Emerging Roles and Strategies to Harness the Host Microbiome in Cancer
SESSION 7: Emerging Roles and Strategies to Harness the Host Microbiome in Cancer
Moderator: Karen Sfanos (Johns Hopkins University)
The Role of the Microbiome in Host Androgen Production and Therapeutic Resistance in CRPC
Karen Sfanos (Johns Hopkins University)
Gut Reactions: How the Gut Microbiome Controls Effectiveness of Anti-Breast Cancer Drug Tamoxifen
Elizabeth Bess (University of California, Irvine)
View the Transcript Below:
Emerging Roles and Strategies to Harness the Host Microbiome in Cancer
Karen Sfanos, MS, PhD [00:00:12] Good morning, everyone. I am super excited to have a microbiome session as part of the annual retreat this year. And I’m super excited to have my co-presenters, Dr. Andrea Alimonte and Dr. Elizabeth Bess presenting along with me. And today I will be speaking to you about some work that we’ve done regarding host androgen production and therapeutic resistance in CRPC. I’ve never been cool enough to have any relevant disclosures. All right, so I want to start with an acknowledgement slide to make sure that I mentioned that the work that I’ll be talking about today is supported in part by 2023 Challenge Award and to acknowledge my wonderful collaborators on this award, who I will be highlighting throughout my presentation, but just wanted to start by acknowledging their wonderful efforts. Alright, so our work in androgen production by gut microbiota actually started back in 2016 with the first challenge award that I ever received. But it enabled us to start collaborating with our wonderful clinical colleagues in the Department of Oncology and to start biobanking specimens for microbiome studies. And yes, that is exactly what it sounds like it is. So, what I’m showing you here is one of our earliest studies in this area. We’re in a small cohort of 30 patients, we asked the very straightforward question. Are there compositional differences in bacterial species in the gut microbiome in individuals with prostate cancer who are either not being treated with any medication shown in gray, who are being treated with a GNRH agonist or antagonist alone, shown in green, or that we’re being treated with oral antiandrogens, shown in orange. And what I hope you can appreciate is that in the individuals that are being treated with oral antiandrogen therapies, the profile of the gut microbiota are similar to each other, and then also somewhat different from the other two groups. And when you take this data of the bacterial species that are present in the gut of these individuals, and you run them through a program called PyCrust, which basically infers what the functional capability of these microbes are, we were really surprised to find that the top two hits for functional pathways that were present in the gut microbiota of individuals on oral antiandrogens were pathways involved in steroid biosynthesis and steroid hormone biosynthesis. Now I do have to admit at the time I actually didn’t realize that there were microbes in the gut involved in steroid biosynthesis. And that’s really led to a whole new area of research from my laboratory that started with establishing a collaboration with a wonderful colleague, Jason Ridlon at the University of Illinois. Whose entire research program is focused on discovering new bacterial enzymes that can convert glucocorticoids into androgens. And so hearkening back to Nima Sharifi’s talk, which was much more elegant than mine will ever be in talking about steroid biochemistry. The human steroid 17,20-desmolase, CYP17A1, is involved in much of human androgen biosynthesis. And I’m showing you a typical pathway here of the conversion of pregnenolone to DHEA with both hydroxylation and lyase functions of CYP17A1. Well, it turns out that bacteria also have a steroid 17,20-desmolase, which I will refer to from here forward as bacterial desmolase. And by the way, the term desmolase just means an enzyme that can break a carbon-carbon bond. And this is encoded in bacteria, but it des-A and des-B genes. So, Jason Ridlon showed a while back, demonstrated this in a keystone species of gut microbiota called clostridium scindens specifically in a strain called ATCC 35704. So bacterial desmolase actually functions quite differently from the human CYP17A1. And the pathways that he has described so far are the conversion of cortisol to 11 beta hydroxyandrostenedione, which I’ll refer to as 11 beta OHAD, and the conversion of prednisone to an androgenic compound called delta 1 androsterone, which I’ll refer to as Delta 1AT. And the important part for the purposes of this talk is that 11 beta OHAD and specifically its conversion to 11 oxyderivatives, as well as this as this androgen metabolite of prednisone, are actually potent AR agonists, either on par with or greater than that of testosterone and DHT. So, this is just what these reactions look like if you look at bacteria that are being grown