01:22:10.000 And the main reason, one of the main reasons we use this horizontal co-culture configuration is it allows imaging very easy. 01:22:10.000 --> 01:22:14.000 And we can simply expand the number of different cocaine chamber. 01:22:14.000 --> 01:22:22.000 Versus the vertical co-culture configuration usually limited to a, uh, two-chain… two-cell type co-culture. 01:22:22.000 --> 01:22:35.000 Next, uh, the second key feature is we have, um… sorry, so we have a localized cell loading, cell migration, biochemical diffusion, all of these key functions enabled by this device. 01:22:35.000 --> 01:22:42.000 The next slide shows the, uh, the second feature where we have a media reservoir sitting on top of their main culture chamber. 01:22:42.000 --> 01:22:46.000 So basically, we have a plastics sitting on top of this, they are bonded together. 01:22:46.000 --> 01:22:54.000 And if you look at the next slide, we will… will it basically have a gradient-driven media diffusion. 01:22:54.000 --> 01:23:00.000 And that enables localized media supply, localized effluent collection, and as well as localized treatment. 01:23:00.000 --> 01:23:08.000 So, whatever factors that are produced by cells, they will be diffused into media compartment, so that we can sample those. 01:23:08.000 --> 01:23:13.000 When we are applying a tox scan, which I'm going to show later on, we can apply to one chamber. 01:23:13.000 --> 01:23:20.000 Uh, and then that Toxic gun can, depending on the situation, that Tox gun can diffuse from one chamber to another chamber. 01:23:20.000 --> 01:23:25.000 So we can use this media reservo block to do, uh, to enable this one. 01:23:25.000 --> 01:23:34.000 So here, uh, roughly the size of this initial device fits in a well over a 6-wheel chamber. Next. 01:23:34.000 --> 01:23:43.000 So, in terms of cell source, um, cell source are one of the most challenging parts of any NPS devices. In many cases, people are using IPSC. 01:23:43.000 --> 01:23:49.000 In the case of a MPS mouse of the phylumatin interface, there are really no IPAs available. 01:23:49.000 --> 01:23:58.000 So, what we do is we, uh, initially, about 6-7 years ago, when we initially started this project, we used primary cells harvested from. 01:23:58.000 --> 01:24:03.000 Uh, just got it, uh, placental tissue from a thrombus. 01:24:03.000 --> 01:24:12.000 The past several years, my colleagues, Dr. Menon's group has now made this immortalized this one and developed this one into a cell line. 01:24:12.000 --> 01:24:19.000 have done a variety of different testing to show that these cell lines are very robust, so now we have a robust supply of a different. 01:24:19.000 --> 01:24:29.000 cell lines derived from primary cells. And, uh, which we are currently using in all of our, uh, different, uh, MPS devices of the female reproductive system. 01:24:29.000 --> 01:24:44.000 Next. So, um, several years ago, um, I started looking into where, beyond, let's say, a mechanistic study, beyond drug discovery. 01:24:44.000 --> 01:24:52.000 Where we can utilize it. And so, in the past several years, I really had a great pleasure of working as part of the Texas A&M Super Fund Research Center. 01:24:52.000 --> 01:25:00.000 This is 2022 is where the renewal started. And so I participated as a Project 3. 01:25:00.000 --> 01:25:11.000 Um, so first of all, the overall goal of the A&M Super Fundraising Center is comprehensive tools and models for addressing exposure to mixture during environmental emergency-related contaminant event. 01:25:11.000 --> 01:25:19.000 So, two things. One is focus on mixture. Second one is focusing on emergency, environmental emergency-related contaminant event. 01:25:19.000 --> 01:25:26.000 So, within that context, we looked into the pregnant woman as one of the vulnerable population. 01:25:26.000 --> 01:25:33.000 And Product C is trying to address that, and here you see a center overview where we have 5 different projects and many different core. 01:25:33.000 --> 01:25:40.000 So the project tree that really utilizes the freedom interface, or on a chip system. 01:25:40.000 --> 01:25:46.000 working very closely with many of the center component to look into the potential hazard of. 01:25:46.000 --> 01:25:58.000 individual chemical and mixture. Uh, especially coming from environmental emergency-related contaminant, and how that may impact the pregnancy and potentially associated with pretembrance. 01:25:58.000 --> 01:26:11.000 Next. So, it's the very first study at the early stage of this project, we have, uh, we have used a very well-known, uh, contaminant, uh, cadmium. 