Urine is Not Sterile: What the Urobiome Means for Pelvic Floor Surgeons - Alan J. Wolfe
Alan J. Wolfe, PhD, Department of Microbiology and Immunology, Loyola University Chicago, Chicago, Illinois, USA.
Conference Coverage: AUA 2019: Urine is Not Sterile: What’s Next?
Alan Wolfe: About 12 or 13 years ago I walked up to a urogynecologist. I said, "We have two words in common, E. coli; I work on it. You want to kill it." Linda Brubaker said, "We should talk," and over the course of a number of one hour conversations, at one point she said, "You know, some people think the bladder is sterile," and I said, "No, flipping way." I'd studied bacterial motility and chemotaxis and biofilms. I couldn't see how in the world, at least in women, that 30 millimeters could be a barrier to a colonization by bacteria, the length of the urethra.
So these are my disclosures. The old dogma is that the bladder is sterile, that bacteria are not involved in most urinary disorders, and urinary tract infections usually occur when a uropathogen crawls up the urethra into the bladder and that it's very E. coli centric. But what we did was to test the dogma that the bladder was sterile. I'll show you how we did that in a moment. But the new discovery is that there are bacteria that normally live in healthy bladders, that these would be called the urinary microbiota or urinary microbiome, we call it the urobiome. And that urinary disorders are likely the result, in many cases, of dysbioses that is a disruption of the normal community of bacteria. And it's way more than E. coli.
So how did we get to the idea that the bladder was sterile? Well, it began in the very beginning of microbiology when Pasteur and others were, in essence, trying to disprove spontaneous generation, right? You all, when you were younger, read about the studies where someone would put like meat in the jar and they'd have two jars. One would be covered and one would be uncovered. And, of course, there'd be maggots in the one that was uncovered after several days and not the one that was covered. They did the same thing with urine. So they put urine in flasks and they covered one flask and left the other flask open. And after several days, the open flask was turbid, and the uncovered and ... excuse me the uncovered flask was, was turbid, and the covered flask wasn't. And they interpreted that as, first of all, there's no spontaneous generation. And second is that urine is normally sterile. It's a misinterpretation of the data.
It's flawed logic. Fact is, is that the bacteria that normally live in the bladder, that are normally in the urine don't grow at room temperature in air during the period of time, especially not in urine. Urine's not a very good growth medium. So the organisms that we've discovered that actually live in the bladder really like to be in 5% CO2. Of course, they like to be at 37 degrees C, many of them are strict anaerobes.
So, if you jump ahead to the 1950s Ed Kass who was an ID doc at Harvard was trying to come up with a test that would allow him to know when a patient who was going into kidney surgery actually had pyelonephritis. He was particularly interested in E. coli, which is indeed the most common ideological agent of pyelonephritis and he set up what's now considered to be the traditional or standard urine culture test that the clin lab does. It's very, very good at detecting E. coli because that's what it was designed to do.
Somewhere along the line it got moved into the bladder. People started using the same test to determine whether someone had cystitis. And you know that the test doesn't work all that well because people have been arguing about thresholds for 50 years now. So, as a microbiologist, I knew that there were lots of organisms that simply don't grow under any of the typical growth conditions. So Linda and I decided that what we were going to do is test the dogma.
So I knew that, now 13 years ago, the term human microbiome was becoming prevalent. The numbers are now been ... it used to be people thought it was 10 to 1 bacterial cells to human cells, but it's actually closer to one-to-one, which is still a lot of bacteria. The way I tell the medical students is that they're walking, talking incubators. They better get over it. We can't live without them. They provide us with vitamins and other nutrients. They provide some of our energy and they definitely protect us. One of the things they do is educate our immune system.
So since other organisms that had been ... excuse me, since other organs that had been considered to be sterile were not, like the lungs. And since we knew that most bacteria didn't grow under the standard urine culture conditions we proposed that the bladder was indeed not ... the bladder was not sterile.
So this is how we went about it. We needed to get urine out of the bladder that we knew was coming from the bladder. And we took a number of different samples. We took voided urine, the typical midstream voided urine that people erroneously call clean catch because it's not so clean. Vaginal swabs, transurethral catheterized urine, suprapubic aspirates. We took a skin sample where the needle went into the lower abdomen and a sham needle stick that went into the lower abdomen but didn't go into the bladder. All of these individuals were going in for surgery. The voided urine was tested by the standard urine culture test. They were all negative. That is, the clin lab reported no growth, okay?
