 I want to call the CHAP meeting to order, and we'll start by introducing ourselves again. My name is Phil Murkis, and contrary to what the little sign says, I'm the chair of the CHAP committee. Good morning. I'm Bernard Schwitz, and I want to confirm that Phil is the chair, and also that I'm the vice chair. Holger Koch, so I am honored to sit next to the vice chair, Ruhr University, Bochum, Germany. Chris Jennings, biostatistics, Virginia Commonwealth University. Richard Sharp from University of Edinburgh, UK, and I commanded a title invited expert, which I find rather disturbing. I pointed a bit as clear, but Kimbuckle High Brown University, and also an invited expert. Russ Hauser from Harvard School of Public Health. Andreas Kortenkamp, I used to be from London University, but now I'm with Brunel University's Institute for the Environment. Michael Babich, I'm the CPSC project manager. Okay, this morning we're going to start off with a presentation from Rebecca Kluwell on the DIMP dose response studies that they have performed, and then we're going to have our two presentations from our invited experts. Mike? Ann, do you have something? Well, let me just remind everyone, this is a public meeting. It's being webcast and recorded, and I think that's it. Thank you all for coming. Rebecca? No, that didn't help. Full screen mode? Good. Thank you. I want to thank you for letting me come today and speak about these studies. I'm pretty excited because we're finally at the stage where we can call them final. The manuscripts have been submitted last week, and so hopefully we can get this out there in the public and let you see what's happening. I am a research investigator at the Hamner Institutes. We have a long history of research in the Thalates. We used to be CIIT, Centers for Health Research, and before that we were Chemical Industry Institute for Toxicology, and our work began with Paul Foster and carried on with several scientists you've probably heard of, including Eve Millkreest and Kevin Guido. I worked with Kevin Guido for a little while before he moved on to the FDA. These studies have been funded by ExxonMobil, and they requested that I speak here today, but I would like you to know that this is work that was performed independently at the Hamner. All right. I'll have to go to the next. That'll work. Okay, so there's a long history in the Thalates, kind of defining the window of susceptibility for exposure and lifelong effects. The studies that I'm presenting here today were aimed at really kind of covering this entire window with as much as comprehensively as possible, including the possible effects in the fetal rat from gestational exposure, and then also in the adult, or at least juvenile rat through gestation and lactational exposure. In addition to that, we did a kinetics study, which pretty comprehensively examined the pharmacokinetic behavior of DIMP at several doses during gestation in order to get an idea not only of the external dose relationship to effect, but the internal fetal target dose. So I'm going to start with the gestation study. The design was to, well, the design was basically put together to determine the fetal dose, the fetal target tissue dose, at several different doses of DIMP. We looked at six different time points. We looked at three different doses, 50, 250, and 750 milligrams per kilogram per day. The dosing design was based on previous studies with other Thalates, which showed the sensitive window to be from gestation day 12 to 19. And we examined, in addition to the metabolite kinetics, we also examined fetal effects at two and 24 hours post-dosing on GD 19 and 20 in the fetal rat. We looked at five different metabolites, and I suppose I'm going to start with the effects study so I won't talk too much about the metabolism, but for all of the hormone and histopathology effects, we looked at an N of eight, which is eight litters. And for AGD, we looked at every single male pup from the eight litters. For the testosterone, we looked at two to three pups per litter, and then measured them individually in an average symbol by litter. So for the developmental effects for testosterone at two hours post-dosing, we did see a significant decrease in testosterone at two hours post-dosing for the 250 and 750 milligrams per kilogram per day doses. When we looked at 24 hours after the dose, the testosterone levels have recovered even at 750 milligrams per kilogram per day. I did include some information here on the other Thalates so that you can compare the potency. This is similar to what Earl Gray's lab has seen, and Hannah said all their recent publication, that DIMP is significantly less potent for the testosterone inhibition, though what we're interested in is not only relating these Thalates on an external dose, evaluating the bicep, but also internal dose because across the Thalates, metabolism can be quite different. For antigenital distance, we did not see a decrease in antigenital distance on GD 20 for absolute or scaled AGD at doses up to 750 milligrams per kilogram per day. That is, I believe, similar to what other people have seen. And for his pathology, we examined the seminiferous tubule diameter, which an increase in seminiferous tubule diameter has been established as a marker of dibutal Thalate exposure. We did not see any increase, and this was measured quantitatively with imaging software, where we measured each individual tubule in all of our slides, and then averaged by slide, and we didn't see any change. And we also had two pathologists actually looked at this slide, and neither of them saw any significant difference in the seminiferous tubules. There was an increase in multinuclea gonocytes, or as Kim Buckelhide has made clear to me, they're actually germ cells, but from a morphological standpoint, they still look like gonocytes, and so the pathologists who looked at these, which Gabrielle Wilson from EPL and Diane Creasy from Huntington Laboratories, or Huntington Life Sciences, I'm sorry. Anyway, called them gonocytes. So we did see a significant increase at 250 milligrams per kilogram per day. And this, we also, in addition to that, the pathologists noted an increase in the number of animals with large latex cell aggregates at 750 milligrams per kilogram per day, and this was in GD 20 rats. And this is, because we know DIMP is a P-PAR agonist and has been shown to cause liver weight effects, we also took the maternal liver during this study and weighed it, and we did see an increase in the maternal liver weight at 250 and 750 milligrams per kilogram per day. And we don't believe that's related to the reproductive or developmental effects, but we did want to check it just for completeness. And so during the study concurrent with the study, we also measured the metabolite disposition. So we looked at four, five different metabolites of DIMP, including the free monoester MIMP, three oxidative metabolites of the MIMP, the primary metabolites that have been identified by silva at all. And in addition to that, we also looked for the glucuronide conjugate, because there was evidence from some in vitro studies that the rat was able to conjugate MIMP in vitro. And so for that, we actually had to prepare our own standard and develop a method, but we were able to measure the glucuronide in blood samples. And we evaluated the maternal plasma, PUP plasma, maternal urine, maternal liver and placenta, as well as the fetal testes and amniotic fluid. The tissues, the time points, and the doses were all based on previous studies that we have performed at the Hamner with DBP, which gave us a good idea of where we might see saturation of metabolism or oral uptake and the tissues that we knew we should be able to see some of these metabolites in. So these are just the plots which show you that we looked at all five metabolites and we were able to see all five metabolites in both the maternal and fetal blood. We actually saw all five metabolites in all of the tissues. In the urine, there was a very, very negligible amount of the glucuronide and free MIMP, which I'll show you in a second. In the blood and tissues, the primary metabolite was the carboxylated MIMP. Then the second most prevalent metabolite was free MIMP itself. PUP levels were about three or four full lower than the maternal blood levels pretty consistently. This is to show you basically that there was a saturation and oral absorption at the highest dose. And you can tell that because the CMAX does not increase with those, the maximum concentration in the area under the curve flattens out with dose. But the elimination was linear with time, suggesting that clearance is not saturated with dose. The maternal urine tabulites, the carboxylated metabolite of MIMP, was the primary metabolite which has been shown before by Silva et al. We did see MIMP and the MIMP glucuronide in the urine, but it was very low, less than 0.1% of the dose. So a preliminary analysis of DIMP, we used a physiologically based pharmacokinetic model, which I developed at the Hamner for DEHP and just applied it to DIMP. The first one we didn't change any parameters because it looked an awful lot like DEHP, the metabolite kinetics. And I won't go into the model, but what we basically found is that the metabolite, the model fits the DIMP pretty much just as well as it fits the DEHP. And I'll admit that the model is really preliminary right now and it needs a little bit of adjustment. But what this tells me is that the pharmacokinetics of DEHP are very, very similar to DIMP, which would indicate that any differences that we see in potency between the two chemicals is probably a pharmacodynamic difference, and so that's where we are with the kinetic study. So for the postnatal effect study, which is probably more of interest to the panel here today, we consulted several experts when we developed this study, including Shelly Tyle from RTI and Earl Gray from the EPA, and they were very helpful, and I'm very grateful to them. We made a lot of changes to our study design, extending from P&D-14 to P&D-49 to include adult or post-puberol effects, and we kept more animals than we intended to. So we ended up with an N of 20 litters per treatment group and at least as many male rats as we could possibly keep because of limitations in how many were born. So sometimes you get five or six male rats per litter, sometimes you get one. We kept as many as we could. So in order to be as comprehensive as possible and as sensitive as possible, we had as many as 24 litters per treatment group. We began our diet, we had dietary dosing in order to keep it more human relevant. We began the dosing on gestation day 12 and we treated through