in anaerobic culture. And if you take a culture of clostridium scindens ATCC-35704, which is desmolase positive, and you feed it either cortisol or prednisone, you can see by LC/MS/MS that these are converted to 11 beta OHAD and the delta 1AT metabolites respectively. If you take a strain of bacteria, for example, here B acidifaciens that does not have the desmolase genes, you do not see conversion to these androgenic metabolites. And just to further indicate that the bacterial desmolase does not function the same as CYP17A1, if you give C. Scindens pregnenolone or 17 alpha hydroxypregnenolone, you do not see conversion to, for example, DHEA. Importantly, we found that abiraterone acetate also does not inhibit bacterial desmolase, even though that is the inhibitor given clinically for CYP17A1 in humans. So, on your right-hand side, sorry, this is an experiment performed in my laboratory where you take C. scindens, you give cortisol or prednisone, and you see the 11 beta OHAD and delta-1AT metabolites, and that this is not inhibited by the addition of one micromolar of abiraterone acetate to the bacterial culture. The experiment on the other side was an experiment done in Jason Ridlon’s lab, where they took C. scindens and another species of bacteria called Propionimicrobium lymphophilum, which is actually a urinary bacterium that also contains this desmolase gene. They show that you can, if you give these bacteria cortisol, they convert to the 11 beta OHAD, and this is not inhibited by the addition of 50 micromolar of abiraterone to the culture. So, with everything that I’ve told you thus far, this is our working model of how we think that gut androgen production might lead to therapeutic resistance, specifically in the setting of treatment with abiraterone acetate. So, in a setting where you’ve shut down testicular production of testosterone with a GNRH agonist and antagonist. Combined with abiraterone acetate, which is a CYP17A1 inhibitor that blocks adrenal androgen synthesis, there is still the capability of the bacteria in the gut to still convert certain precursors into androgenic compounds. The reactions that I’ve shown you are the conversion of cortisol to 11 beta OHAD and the conversion of prednisone, which is given as the replacement glucocorticoid of choice along with abiraterone acetate into this androgenic compound called delta-1AT. So, we think that our hypothesis is that these androgens that are produced in the gut can be absorbed and circulate and contribute to therapeutic resistance. So again, with that challenge award starting in 2016, we’ve put a great deal of effort into biobanking samples from individuals undergoing treatment with abiraterone acetate longitudinally. And we’ve created sample cohorts where we have samples taken from when patients were stable on abiraterone acetate treatment, meaning their PSA levels were stable, versus at times when individuals were progressing on treatment, so their blood PSA levels were rising. And if we do metagenomic sequencing of the gut microbiota in these samples, I’m just giving you three examples here. We do find certain species, strains of bacteria that are differentially more abundant in the individuals that are progressing on treatment with abiraterone. Including this one species of clostridium scindens called VE202-14. I want to make the point though that when interpreting data such as this and what the next steps are, this is really hard because we don’t think that it’s just one or a few species or even strains of bacteria that are capable of producing androgens in the gut. And in fact, the carriage of the enzymes that can do these androgen conversions are specific down to even the strain level. And we don’t know the full functional capacity of bacteria in the gut to produce androgens, and we certainly don’t know all of the enzymatic pathways that are present and the genes responsible for this. So recently we’ve taken an alternative approach to trying to study this in actually measuring the levels of the metabolites. And what I’m showing you here are measurements of certain metabolites by LC/MS/MS in fecal samples. And this isn’t a cohort, again, of individuals being treated with abiraterone, where in blue are control samples of individuals with prostate cancer who are not undergoing any treatment. In green are those being treated with abiraterone acetate whose blood PSA levels are stable, they’re responding to treatment. And in red are samples taken during a time when individuals are progressing on treatment with abiraterone. So, in the control samples, I hope you can appreciate that we can actually measure levels of androgenic metabolites such as testosterone and DHT. And I was originally a little bit surprised by this, but I’ve since learned that we would actually expect that due in part to enterohepatic circulation. What I hope you can also appreciate is that the samples shown in green when individuals are responding to