So the question now is, can we detect bacterial DNA in culture-negative urines? And we used, what's called, 16S ribosomal RNA gene sequencing, and next-generation sequencing technology. So you use the 16S ribosomal RNA gene because it's part of the ribosome. The ribosome is essential for life and so there are parts of the gene that don't evolve at all and other parts that evolve, but they evolve very slowly. The importance is that you can use the differences in the sequences to identify bacteria at the species level, okay?
So what we found is that the aspirated urine and the catheterized urine resembled each other in terms of the bacteria that we detected, and the amounts that we detected. And that the voided urine did not look like cath urine and it actually looked more like vaginal swabs at least too often to use voided urine to ask anything about what's going on in the bladder. Since the aspirate bypasses vulvo-vaginal contamination we were able to conclude that the bacteria that we were detecting had come from the bladder. So the other thing that was really important from this study is that we now knew that we could use catheterized urine to study the microbiome of the bladder. So most of our studies have used cath urine.
So the next question was, are these bacteria live? Because you can sequence the DNA of bacteria that are dead. So I went up to Paul Schreckenberger, who at the time was the director of Loyola's clinical microbiology lab, and I said, "Paul, I'm looking at the genera that we can sequence, right? And I know you can grow these guys, but they don't grow by standard urine culture." So he went back to the clin lab, sat down with a master's student at the time, Evann Hilt, who Paul and I were both mentoring. And they developed something that, they call it, expanded quantitative urine culture or EQUC.
So this is the standard urine culture. It's one microliter of urine on blood agar and McConkey agar grown aerobically for 24 hours at 35 or 37 degrees. So what we did was to plate 100 times more volume on a number of different media, a number of different atmospheric conditions, CO2 air as well as anaerobic. And we allowed those plates to incubate for 48 hours. So we published this, Hilt et al., about five years ago. Now, I'm going to show you one example. It's going to be a blood agar plate. It's going to be the same urine, okay? Actually two blood agar plates. The first one that urine's going to be plated as if it's the standard urine culture, right? One microliter blood agar for 24 hours in atmospheric conditions. And then the next plate's going to be 100 microliters on a blood agar plate, the same urine but incubated for 48 hours under air, okay?
So here's the first one. That would be reported as no growth. And if you didn't know any better, you'd say that urine is sterile. It's not, it's not at all, okay? Most of these got lots of different colony morphologies up here. Some of them are tiny and there are lots of them. They're slow-growing. They wouldn't be picked up after 24 hours, but they're clearly there at 48 hours. And some of these other morphologies they're just not there in very high numbers. And so if you only put one microliter on a plate you'd never detect them.
Okay so we now know that the urine taken from the bladder is not sterile, and it has a mixture of organisms. And so now the questions are, all right, so who's there? What's in normal urine? Does the bacterial composition, the diversity associate with any disease state? What are the bugs doing? Do they differ amongst the various different pelvic floor sites? When is the microbiome established? And how do the bacteria change over time? I don't have time to answer most of those questions, but I'm just going to pick out a few of them.
So the first is what's part of normal flora? I'm going to show you studies from women and then from men. So this is 224 women that are asymptomatic so it's all of our controls. In this case, we used EQUC to characterize the microbiome. We also do 16S sequencing and something called shotgun metagenomic sequencing as well. All right, so the blacks are negative. They come up culture-negative by EQUC. About half of those will be positive by sequencing, and I can't tell you about the other, about the quarter that are negative, negative. We don't necessarily think that they're sterile, we just call them below the threshold of detection.
But what you can see is that in women, the largest group of individuals are dominated by Lactobacillus, which in our color scheme will always be that color blue. Another group is dominated by Gardnerella, another by Streptococcus. And then there's a little bit of Staphylococcus. And you see these are asymptomatic controls and some of them have E. coli. And then all those other colors, they're mostly organisms you've never heard of before.
Here's men. We took voided urines and cath urines from men with BPE. And what we found is a number of things. First, you can see that in the catheterized sample we can detect bacteria in at least the bladders of some men. And if you look carefully, you can see that the voided urine and the catheterized urine doesn't look the same in many instances. We think that the voided urine is sampling primarily the urethral microbiome and that it's distinct from the bladder.
Okay. Oh, you'll also notice that men don't have a whole lot of Lactobacillus or Gardnerella, there's some there though. Okay, so does the bacterial composition of diversity associate with disease? I'm going to show you one study that we did with women that suffer from urgency urinary incontinence. So we had women that had UUI, women who were asymptomatic, and we did EQUC and 16S ribosomal RNA sequencing on them.