P&D-14. We kept all necropsies and observations. We're completely blinded, so we had a system where we gave the animals letters and anybody who was in the necropsy room did not know what those letters goaded to as far as treatment group goes, and that includes a histopathology as well. So for endpoints on P&D-2, we looked at AGD. We took testis testosterone from one male pup per litter. We don't want to do more than that because we would have reduced the number of animals per P&D-49 observations. And we looked at testis and epididymis histopathology. On P&D-14, we measured AGD and nipple retention. This was not a necropsy time point. This was just external observations. And then on P&D-49, we looked at AGD, nipple retention, testis, testosterone, hypospadias, or any fallous malformation, and we measured pre-putial separation. We also looked at the morphology and tissue weight of 10 reproductive tissues, including the liver and kidney. So this just shows you the actual or calculated DIMP dose, which was calculated from the amount of food that was taken in by the animals, which was measured four times a week, and the concentration that was measured analytically in our food. So during gestation, our target doses were 50, 250, and 750 milligram per kilogram per day. And during gestation, we were pretty much right on target. And in lactation, due to the increase in food intake and the decrease of maternal body weight, the dose has reached as high as two-fold higher than our, or actually a little bit more than two-fold higher than our target doses. So we were at or above our target doses throughout the study. So on postnatal day two, we did the decrease in DIMP at the highest dose, 750. None of the other doses showed any body weight effects on P&D-2. This is different from GD-20. There was no body weight effect in the fetus on GD-20. AGD, only DBP, and I should have mentioned that we included DBP as a positive control for this study, mostly because we didn't see much happening in the gestation study, and so we wanted to have a positive control to compare all future effects to. So the dose of dibutal phthalate was 7,600 parts per million, which equates to about 500 milligram per kilogram per day during gestation. So we did expect to see effects at this dose. So for DIMP, we did not see AGD effects, but we did see them with DBP. For multi-nucleated gonocytes, we did see an increase both at DIMP at 750 milligram per day, and we saw an even greater increase with dibutal phthalate. In addition to that, we did see an increase at the highest dose of DIMP and at DBP for latexal aggregates. And this is an example, the histopathology, which you probably can't see very well, but it's just highlighting what our pathologist was calling multi-nucleated gonocytes and latexal aggregates. On PND-14, we saw a pretty substantial reduction in neonatal weight at the highest dose. It's actually about 30% of their body weights. I believe this is probably due to palatability of the milk. Some of the data with DEHP shows that at high doses where metabolism or hydrolysis is saturated, you actually get the diester at pretty high concentrations in the milk. So because this drastic weight reduction is only present at 14, none of the other time points we looked at. DBP, there was still no weight effects. So for AGD, we saw a decrease on PND-14 in both the high dose, DIMP, and the dibutal phthalate-positive control group in women's scale by body surface area that reduction in AGD was maintained. I believe that it's possible that there's still some confounding factors there with the neonatal weight because it is so drastically low, but right now it's difficult to tease that out. We also looked at nipples and areola. We did not distinguish between nipples and areola at PND-14. Everything that was there was counted, so the only increase we saw was with dibutal phthalate. On PND-49, where we're looking for permanent effects, we used PND-49, assuming that the preputial separation would be finished by then, given some expert advice, and it turns out that even in our control rats, some of the rats hadn't reached fully complete preputial separation by PND-49. So we ended up bringing an expert from Earl's lab to discuss this with him, and he suggested a scoring scale of zero to three, where three is complete separation, and one is no separation at all, and zero is no separation at all, and one and two is in our partial separation, and then almost completely done. So that's how preputial separation was scored in all of the remaining necropsies. Based on this preputial separation score, only DBP caused a delay in preputial separation. Only DBP caused permanent nipple retention, and none of the treatments, including DBP, caused a permanent change in inogenital distance, whether we used the absolute or scaled. We also performed histopathology on PND-49. We did not see any permanent effects from any of the DIMP treatments. For DBP, there wasn't any statistically significant increase in the changes in the histopathology on PND-49 however there were some really dramatic effects like complete atrophy and mineralization, and we did have a genesis of one testis. And these are the different tissues that we looked at, their weights, and the DBP, we saw a decrease in seminal vesicles, ventral prostrate weight, LABC weight, and the kidney weight at 500 milligrams per kilogram DBP. We did not see a change in any of the reproductive organ weights at any of the DIMP doses. The only significant effect we saw with DIMP was an increase in gubernacular cord length in the lowest dose, but not at the higher two doses. I don't believe this is a treatment related effect because we don't see a dose response trend, and in addition, a lengthening of gubernacular cord is not what you expect to see with a thallitude expected to be lower. All right, and then as far as the morphological effects, only statistically significant effects that we saw were in DBP, and that was incomplete epididymis, which means it was missing a part of the, and it was incompletely formed. Generally, that was the body of the epididymis, sometimes it was the head or the tail, and these were confirmed by histopathology. And the flaccid epididymis, which is a term that we kind of coined, I haven't seen it around, but basically what it meant, it was really mushy and squishy, and it should not have been. When we gave that to the histopathologist to look at what she said, or the pathologist, I'm sorry, to look at, she said it was, it had more fat cells, it had less actual epididymal tissues, and the tubules were generally larger, so probably fatty epididymis might be more descriptive. And for DBP, we also saw one incidence of undescended testis. We did see one animal where it had complete atrophy of the testis and epididymis unilaterally, and we saw mild or slight hypospatias. We didn't see any really dramatic hypospatias, we didn't see cleft palate, but we did see slight hypospatias or mild hypospatias in five out of 21 litters in the DBP group, and exposed ozpenous, which just suggests it's a little bit more severe thallus malformation in one out of the 21 dibutal phthalate litters. So none of these effects were statistically significant for DIMP. We did not see an increase in the flaccid epididymis, and we didn't see any incidence of incomplete epididymis at the higher doses, though we did see a couple at the very lowest dose. And so, and this is, last slide is just a comparison for, for you guys to look at, summarizes the differences in the effects between di-isoninophthalate and dibutal phthalate. Di-isoninophthalate clearly causes effects on body weight, which is not seen with DIMP. The, the dibutal phthalate causes nipple retention, changes in AGD at pretty much every time point except PND49. It alters thallus development, epididymal development, preputial separation, and for the reproductive organ whites were decreased in our studies, and we obviously hiss pathology effects. And then for DIMP at doses greater than 250 or milligram per kilogram per day, we do see decreases in body weight, and we did see a decrease in AGD, but only on PND14, and we did see some changes in multi-nucleated gonocytes and latex cells. So in conclusion, we believe we've devised, we devised or designed and with the help of several people, several experts, I'm present here today, including Cam has helped quite a lot in at least advising on who we should seek out for help with the pathology. So, so that was helpful. And then also Earl Gray was involved in this, at least gave us advice or reviewed our, our original plan and Shelley Tile too. And anyway, so we've developed a pretty comprehensive suite of, of studies. I think we've established a clear no well of 50 milligrams per kilogram per day. We just don't see anything, whether it's ingestation or lactation at 50 milligrams per kilogram per day. At 250 milligrams per kilogram per day, we did see a significant increase in multi-nucleated gonocytes and testosterone reduction on GD19 and the AGD on PND14. All of these effects were recovered later time points. And so I guess it's a matter of debate whether you would call these adverse or not. The role of testosterone is part of the leading mechanism to the male effects, I believe, is unclear at this point. But we did establish a change there. I think probably most importantly, I believe this is a third bullet of the slide, is that we did do a global analysis based on, similar to what Earl Gray has, has recommended before in the living and Paul Foster recommended previously in the literature. So for that analysis, what we basically did was if they showed any of the suite of effects that are possible with the phthalate exposure, we called that rat a positive. So it was a positive responder. If it did not, it was a negative responder. And then we did a global analysis. And only DBP was statistically significant for what could be called the rat phthalate syndrome. So DIMP at 750 milligrams per kilogram per day was not statistically significant from control, even when you included the entire suite of effects, which includes pre-putial separation, nipple retention, all of the malformations on PND49. So that analysis was performed on PND49 only. And so that would be the end of my talk. If you have any questions, I'd be happy to take them. Or if you want to send questions later, I'm of course available. So thank you. We'll take any questions for clarification. CHAP committee? Andreas? It wasn't clear from me from the details which you provided whether you corrected for litter effects in your statistical analysis. So litter are the random factor influencing what you saw. Okay. So what we actually did, and this