therapy with abiraterone acetate, with the exception of prednisone, which actually accumulates during abiraterone acetate treatments, all of these gut androgens go away. But then what I hope that you can also appreciate is that in a subset of individuals that are progressing on treatment, we can start to measure these androgens again. And some of them are at pretty high levels. Also, interestingly, our metabolite, the 11 beta OHAD, we can only measure in fecal samples from individuals progressing on treatment with abiraterone acetate. And I’m going to pause here just for a moment to say that we’ve been very careful about developing these targeted mass spec assays for these fecal androgens. So, we’ve been going one at a time, but we can only measure the androgens that we have on the panel. All of the androgens that are metabolites that I’m showing you here can be converted to other things either by bacterial enzymes or human enzymes. So, we won’t be measuring them if they’re not on our panel. So, as we move into starting to try to measure some of these in the circulation, we are in the process of expanding our panel to as many androgenic metabolites as we can. All right. So, I mentioned to you previously that this one strain of clostridium scindens, VE2O2-14, was more prevalent in individuals progressing on therapy with abiraterone. And our colleague Jason Ridlon looked at this data and was like, well, that’s really interesting. However, that particular strain of clostridium scindens doesn’t have bacterial desmolase. However, many years ago, it was reported that that bacterium could convert androstenedione to epitestosterone, indicating that that strain of bacteria contains a 17 alpha hydroxysteroid dehydrogenase. And then Jason did this really interesting experiment where he pulsed the bacteria with 17 beta OHAD and looked at differential gene expression, and he found exactly one gene that was overexpressed in bacteria treated with 11 beta OHAD. Did further studies to confirm that that is indeed a 17 alpha HSDH and has now named the gene DESF. So DESF can also convert our 11 beta OHAD to 11 beta hydroxy epitestosterone and our prednisone metabolite to epi delta 1AT. All of which the 11 beta hydroxy epitestosterone, epitestosterone, and epidelta 1 AT, we’ve now shown in in vitro studies serve as potent AR agonists. Interestingly, for epitestosterone, in the past that has been reported as an AR antagonist, and we’ve now done work to show that that’s not the case. It’s an AR agonist. And I don’t have the time to show you all of that in vitro data, but I have the reference listed here if you want to read more about that. We subsequently developed a targeted quantitative PCR assay to measure fecal DesF levels in our patient cohorts. And what I hope you can appreciate here is that there are significantly higher levels of fecal DesF in individuals progressing on abiraterone acetate versus individuals with hormone-sensitive prostate cancer not undergoing any treatment. And then the fecal DesF levels are also elevated versus individuals that are stable on abiraterone and acetate treatment. On the far graph, these are patient match samples for when they were stable on treatment versus progressing. And whereas the trend is not the case for everyone, I hope you can appreciate that in a subset of individuals, we do see quite a significant increase in the fecal DesF levels when patients are progressing on treatment. So, in the last few seconds that I have left, I just want to introduce a clinical trial that is ongoing from this work. And this is specifically involving this prednisone metabolite, the Delta 1AT, or its further conversion to the epi Delta 1AT. There have been two prior clinical trials that have shown that if you just switch the replacement glucocorticoid given with abiraterone acetate from prednisone to dexamethasone, you get a renewed clinical response to abiraterone acetate in up to 30% of individuals if you’re looking at a PSA 50. Importantly, one of these studies also showed that you can even see a radiographic response to just switching the glucocorticoid from prednisone to dexamethasone. So, the bacterial enzymes that I’ve told you about, whereas prednisone serves as a substrate for these androgenic conversions, dexamethasone actually does not. So, our hypothesis is that perhaps this renewed response to abiraterone acetate is actually due to eliminating the substrate for androgen conversion with prednisone. And so, we are currently looking at this very carefully in a clinical trial led by Cathy Marshall at Johns Hopkins. I don’t have the time to tell you the details of the trial. Happy to chat with anybody further about it, other than to mention that the current Challenge award that we have is funding some beautiful correlative work, and we are taking great care to have this be a very biospecimen-rich clinical trial. And hopefully maybe I can come back to a later PCF meeting and tell you all