01:26:11.000 --> 01:26:14.000 There's been a bunch of etiological studies out there. 01:26:14.000 --> 01:26:22.000 So, we're using this 4-chamber fetal motor interface device. We basically apply cadmium into the maternal chamber and looked at how. 01:26:22.000 --> 01:26:32.000 cadmium transport from the maternal chamber to the 3 different layers of the fetal chamber, so we can look into propagation. 01:26:32.000 --> 01:26:38.000 And most importantly, we can look at ads academy propagates through the different cellulaya. 01:26:38.000 --> 01:26:44.000 Which you can sort of see in the illustration, what I see in the figure C here, inflammatory response. 01:26:44.000 --> 01:26:49.000 We can look at how the different inflammation and cell death responding to. 01:26:49.000 --> 01:26:56.000 to as can even propagate. So, as ketamine propagate from the lower side, which is the maternal descendior chamber. 01:26:56.000 --> 01:27:01.000 to the upper side, to a different fetal layer, we can look in how a cell does occur. 01:27:01.000 --> 01:27:08.000 how inflammation occurred, so this allows, this shows that this fetal maternal interface ownership. 01:27:08.000 --> 01:27:17.000 can assess chemical transport kinetics using this device. And then, uh, also look into toxic and exposure-mediated adverse effect at the fetal maternal interface. 01:27:17.000 --> 01:27:23.000 And so here we can see both direct result coming from a direct exposure to caladium. 01:27:23.000 --> 01:27:28.000 As well as indirect exposure coming from other factors or other inflammatory factors. 01:27:28.000 --> 01:27:32.000 That are diffusing from. one cellular layer to another layer. 01:27:32.000 --> 01:27:42.000 Next. So… If I'm looking at the microphysical system, which, uh, which sometimes people call organoid shape or tissue shape. 01:27:42.000 --> 01:27:50.000 Uh, as I sort of… shown partially, and you alluded to, there are many unique features why people are using this MP system. 01:27:50.000 --> 01:27:57.000 essentially more in vivo-like in vitro system can control the cellular microenvironment very well. 01:27:57.000 --> 01:28:02.000 Can co-culture multiple different cell types in a tightly controlled manner. 01:28:02.000 --> 01:28:08.000 Uh, in three-dimensional or 2.5D. depending on how the in vivo situation is. 01:28:08.000 --> 01:28:15.000 My Costco… my cost could be comparable, and we also can be integrated with a variety of different feature… a sensor. 01:28:15.000 --> 01:28:22.000 However, next, however, if you look at, uh, whether we can use them as a routine to learn model. 01:28:22.000 --> 01:28:26.000 There are many challenges that we need to overcome. Next. 01:28:26.000 --> 01:28:32.000 So, some of the challenges are it require, oftentimes, not always, but oftentimes require. 01:28:32.000 --> 01:28:46.000 dedicated, uh, tool and instrument. Uh, to be able to use this MPS device, and sometimes doesn't really compare… is not really comparable with a traditional, uh, workflow in a lot of. 01:28:46.000 --> 01:28:59.000 medical lab, user-friendliness, relatively low, and uh. Some, like, most importantly, throughput, whether it's a fabrication, throughput, operation of throughput, that throughput is relatively known. 01:28:59.000 --> 01:29:10.000 low. So, as I work with my colleague in the Superfund Center that is led by Dr. Ivan Rusin here in the vet school at A&M. 01:29:10.000 --> 01:29:16.000 You know, by the time you have a couple different replicates, couple different dose response, positive control. 01:29:16.000 --> 01:29:24.000 Even for a single chemical, uh, chemical toxicon, the number of experiments that you have to run easily in the several tens of. 01:29:24.000 --> 01:29:31.000 So, next slide. Well, we're really looking to, can we really improve this one? So, if you look at the next slide. 01:29:31.000 --> 01:29:36.000 Uh, we have been working on increasing throughput to improve the MPH utility. 01:29:36.000 --> 01:29:42.000 So, in an example of our two-chamber unit, uh, which. 01:29:42.000 --> 01:29:51.000 Uh, we look into both from fabrication side as well as operations side. So, if you look at the fabrication side, initially, this is all a. 01:29:51.000 --> 01:30:05.000 Microfabrication process. Uh, called Replica molding. So we have a master mold from which you are replica molding a variety of different polymer devices. So, in a simple step is we went from a 4-inch. 01:30:05.000 --> 01:30:13.000 way for Massimold into 6-inch wafer mode, so that in a single fabrication run, we can fabricate many more devices easily. 