was I worked with a statistician on this because I am not a statistician. And I've actually, right now, the approach that he is is slipping my mind. But what we did was actually control for multiple pops per litter. So you looked at the individual effects and then we looked at the number of animals per litter and we did account for the number of animals per litter. Is that what you're asking? No, it is well known that if you compare animals from the same litter, they're quite similar to each other. But if you compare animals from different litters, they are not necessarily very similar to each other because you have something like a litter effect. Or in fish biology, it's called a tank effect. If you keep fish in the same tank, the biological response will be similar, more similar to say the fish in a tank next door. With litters, it's exactly the same. For a statistician, that means that if you have the responses from each of the animals from one litter, you can't pool directly with all the data from the second, third, whatever litter. You have to adjust for the fact that animals from the same litter are in fact more similar to each other than if you make the comparison between litters. So that's my question. Did you adjust for litter effects in that way? Well, I believe so. But I would definitely have to check with our statistician. I could show you the spreadsheet that he used. And I believe it was the jackknife method for comparing between litters. And I could give you the reference also that he used. But yeah, that was certainly something that I was asking him when he needed to control for variations between litter and within litter. So yes. I'm sure you were aware of the paper by Borch et al, which came out in reproductive toxicology, I think last year, where they also did a DINP study. Yes, Bo Berg now, yeah. Bo Berg, yeah. Of course, yes. She got married? Yeah, she did. And I looked at their data in comparison with yours. And there are actually a lot of similarities. But on one point, when they did what you did with testosterone production, no levels, they also measured testosterone levels, exactly what you did. But in addition, they measured testosterone production and saw a much more pronounced effect of DINP. Did you do the same experiment or we did not? We did not do the X vivo testosterone production. So that's how Hannes et al did it too. They do the X vivo testosterone production. So Bo Berg et al actually did not see the level of effect that we did with the concentration or testis in vivo concentration. So they didn't see a consistent testosterone reduction that way. They did see it when they did the X vivo production. But I believe ours actually showed more of an effect than theirs on testosterone. They also saw, I think if I compare the testosterone data, they published with what you have presented here, I think a fairly good agreement also with anal-gentle distance. However, they observed retained nipples, but you didn't. And now it comes, they used with star rats, you used spagdolles. It's a more general question, bearing all this in mind. How do you see your results in relation to theirs? Can we attribute this to species differences? Will this alter the overall conclusion or what's your view? I believe the results are very, very similar. I have actually spent quite a lot of time looking at her paper because there are a lot of similarities. We are very, very similar. I believe the doses where we see effects are very similar. We actually saw AGD effects on P&E 14 at 750. They didn't see them till 900, but really at that point we're at really high doses. I would say in general that these papers could be used together. I believe that they're together. It's a pretty constant and intensive suite. What differences there are? It could be strain. It could also be number. Some of what we looked at, we had a much higher statistical power, I guess, because we had an N of 20, where some of their end points varied per end point, but some of their end points, they had as low as five. I think that could play a role. We did see the same trend that they saw with the Nipples at P&E 14, but the change was so slight. We're talking a control of 1.5 in a treatment effect of 1.8% of the rats. I would say with you have 20 litters or more that at that point it's just not statistically significant effect. I suppose it is possible that it's a Wistar effect. I haven't done too much evaluation of the Wistar versus Sprague Dolly. I think at the end you suggested a disconnect between the fetal testosterone levels and the ultimate phenotypic manifestations that you might anticipate from low fetal testosterone levels. I did, which makes me sad because I've been pursuing the protestosterone for quite a while. Given that you actually observed quite a significant fall in fetal testosterone levels with the I&P, so do you have an explanation, an alternate explanation for DbP's effects, where that connection does seem to be present between lower fetal testosterone and phenotypic outcomes and the lack thereof with the I&P? Well, I have two hypotheses and this is what I present in my paper and I don't know if either one of them is right. I believe either you need more than 50% in addition of testosterone to manifest all of these effects. DbP, where you see most of these effects, actually causes 80% inhibition or more. It causes at 500 milligram per kilogram per day. DbP causes 97% inhibition of fetal testosterone, so that pretty much wipes it out, right? Or at least that's what Kevin Guido and Susan Borgoff studies both showed. So that's significant. That's almost complete loss of testosterone while the DbP is in the system, but DIMP doesn't seem capable of producing that much inhibition. Even in the New Hannas at all study, they went up to 1.5 grams per kilogram per day, so 1500 milligram per kilogram per day and they still were not able to reach that level of testosterone inhibition with DIMP. So, and I believe that's a kinetic reason. You just can't get that much DIMP into the system. So that's one possibility is that DIMP just can't produce enough testosterone inhibition to get you there, to the full complement of effects that DbP causes. Or the other option is that there is actually a mechanistic difference. Perhaps there's more than one target there with DbP. We may not have discovered it yet. Though it's clear that some of the testis effects that happen with DbP are not related to testosterone. I mean, some of the papers that you and Kevin Guido worked on together have shown that testosterone is not related to some of the histopathology endpoints and also cryptorchidism seems to require additional targets besides testosterone alone. So it's possible there's another target that DbP is hitting that DIMP is not. I don't know that I could, I can't say that for sure because I'm not even sure what that target is. If, yeah. I just make a comment that I think an alternative explanation would be that DIMP is not hitting testosterone in what we call the masculinization programming window. That's up to E18.5 and it's only in that period that you can, that suppression of testosterone will lead to overt male reproductive disorders. If you treat beyond that time and suppress testosterone at the end, there's absolutely no measurable consequence. And AGD is a readout of that early effect, not of the late effect. All right, but we did treat from gestation day 12. Oh yeah, so as what I'm saying is that in the data that I'll show in my talk, then the effects of even of DbP are relatively modest in that critical window. They're much more profound afterwards. And there's an explanation for that that I'll also show. So I think that your DIMP, based on its lower overall potency, would probably be having a negligible effect in that programming window, but in a more pronounced effect later on. Okay, I guess I'm not entirely sure what you're saying, because I mean, we did measure out further. I mean, we looked out to PND 49. No, I'm saying that the effect is when you when you affect testosterone in the fetal testis, the critical period is up to E18.5. After that, quite honestly, it doesn't matter what suppression you get after that. The only thing you're going to have an impact on is something like satony cell proliferation or penis growth. Okay, but if the if the DIMP is present earlier in gestation, why wouldn't it be affecting testosterone? Well, I'll show you why in my talk. Okay, I'm happy to look at that. Chris. So there's there are a lot of endpoints that you looked at. So did you look at sort of thinking of an individual pop as sort of a single pop and look at sort of the combination of endpoints so that a pop would be designated as having an effect if one or more, you know, taking into account the sort of syndrome approach? That's what we did. That was a global approach that I described at the very end. I didn't get too far into it, but yes, we did do that. Yeah. So and when we did that, a DVP was statistically significant from control. In fact, our P value is like point zero zero zero zero one. So it was like three zeros and a one. I'm not sure how many zeros I said, but it was very significant that but DIMP was not statistically significant at any of the doses. And that was that was assuming any effect meant you were positive. Kim. Just to follow up on Richard's comments. So you may have treated on GD 16, 17, 18, but you actually didn't measure. We didn't measure tests on 16, 17 or 18. And you know, that's the window in which the phenotypic endpoints are determined. We did. Okay. So we didn't measure testosterone on GD. I mean, okay. So we treated from GD 12 to 19. We measured testosterone on 19, but we didn't measure testosterone earlier than that. We did actually measure some blood levels in the fetus and dam on GD 16. I can tell you the DIMP or the MIMP was at equal levels in the fetal blood on that day than it was on 19. So earlier, the MIMP and all of the other metabolites were present in the fetus, but I did not measure the testosterone that early. So I hope Richard's talk on will will elucidate the issue with the testosterone production. But we early on agreed in chat that we would see fetal reduction of testosterone production as an important endpoint to consider. So having a look at your talk, you kind of confirm the study by Hannes that DIMP significantly reduces fetal testosterone production. And you also confirm Hanna's work that you would regard DIMP about two to three times less potent. That respect. I think from our studies. So our in vitro studies with DPP, which are our history of DPP is quite a sense of and our current studies with DIMP, I would call it probably about four, but three or four fold. So pretty close. 