Transcription
This week in microbiology is brought to you by microbe.tv and is supported by the American Society for Microbiology. To learn more about microbes, visit microbe.tv and asm.org.
[Music]
This is TWIM, This Week in Microbiology, episode 342, recorded on October 2nd, 2025. I'm Vincent DrakandYellow, and you're listening to the podcast that explores unseen life on Earth. Joining me today from, I don't know, Ann Arbor, Michigan, Michelle Swanson.
>> Yes. But, but did I hear you correctly? Did you say it's October?
>> Oh my gosh.
>> How did that happen?
>> September just, poof.
>> Just like that.
>> And August.
>> The whole summer's gone.
Also joining us from Charleston, South Carolina, Michael Schmidt.
>> Well, hello everyone. And we're blessed. We dodged two hurricanes this past week. So, we're very thankful.
>> Good for you, cuz water is a problem.
>> Oh, yes.
If you enjoy these podcasts, we we have a bunch of podcasts besides TWIM and TWIV and TWIP and TWEO and TWIN and IMMUNE and lots of others, all done by scientists in their fields who are experts, by the way. So on these podcasts, we do believe that expertise is valuable, and in fact, it's essential.
>> And data, data counts. Data. We we we look at data. We make conclusions based on data, and we discuss papers from that viewpoint, right? We we don't say, "I think this is the way it should be because I feel it."
>> We say, "This table one says this." And if you like these programs and you can learn science from them, we'd love your support.
>> Because we don't put ads in to disrupt your learning. Microbe.tv/contribute. Your contributions are US federal tax-deductible because Microbe.tv is a nonprofit entity.
Today, we're going to explore the microbiome. We have two papers that are going to look at links between the microbiome and various conditions. And Michael will start us out.
I'm going to start off with a paper that appeared in JAMA Oncology on the 15th of September, and it's authored by Unuyang Qu, A. Chen, J. Wong, and M. Usyk, N. Freriedman, W. Yi Hang, C. Um, T. Gunda, P. Oberston, H. Lee, R. Hayes, and Y. An. Title of the paper is "Oral Bacteria in Fungal Microbiome and Subsequent Risk for Pancreatic Cancer."
So this paper sits right at the intersection of microbiology, immunology, and oncology. And it tackles one of the most intractable problems of modern medicine, namely pancreatic cancer. Anyone with a family or friend who has ever received a diagnosis of pancreatic cancer appreciates how devastating a diagnosis that that disease can be. The 5-year survival rate for individuals diagnosed with pancreatic cancer is abysmal, hovering around 13%. And in the show notes, I dropped in, uh, some data from the American Cancer Society highlighting survival rates for cancer, the 2025 data set. So, you can take a look at that if you're interested in this.
>> And Michael, the reason is that by the time you feel pain, it's too late, right?
>> It's too late. It's it's advanced to a stage where medicine and the chemotherapy, unfortunately, really doesn't have all that great of effect. And the chemo and therapy that they offer often is so debilitating to, uh, the patients. Now, a huge part of the problem is that the medical community, and as Vincent just said, usually appreciates that an individual has pancreatic cancer when it's already far too late. The known risk factors are things like smoking, obesity, family history, but they only account for a fraction of the cases. So the holy grail here, and this is what these authors have set out to attempt, begin the journey for finding a way to identify high-risk individuals before, before, underscore, before they develop the disease, when the medical practitioners might actually be able to do something about this form of cancer.
>> Michael, can I ask you, can a CAT scan diagnose pancreatic cancer?
>> If you're lucky to have one, yes. And in fact, that happened to my colleague's son. One Christmas, he developed pneumonia. The X-ray technicians were on Christmas break. The only available radiography instrument that they had staffing for was the CAT scan. So they put his son into the CAT scanner. And when they were cat scanning his lungs to look where the pneumonia was, they spotted the lesion on his pancreas. This was in the earliest of stages, and that young man is now an old man, and so he,
>> Wonderful.
>> He was able to get that cancer taken out before anything happened.
>> I was, I was talking to some people about using CAT scanning for diagnosis, and the insurance companies won't pay for it unless you have a risk. Like, if you're a smoker, you can have an annual CAT scan, and they'll pay for it. But if you're a non-smoker, forget it. And that's, we're letting them run our health. That's not right.
Yeah. And now, again, for years, there's this been this tantalizing, if somewhat murky, connection between poor oral health, especially periodontal disease, and pancreatic cancer. But the evidence has often been based on smaller retrospective studies where you can't be sure if it's the oral dysbiosis, the bad actors of periodontal disease, is the cause or the consequence of pancreatic cancer. And this brings me to this impressive paper that was published online in JAMA Oncology. Uh, what they've done here, in my opinion, is the definitive study on this topic to date, whether or not the perio bugs are really implicated in it.
