respiratory tract infection drug that is safe for pregnant women

>> good afternoon, everyone. welcome to the wednesday afternoon lecture on a tuesday. and thank you all for being here, those of you in masur and hello to those watching by video. i think we're in for a real treat here.

this is the annual lecture named for rolla e. dyer, the nih director from 1942 until 1950. he presided over the time when nih became plural. he started with national institute, singular, of health and by 1950 we were plural because of the introduction of

things like the heart institute and the dental institute and the mental health institute. he was also an expert in infectious disease so we always try to choose a dyer lecturer who is in that space of microbiology and we certainly have a wonderful example of that

today. claire fraser, is a distinguished investigator in the area of the microbiome and her topic today is functional dynamics of the gut microbiome in health and disease. she's currently professor of medicine, microbiology and

immunology, and director of the institute for genome sciences at the university of maryland school of medicine. she did undergraduate work in rpi got her ph.d. in pharmacology at suny buffalo. then on the faculty atlanta roswell park until coming --

faculty at roswell park yes, she was from 1985 to 1992 in ninds and niaaa arising to become section chief of molecular neurobiology. in '92 she moved out there not far from here, to be vice president and then by 1998 president and director of the

institute for genomic research, where many exciting things happen related to uncovering genomic aspects of microbes. since 2007 she has been in a position i just named a moment ago at the university of maryland in baltimore. she's elected member of the

institute of medicine, she has been inducted as member of the maryland women's hall of fame. she's a bureau's wellcome fund visiting scientist professorship awardeend it's a great pleasure to have her here to talk to us about something is happening and i think causing a great deal of

interest across many different field it is way in which the gut microbiome plays a role in health and disease. there's hardly anybody better to talk about that than dr. fraser. join me in welcoming dr. fraser. [applause] >> thank you very much,

dr. collins. i'm absolutely delighted to be here. that was a very generous introduction. and as was mentioned in my introduction i have been involved in microbial genomics since 1995.

this is actually the 20 year anniversary of the completion of the influenza genome, when i think back to those days, it's in many ways seems like yesterday and other ways seem as lifetime ago. but what i'm going to talk with you about today is work from my

lab that represented a natural evolution of the studies from focusing on single microbes that we can grow in culture to now studying complex communities in various environments, the environment that's most interest to me is the human gi tract. this does not bode well, if i

made it past the first slide. while we're getting technical help here, let me continue and just ask you to think for a minute when do you think the notion that we co-exist with complex community of microbes first appeared? i won't ask a show of hands or

ask for people to throw out time lines but you maybe surprised to know i'm going to assume -- yes. great. thank you. i won't touch, get rid of that, i don't want to touch that. so here we are. here is the the answer to that

question that i just posed. it was actually an ton von lavenhook that first described the microbe that live in association with humans. one of the first samples that he examined after his construction of light microscope was material he scraped off the surface one

of his teeth and suspended in solution. he wrote about what he saw in a letter to the role society of london. he was in awe of the large number of what he called small animal calls that he saw in this sample.

in fact that there were so many in the water that the water seemed to be alive. so that was a long time ago and it goes to point out very often in science it can be difficult to have an original idea. for the better part of the 350 plus years since von lavenhook's

first description of the oral microbiome, morphological description, i think we were very much in the dark about what this community of microbes that lived in association with humans does. it was all thought rather mysterious, this reflects the

fact that we didn't have the tools to study these communities at molecular level like we do for a long time this was described as the normal flora, which we don't use that term any more, flora refers to plants but the thinking was that our microbiome was mostly inert,

relatively stable, it wasn't relevant to disease and when we were thinking about microbe important for human health, the only thing we needed to focus on were the pathogens. all that has changed in the past ten years or so. i think we owe a great debt of

gratitude to the vision of dr. collins when he was then still director of the genome institute in assembling a group of people to begin to put together a plan and construct a vision for what the project might look like to study the human microbe.