01:30:13.000 --> 01:30:25.000 Um, and then in going from a putting individual devices, uh, into a 6-wheel, well over 6-wheel device, we created some custom devices so that we can have an operational improvement. 01:30:25.000 --> 01:30:36.000 Especially in terms of device configuration. So that way it fits in a conventional standard plate assay format for higher throughput operation. 01:30:36.000 --> 01:30:42.000 Next, in a second example, this is how we configured our 4-chamber order chip device. 01:30:42.000 --> 01:30:47.000 Obviously, the fourth chamber device require a little bit more larger footprint. 01:30:47.000 --> 01:30:59.000 Uh, so we modify the device configuration a little bit so that we can still fit many of these full chamber, full cochle chamber device into a plate format. 01:30:59.000 --> 01:31:09.000 So here, each of the individual array has 5 independent devices, and we configure this one so that 4 of these array fits in a standard, uh. 01:31:09.000 --> 01:31:15.000 96 world plate format. So, we worked on some design optimization and geometry. 01:31:15.000 --> 01:31:21.000 to really be able to fit this one, to make not only a fabrication process easier, but. 01:31:21.000 --> 01:31:24.000 The operation price is also easier. Next. 01:31:24.000 --> 01:31:30.000 So, then, uh, we also wanted to adopt this one to an automated testing workflow. 01:31:30.000 --> 01:31:38.000 So, um, oftentimes, fabrication of a large number of devices is one thing, and one challenge has been overcome. 01:31:38.000 --> 01:31:42.000 an operation becomes still a bit of a labor intensive. 01:31:42.000 --> 01:31:49.000 So, main reason the operation is a little bit labor intensive, even though all of our operation is pipette-based. 01:31:49.000 --> 01:31:55.000 These are individual devices, so we basically made them into an already format so that, uh, next one. 01:31:55.000 --> 01:31:59.000 So that we can use a multi-channel pipeline or a robotic liquid handling system. 01:31:59.000 --> 01:32:06.000 So we use a low-cost, open-shorn robotic handling system and made our device comparable here. 01:32:06.000 --> 01:32:13.000 And here, basically, about 180 MPS devices can fit in into a single, single run. 01:32:13.000 --> 01:32:18.000 We have, and using this one, next slide, using this one, we were able to basically load. 01:32:18.000 --> 01:32:23.000 or cells into all of these 180 devices with minimum error. 01:32:23.000 --> 01:32:28.000 And then we'll also be able to use media loading, media exchange, as well as. 01:32:28.000 --> 01:32:38.000 eflorn sampling of this operation. So that was a big improvement for us because the big challenge of our operation is the cell loading process. 01:32:38.000 --> 01:32:43.000 And then the every 48-hour, or 48 to 72 hour media exchange process. 01:32:43.000 --> 01:32:49.000 And then the 24-hour media sampling, effluent sampling they're conducting. 01:32:49.000 --> 01:32:54.000 So that we can do a downstream inflammatory cytokine analysis. 01:32:54.000 --> 01:33:06.000 So this really helped us in terms of the operation. And finally, we tested a variety of different polymer to polymer, and then transitioned into an injection molding process. 01:33:06.000 --> 01:33:14.000 of this device so that we can have the fabrication process even much more easier. 01:33:14.000 --> 01:33:28.000 Next. So now, um… Uh, I'm with this device, uh, as I previously mentioned, we have this fetal membrane device, and we have the placental oral chip device. 01:33:28.000 --> 01:33:33.000 So, we were very much interested in looking into the, uh, hazardous. 01:33:33.000 --> 01:33:45.000 how the environment has substances. impact both interfaces to really look into this one comprehensively. So we did, uh, we used both interfaces and, uh, made a comparison. Next. 01:33:45.000 --> 01:33:58.000 So, um… The… in our case, in addition to the early studies of using ketamine, we were very much interested in the PFAS family chemical. 01:33:58.000 --> 01:34:03.000 Because as this audience knows very much, it's a very sensitive chemical major public health concern. 01:34:03.000 --> 01:34:08.000 And, uh, may contribute to preterm growth space in some of these studies coming out. 01:34:08.000 --> 01:34:22.000 So, uh, next. The, uh, hypothesis here is that MPS models can be used to study environmental toxic scan, and they are potentially harmful effect on both the placental as well as fetal membrane interface. 01:34:22.000 --> 01:34:28.000 So specifically, we use 4 different PFAS family chemicals on both placental and felon membrane interface. 