So, what did they do? Right off the bat, the strength of this paper is in its design. This is a prospective cohort study. They didn't find people with cancer and then look at their mouths. They took oral wash samples from over 122,000 healthy people enrolled in two massive, long-term US cohorts, the Cancer Prevention Study 2 and the PLCO trial. And then they waited. And so, over a median follow-up of almost 9 years, from those 122,000 samples that they collected, 445 of those participants unfortunately went on to develop pancreatic cancer. Now, the researchers then went back to their freezer farm of 122,000 samples, pulled the pre-diagnostic oral samples, and matched each one with a sample from a control who remained cancer-free. Now, this prospective design is critical because it means any microbial differences they found in the cancer group were there before the diagnosis of pancreatic cancer, which gets us much closer to assessing risk rather than just association. In other words, they,
>> Can I, can I just interject? I am so proud that we as a country were funding such a large prospective study. I mean, this was a lot of work.
>> 122,000 people.
>> I mean,
>> Nine, nine years, too. It's a lot.
>> It's a lot. Uh, this is not a project you just scratch your head and do the next week. Just think of the freezer farm that they had to have and the inventory control that they had to have. I mean, it's,
>> But it can be quite powerful, right? Pay me now or pay me later.
>> Pancreas.
>> Well, you also,
>> Terrible problem. You need nine years of continuous funding. Yes. Right. You can't have any of this baloney where you, you withdraw funding for a month or two months or three months because you feel it's politically correct. No, you have to have consistent funding. And that's why it's ridiculous for this administration to,
>> And I'm sure everyone listening has lost someone, knows of somebody who's passed from pancreatic cancer. So,
>> And in fact, what prompted me to look at this paper, one of my close friends and one of my coaching partners for TEDx Charleston passed away from pancreatic cancer, and she was only diagnosed in April, and then I went to her funeral a few weeks ago. And so, this, this touches everyone.
Um, but let's get back to the science. Um, they just didn't do 16S sequencing to survey who was there. For bacteria, they actually used whole genome shotgun sequencing. And for fungi, they used a technique called ITS sequencing, where the ITS is a fancy abbreviation for internal transcribed spacer, which is the region of DNA located between the genes that code for ribosomal RNA, specifically the small subunit ribosomal RNA in the fungus and the large subunit ribosomal RNA. And these regions, just like the loops in 16S RNA, evolve rapidly because there's no selection pressure on them, and are, you know, as important phylogenetic tools to analyze, uh, and speciate fungi. And so, this, this fact that they did whole genome shotgun sequencing of the m of the bacteria, as well as the ITS for the fungi, this gives them a much greater resolution as to the identity of the microbe down to the species level of actually who's actually living in these oral communities.
And what they found, candidly, their findings were pretty striking. First, they confirmed what others have suspected. Three well-known bacterial pathogens were linked to periodontal disease and were also found in the cancer, uh, subset that they evaluated. The first one was *Porphyromonas gingivalis*, which, in addition to causing periodontal disease, has been linked to a myriad of metabolic disorders. Everything from cardiovascular disease, Alzheimer's, type 2 diabetes, obesity, inflammatory bowel disease. That's one creature you don't want in your mouth. *Porphyromonas gingivalis*. Uh, they also found *Tannerella forsythia*, which is one of the red complex microbes associated with severe gum disease. And I dropped into the show notes a review on what this red complex and purple complex mean about, and it has to do with how periodontists identify periodontal microbes. And they also found another one, *Parvimonas micra*, and they're all associated with an increased risk of developing pancreatic cancer. So, we have three guys, but the microbiome wine scan revealed a whole lot more. They identified a total of 21 separate bacterial species, and for the first time in a prospective study of this scale, four fungal species were significantly associated with cancer risk, specifically *Candida albicans* and *Candida tropicalis*. And those microbes are known to produce nitrosamines, which, anyone who's ever done an Ames test, know that that's what actually causes hypermutations.
Now, here's the really powerful part of the paper. They just didn't stop at individual microbes. They combined the signals from these 27 risk-associated bacteria and fungi to develop a single microbial risk score. Now, Vincent's going to talk about microbiome and sleep, and they too develop a similar microbial risk score. They called something else, but just begin to think about this microbial risk score. And what they did is they found that for every one standard deviation increase in this MRS, or microbial risk score, an individual's risk of developing pancreatic cancer went up by a factor of almost three and a half. This is a huge effect size. And they present it by odds ratios. And odds ratio is one of those fancy statistical terms. And an odds ratio simply measures the strength of an association. And that's what they're seeing here is an association of the microbe and the cancer. So the odds ratio went up, and it measures the association between an exposure and an outcome by comparing the odds of the outcome occurring in the exposed group. So, if you have an odds ratio greater than one, that's a positive association. So it means your likelihood of developing the condition they're investigating is increased. If it's one, it means there's no association. And if it's below one, it means there is a negative association. And so, effect size is a statistical measure of simply magnitude or the strength of the relationship. And an effect size of 3.44 is a huge effect size, and that's likely due to the fact that they had so many individuals in the samples that they provided.