3.8. If you look on a, if you look on 3.8, okay. But that's if you look on an external dose basis. And so I mean, I didn't talk about that much here, but I think, you know, it's really important to consider metabolism here. And there is an extensive kinetic database for DPP in the rat in the rat fetus. And there's, you know, this new PK study with DIMP. And so I think it would be really wise to actually look at fetal blood levels for the active metabolites when you're going to make this this comparison. There's actually a group at the Hamner right now that's working on that as part of the EPA star grant. So they're not including DIMP in their analysis right now, but they are looking at five phthalate metabolites at comparing the pharmacokinetics using PBPK modeling and of the five different metabolites. And then also looking at potency of testosterone inhibition using in vivo in vitro assays and, you know, developing a risk assessment based on internal dose and pharmacodynamic potency. So I don't know if you're interested in hearing from them, but that's my fellow group at the Hamner. I still try to distill out the important messages from your talk. So you have no well regarding testicular testosterone reduction of 50 milligrams per kilogram body weight per day. Yes. And you have a no well regarding germ cells, morphological changes also at 50 milligrams per kilogram body weight per day. And you say that those two end points seem to be triggered by different mechanisms mechanisms. That's possible. Me saying that is coming from work that has been performed previously with Diabeal Phthalate from Kevin Guido and Kimbuckle Hyde was on that paper. So as far as that difference in mechanism goes, he might actually be the person to talk to. So he's not. Yeah, he's the multinucleated, going aside, germ cell expert. So Kimbuckle Hyde will elucidate the issue with the germ cells and Richard Sharp will elucidate the issue with the field testosterone. So I'm looking forward to it. But again, you confirmed the nobles and the effects that have previously been observed for DINP in your study. I believe that we established a no well that had not previously been established. So I mean, we measured at lower doses that have been measured. Nobody else has measured that low. Everybody seems to think you can start at 750 milligrams per kilogram per day. I disagree. I think you need to go lower so that you can find out where your true no well is. And also 750 milligrams per kilogram per day is not human relevant. It's just not. That's a hugely high dose. No well has been 50. Our no well is 50. I can absolutely stand on that. Yes. No more questions. Thank you for your presentation. Thank you. Thanks for letting me come. Okay. Kim, are you ready? Well, good morning, everybody. It's a pleasure to be here. I'm very happy to be interrupted at any time during my presentation. And, you know, I noticed in Becky's presentation, no one asked any questions during the talk, but I hope if there's any points from the slides that you have questions about that, you'll stop me so I can provide clarification or at the time where we can go back and look at those points. I'm really going to speak about exposure during the fetal period. I'm, you know, I've looked at both Lydic cell endpoints and at germ cell endpoints. And, you know, one of the things that we've looked at is human xenotransplants to try to look across species for effects. And the focus of that has been on Lydic cell endpoints and also be speaking to multi-nucleated germ cell endpoints. But this is all in the context of fetal exposure. You don't need this background for your information. You're all aware of this. But just for the general audience, you know, there have been increases, apparent increases in the rates of some male reproductive tract endpoints over time over the last 50 years. Crypto-Orchidism, hypospatias, sperm quality, testis germ cell cancer. And that those changes have been in the context brought forward by Richard Sharp and others of changes in the vulnerable period of development in the fetal testis. So, if we look at the cell types involved in this period, we have sertoli cells, germ cells, and Lydic cells as potential targets. Altered Lydic cell function, lowering testosterone and INSL3 production can contribute to the formation of cryptorchidism and hypospatias. And altered sertoli cell function can lead to impaired germ cell development and then the manifestations of altered sperm quality and testis cancer. And this has been contextualized in the context of a syndrome known as testicular dysgenesis syndrome that encompasses a spectrum of disorders that are interrelated and may share a common origin in developmental exposure. So that's all I'm sure familiar to you. And phthalates have come to the fore in this regard because they recapitulate some of the changes that are seen in the human population in this regard in rodent models. And there's significant exposure to phthalates, some of which happens in critically ill neonates, or at least has historically that, you know, the phthalates are being removed from much of those plastics at this time, but there's still significant exposure there. So that is in general background. And I'm going to speak primarily in the first part of the talk to the Lydic cell effects and then I will turn in the second part of the talk to germ cell effects. So if we look at cross species, so you'll see on the bottom here we're looking at rat effects, mouse effects, and human effects in the context of targeting seminephrous cords and the Lydic cells. And so the seminephrous cord effects that we measure are multinucleated germ cell induction and changes in seminephrous cord diameter. The Lydic cell effects that we measure are changes in steroidogenic genes that then manifest as alterations in testosterone levels, both within the testis or secreted by Lydic cells. And then endpoints dependent on testicular testosterone levels in the fetus. And those endpoints would be antigenital distance, hypospeedius, cryptocutism, Lydic cell hyperplasia, and nipple areola retention. So in the rat, with phthalate exposure, you get all of these effects as you're very aware. In the mouse, you do get the effects on the seminephrous cord. So you have induction in the mouse of multinucleated germ cells and increased seminephrous cord diameter. But you do not get the effects resulting from that you see in the rat due to Lydic cell suppression of steroidogenic gene expression and testosterone secretion effects. So you don't see this in the mouse. So the question for us and why we got involved in the work I'm going to tell you about, which is transplanting human fetal testis, is to try to resolve this species discordance that we see between the rat and the mouse. And I actually got involved in the mouse work during a sebatic leave, which I think was in 2003. So this goes back a while and started to do the mouse work because I wanted to be able to use the genetic manipulability of the mouse to ask questions about targeting of the phthalates in the fetal testis. And particularly I was interested in the Lydic cell effects. And then it was actually quite enormously frustrating that the mouse did not turn out to behave the same way as the rat. So that one could sort out some of this targeting information using transgenics and knockouts in the mouse. But given that species discordance, we started to ask the question, what happens in humans? Are human, is the human fetal testis response more like that of the rat or the mouse? And the approaches that one can take to answering that question are epidemiologic approaches. There have been, as far as I know, three papers now that have looked at exposure in the moms in humans and then outcomes in newborns in terms of predominantly in a genital distance as a measurement of effect. This turns out, I think, in my mind, at least to be a very difficult kind of study to do from an epidemiologic point of view. These studies have been small numbers of individuals. You're making single point measurements during a critical time in the mom distant from the effects on the fetus and measuring mom exposures but not fetal exposures. So these are, in my mind, fairly exploratory, the epidemiologic data so far that's been available to look at this question. In vitro cultures have been problematic in this area. And I say that predominantly because in the rat, which is our most robust, robust model, the rat in vitro cultures of lytic cells have not, both as organ cultures and as single cell cultures, have been really variable in their ability to report the phthalate effect in terms of a suppression of steroidogenic genes and testosterone secretion. Some of that problem has to do with primary cells when placed in culture even in the context of the organ begin to undergo a change in their function. That's actually relatively rapid. So by the time 24 or 48 hours after transfer into culture media, you've actually lost a lot of the differentiated function in those cells. And cell lines tend to be not responsive in the same way as in vivo primary cells. So in vitro cultures have been problematic given the positive control systems that we've been able to look at. So that led us to look at xenografting as an alternative approach. And the model that we've used, you know, we first have validated our xenografting approach using rat and mouse xenotransplants and then moved to looking at human xenotransplants. So I'm going to tell you about those experiments. What we have done for the rat and mouse is to take, this is a uterus full of fetuses from a rat or mouse. So we take GD16 rat fetus testis, transplant those into a host. And we do our transplants into the renal subcapsular space. And we've chosen that space because it's highly vascularized. It takes really quite well. The fetal testes are easy to find there. They don't move around. And we then expose to a phthalate. And we've been using DBP or control and then collect the fetal testes. And as you'll see from the data I present, we've been doing time course exposures. So we do the transplant. I'll show you a slide in a minute. Let them go and then look over time at what the changes are. One of our first questions was, you know, is this difference between rat and mouse? So rat is sensitive in terms of lighting cell suppressive effects and mouse is resistant. Is this an intrinsic property of the testis itself? Or is it a host species property? So to address that question, we've taken fetal testes from rat and mouse and transplanted them both into rat immunodeficient rat hosts and immunodeficient mouse hosts. All right. So this is our first set