the human microbiome, this one a project that make use of all capabilities all the fire power, all the human capital that had been assembled to complete the human genome project and then go on to begin to look at additional genomes, the functional units of the human

genome and the encode project. we knew that these same approaches that my colleagues and i had been using could aptly be applied to the study of microbial species on earth, estimated at 10 million of those species, few are pathogens, most are beneficial.

so i remember being present at what i think was the very first human microbiome planning meeting, it was at a hotel in rockville on the eve of valentine's day, many people who were there were bemoaning the fact they weren't home with loved ones but it was a

collection of really interesting people with expertise in microbiology, genomics and ecology. and it really began to construct a framework for what tease human microbiome project officially launched this 2008. this project was designed to use

culture independent methods to characterize the microbial communities living within us both on all surfaces and cavities of the human body. note on this slide that i have culture independent highlighted in red. that was a very are important

aspect of this project because we know, and field of microbiology has known for a very long time if you go into the natural world and try and take complex communities from their habitat and grow them in the lab you fail miserably. at best you can expect to

culture ten percent, sometimes not even that large a representation of what is there because these are interdependent communities of organisms and we don't now how to capture the conditions that are necessary for growth of all of them. but tools and technology, the

approaches in 2008 had absolutely matured to the point where we can take a starting sample, whatever it be whether human sample, soil, ocean water, extract nucleic acid and use the tools of genomics to begin to get tremendous insight who was living in that community.

and so that's what we began to do collectively and i should mention at the same time there was another project going on in europe, the meta hit project focused just on the human gut that also had added tremendously to our body of knowledge. but the human microbiome project

focused on five major body sites; i will get to those in a second. this was a fantastic consortium that functioned well for five plus years. there was a great focus on understanding the communities, in a significant number of

healthy subjects, also demonstration projects that i was involved in that before focused on trying to demonstrate whether or not there were links between altered microbial communities in particular diseases. there was a data analysis and

coordination center led by owen white. as is -- was the case with the human genome project, focus on technology development computational tools and a number of lc projects as well. by all measures we can say this was a resounding success and i

was delighted to be part of the project. so i think we now -- we have a different view of the human microbiome than we did prior to all this effort. i would say today we really understand the that this is a co-evolve ecosystem that results

in a mutual benefit for both us as the host, and benefit of these commensals that share our space. i tend to take a gi centric view of the human microbiome but there's also focus on oral cavity, the respiratory tract, skin, vagina and all these are

important. in the gi tract we know now that our microbial partners contribute to the digestion of our food, provide energy, for our metabolism, make a number of essential compounds such as vitamins, short chain fatty acids can also make a number of

toxic compounds and i thought it was interesting in the news yesterday the who recommendation to avoid processed meat because they have been deemed to be carcinogenic. well i would argue that a large reason they're carcinogenic is because of the action of the gut

microbes in breaking down the components that you find in processed meat. a really important part role for the gut microbiome is development and maintenance of the immune system, weal touch on this in a bit. there's a growing interest in

better understanding the gut brain access because in a number of animal studies they -- the gut microbiome has been shown to influence neural development and behavior. in all human environments a shared function of the microbiome is acts first line of

defense against pathogens. in the spirit of dr. dyer, this is the topic here that i'm going to focus on today and talk to you about some of the work going on in my lab. couple of thoughts about the microbiome before we move on, it's believed only 10% of the

cells in the in the human body are human, the rest are microbial cells and 99% of that microbial mass is located in the gi tract. if you make some assumptions about genome density and genome size and bacteria and other microorganisms we can do

calculations and quickly realize that the microbiome likely contributing more than hundred time it is the number of genes to our biology as there are genes in the human genome. what's interesting is that we know a number of the genes that are contributed by our microbes

encode activities that are not present in the human genome. so if anything the microbiome project has forced me, somebody who has made a career as a genomics person taking a someone a reductionist view of biology to rethink all of that and consider human health and

disease from a much more holistic perspective that includes not only us but our microbial partners. just a couple more slides to set the stage. all these studies have benefited tremendously from multipleomics approaches, meta genomics, meta

transcriptomics, proteomics, metabalomics. we are interestingly going back to culture. i made point we started the project realizing we couldn't useture as first step but there's an interesting going back into culture to try to

better understand the properties of some of the interesting bacteria that have been -- that have come to light as a result of these efforts. one of the most important approaches used almost universally and i just bring this up because i'm going the

make reference to this kind of data throughout the talk, is the use of a 16s ribosomal rna gene as a marker. i like to think o of this allowing us to take a molecular sense and count how many of what different types of organisms are there.