Now, their finding was also consistent across both cohorts that they measured. Remember, they took their 122,000 specimens from two distinct long-term cancer prospective trials, and it held up during the cross-validation, which is a statistical measure, which again tells you it's, it's a robust signal, or likely as correlative a signal might be. And their figures are extremely easy to interpret given the amount of data that they have presented to us.
Now, my perspectives as to the strengths and limitations of this study are as follows. First, the prospective design, it's a strong strength. They had two large, independent cohorts. This is the gold standard for clinical trial, and it really gets at this type of epidemiological question that they're asking. It minimizes the chicken and the egg problem.
>> Amplify that, Michelle. You could imagine a cancerous environment, the physiology would be different, and it might, um, select or provide a favorable environment for certain bugs. But since they, um, isolated these bugs well before cancer appeared, they have to worry a lot less about that.
>> The second positive aspect of their study, and I really commend them on this, is the high-resolution sequencing methods. Now, whole genome sequencing obviously has more information in it than just 16S, and they went the extra bit to do that. And this gives them, as well as us, a more detailed and accurate picture of the microbiome than we ever had before. But more importantly, it gives us an inkling of what microbes we may want to test in either an organoid model or an animal model to effectively get at mechanism. The third strength is the development of this microbial risk score. This is a key step in translating this research into a potential clinical tool.
But of course, no study is perfect. And Vincent and Michelle will be the first to say, it's still correlative. It's not causation. While the prospective design is a huge leap forward in their findings, it doesn't prove that these microbes are the beasts responsible for the cancer. They could be a marker for some other underlying factor, like a specific type of low-grade systemic inflammation that could be the real culprit. After all, the adverse consequences of periodontal disease is the result of chronic inflammation caused by the gram-negative microbes in your mouth and the subsequent bone loss that lead, that in turn leads to tooth loss. Even if it's not the direct cause of the neoplasia, it, as they point out in their conclusions, um, a particular species that that is more often associated with cancer could be used as a biomarker. You go in,
>> Absolutely.
>> Check the dental, a dental sample, right? And, uh, if you see this bug, then this person needs to be watched more closely,
>> Proactively for cancer. We need not forget that dentistry is a primary healthcare discipline.
>> And you can well imagine if this test bears fruit in identifying precancerous lesions when we can do something about it. You can imagine going to your dentist at age 25, they do a paper point, uh, survey the microbes associated with that paper point, and then tell you, "Go see your primary care physician. You're justified to have a CAT scan to see if there's anything on your pancreas," or if you have any of the other biomarkers often associated with p, the development of pancreatic cancer.
>> Or if that person was starting to feel some abdominal pain, right? They would more absolutely quickly go and talk to their healthcare.
>> Pancreatitis is one of the indicators. And the second limitation is samples were only collected at a single point. The authors rightly argue that the adult oral microbiome is pretty stable, but one spot, one sampling is only a snapshot. So again, but they had 122,000. So that, you know, leads me to believe that this is, is, is really solid evidence. And the third one is the study population was predominantly white. So we need to be cautious about generalizing these specific microbial signatures to other populations because no two humans are alike from the perspective of their microbiome. They may have general trends or not.
Now, what this means for medicine, or what happens next, or more importantly, when might you expect your dentist or primary care specialist to ask if they might take a swab of your periodontal tissue to test for pancreatic cancer? The answer is not tomorrow. Right now, this is a research finding, not a clinical test, but it maps a clear path forward. In the near term, 5 to 10 years, I can see this microbial risk score being incorporated into a broader risk assessment panel. As Michelle said, it's a biomarker, folks. A physician could combine your MRS with your family history, your genetic data, lifestyle factors to get a much more accurate picture of your risk of developing this sort of cancer. For those who land in the highest risk category, it could also justify using, as Vincent said in the opening, using the more invasive and expensive screening methods because the treatment of pancreatic cancer is not an inexpensive endeavor. The chemotherapy, the surgery, and the human cost are all extremely expensive. And they may then justify doing endoscopic ultrasound or the CAT scan. And the goal here is not diagnosis, but rather risk stratification. It's about finding that needle in the proverbial haystack, that small group of people who would benefit from the more intensive surveillance system.