of experiments. So after we, you know, sort of address this question, I'll show you the data on that. And it, you know, our data suggests that in fact the response is intrinsic to the testis and not host species dependent, which makes sense. Then we went forward and looked at the human response. So here's our model. GD16 rat testes xenografted into the subcapsular renal space in either a rat immunodeficient host or mouse immunodeficient host. Let them go for 24 hours and then exposed to dibutal phthalate over a range of doses over a period of three days, harvesting six hours after each exposure. So our time course experiments that I'll show you are by collecting testes six hours after an exposure using this very acute transplant model. And so we initially did that by transplanting fetal rat testes and I'll show you the data on that and then the data on transplanting fetal mouse and fetal human testes. You can't see this very well because of the lights, unfortunately. But here is kidney in the host with the capsule and here is the fetal testis transplanted into that space. And, you know, actually these transplants look remarkably good. These arrows are pointing to multinucleated germ cells that are induced in the dibutal phthalate exposed hosts and transplants as compared to the control. And here's the kind of data that I'll be showing you for each of these species in terms of the transplant effects. So in this panel, so treated as always in the light blue bars, the control is in the gray bars and this is multinucleated germ cells per total germ cells present. So we use the total germ cell denominator just to provide a base of reference. And we're looking here at two doses comparing 250 milligrams per kilogram per day dibutal phthalate compared to control in these fetal transplants. We've done this over a time course and we also are looking in this panel at steroid eugenic gene expression and it's a set of genes, CIP11A, 17, SCARB, STAR, and INSL3. And you can see with increasing dose, so we have a dose response here of 250 and 500 milligrams per kilogram per day that there's an effect, what becomes a consistent effect on steroid eugenic gene expression at the 500 milligrams per kilogram per day exposure level. We also have testosterone secretion data, so this is then removing the implant, putting in into short term three hour culture and measuring the level of testosterone secreted by those implants after exposure in the host setting and measuring what the testosterone production is. So this set of data is consistent in our transplant model with what one would expect to see in the intact in vivo setting in the rat. We have an induction of multi-nucleated germ cells, we have an effect on steroid eugenic gene expression, and we have an effect on testosterone secretion in this rat setting. All right, so in the mouse we have the same histopathologic picture, again this is in kidney, this is in the rat host. We have induction of multi-nucleated germ cells in the, that helps, thanks Mike. And then we have this similar kind of data, so again looking at 250 milligrams per kilogram per day and 500, we have multi-nucleated germ cell induction at both doses and essentially no effect on steroid eugenic gene expression at these two doses in the mouse and no effect on testosterone secretion in the testosterone production assay. So the mouse data is also consistent in this transplant model with the in vivo fetal testis response in the mouse. So you know this was all into a rat host and you know one could think of the rat as being the permissive species in essence because the rat responds. So we also checked this effect by going into mouse hosts. This model took a lot of work to develop, you know we spent a lot of time optimizing the timing, graph location, the host hormone environment, so we're actually going into intact hosts. We looked at castrate hosts, we looked at hyperphysectomized and castrate hosts and in this short-term acute model we did not see at least at a crude level a difference between those host environments. So what we are looking for was effects on the seminefused cords and leitig cells and you know we repeated the studies that I just showed you with the mouse host and it gave the same in vivo consistent responses for the rat fetal testis and mouse fetal testis as I showed you for the rat hosts. So the response in our hands is an intrinsic response to the fetal testis itself. It's not a host dependent response and you know we consider this then our proof of principle we have a sensitive species response and a resistant species response for this technique and proceeded from that point to start looking at human fetal testis and their responses to phthalate exposure. So this work has been ongoing for several years in terms of collecting samples. What I show on this table is the range of gestational age of the samples that we have so it goes from 10 weeks gestational age up to almost 24 weeks gestational age. One of the issues to keep in mind is the postmortem interval that is the time between delivery of this material and our getting it and implanting it. We found no relationship between this period of time and the results but it's an important thing to keep in mind and we've minimized or we've kept that length of time to less than 36 hours for our work. The other important aspect of this to keep in mind is these are spontaneous abortions only. So we have received material only from fetal losses that would happen anyhow. So there's typically an underlying cause here. In many of these cases there's an infectious process going on that leads to the spontaneous loss. There's something going on in the background some of which are noted here so there may be genetic anomalies in these fetuses or other aspects of physiology or infection that have led to this outcome. We tried to take a look at whether these other effects might be influencing the results that we have seen and we haven't been able to see any consistent effects of these other ongoing processes on our outcomes. This is a human fetal testis transplant I believe from an 18-week gestation fetus and we did labeling with bromideoxyuridine in this sample for a three-day exposure or a three-day time period and a three-day BRDU exposure and what's remarkable really is the amount of proliferation that is taking place in these fetal testis after xenotransplantation and it's in all the cell types. So you can see germ cells, sertoli cells, and lighting cells are proliferating in this setting. This is a CD31 stain for human vasculature in the transplants and as you can see there are vascular elements in the human fetal testis and in fact moving into the renal parenchymal space from the transplant and these seem to be patent and viable vascular spaces. So based on this data these are proliferative, the human fetal testis transplants, proliferative and vascularized at the point that we're looking at. With the exposure, again we're comparing dibutal phthalate exposures and control, we do get induction of multinucleated germ cells and the data is consistent across day and across dose in terms of multinucleated germ cell induction. So again it's the same legends here. The blue is that dibutal phthalate exposed compared to control a dose response of 100, 250, and 500 milligrams per kilogram dibutal phthalate and one day, two days, and three days and this is an initial 24-hour period so we do the transplant, we wait 24 hours, then our one-day treatment is an exposure followed by a removal of that fetal testis six hours later. So this is 30 hours after initial transplant, 30 plus 24 for two days, etc. And we've looked at the steroidogenic gene expression and see essentially no effect on the light excel, steroidogenic gene expression with this exposure again across doses and this is after two days of exposure. We have not looked at testosterone production in this model so that's a limitation of the results that we have to this point. Okay so summing up this set of data, our human fetal testis response is similar to the mouse and we see a consistent effect across species on multi-nucleated germ cells and I'm going to turn to speaking to that in a moment but a species discordant effect on light excels, steroidogenic gene expression and the human fetal testis response is similar to the mouse response in being resistant to effects on the light excels. Is there a reason why you haven't done the testosterone production? So we're you know that we we tried looking just at implant testosterone levels not the production in culture and those levels were very low first of all but also all over the place it was just extremely noisy and we just haven't had the samples to try other approaches and we plan to make use of a technique that Richard has used quite successfully and published on in terms of looking at physiologic endpoints like seminal vesicle weights in the host as a readout for integrated testosterone production and I'm hopeful that that will work we just haven't gotten to that point yet. So the implications of you know our results are that you know the human effect would be similar to what you would anticipate in the mouse and in fact then you would not anticipate having those testicular dysgenesis syndrome consequences that are testosterone dependent such as hypospatial syncryptorocatism appearing in humans. I want to sort of step away from this for a second and just look at the comparison of what's going on in rodents versus humans in terms of windows of susceptibility and secretion of fetal testosterone. Before you Yes. You want me to go back? Yeah. Okay. One of the things that strikes me is that we don't know in any of these systems is how much DBP is getting in to the testes in all three of these situations. No that's so that's so we don't know that in human although we're looking at that right now but we have published in the mouse work and this would be Kevin Gatto's paper the levels of DBP that are in mouse fetal testes and they're similar to what get is present in rat fetal testes. So we know that the exposure at the level of the fetal testes is actually the same in the mouse and the rat. We just don't get the response in the light excel in the mouse that you see in the rat. That is no. That is no. That is published. That's published. So that's Kevin Gatto's paper 2007. So that pharmacokinetic that is verified pharmacokinetic exposure levels are similar in the two species and what for the transplant studies it's all the same species right. So the host is you know it's all rat host. Is that that question? Yeah. So you consider DBP as the active species? Excuse me? So you consider DBP as the active species or? No. No. Mbp. Mbp. Yeah. And