The first thing I'm showing you is, what we call, a species accumulation curve. So it's a rare faction curve. An individual comes in and you ask the question, "Well, how many unique species did we identify?" And you put that down as number one on the X-axis. And, let's say, there were three unique species, put three on the Y-axis. Next person comes in, you test their sample, and there's one unique species. Maybe there were two that we'd already seen in the first person, but one unique. So now it's at number two on the X-axis, and now we have four unique. And what you're looking at then is the path or the trajectory of those curves. And what you can see is that the cohort of women with urgency urinary incontinence had more unique species than the individuals that did not have UUI. So what we would say is that the cohort of women with UUI is richer in bacterial species than those that are asymptomatic.
So what's interesting is that we can go in and ask, which organisms are most associated with individuals with UUI versus those that do not? And you can see some organisms that you might recognize Corynebacteria, Gardnerella, lactobacillus, Streptococcus. But most of you probably have never heard of Actinobaculum, or Arthrobacter, or [Al-la-gel-a 00:14:42]. I've got a red arrow pointed down, this is to show that all Lactobacillus are not created equal. Lactobacillus gasseri is associated with individuals with UUI. Lactobacillus crispatus is associated with the controls that were asymptomatic. As I put it some times, if I was a prepubescent girl standing in line to get my vaginal microbiome I'd pick Lactobacillus crispatus every day and twice on Sunday. It is good there. It is also good in the bladder. It appears to be a protective organism.
Okay, just to show you that UUI, SUI, and UTI communities are different I've got these pie charts here. And what you can see is individuals that are asymptomatic have a different composition of bacteria, just look at the colors, than the individuals with either of the two urinary incontinence conditions. And that they're really quite different from those with urinary tract infections. The yellow, the goldish color, that's E. coli. And when I put UTI up against recurrent UTI, those pie charts look quite different. E. coli becomes much less dominant.
Okay, so what about men? Is there an association with symptoms? So in that same study, what we did is use the IPSS score, and we ranked the individuals from low to high, and showed that there's more color to the right so, as the score goes up it's more likely that we'll detect bacteria in the catheterized sample.
Okay, so far what I've shown you is that microbiota are distinct between individuals that have UUI and no symptoms, that microbiota are distinct between UTI and non-UTI. A study we did showed that we could probably phenotype individuals before giving them treatment for OAB because some individuals simply don't respond and it's associated with their pretreatment microbiome composition. We have evidence that post-operative UTI risk is associated with the pre-surgical microbiome. And there are microbiota associated with urinary stones of every type of chemistry.
And in the last minute and a half, what I want to show you is that bacteria do change over time. So these are eight young women that were recruited and paid to sample themselves every morning and to answer all sorts of very invasive questions about just about everything they did. And my former graduate student Travis Price lived in the BSL 2 room for a year. That's a special room where you can work with pathogens.
So we found three patterns. Four of the individuals were dominated by Lactobacillus throughout the three months. Two individuals varied between a Lactobacillus-dominant and Gardnerella-dominant microbiome. And another individual had various ratios of Streptococcus, Staphylococcus, and Corynebacterium.
So what alters this pattern? And what we found is that there were two things in a 20 or 30 something-year-old woman's life. And the first, of course, is menstruation. So the red dots over the top are the days of menses. And what you can see in the individual that was dominated by Lactobacillus is that the composition of the lower urinary tract urine was really quite different. And that the individual that had Strep, Staph and Coryne was the brown showed up. It's Actinomyces, and it was very repetitive.
And the other thing that young women tend to do that altered their microbiome was to have vaginal sex. And so almost every time someone had vaginal sex, this green showed up. The green is Streptococcus, right? And what Travis did was to re-up the woman on the right and her partner and they swabbed themselves in all sorts of different places multiple times across a three week period. And found that they were definitely sharing microbes, that the green was actually Streptococcus mitis, which is an oral organism. And the Streptococcus mitis is definitely something that lives in her lower urinary tract.
So what's next? We want to know when the urobiome gets established. We have some evidence that we can detect bacterium bladders of little boys and girls that are no older than two months that we think very early. Does the host environment influence the urobiome and vice versa? How do members of the urobiome interact with each other? How do they interact with the host like the urothelium? Which bacteria are protective, which ones are pathogenic? What mechanisms permit these bacteria to reside in the bladder? And how do we translate our discoveries into the clinic?
This work is done by a large group that we call the Loyola Urinary Education and Research Collaborative or LUEREC. It involves basic scientists like myself and my group, clinical microbiologists, clinicians in urogynecology, and urology in maternal-fetal medicine, and nephrologists.