In the long term, 10 years plus, if further research, especially mechanistic work in animal models and organoids, can establish the causal link, the implications are even more exciting. It raises the possibility of prevention. And so, could targeted interventions, from something as simple as improved dental care to more advanced strategies like probiotics or prebiotics designed to shift the oral microbiome away from the high-risk profile, actually reduce a person's risk of developing this terrible disease? They're still a ways off, but this paper provides a solid evidence-based foundation to start asking those questions. It's a fantastic piece of work that underscores a fundamental tenant of microbiology. The bugs in one part of our body can have a profound and unexpected consequence in another.
They point out that, um, some of these microbes that are common in the, are found in the, um, mouth migrate and are found in the, um, inflamed pancreas, right? So it's not that they stay there.
>> Especially the, the, the *Candida* abundance. And they've been able to isolate *Candida* from normal tissue bio, uh, specimens from patients with cancer in an animal model. Specifically, *Candida* has been reported to drive, um, pancreatic oncogenesis through the modulization of the tumor immunity. And so, I, I think their best, the one path I would chase is the *Candida albicans* and the *Candida tropicalis* link in pancreatic cancer. And I'm sure they're, they're developing models to specifically look at it. And I was really excited to see their MRS score and then to transition to Vincent to his paper when we're learning about a similar, uh, risk development score when it comes to the topic that Vincent's going to walk us through.
So, Michael, these three pathogens, *gingivalis*, *tannery*, and *micra*, these are only present in in diseased, uh, oral tissues?
>> In large numbers, yes. But you can get transient, like everything else. You are what you eat, and they can be introduced. But *Porphyromonas gingivalis* only eats protein, it does not eat sugar. And so, you're, you're the food source. And the *Porphyromonas gingivalis* linked to Alzheimer's with its proteases is the gingipains, and that's principally how it's leading to gum disease. And all it starts with gingivitis and then goes to periodontitis.
>> And the same with the *Candida* species, right?
>> *Candida*, you develop oral thrush, and that's principally seen more often in cancer chemotherapy patients because of the profound consequence of the cancer and the way they, they assess, um, for thrush, is they can wipe it off of your tongue with a 4x4, a cotton wipe can pull the, the biofilm off of your tongue surface. And the other individuals that are susceptible to thrush are, of course, uh, uncontrolled diabetics where they have large concentrations of sugar, uh, in their, uh, mucosal secretions. So, if you, if you did the screening that you were talking about, you identify people with high levels of these bacteria or fungi, you could, you could either do screening, as you say, or try to eradicate them with antibiotics.
>> You could. Some of them are very susceptible to antibiotics. And in fact, there's even probiotics out there to actually improve periodontal health. And we know when periodontal health improves, these microbes can go down in their concentration. I'm certain it's a concentration effect.
>> You know, the more you have, the more likely you're going to have a bad outcome.
>> Sure.
>> It's like smoking.
>> And then, I mean, let's say we, we eliminate, uh, this fact, this risk factor, would there still be pancreatic cancer caused by other causes?
>> Yes. Yeah. I, I'm sure. Because what they didn't have in this study is the genetic profile of the patients. And you know, 9 years ago, we didn't have whole genome sequencing of humans as inexpensively as we do today. It'd be ideal to be able to go back to those 122,000 and say, "Were, were you in Ancestry.com 9 years ago or 23andMe? Or could we sequence your genome to figure out what's going on?" But as we discussed when you were having your holiday with Mark Martin, it's all about epistasis and what's happening to our genome in addition to what's happening to the microbial genomes.
>> Sure.
>> But this is,
>> I think this is a great example of the power of next-generation sequencing, right? Yeah.
>> Because on its own, it needs a problem to solve. On its own, it's just, "I want to see sequence things and look." No, but if you have a focused problem like this, it's beautiful. And this meeting I just attended, the European Society for Clinical Virology in in Greece, it's all about next-generation sequencing in in in the clinical setting. Absolutely. And everyone's so excited about that and and the power and so forth. And it just shows, techni, a technique is only as good as the questions that you give it.
>> And this was a fantastic question. And as Ilio pointed out when we started doing these these papers, you know, a sequence is nothing more than the New York telephone directory. If you don't know who to call, what, what good is it? You, you need someone to call, which, in other words, in Ilio's parlance, was the question. What question are you going to ask of the telephone directory?
>> Yeah. And then, again, returning, um, the patient cohorts from the American Cancer Society Cancer Prevention Study and the Prostate, Lung, Colorectal, and Ovarian Cancer, um, screening trial. Thank you so much for your work.
>> It's a great example.
>> And your support of this project.
>> Yep.