so. And that was measured as well. So we would say the kinetics of Mbp would be the important point to consider. Absolutely. Mbp in and Mbp out. Yeah. Well Mbp at the level for these effects at the level of the fetal testes. So you are in in the column for the humans investigating the effects in human testicles with rat or mouse metabolism and metabolism kinetics. Correct. Would you consider human and rat metabolism similar? In terms of the capacity to generate mono ester from diester exposures? In this term. Yeah. I think there probably are differences but the underlying esterase activities are present in both species. I'm not sure that we quantitatively know the difference but I would say at very low levels of exposure there's going to be sufficient esterase around to convert diester to mono ester to a large degree in both species. What about the kinetics? Because you said both in and out is important. Would you consider kinetics similar to in rats and humans? Would you consider rats being faster or slower than humans? In terms of secretion? Yes. I think we're in the same ballpark in both species. Based on what data? Because now we're also talking about placental transfer. Right. No, I understand. Right. So I don't know that we know the placental kinetics in humans. We know that data well I'd say in the rodents but we don't know that in the humans. So it's more of a extrapolation in the human center. Just one more point though on this exposure situation. The fact that we have multi-nucleated germ cell induction in all the species in all the models that we've looked at means that exposure is getting there. Because the phthalate has to be getting there to be producing that effect. So there is an exposure that is happening at least sufficient to induce multi-nucleated germ cells. The reason I asked my question is that that could be caused by a very low exposure. That's not true in the rat. So we know in detail what the dose response for multi-nucleated germ cell induction is in the rat. And in the rat the low L effect for multi-nucleated germ cells is actually quite similar to the low L for lighted cell testosterone effects. They are essentially identical. So that you know that just you know that's a nice sort of positive control for exposure. Another question would you consider the C max or the area under the curve more important? The data that's so the hamner has done an exposure with dbp, bolus, gavage dosing and food and interestingly and I think surprisingly to me the dose response effects were really quite similar for the endpoints that were being measured. And so that would suggest that area under the curve exposure is actually more significant or more important than C max because C max is going to be much lower in a feed exposure setting. All right. So we're going to go to this diagram and just speak generally about some of the vulnerable windows and differences between species. So Richard Sharp raised the point that the you know the development of the male reproductive tract is actually in the rodent happening in the first part of this window when testosterone secretion is actually being initiated in the fetal testis. So it's early on in this wave of testosterone production. The responsivity to phthalate exposure actually happens throughout. But the determination of end organ physiology like what can be disrupted to give rise to cryptorchidism and hypospadius is actually relatively early on. And that's true in the human as well. So the events that are happening that are important are occurring in gestation week and you know 8 through 12 or so in the human but the actual peaks of testosterone secretion are happening out at gestational weeks 16 and 18. In humans this developmental window is much much broader and the secretion window of testosterone is much much broader. There's a lot of parallelism in the events that are taking place across species but the compression in terms of time is you know remarkably contracted in the rodent. There's also in humans an interesting postnatal early postnatal period and I think Richard is going to speak to this in more detail when there's sort of what's called a mini puberty in humans which is in a window that happens you know within the first six months of life. And it's sort of been recognized to encompass testis enlargement, a mini sort of spermatogenesis spurt that takes place at that time including some increases in testosterone levels during that window. And that may have important physiologic consequences during that time and actually relatively little is known about that time period from the perspective of underlying biology and physiology. So you know what I'm going to turn to now is work that we have published recently looking at the multinucleated germ cell effects resulting from phthalate exposure. Yes please. Before we go there to a question relating to the windows which you explain. In the rat as far as I know the male programming window starts at around 15 and a half embryonal day. Is that right Richard? They're about. If I translate that into the human we would talk as you said weeks 8 to 12. Now your earliest material from spontaneous abortions is from week 10. Right. So you would be probably out of that window or so I have two questions. First question is how do you judge your material with relation in relation to the window of sensitivity in the human and secondly would you guess that if you work with material older say closer to 23 weeks outside that window that you can draw any meaningful conclusions from these experiments. All right so I'll first speak to what we know about in the rat. So in the rat the phthalate effect on steroidogenesis and testosterone production occurs throughout the time period of fetal testosterone production. So there is sensitivity in the lighting cell to phthalate induced suppression of testosterone production and steroidogenic gene expression throughout this entire window. Now the you know the important sort of end organ event steps are early in this window but the phthalate effect is constant throughout that window or at least persistent throughout that window. We don't know of course in human whether that's true whether there's a difference in the sensitivity of the fetal lighting cells in human out here as opposed to early on. What we you know our average sample is about an 18 week gestation sample. We do have some early on samples I think 10 and 14 are earliest samples. We don't see any difference in the response dependent on age. There's not a sensitivity here that is lost it's you know there's a resistance throughout in terms of lighting cell effects throughout that window. I can't really say more than that I know Richard has worked with much earlier cells in fetal testes than us. Well not strictly we have done some first trimester but nothing in any great detail that we could draw conclusions on. But I think I mean your point is extremely important but if we take fetal testis explants as well as a readout then there have been fairly detailed studies done by Rene Haber and also by Ben Arzegu using human fetal test is first trimester explants and there's no effects on sterologenesis. So at least there's a consistency but I think not xenografts but Rene Haber is doing the xenografts with first trimester and he's using the same protocol that we're using. I don't know what results he's found but I would be surprised if he finds effects but science constantly surprises us. So I think there is always a possibility that it could be different but as I'm going to show then if the mechanism via which DPP affects the rat if that's the vulnerable mechanism across species then I don't see why if you don't see an effect late on that you would expect to see an effect early on but you know there are presumptions in that. Can I can I summarize can I check whether I understand this correctly okay we agree that the male programming window in the rat is earlier in the earlier part there it corresponds to about let's say week 8 to 12 thereabouts in the human. So we have this question or this question mark but your argumentation to say the xenograft material from the human fetal testes that comes from later week that these experiments are still valid or meaningful because if we did the same in the rat we would see salate effects on testosterone production throughout the male programming window and beyond i.e. at later stages is that correct correct okay thanks but again we have you didn't you didn't look after these effects could you again please show the slide where you have shown the results the other yep here um what is the insulin like three here that's a light excel um factor that is important at cryptorchidism are there effects or no effect no so this is the 100 milligram per kilogram dose and you know I think that's just an anomalous finding there in our PCR results so this is 100 250 and 500 so I'd say that's noise in our system just um Kim in terms of the I don't know you had maybe 20 or 25 different samples right us just how did how did you so each one was an experiment in and of itself a transplant right a xenograft how did you then combine them to look at whether there were or were not differences based on gestational week I mean did you say all those before 12 weeks versus those you know 18 to 22 or so what we can't look at them really each individual right so for each of our human fetal test is we took the fetal test is diced it into you know millimeter cube kinds of fragments put them into control rats and put them into tally treated rat hosts and you know then compared the results in treat with treatment or not and we tried to do linear regressions with age we tried to do linear regressions with postmortem interval for anything else that we could think of that might be significant determinants of response and we didn't see any tendencies across the one to one comparison what kind of the control to so we did like made ratios of any parameters that we thought might be of interest in the control tried to say you know all those taken within the we tried to clump them we tried to clump as well and we never saw anything that was consistent across any of the parameters that we thought might be informative like age differences by gestational yeah let me just go back one more and then you can see here is is that right I'm having trouble seeing those numbers though okay could you just point out how many numbers are on the timing that's less than say 12 so we've got 10 and 14 13.4 16 and then sort of goes up from there so it's so okay I couldn't even tell I was it was ranked right it is sort of ranked except for 10 and 14 so there's only one value less than 12 yeah and then we you know our