about the results of this. And I also wanted to mention that we are in the process of looking, searching for small molecule inhibitors that are specific to these bacterial enzymes in the gut. All right. So, to summarize, I’ve told you that bacteria in the gut, and even though I didn’t have time to really go over it, some of these are also in the urinary tract, can produce androgens that support prostate cancer cell proliferation. And as you will be hearing about by Elizabeth Bess, and in this talk, the gut bacteria can also metabolize oral drugs with important consequences potentially. And then that epitestosterone and other DES A, B and DESF metabolites may be unrecognized androgen receptor agonists. And we’ve shown that at least DesF carrying bacteria are enriched in advanced prostate cancer patients that are progressing on therapy with abiraterone acetate. And then finally, again, I didn’t have time to talk about the urinary bacteria, but they are contributing to androgen production in the urinary tract. We have no idea if this is important at all in prostate cancer biology, but it’s a provocative thing to think about. So, with that, I will end and acknowledge my wonderful lab members who have contributed to these studies, my wonderful colleagues at Johns Hopkins University. I’ve mentioned Jason Ridlon who’s been fundamental to all of these studies and his lab members and funding from multiple sources, most importantly of which has been PCF, that has indeed funded every part of my microbiome work from the very beginning. Thank you and I’d be happy to take questions.
Michael Freeman, PhD [00:18:16] Michael Freeman, Cedar Sinai. This is really, really interesting work. I had two questions. I didn’t know anything before your talk about the gut androgens, so that’s really fascinating. But so, two questions. So how many of these androgenic compounds are really out there? I mean, given the complexity of the microbiome. And the second question is the affinity for the AR seems to be a really important variable. And so, my conception of the androgen receptor is that it’s not a promiscuous binder among nuclear receptors. So DHT, for example, has a higher affinity than testosterone. So, can you talk about the relative affinity of these guys and the biological relevance of that?
Karen Sfanos, MS, PhD [00:19:11] Yeah, so your first question about the full spectrum of androgen metabolites in the gut, my very honest answer is we don’t know. Because we’re discovering new bacterial enzymes every day that might contribute to this. And we don’t have comprehensive assays to measure every androgen metabolite that there is. That’s been the very difficult part is trying to put together this panel one by one. So, to be determined. In terms of your question about androgen receptor affinity for these different metabolites, all I can really say is that a lot of the work that we’ve done is in cell culture. And we show both cellular proliferation and different prostate cancer cell lines that are responsive to androgens. We show that these metabolites lead to increased expression of androgen responsive genes such as TMPRSS2, PSA, and we show that we can inhibit all of that with treatment with enzalutamide, indicating that it is androgen receptor mediated. I certainly can chat more about this with you. There has other been other studies beyond ours that have shown that epitestosterone, for example, is an AR agonist using different assays. But that’s what we’ve done.
Daniel Petrylak, MD [00:20:37] Petrylak, Yale. So fascinating presentation. You mentioned that some of the urinary bacteria may also contribute to this as well.
Karen Sfanos, MS, PhD [00:20:44] Yeah.
Daniel Petrylak, MD [00:20:45] And we often have patients with urinary tract infections. What is the effect of treating a patient with antibiotics on androgen synthesis and should we be wary of this when we are treating patients with abiraterone or other agents?
Karen Sfanos, MS, PhD [00:21:00] I think so that’s a great question. In terms of, so we don’t know if androgen production in the urinary tract, like we just don’t know if it might limit urinary tract infections. It’s an interesting idea. I don’t really know. But if it does, then yes, antibiotic treatment might exacerbate that. In the gut, I didn’t have time to talk about it, but one of the arms of our trial is to add an antibiotic called Flagyl, which covers the gut anaerobes that we think are the best, you know, idea that we have for what might be responsible for this. And then if you eliminate those gut anaerobes, maybe you can help to confer treatment response to abiraterone, even with species that we haven’t figured out yet in terms of their capacity for androgen biosynthesis. And it’s also why we’re doing the small molecule inhibitor work, because just broad-spectrum antibiotic use in is never gonna be the best solution for trying to confer renewed therapeutic response. So yeah, great question. Hi.