Now, my paper is also on the microbiome, but has to do with sleep quality. And this is a paper in Cell Reports, and it is, um, called "Gut Metagenome and Plasma Metabolome Profiles in Older Adults Suggest Pyruvate Metabolism as a Link Between Sleep Quality and Frailty." So, this has four co-first authors: Yani Pu, Shinji Xi, Liy Huang, and Wei Shang Wu, and one, two, three, four corresponding authors: Xiao Fang Wang, Xiao Shan Lu, Chang Zhang Wan, and Yan Zhang. I think all of us need to get good at pronouncing Chinese names because we're seeing more and more really high-quality papers coming out of China. Uh, and every time I, I look at an interesting finding, it, it's out of China. So this comes from Fudan University, Shanghai, Shaogong University, School of Medicine, Fudan University, Fudan University, University of South China, Central South University, Xia Jang University, Harvard Medical School, Hey, um, Shanghai University, and Fudan. So, uh, this is all about sleep quality, which, which is what attracted me to it.
>> Because I don't sleep very well, and I don't know why. But it turns out that, uh, poor sleep quality affects between 17% and 50% of adults over 60 years of age. And you may say, "Oh, so what? You don't sleep, you fall, you fall asleep during the day." No, it's way more than that. If you don't sleep well, you can have other things happening like cardio-metabolic diseases, cognitive impairments, cancer, frailty, frailty, and increased mortality. So, what is frailty? It's an age-related physiological decline in physical, social, and cognitive dimensions. And so, you're more likely to fall, you're more likely to have disabilities, hospitalizations, and deaths. And other studies have linked poor sleep quality with frailty, but we don't know what the mechanism is. Why is it that if you don't sleep well, you get increased frailty?
And before you go there, Vincent, I want to, if you don't know what poor sleep quality is, into the show notes, I've dropped the Pittsburgh Sleep Quality Index survey that you can actually take yourself. And I also put the primary reference in so you can figure out if you qualify as a poor sleeper or a good sleeper based on the sleep quality index. Because Vincent's going to talk about the sleep quality index because it's so vital in the analysis of their data.
>> Yeah, I appreciate that. I'm going to take it. I, I suspect my sleep quality is okay. I, I mean, I wake up like two or three times, and I consider that crappy sleep, but it's probably worse than other people.
>> Well, they got the scoring platform in there so you can, you can see what goes into good sleep. And I, I looked at the survey this morning and I said, "This has to go into the show notes for the people."
>> Yeah, I'm going to take that. I'll tell you next time on TWIM what my sleep quality index is.
>> And consider caffeine and alcohol, alcohol intake, both of which will disrupt your sleep.
>> Right. Well, so I don't, I, I have one cup of coffee in the morning. I don't drink. So those are out. But, you know, one thing about sleep, I was reading somewhere else, uh, the kind of light you have just before you go to sleep is really important, right? Blue light, like on screens, is really bad because your body takes, uh, a different kind of light as a signal to start making melatonin, which makes you go to sleep. And so, if you get up in the middle of the night to go to the bathroom, you turn on the bright light, that's going to screw up your sleep cycle. Oh my gosh. They said have a little nightlight there that's just glowing. It gives you enough light. You know, I didn't know any of this. So, boy, science is just great, isn't it? I love it. Anyway, um, I don't feel frail, right? I have a lot of energy. I exercise. I eat properly. So, I'm not too worried about getting frail, but I, I just want to, you know what? Self-care is healthcare. Yes. You've got to take care of yourself if you want your health to be good. You can't take drugs and and this and that. You have to take care of yourself. And so, this is part of it.
So, what's this mechanism? There's some evidence that the microbiome is involved. Uh, some research has shown that frail and non-frail older adults have different gut microbiota, and in particular, they differ in the abundance of butyrate-producing, uh, microbes. But we don't know the role of these microbes. We don't know the role of the metabolites in the connection between sleep and and frailty. And in fact, there's limited research on older adults, which is where the problem exists.
Now, besides the studies of the, the microbiome, as in, um, whole genome sequencing, as Michael mentioned for the, for the pancreatic cancer study, we can also do metabolomics. We can identify all of the molecules that are present, uh, in say, blood or intestine, by mass spectrometry. And this is a new addition to the through our "omics," right? Where you can get even more information. And we have others have seen that there are alterations in metabolites related to sleep. People with insomnia, alterations in metabolites like proline, isoleucine, leucine, and glycerol have been found in older adults without frail frailty. So, uh, it's not just a microbiome, you have to look at the metabolites as well.
So, what they did, uh, in this study is to do an analysis of the gut metagenome and plasma metabolome in a cohort of 1,225 older adults. Now, this pales in comparison to to the first paper we did today, but it's what it is.
>> It's, it's a big study. 1,200 is a big cohort.
>> Yeah. It's not 122,000.
>> No, but it's not nine years either.