average is 18 actually 18 weeks so the window is an important question then it is potentially an important question I if you believe the rat argument the response remains throughout gestation in the rat to thallid exposure does a change in magnitude at all um yes yeah yeah it you know I think Richard is saying yes it does it gets more pronounced in later gestation the levels of testosterone actually go up quite significantly with time the magnitude of the change in testosterone the magnitude of difference between the control and the thallid exposed gets much larger later in gestation in fact at e15.5 we find no significant inhibition at all it's only towards the end of the programming window that you begin to get effects and there's a logical reason for that mechanistically um and I think that explains why dbp actually or other thalates have relatively modest effects in terms of causing um things like hyperspatias because they don't cause much suppression of testosterone in the programming window they cause much more pronounced suppression later on again I'm looking forward to this talk um how did you make sure that you measured the gene expression and the and the hormone like effects in the right window when did you measure it because we've seen that two hour post those there are effects for dinp and 24 hour post those there are no effects anymore how did you ensure that you measured in the right window okay so we only measured one time point after exposure and that was six hours and that is based on the pharmacokinetics that we know for dbp and the rat right so you know something about the kinetics yes so that would be in between the two hours and the 24 hours of the clueless study excuse me that would be in between the two right so for each of the days so we measure you know we've done time courses day one day two day three after each daily exposure we sampled six hours later consistently six hours later but you didn't test if it would there would be an effect observable let's say two hours or 24 hours you just measured at six just measured at six and that's based on what we know about the rat response to dbp for steroidogenic gene expression from studies that kevin guide was done so you would assume that the effect at the six hour post those in the clueless study would be bigger than well she was looking at two hours and 24 hours those what was her endpoint at two hours i'm trying to remember what is testosterone so we were looking we're looking at gene expression you know which is a different endpoint which is presumably earlier as an endpoint than testosterone so the gene expression is earlier should be so if she actually that's actually complicated but yes so there is an immediate testosterone suppression effect after phthalate exposure that is probably unrelated to gene expression decreases but then gene expression typically also declines and you get a later much more significant decrease in testosterone levels but your rationale behind choosing the sampling time point of six hours post those was based on kinetics kinetics in the rat yep okay so moving on to we're now going to look at multi-nucleated germ cells that question and i'm going to tell you about and this is published work tell you about a set of experiments that we've done to look specifically at this question because this is the this is the endpoint of phthalate effect that is conserved appears to be conserved across species so this is the publication reference here saffroni at all in journal of andrology and this this work is has taken place in the p53 null mouse all of the work that i'm going to talk about here um and the rationale for using the p53 null mouse is we know a lot about the role of p53 as an apoptosis inducing gene in mouse for germ cells you know so we've looked at that question for a long period of time and you know so our rationale in using the p50 so using the p53 null mouse was we know the mouse is responsive in terms of the induction of multi-nucleated germ cells so we know we can induce multi-nucleated germ cells in this model what we wanted to do was remove any drivers for the apoptosis of those germ cells and then see what happened so when we started this project we're thinking okay we're going to get rid of this very potent apoptosis inducing germ cell gene and then we'll see what the life experience of those multi-nucleated germ cells is in the absence of their dying after their induction which is what typically happens in the rat and the mouse you know these multi-nucleated germ cells they all go away they all die by apoptosis by a couple weeks postnatally and so the concept for this study was let's get rid of this you know important and powerful driver of apoptosis and see what the life history of these cells is so the model was to take these mice so pregnant dams exposed from gestational day 12 through birth in the p53 null we actually it's a heterozygous dam and so there are pups that are delivered from these moms that are both homozygous deficient in p53 and heterozygous in p53 and then we looked at various time points gd 19 p and d 1 4 7 10 and then later on in adulthood for the presence of these multi-nucleated germ cells the numbers of them and their life history so at gd 19 we induce we see the presence of and we're using the exposure of 500 milligrams per kilogram per day double phthalate we see the induction of these multi-nucleated germ cells and this is what the you know the kinetics of the presence of these cells looks like so again we're comparing p53 null to p53 heterozygous pups and one you know there are a couple interesting points here one is if you look at the control so this this is a p53 null pup that is has not seen any phthalate it has actually quite a higher incidence of multi-nucleated germ cells than a than the heterozygous control so being null in p53 means that you have more multi-nucleated germ cells and that might make some sense if these cells are predisposed to die if you remove that driver of apoptosis then they can persist and you get more of them whether you have phthalate on board or not in the presence of phthalate so that's these filled markers you get a you know another quite significant induction in the numbers of multi-nucleated germ cells so in the null that's the circles here you know you get a lot more multi-nucleated germ cells with phthalate treatment than without but overall the null environment is conducive to the presence of these multi-nucleated germ cells so one of the points that I think is an important take home message here is that multi-nucleated germ cells normally occur you see them they happen at a certain frequency but as you can see from you know the time course here all of these or at least a majority of these multi-nucleated germ cells are eliminated fairly rapidly after birth so in the mouse they're gone you know to almost a full extent by day 10 and here's some of the quantitative data on them one you know one thing that we observed in this model however was that there was a persistence of very abnormal and large germ cells that were multi-nucleated into adulthood in the p53 null setting so only animals that were p53 null that had also received phthalate manifested the presence of these you know relatively bizarre very large cells into adulthood so in fact the experiment worked in that sense that we removed this driver of apoptosis and you know we don't have formal proof that these are in fact the same germ cells that were induced in the fetus that are then persisting into later life but there you don't see these kinds of cells normally in the testis they're very abnormal and our presumption is that in fact these are cells induced by phthalate exposure in p53 null mice that are persistent and this is the kinetics of the persistence of those cells and they're very few of these cells they're there essentially they're all gone and these by the time these mice you know are less than a year of age so the window of the survival of these abnormal cells seems to have been extended by the fact that we've removed p53 as a driver of apoptosis but it's not infinitely extended and you know the suggestion is by these kinetics that in fact these cells do not divide they were induced in the fetus they persist they're allowed to live because you don't have p53 around anymore but they under they do undergo death and apoptosis at some rate and ultimately are all eliminated so we used markers that are known to be present on gonocytes oct 3 4 and placental alkaline phosphatase so oct 3 4 here this is an gd 15 fetal testis where the gonocytes are labeled with this marker you don't see that on the germ cells at gd 19 that are multi-nucleated and you don't see it on these very abnormal germ cells that are present in the testis later in life in these p53 null mice uh and placental alkaline phosphatase this is placenta syncytrotrophoblastic staining you don't see that as a marker gonocyte marker that's not present on these cells at gd 19 and it's not present on the these abnormal adult cells either and so as shown previously these cells do not appear to be gonocytes and that is at least important us conceptually because when we think of testicular germ cell cancer we associate that with the presence of carcinoma in situ cells which are seen as precursor cells for the production of testis germ cell cancer and carcinoma in situ cells carry with them gonocyte markers the presence of gonocyte markers and uh so carcinoma in situ cells are abnormal gonocytes the hypothesis then for testis germ cell cancer in humans is that there's some abnormal event during fetal development that sets aside some cells uh that sort of freezes them in a gonocyte stage of development which then manifests later as carcinoma in situ cells they express these markers and then there's some other induction events that lead to the development of testis germ cell cancer in young adulthood in these individuals and the suggestion from you know many studies richard sharps and this work as well is that in fact these multi-nucleated germ cells are not gonocyte in origin do not then share with carcinoma in situ cells the kinds of markers that one would anticipate um and even in the setting in which you're removing a major driver of apoptosis these cells are not proliferative and don't persist uh they persist for a while but they don't persist or proliferate um later in life so i'd be happy to answer any questions about about this um we also because we do xenotransplants we also did xenotransplants of these p53 null phthalate exposed fetal testis from mice into wild type mouse hosts in order to allow them to even live longer um because we're trying to give them as much of an opportunity