01:34:28.000 --> 01:34:34.000 And we looked into both how the direct exposure as well as indirect exposure. 01:34:34.000 --> 01:34:43.000 Uh, can affect the potential. potential pregnancy risk. Next. 01:34:43.000 --> 01:34:55.000 So, uh, in terms of the first thing we do is we tested this PFAS family chemical to make sure that, uh, device surface is not really absorbing, so we tested this one on both. 01:34:55.000 --> 01:35:04.000 The phyto-motonin interface device, as well as the placenta device, looked into absorption. Next slide, we quantify this one by a LCMS. 01:35:04.000 --> 01:35:09.000 Basically, the quick 48-hour study confirmed that there's no chemicals. 01:35:09.000 --> 01:35:25.000 Next one, um, we looked into… next… We looked into, then, direct exposure in the 2D to do a dose range finding. So this is typically how we use a CD for well plate to do a dose range finding. 01:35:25.000 --> 01:35:35.000 Uh, accessibility after 48 hours, next one. Uh, we'll use a cell title, so here you see the fourth one, the seizure, the maternal. 01:35:35.000 --> 01:35:41.000 cell, as well as 3 different fetal cell, and how they respond to the exposure in terms of viability. 01:35:41.000 --> 01:35:53.000 Next slide shows the, uh, next step shows how the placenta cells respond. So you can see that they are responding differently with placenta cells a little bit more resistant to all of the PFAS chemical. 01:35:53.000 --> 01:36:03.000 Next one, then, uh, we looked into, uh, actual device, and, uh, based on the previous dose range finding, we applied the. 01:36:03.000 --> 01:36:08.000 chemical into the seizure chamber, and then in the case of. 01:36:08.000 --> 01:36:12.000 in the case of a placenta, we apply this one to an STV chamber. 01:36:12.000 --> 01:36:18.000 We looked into relying on gravity-driven flow, and then we looked into the response. 01:36:18.000 --> 01:36:26.000 So, if you look at the next one. Uh, what you're seeing here is, um, an incubator for 48 hours. 01:36:26.000 --> 01:36:35.000 We can look into endpoint assay, LCMS, LDH assay, and then cytokine analysis. So, next slide will show the actual result. 01:36:35.000 --> 01:36:43.000 In the case of PFAS on a fetomotin interface, you'll see how the different, forcible propagation occur. 01:36:43.000 --> 01:36:47.000 Next one. And so the result is that. 01:36:47.000 --> 01:36:56.000 PFAS propagate across the device, about 10% by the time it reaches the final, uh, fetal chamber. Next. 01:36:56.000 --> 01:37:06.000 Um, and then, uh, we also looked at the placenta, uh, how the PFAS propagate. Pfas has a bit more of a stronger barrier, so they are retaining PFAS a little bit more. 01:37:06.000 --> 01:37:20.000 Next slide. Uh, next slide will be a viability assessment, so… If you look at the viability, we… this is what you're seeing in comparison to next slide, uh, you can see how placental. 01:37:20.000 --> 01:37:26.000 Uh, placental cells respond. Next. 01:37:26.000 --> 01:37:39.000 So, um, this shows how the viability changes, and if you just go straight to the, uh… entirety of the 27, page 27. 01:37:39.000 --> 01:37:47.000 Yeah, just keep going. So here you are looking at the cytokine analysis of the phenylumetone interface. The LPS was a positive control. 01:37:47.000 --> 01:37:55.000 And depending on the different PFAS chemical that you're seeing a variety of different responses, so next one. 01:37:55.000 --> 01:38:02.000 You look into how different cytokines respond to, in this case, PFDA had the most, uh, induced the most significant fetal inflammation. 01:38:02.000 --> 01:38:06.000 response, especially in the AC, compared with other PFAS. 01:38:06.000 --> 01:38:15.000 Chemical, next one. Now, when we apply this one to the placental interface, you'll see a different response. 01:38:15.000 --> 01:38:24.000 Next one, uh, here, um, so what you're seeing here is that these two interfaces respond, uh, quite differently. 01:38:24.000 --> 01:38:34.000 So, in summary, if you just can continue with the slideshow here, the fetal membrane, direct exposure on the fetal AC showed cytokine responses. 01:38:34.000 --> 01:38:42.000 more reactive to PFAS and a short micellular inflammatory and cytokine cytotoxic response. 01:38:42.000 --> 01:38:49.000 In the case of placenta, drug exposure where trophoblast cells and Huvex cells were more resistant. 01:38:49.000 --> 01:38:55.000 Um… and then a robust barrier to PFAS, where we saw limited propagation. 01:38:55.000 --> 01:39:02.000 And then, uh, finally, uh, truthful is resistance to cell death and inflammation, uh, etc. 01:39:02.000 --> 01:39:07.000 So if you look at the quite summary, cellular and physiological impact. 