Unknown [00:22:15] Great, great, work. Hopefully we have time for one more. I was just wondering, you know, things like DHT get UDP-glucuronated and secreted. Do you think that some of these guys could start, you know, metabolizing that to allow for it to come back into circulation or…
Karen Sfanos, MS, PhD [00:22:37] Hundred percent. You know, in the urinary tract, you know, that’s something it’s like we’ve been discussing this, like even if it is, like, is it gonna actually be absorbed in the circulate or is it just going to be excreted, like you know, to be determined. But in the gut, absolutely. It’s something we haven’t studied yet, but that’s a great question. Yeah. All right. Next, we’re gonna hear from Dr. Elizabeth Bess, who is going to tell us some more interesting work on drug metabolism by gut microbiome, but now in the setting of tamoxifen and breast cancer and think about enzalutamide while she’s talking.
Elizabeth Bess, PhD [00:23:30] Good morning. It’s such a treat to be here. I’m a chemist who studies Parkinson’s disease and breast cancer, and I’ve really been enjoying learning more about prostate cancer through this meeting. And the story that I’ll tell you about, I think we think of as sort of a blueprint for thinking about ways to examine drug metabolism in the context of the microbiome. So first, I think an important sort of context for what I’ll share with you today is the idea of where does our microbiome even come from and what sort of functional implications does that have on our health and on how we treat disease. And I like how Ed Young said this the microbiome is a sum of experiences throughout our lives, the genes we inherited, the drugs we took, the food we ate, the hands we shook. So, a combination of both the unique genes that make up our human cells, as well as our unique lived experience is shaping which bacteria are present in our gut and the abundance of all of these bacteria. And because all of these factors are different for every one of us, the result is that we all have a microbiome that is as unique as a fingerprint. And each of these trillions of bacteria that are living in our gut are encoding millions of genes that are encoding for a variety of different chemical functions. And so, we have an impact, many impacts of this, but one of them is that which medication can effectively treat an individual can be impacted by which bacteria and which chemical reactions are encoded in this environment. And so, the story I’ll tell you today is about breast cancer, a disease affecting one in eight women, and particularly about the drug tamoxifen. This is a drug that is the most prescribed drug to prevent breast cancer recurrence. And it’s prescribed to people for usually five to ten years to keep breast cancer at bay following treatment. However, tamoxifen is only effective for about half of people, and it’s not clear why. There’s some evidence that various SNPs in the liver, cytochrome P450 enzymes are playing a role in the variable efficacy of this drug, but it doesn’t tell the whole story. And we’ve been curious about the role that bacteria may be playing to help fill in that gap in fully understanding the role of variable efficacy of this drug. And in particular, we’ve been examining how bacteria may be responsible for controlling the level of drug in the bloodstream. One of the primary reasons that tamoxifen has variable efficacy is that people for which it’s not effective are not getting enough of the drug into their bloodstream to have a biological effect, so not entering this therapeutic window. And we suspected that bacteria could be playing a role in this. This was a study that was led by my lab and also in collaboration with wonderful collaborators, Cholsoon Jang, a mass spectrometist, and Katrine Whiteson, a bioinformatician. And the students shown here were performing the experiments. Liz Ortiz, Yasmine Alam, Sheron Hakopian, and Julio Avelar-Barragan. So, this story begins in a glove box. This is where we were rearing germ-free mice that have no bacteria in or on them. And this allowed us to very selectively and controllably examine the variable of the microbiome in thinking about drug metabolism. So, we took germ-free mice, and we first simply colonized them with a collection of fecal samples from five healthy individuals and put them in one glove box. And in a second glove box, we kept a group of mice germ-free. And both of these groups were dosed with tamoxifen. And what we see is that the mice that had a microbiome were able to get tamoxifen into the bloodstream. The mice that did not have a microbiome did not get tamoxifen into the bloodstream. But I think that this story gets a little bit richer as we think about a third group that we actually included in this experiment. So, we didn’t just perform a PK study, but first we provided these mice with tamoxifen for a 10-day exposure, trying to get towards the idea that people given this drug take the drug for five to ten years, and there could be a long-term impact of this drug for long-term exposure. So, what we see is that after