>> Yeah. Yeah. That's right. Yeah. This is a snapshot kind of, uh, analysis. Um, and then they, they assessed their sleep, their sleep quality with the Pittsburgh Sleep Quality Index, PSQI, which you can do for your own, on your own, uh, as, as we'll put the link, uh, in the show notes. Uh, and they tried to make an association, and again, it's an association, just like the previous study, an association between sleep quality, microbiome, and metabolites.
So, 1225 participants from the RUGAL Longitudinal Aging Study, average age of 78, 46% male, and 515, 42% of them were categorized as having poor sleep quality by the Pittsburgh Sleep Quality Index. So, the people with poor sleep quality were more likely to be female. Women worry a lot, I think. No, no, you don't think so? I think they do. More likely to be have a lower education level, and they're more likely to use cardiometabolic medications. They, they also, people with poor sleep quality have a higher prevalence of impairment in basic and instrumental activities of daily living. They have a poorer grip strength. They have lower cognitive function and depressive symptoms. And then, of course, they have an overall higher frailty index. All right. So that's another thing you can calculate, a frailty index. Okay. So that's the background for this study.
So, what they find looking at the microbiome is that the overall structure of the microbial species at the species level differs significantly between participants with good and poor sleep quality. Right? There, you have to do statistical analysis and alpha and beta diversity. All right? So, both alpha and beta diversity of the microbial functional pathways, where you can look at the microbiome and just and say, "What functional pathways are encoded by the microbes?" They're assoc, they are, uh, significantly associated with sleep quality. So, better alpha and beta diversity. They also identified 14 species and nine genera showing differential relative abundances between good and poor sleep quality groups. And the poor sleep quality group had a higher abundance of potentially harmful bacteria like *Erysipelatoclostridium ramosum*, *ramosum*. That's a bad one you don't want to have. It pops up in many studies. And *Violinella infantium*. These have been associated with cognitive dysfunction. On the other hand, the species enriched with good sleep quality include *Roseburia faecalis*, *Faecalibacterium prausnitzii*, and *Prevotella copri*. All known producers of short-chain fatty acids, which we know promote good gut and systemic health. You want those short-chain fatty acids.
>> You want a lot of butyric acid, folks.
>> Should we go out and take,
>> Oh, your, your gut,
>> In general, right? Dampens inflammation.
>> It dampens inflammation, and your, and your microbes need to make it because butyric acid is a tremendous energy source that everyone will eat. And so, it's like giving candy to a baby. They'll,
>> Where can you, where can you eat things rich in butyric acid, or does it have to be metabolized?
>> It has to be metabolized. And that's why every, every time you see a GI doctor, they tell you to eat more fiber. Eat more fiber. That's what actually facilitates the good microbes to come up.
>> It's what our grandmothers told us decades ago, right?
>> Yes. Oatmeal. Oatmeal.
So, they adjust all of these data for age, sex, BMI, smoking, drinking, education, and cardiometabolic medication. So that's, you want to try and get rid of confounding factors. You do your best. You can't get them all, right? You never can. People are complicated.
>> Uh, and then,
>> Then four species are associated with sleep quality. *F. prausnitzii* and *P. copri* remain enriched in the good sleep quality group, while *E. ramosum* is enriched in the poor sleep quality. *F. prausnitzii* gives you anti-inflammatory effects and is positively correlated with delaying frailty, and *P. copri* gave benefits related to carbohydrate metabolism and immune, uh, modulation.
So, then they want to quantify this. So, similar to what was done in the, the pancreatic cancer paper, they, they develop a gut microbiome sleep quality index score, GMSI, uh, based on these 14 species that are related to sleep quality, right? And what they found was the participants with good sleep quality had higher GMSI scores, right? So, the higher the number, the better your sleep quality. Uh, and higher scores were inversely correlated to cardiometabolic medication use, hypertension, impaired gait, and difficulties in basic, uh, ADL and frailty. ADL is activities of daily living, right? So, inversely, so higher score is inversely correlated. So, you're better off with a higher score.
Uh, what about sleep in this, this cohort? Sleep duration, sleep efficiency, and sleep latency. So, they're all three ways you can characterize sleep. Uh, were the major contributors to the overall PSQI score, right? The score that, it's not the one we've just been talking about, but that's their third figure in this paper. It's the relationship.
>> Yeah.
>> Yeah. Yeah. The,
>> Sleep Quality Index, right? The Pittsburgh Sleep Quality Index. Thank you. Sorry.