01:39:07.000 --> 01:39:11.000 So if you look at the summary of these two differences. 01:39:11.000 --> 01:39:16.000 you'll see how the fetal membrane and the placental respond very differently. 01:39:16.000 --> 01:39:22.000 And that really shows the importance of looking into both the fetal membrane as well as the placenta. 01:39:22.000 --> 01:39:31.000 Um, next one. For that, um, so I'll just skip the summary here. 01:39:31.000 --> 01:39:37.000 So, here, I would like to kind of maybe take a last minute to look at what the future direction is. 01:39:37.000 --> 01:39:45.000 So, we have demonstrated that improving ethere is quite important to be able to do this kind of a study. 01:39:45.000 --> 01:39:52.000 Uh, so our high-through automated NEMS model, uh, to have a practical routine utility. 01:39:52.000 --> 01:39:59.000 in a potential hazard of environmental chemical. really having a hierarchy with automated operation is critical. 01:39:59.000 --> 01:40:06.000 And that's really both an ongoing and future incentive theme, not just for our project street, but in many other. 01:40:06.000 --> 01:40:13.000 projects. The second one is… That enables testing broad ranges of PFAS chemical. 01:40:13.000 --> 01:40:18.000 I'm referring to tens to hundreds of different PFAS chemicals. 01:40:18.000 --> 01:40:25.000 Uh, we are ongoing effort in testing both individual chemical and a mixture, and how their response are different. 01:40:25.000 --> 01:40:37.000 That's another key Texas A&M Superfund Center theme. The next one is real-world testing, real-world sample from disaster on other… sorry, can you go back? Another key temperature theme. 01:40:37.000 --> 01:40:43.000 And then, uh, we'd like to establish this one. It has been establishing this one as a mechanistic children study preterm birth. 01:40:43.000 --> 01:40:49.000 And finally, we are trying to incorporate population variability into the MPS model by. 01:40:49.000 --> 01:40:54.000 Using cells from different individuals and test their differential impact. 01:40:54.000 --> 01:40:58.000 And that's another feature. There will be a key feature. 01:40:58.000 --> 01:41:02.000 ongoing, somewhat ongoing in future time, we should perform Sanapim. 01:41:02.000 --> 01:41:10.000 So what we're showing here very much aligned very well to our current and our future theme, and how we. 01:41:10.000 --> 01:41:17.000 How, uh, thematically it fits with the rest of the project. And then, uh, with that, I would like to thank. 01:41:17.000 --> 01:41:25.000 the, uh, all of the team members, so this is really in close collaboration with Dr. Menor's lab at UTMB, Dr. Ivan Rusen's lab at A&M. 01:41:25.000 --> 01:41:32.000 Followed by NIH's Superfund Center as well as the NCATS, some of the NCATS clinical trial on Trace and PS. 01:41:32.000 --> 01:41:42.000 device as well as the NICHG. So with that, I would like to happy to answer any questions that you may have. 01:41:42.000 --> 01:41:45.000 Hey, thank you very much, Dr. Hahn. Fascinating presentation. 01:41:45.000 --> 01:41:52.000 So we have a question here in the question bot, um, what is the film that you use… that you put in the 3D printer made of? 01:41:52.000 --> 01:41:58.000 Have you tested your construction arrays for chemicals that might affect assays? 01:41:58.000 --> 01:42:05.000 How do you background chemical levels from your construction equipment compare with the purchase wells? 01:42:05.000 --> 01:42:10.000 And is there batch-to-batch variation related to the film sources? 01:42:10.000 --> 01:42:20.000 Okay, so first of all, I would like to clarify, the actual device itself is either replica molded molded polymer ore. 01:42:20.000 --> 01:42:33.000 The or the injection molded plastic, which is a reservoir, the 3D printed structure is only for the guide structure, so that's really printed polymer is not in direct contact with the cell. 01:42:33.000 --> 01:42:42.000 Uh, we are very much aware of many of the 3D printed material, especially the photopolymer resin has a potential toxicity. 01:42:42.000 --> 01:42:48.000 So just want to assure that there's no direct contact with the cell and a 3D printed structure. That is more of a. 01:42:48.000 --> 01:42:57.000 sort of more of an experimental fixture that eventually will be anyway made into injection moly plastic. 01:42:57.000 --> 01:43:11.000 Thank you. Um, let's see, there's no other questions. I have one. So, for your experiments, you demonstrated that you can integrate the operation into a liquid handling robotic system. 