we treat mice for 10 days with tamoxifen or with the vehicle in which that drug was delivered, we see very different results for these mice that did not have long-term exposure. So, in the teal color, we’re seeing that when mice were only dosed with tamoxifen for the PK study, they are having a microbiome, but they do not get tamoxifen into the bloodstream. They only have approximately the same level that we see with mice that were germ-free to begin with. We were able to do this experiment because that long-term exposure was with regular isotope abundance tamoxifen. And for our PK study, this was C13 labeled tamoxifen, so we can very clearly differentiate the tamoxifen that is in the blood as a function, not of long-term exposure, but of that acute dosing for the PK study. So, what might be going on here? If we take a step back and think about what is known about metabolism of tamoxifen, first off, this is a drug that is orally dosed. It itself is inactive, it’s a pro drug. It enters the intestine where it is absorbed into the bloodstream and then makes its way to the liver. Where the liver does chemistry to make this drug more hydrophilic. It demethylates the amine and also adds a hydroxyl group on the phenyl ring at the bottom there, forming two different bioactive metabolites of tamoxifen, endoxifen and 4-hydroxytamoxifen or 4HT. So, these are considered the bioactive forms that have a hundredfold better binding to the estrogen receptor than tamoxifen does. But the liver doesn’t stop here. Its goal is, of course, to get rid of drugs, and so it adds a sugar unit, a glucuronide to tamoxifen, and that slates this molecule for biliary excretion from the liver into the intestine. And adding that sugar, again, inactivates this drug, can no longer effectively bind the estrogen receptor. Once this glucuronidated tamoxifen derivative is in the intestine, this is where we thought bacteria may be playing a role. We thought, and this goes to one of the questions that was just asked in the previous presentation, we suspected that gut bacteria could be responsible for removing this sugar. Bacteria need sugar to grow, and they’re very good at finding sources of sugar, removing it from a larger molecule, and then using it as a carbon and energy source. And so, we suspected this could be happening in this process, and that this would then return endoxifen and 4-HT to its bioactive form and enable absorption into the bloodstream. For distribution to tissues. So, as we further examine this mechanism, our next experiment was to look at mice that were normally colonized mice, so in a typical mouse facility, and then we treated them with antibiotics. And we see again that suppressing the microbiome also suppresses level of tamoxifen in the bloodstream, but not even as much as not having long term exposure of tamoxifen over that 10-day period. So, we tend to see this adaptation of the microbiome to tamoxifen therapy. When we look for bioactive metabolites of tamoxifen, 4-HT here, we also see that we certainly get 4-HT into the bloodstream with mice that were colonized with bacteria, but when we add antibiotics, we don’t see 4-HT. And furthermore, when we don’t have long term exposure, we have no detectable 4-HT entering the bloodstream. So, the big question that we were asking at this stage is how are microbiomes adapting to their environment? What are some of the big strategies that bacteria can use? And certainly, one of them is antibiotics. When we think about bacteria that are exposed to antibiotics, they have a strong driving force to figure out a way to avoid being killed. And so, they can adapt and evolve. And indeed, about a quarter of drugs that are not intended to be antibiotics exert an antibiotic effect on gut bacteria when they are tested in vitro in a one at a time way. And for this recent study, tamoxifen was one of the ones that was examined in a cohort of these drugs and was found to exert an antibiotic effect on individual gut bacteria. When we took human fecal samples and incubated them with tamoxifen at increasing concentrations up to the concentration around 40 micromolar that we see of tamoxifen in the human intestine, we have decreasing amounts of bacterial viability with increasing amounts of tamoxifen. However, it seems that these bacteria are adapting as well. With long-term exposure, even up to only three days, we see that bacteria are rebounding and able to overcome these seeming antibiotic effects of tamoxifen. When we look in our, those were in test tubes, when we look in our mice that were exposed to tamoxifen for, not for 10 days or for 10 days, so vehicle or tamoxifen, we see that the microbiome community is not altered at the general phylum level, nor at more specific levels of genera or species level. And looking at alpha diversity, a measure of the diversity between these populations, we don’t see any significant difference between the microbiomes of mice treated with tamoxifen for 10 days versus vehicle. So, something else is going on here. We’re not getting gross metabolism or alterations to the microbiome, but