>> So, those are the sleep duration, efficiency, and latency, the major contributors to the overall PSQI. Most of the sleep-related microbial features are also correlated with, uh, PSQI, particularly sleep duration, sleep efficiency, and sleep disturbances, but not with sleep medication use. Participants with early sleep onset paired with early wake-up time. So, this is a common pattern, right? The older you get, you tend to go to sleep earlier, but then you wake up earlier. So, it doesn't help. Best thing is to go to sleep later and and wake up later, but that's hard to do. Also, in fact, I, I always years ago I used to go to sleep at like 10:00 and wake up too early. So someone said, "You should go to sleep later," and it helps. You can sleep later, but that's just an N of one, so don't listen to that.
So, participants that had early sleep onset paired with early wake time, and those with late sleep onset and early wake time, which is not good, not enough sleep, they had higher GMSI values compared to those with either early sleep onset and late wake time, or a late sleep onset and late wake time. I know it's a little confusing, but remember GMSI, the higher the better. Uh, and, um, they did an analysis based on the sleep midpoint, which is another, uh, indicator. They showed that participants with a late sleep midpoint, which is a reflection of both the sleep onset and the sleep and the wake time, demonstrated significantly lower GMSI values. Okay. So, if your sleep kicks in at the midpoint, it's late. The midpoint should be earlier. It means you have a low, low GMSI score. Okay.
So, what's going on? What's the relationship between sleep quality and the gut microbiome, and also the, the metabolic pathways encoded in different species in the gut and the metabolites? So, they, they, they asked whether circulating metabolites could distinguish between people with good and poor sleep quality. Right? So, they did the, the metabolites of the, of the blood, and they looked in the two groups, and in fact, they, the two groups can be distinguished to some extent by the metabolite profile. All right. A total, total of 85 metabolites showed variable importance, uh, in, in these properties. Seven showed associations with sleep quality independent of age, sex, BMI, smoking, drinking, education, and cardiometabolic medications. And of these eight, pyruvic acid showed the largest magnitude of associations with poorer sleep, and the bile acid deoxycholic acid was associated with better sleep quality. Uh, this deoxycholic acid also had positive associations with seven sleep quality-related, uh, bacterial species, right? And so, they think this may be boosting lipid metabolism and immune, uh, modulation.
Now, they found that 44 microbial pathways, so these are metabolic pathways in the various bacterial species, uh, had distinct abundances between groups with different sleep quality. For example, three pathways related to vitamin biosynthesis, including pyruvic acid metabolism, pyridoxal phosphate biosynthesis, etc., showed strong associations with poor sleep quality. These, these pathways all have a, an essential step in metabolism. That is synthesis of 1-deoxy-D-xylulose 5-phosphate from pyruvic acid by an enzyme called 1-deoxy-D-xylulose 5-phosphate synthase. And this the abundance of that enzyme was positively related to sleep quality. So, this enzyme DXS is ubiquitous in microbial metabolism, and *F. prausnitzii* was the most abundant contributor to this DXS enzyme in their study, uh, population. So, in other words, the gut microbial functional capabilities, and particularly those involved in pyruvate metabolism, have a, a tight relationship with sleep quality. It's tantalizing, right? Because our gut microbiome is something we actually can shift by what we,
>> Can. Yeah.
>> Oh, yes. I can tell you that it, you can, you can do very interesting things with it.
Uh, finally, um, they found that both sleep quality and PSQI subscores, including sleep disturbances, latency, self-reported sleep quality, daytime dysfunction, were positively associated with frailty severity, as as mailed by the, as measured by these FI scores. And after you adjust for all these things like age, sex, BMI, etc., the frailty index was 0.18 units higher in the group with poor sleep quality compared to the group with good sleep quality. So, this, these are things that have been measured before and shown to be correlated, but this, they just wanted to see if it was the case in their own cohort. Among the sleep quality-related gut microbes that we've talked about, the abundance of *F. prausnitzii* was negatively associated with frailty severity, and also 35 of the 44 microbial pathways were related to PS, excuse me, were also associated with frailty index.
>> And in partic,
>> I was going to say, you can, you can actually witness this in one of their really neat figures, figure 5C, and it shows how the, uh, the fecal bacteria and *F. prausnitzii* really impacts this whole equation of whether or not you're going to get sleep duration, sleep disturbances, and your overall score. So, it's, it's, and the metabolites. So,
>> As, as my med students complain that we teach them metabolism, this, this is why we teach them metabolism. So they'll be able to, you know, figure this out when their patients bring them this paper and ask them, "Hey, what does this mean?"
>> Yeah. Ah, metabolism. It's come, it's, it's good to learn it because you may forget it,
>> But it will come back when you look at a study like this.
>> It, it really does. And, you know, their figures are are well done. In this, this could be a very complicated manuscript to dissect, but they've done an exemplary job taking the complex and explaining it.
>> And, and to generate a frailty score, you know, that, that took a lot of effort. And to, um, come up with a strategy for generating a sleep quality score because there are so many aspects of sleep. Yeah.