as possible to develop into germ cell cancers that was our concept and we never saw anything that would suggest that outcome uh in this model which you know one might imagine would be predisposed to this kind of outcome so that in the human transplant tissue the mngs do they have these you know histochemistry markers the oct three four um yeah we haven't actually done that work yet i think richard you have um sorry russ i missed the question so the the human um xenotransplants xenotransplants in terms of their immunohistochemistry is it so some of the germ cells will be expressing oct four and some won't age dependent and it will be age dependent the earlier ingestation the greater the proportion of oct four expressing cells for the human for the human but there will be some there throughout the whole of gestation they're still there till up to six months after birth so is is that different than what's occurring in the world completely different completely different so it's a big the process of germ cell differentiation is um is fundamentally different in certainly the human and the marmoset we don't know about other primates from rodents um it appears i was discussing this with with kim and uh and just saying it it just looks like it's haphazard in the human like it was just thrown together whereas it looks a sort of a Rolls Royce job in rodents okay um so i want to just give you uh just conclude with a word about where we're going with this uh what we'd like to do is separate out the compartments so the semaphase chord compartment from the interstitial compartment and then compare you know that's using laser capture and micro dissection compare a cross species to try to see what the differences are in what we're observing in rat versus mouse versus human um so just a little bit of general discussion about um these results so george box you said all models are wrong but some are useful he was a statistician so as far as i'm concerned all status statistical models are wrong um i would say for biology for our work sorry about that um all models are limited we need to know what the limitations are um they all generate information and one of the issues that you are grappling with is that in fact with the phthalate literature there's a huge huge literature in fact this is by far the best studied compound in terms of the effects that you're looking at of any that i know of and that massive information generates both clarity and noise um so and there's just lots of inputs now into as you well know into the phthalate literature so we have animal models that are traditional toxicology models knockout transgenic strains and xenotransplant models now in the mix of information information that you're dealing with in vitro models cell culture organ culture and human models epidemiology databases and exposure surveys and phthalates are part of you know every one of these and you know an enormous amount of data and i um wish you the best of luck in sorting through you know this this mound of somewhat discordant information i'd really like to thank my group i do want to let you know that i have received funding from the american chemistry council for some of the xenotransplant model development work so that money went into the creation of the model but it was specifically targeted to the animal development work and not to any of the human fetal transplant work and you know this work has been supported also widely by NIH and could not have taken place without the help of the folks at the hospital uh women and infants hospital so i thank you and i'd be happy to answer any more questions now or later right thank you kim some more questions or should we i think we should go on and do uh richards talk so we can then break for lunch and come back and have discussions afternoon that makes sense short break can i just ask one sort of a at an overview level i mean based on your final slide there i mean if you were in our position where we and i have been by the way i was on the previous chat panel so on phthalates where we had to make a decision um i mean how would you come down on the whole concept of a rat model for reference doses uh in a risk analysis um well it's interesting i i i think arts so if you only had that data that's of course what you do right you have to do it that way it's become much more nuanced with the information that we have so the certainty with which you feel comfortable relying on that data has i think changed significantly as we learn more and more so for me actually the exposure information becomes really the driver for the risk assessment that you are sort of conceptually doing here so you have to really put your wrap your mind around whatever exposure information you can get at target and interpret that um i think there i mean i am more and more impressed that there are really significant differences across species both in biology and in sensitivity and i say that not just based on this fetal test this work that we're doing but we're we've also now expanded um the number of tissues that we're looking at to other tissues and the responses that we're seeing in other sort of classic rodent models of effect are just very very different in the human developmental field exposure setting so i'm impressed that there are very very significant difference across species i'm sort of dancing around dancing your question uh directly um let me think about it and we can talk about it maybe at the end of of richards talk because that's a very important question but another question just from a from a you know statistical analysis perspective and which of course is a model with a you know yeah of course it's wrong yeah or maybe useful um the the question is you know when you say there is no effect i'm not seeing anything i mean i that was a lot of sort of the the points that i was taking from what you were talking about not knowing all the biology there so i mean can you can you put limitations on your own thinking on that i mean what are the uncertainty of that observation sense of what if you have the wrong window what if you have not a big enough sample size what if it's not as large of an effect and you're just missing it i mean i would imagine there's a list of those sorts of limitations oh i think there are lots of limitations um i so i think there are a lot tremendous limitations in the model that we're using you know so the host hormonal environment matters even though we're doing very you know acute short-term exposures and that's i think somewhat limits the impact impact of the host hormonal environment the timing in terms of you know where the fetal test is in terms of its own development probably matters all i can tell you from what we've done is that we i've been impressed consistently by the lack of response at the light axle to the phthalate exposure that you know consistently we just do not see a response in the human and that's so unlike the the germ cell response in terms of this multi nucleation effect we see that always and you know so there's a dramatic disconnect there which means the response to me looks so much more like the mouse um in the human but sorry the whole concept of the of the spontaneous abortion tissues not i mean i don't know much about those kinds of situations uh could they actually have an effect on the light excel development or responses or i mean you talked about potential uh so we yeah we've looked at that carefully and and haven't seen anything so the cases in which that you would imagine that might matter is if you had an an encephaly for example so you have no uh input uh you at least conceptually input of you know which isn't maybe very important anyhow but input of hypothaline pituitary gonadal axis in the fetus you know so that's not very important early on in this process anyhow but we have not seen anything related to maternal environment that resulted in the spontaneous abortion affecting this outcome you know so and um we looked do you think you have the numbers to see it yeah i mean you know so we're up in 26 different kinds of we're up in 26 samples and there's been a real consistency in terms of the lack of light excel response and there's been a consistency in terms of the semidefuse cord response but are you treating the um the spontaneous aborted um tissues as if it as if they're similar i mean could there be distinctions in those that you we tried grouping so we've tried grouping across genetic disorders that we know about that are present in these fetuses infection if we know it's there an encephaly or other sort of nervous system disorders we tried any kind of grouping and we don't see actually any difference in the outcomes that we're looking at even numerically yeah yeah how about with the postmortem time do you see no there's no there's no apparent effect of postmortem time and in the animal to animal model do you we have not done a different difference in postmortem time in the animal yeah we haven't done that for the animal transplants but you pointed out to us the significant noise you were observing yes there is significant noise in the system there's no doubt about that ken thank you for sharing very important information it is reassuring to see somebody working in an area that broadens our knowledge rather than makes it deeper but in a narrow cut so thank you for your contribution how important do you think it would be to expand your existing model of the transplantation with other phthalates so that we see the response of your system to phthalates of different potencies i think that's a great question um everything that i know about phthalates would suggest that the active phthalates act in concert act in the same way act at the same targets uh you know so what we know from the rodent models is that DEHP for example should have the same kind of response pattern you know we haven't looked at it in human but the anticipation is that it would be similar it's a formal question it's an important question you know that one look at something else but there's been such a consistency in terms of molecular response you know so Kevin Guido looked at the patterns of molecular response it's really quite persuasive in my mind can you offer a mechanistic explanation for these species differences which you observe so the rat on the one hand seems to be sensitive mouse and human not so um can you offer any mechanism that might explain that can we postpone that until after lunch let's take a great question by the way let's take a 10 minute break and then we'll come back and and have Richard's presentation then we'll have lunch and then we'll come back and answer that very important question Richard's going to tell us the answer