we are seeing really important functional changes. When we perform untargeted metabolomics, we see that just for the top 25 of metabolites that are detected, we have a really strong pattern of metabolites that are upregulated in the tamoxifen treated group and much less present in the mice that were treated with the vehicle. So really pointing to the importance of looking at the functional levels of what’s happening in these microbial communities. But we were still really curious to figure out what specifically is going on here. So, we see no changes at the gross level of the microbial community, but we do see changes at a functional level, both in terms of drug and bloodstream and also metabolites. We wanted to next zero in on that deglucuronidation pathway, the cleavage pathway, and ask what role could this be playing in getting drug into bloodstream. And so, in order to do this, we took fecal samples from nine different people, and we extracted all of the enzymes for these fecal samples and incubated them with the glucuronidated form of endoxifen and measured subsequent formation of sugar removal for formation of endoxifen. And what we see across this panel of nine fecal samples from healthy individuals is wide variability in the capacity of the sample to bioactivate this drug. So, sample six was excellent, sample four not so much. So really highlighting the inter individual variability here. The range was also really important for technical reasons because it allows us to ask a question about correlation analysis. We were wondering if because we have this range of endoxifen activation, could we use it to ask which genes have an abundance that correlates with extent of endoxifen activation, and could that actually be responsible for the activation that we see? And we were specifically focusing on beta-glucuronidases or GUS. This is a family of enzymes that is known to remove sugars from molecules, although it hadn’t been studied for tamoxifen. And this is a complex enzyme family, it’s huge. There are at least six broad classes of beta glucuronidases. And over 270 different types of these enzymes have been detected. And in each one of us, we have four to 38 flavors of beta glucaronidases, all with different substrate specificity. And so, our challenge was to see if we could find which of these enzymes could actually be responsible for bioactivating endoxifen. And so, what we did was take those fecal samples from people and do whole shotgun metagenomic sequencing. So, we’re measuring all of the genes that are present across all of the bacteria. And then we bioinformatically pulled out all of the genes that were mapping to GUS families. And that’s what you’re seeing in the image here. And we’re also stratifying across taxonomy. And so, we can see that there is a lot of variation across GUSes in these samples. And we use this information to perform that correlation analysis that ended up working out really well. So, as we have increasing amounts of endoxifen activation, there were two GUS enzymes from Bacteroides Fragilis that had increasing abundance of the gene mapping to increasing amounts of endoxifen activation. And we are really curious about this because. B. Fragilis is an organism that is known to bloom in the gut with consumption of a high fiber diet. So, it’s an organism that is very easy to manipulate in the gut through diet, and fiber is considered an important part of a healthy diet, but most Americans get far too little fiber. And so, what we are currently in the process of investigating first in mice and then moving towards people is the ability to provide a high fiber diet in conjunction with tamoxifen to ask whether this is sufficient to be able to drive an increase in endoxifen levels in people where that level is not high enough to have a therapeutic effect. Overall, what this study is adding to the emerging view in the microbiome community, and what Karen spoke about earlier is the importance of understanding the unique collection of molecules that are present across people’s microbiomes, and that’s seeming to be playing an important role in having more of an individualized approach to thinking about treatment for people. I’ve highlighted throughout the folks who’ve been involved in this work and acknowledge our sources of funding and happy to answer any questions if there’s time.
Felipe Eltit, PhD [00:39:05] Hello, I’m Felipe from UBC and I think that my question is kind of obvious. Because we think about prostate cancer and you’re working in breast cancer, is it the baseline difference between male and female or men and women in bacteria that can from the baseline explain any difference of what we will find in a prospective study in prostate cancer?
Elizabeth Bess, PhD [00:39:30] Yeah, so there are there are differences across sexes and certainly the hormones are circulating through the microbiome and are getting metabolized and also produced as we’ve seen. But the sex effects tend to be relatively minor compared to many of the other environmental effects that bacteria are being exposed to. Thanks.