>> So, I really appreciated the thoughtfulness that went into that. So, one more experiment they did. They wanted to look at the role of the gut microbiome and metabolites in the relationship between quality and sleep quality and frailty. Again, the same question they've been addressing, but this time they do what's called a mediation analysis. Do what compounds mediate the, where you're going to be on this index. And pyruvic acid and multiple pathways involving it, and *F. prausnitzii*, which is playing a major role in pyruvate metabolism, had significant mediatory effects on the association between sleep quality and frailty. And so, GMSI, pyruvic acid, pathways, accounted for 10%, 4%, and 3.9% of the association between sleep quality and frailty, uh, index. So, um, you see now, overall, this study shows a mechanistic role of the gut microbiome in the sleep, uh, frailty relationship. And these findings may tell us that there is a significant mediatory role of these factors in the relationship. In other words, you may be able to somehow target them to treat the sleep problem and consequently the frailty problem.
Now, they do, they do say that this is just a preliminary study. We need to study, uh, more people. There may be other aspects of the, the control of sleep quality. They say, and this is very good. I like this statement. Uh, "The findings should be interpreted with caution as exploratory and hypothesis-generating rather than causal."
>> So, it's a very nice way to, uh, to, to summarize it. And they say, "We need further prospective studies. We need investigations into the mechanisms before we can come up with, uh, with any treatment." But boy, wouldn't it be great to take a probiotic and sleep better at night?
>> Yeah.
>> It's great motivation to pull more people into this field and let's, uh, let's get her done.
>> Sure.
>> Well, this is, so this kind of topic is people are interested because it relates to them,
>> Daily. Absolutely.
>> Nightly. I picked the paper because I was personally interested in it, and I try and find papers like this for all the podcasts that address things that are real-life problems, right?
>> Everybody can.
>> So, both papers today are really at that remarkable intersection where it affects you, the individual listening to this.
>> Yeah. And it took a large group. There were 19 authors on this paper, um, and four of them contributed equally. And I'd just like to highlight, um, their background. Yanni Pu obtained her PhD degree in bioinformatics from Fudan University School of Life Sciences. She's a postdoc, um, fellow right now, focusing on the human microbiome and healthy aging. She does have some advice: "Start before it's perfect. Just take action. Also, talk to people, share ideas, build a support network. Science grows faster when we connect with others." When Yanni is not doing their science, they enjoy, um, travel, badminton, and movies. Another co-first author is Shinji Xi, um, who is a PhD candidate in biostatistics at Fudan University, focusing on statistical methods for multiomics data and the microbiome-host metabolism interface. So, clearly, um, played a big role in this analysis. Her advice is, "Don't fear mistakes. Trying is progress in and of itself. Ask for help when stuck. Remember to care for your own well-being. A good run or workout can clear the mind." Wei Yang Wu is an associate professor at the Human Phenome Institute of Fudan University, managing cohort resources and biobank development with a focus on immune microenvironments and multiomics research. And their advice is to, "Stay curious. Enjoy small discoveries. Failures are normal. That's why we call it research and not search. Breakthroughs, breakthroughs come after many tries." And, um, Wei enjoys hiking and stargazing in their free time. And then Lily Hang earned her PhD in medicine at Shanghai Jiao Tong University School of Public Health, is now an assistant professor there, and focuses on environmental pollutants, lifestyle factors, gut microbiota, and human health. And her advice is to, "When entering a new field, build from what you know, add step by step, and don't work in isolation. Talk with your peers. It makes research clearer and faster." Great advice.
>> Last night on Office Hours, the live stream, I gave them a quiz, and, you know, some people got it right, some people got it wrong. And one person wrote, "Fail," because they got it wrong. I said, "No, it's not a fail to get the wrong answer. You learn something."
>> Yeah.
>> You'll never forget.
>> You'll never forget it. You've, you've learned, so it's not a failure ever.
>> Right.
That is TWIM 342. You can find the show notes at microbe.tv/twim. If you have any questions or comments, twim@microbe.tv. And if you'd like to support us, support this work of disseminating good science, go to microbe.tv/contribute. Michelle Swanson's at the University of Michigan. Thank you, Michelle.
>> Thank you. It was my pleasure.
>> Michael Schmidt's at the Medical University of South Carolina. Thank you, Michael.
>> Thanks, everyone.
>> I'm Vincent Rakinello. You can find me at microbe.tv. I'd like to thank the American Society for Microbiology for their support of TWIM and Ronald Jenkees for the music. This episode of TWIM was edited by Ray Ortega. Thanks for listening, everyone. We'll see you next time. And I'll give you my Pittsburgh Sleep Quality Index score.
>> We hope we all take the test.
>> See you next time on This Week in Microbiology.
[Music]