01:43:11.000 --> 01:43:17.000 So how difficult would this transition be for other MPS devices? 01:43:17.000 --> 01:43:25.000 So, first of all, in our case, our MPS device, from the very beginning, has been a pipette operation-based for two reasons. 01:43:25.000 --> 01:43:29.000 In most cases, having to use a fluidic pumping system. 01:43:29.000 --> 01:43:37.000 I think it makes it much more difficult. Second, in our case, these cell types are typically not exposed to any sort of a high shear stress. 01:43:37.000 --> 01:43:43.000 We actually did a comparison between a dynamic flow versus static no flow. 01:43:43.000 --> 01:43:49.000 And we didn't see any effect. Actually, dynamic flow is having a negative effect. So in our case. 01:43:49.000 --> 01:43:56.000 Uh, any, uh… Transitioning from a pipette-based operation into a multi-channel robotic pipetting system, relatively easy. 01:43:56.000 --> 01:44:02.000 And I think many other MPS devices that has a pipette-based operation and an open well structure. 01:44:02.000 --> 01:44:11.000 will be very easily amenable to transition. Having said that, MPS devices that, for a variety of reasons, have to have a. 01:44:11.000 --> 01:44:17.000 Um, a shear stress, relatively high shear stress, such as, like, a blood vessel, etc. 01:44:17.000 --> 01:44:31.000 Those will be a lot more difficult to transition into a liquid-handed robotic system because of the assurance needed. So I would say some are relatively easy to transition, some might be a little bit more challenging to transition. 01:44:31.000 --> 01:44:36.000 Hey, great, thanks for your response, and thanks for the presentation. 01:44:36.000 --> 01:44:54.000 Um, so, um, I want to thank all the speakers for, um, their great presentations today, and the, um… The folks who are joining in on the webinar, we appreciate your attendance, and I want to also thank Molly for. 01:44:54.000 --> 01:45:08.000 making… hosting this event so seamless, so… Um, thanks again, everyone, and I'll turn it back over to Molly to close this out. 01:45:08.000 --> 01:45:13.000 Thank you. So, before we conclude, let's take a look back at the seminar homepage. 01:45:13.000 --> 01:45:19.000 This website will be active from today on, and contains important resources. 01:45:19.000 --> 01:45:24.000 Such as links to download our presentation materials, contact information for our speakers. 01:45:24.000 --> 01:45:35.000 And a link to our feedback form. We ask that you consider filling out the online feedback form. We do look at your comments as we continue to try to improve the content and delivery mechanisms. 01:45:35.000 --> 01:45:42.000 You can also request a confirmation email from the feedback page as a record of your participation in today's events. 01:45:42.000 --> 01:45:55.000 just look under tips, questions, and support. So here are a few ways to learn more about the Superfund research program. The SRP website homepage link will be posted in the chat. 01:45:55.000 --> 01:45:58.000 And I encourage you to check out the various components of their website. 01:45:58.000 --> 01:46:09.000 To learn more about the program and initiatives. I would also like to remind participants of the various ways to stay connected to ensure that you don't miss any webinars. 01:46:09.000 --> 01:46:15.000 For more information on EPS cooling webinars, please refer to the cluen.org website. 01:46:15.000 --> 01:46:19.000 You can also subscribe to the monthly newsletter, Tech Direct. 01:46:19.000 --> 01:46:25.000 We also encourage you to follow NIEHS on X, YouTube, LinkedIn, and Facebook. 01:46:25.000 --> 01:46:29.000 You can also subscribe to the SRB Risky Learning Listurf. 01:46:29.000 --> 01:46:43.000 For information about upcoming webinars. And with that, I would like to encourage everyone to join us for the next session in our series on January 21st at 1pm Eastern Time. 01:46:43.000 --> 01:46:51.000 Available on this slide is a QR code that will take you to the webinar series homepage for you to register for the session. 01:46:51.000 --> 01:46:57.000 If you are unable to attend live, an archive will also be available shortly after the session. 01:46:57.000 --> 01:47:01.000 And we'll include session recordings and our speakers' presentations. 01:47:01.000 --> 01:47:04.000 With that, it is time to conclude today's session. 01:47:04.000 --> 01:47:10.000 Thank you so much for your participation, and thank you to our presenters for their time spent preparing. 01:47:10.000 --> 01:47:25.000 And sharing their presentations with us today.