what the properties of this gene that allow this to happen, you can see the secondary structure here, it's a gene of 1500 base pairs that contain a number of variable regions interspersed with highly conserved regions being the peaks here. we can design pcr primers

against conserved regions and am pretty across one or more variable regions and with sequence analysis what that does is essentially provide us a bar code. that allows us to distinguish with a very high degree of resolution the different types

of organisms that are present in the sample. so we start with dna, pcr amplify 16s gene, we get the sequence and then search that against a number of databases growing all the time to do phylotype classification and compare community composition

across samples. what we have come to understand is that our body represents the microbial ecosystem, serves as a scaffold for wide range of different microorganize. s, primarily bacteria but also key of ours is microeukaryots and data from first series of

studies funded by the human microbiome project reported in great detail. the difference is in the composition of these communities, that are associated with different body sites. these become predictive of a body site i would venture to

guess that you could show somebody who is familiar with this data overall composition and they can tell you with high degree of certainty what samples was -- has been studied. and i think also when we consider thinking about all these results we really need to

remember that what we are studying is an e co-system. a number of principles of microbial ecology have been shown to apply and we're going to talk a little bit about the idea of stability and resilience as a property of these microbial communities.

other data from the hmp demonstrated there is -- while there could be a tremendous amount of variability in who is there and this is data shown as the phylum level in different samples, from different environments many the human body and these represent just a

series of bar graphs from hundreds of samples, when you look functionally at the predicted metabolic pathways that are encoded in meta genome data you see a lot of this variability tends to disappear, while these communities differ in terms of composition they are

much more similar in terms of function and that makes sense as we believe that these functions are critically important for human health. another idea that has come out of studies in a number of human environments but still focusing on the gut is idea that the this

tremendous extent of interindividual variability can be simplified somewhat and it was first described in 2011, the idea of community types or intratypes as a framework that can be useful in studying differences in these communities and what's captured in this

intertype concept is the idea that if you were to sample the gut microbiota in us you find each of us was dominated by the presence of a single bacterial genome. the original report suggested that there maybe three intertypes, i think the number

is probably greater than that but it's not infinite. and that i think that can be helpful these intertypes tend to be stable over time and there's factors that seem to be at play determining what your intertype might be, these include host genotype, diet, intestinal

transit time is one of the most recent reports that i have seen. and there are likely other factors that contribute that we haven't figured out yet. we know in the young these communities that are developing, they are fairly low diversity. very dynamic, you reach

adulthood, these communities mature, they tend to be more diverse fairly stable and with old age these communities tend to lose diversity again. we also know that there are a number of factors diet as mentioned before, antibiotic treatment is another big one

that can influence these communities, disease, there are many, many links between changes in the microbiota and disease, i think this gives us a framework for how we interpret data in the context of a life span. so now with that as background, let me get into the meat of my

talk where we're going to focus on pathogens and their interaction with the gut microbiota. this is a fabulous guinea gur from a review article published a couple of years ago. what it is trying to depict here are what we think some of the

differences might be between the healthy or intact microbial community in the gut and disrupted community. let's begin on the left, healthy microbial community, there is -- we know that the microbiota sits in fairly close association with the epithelial cells that line

the gi tract, those cells are covered with a thick layer of mucous, some commensal microbes are present in association with the mucous, some in the lumen. and there is a balance that exists between our microbes and the gut associated immune system that lies right below epithelial

cell layer. we know there is cross talk in both directions that the microbes can have impact on adaptive immune response host can stratify microbes in the lumen particularly through secretion of iga. and when the system is in

balance though there's exposure to exogenous pathogens, they are of little effect. contrast what we see on the right hand side, disrupted microbial community could be disrupted through the administration of antibiotics, there's been focus on the

effects of antibiotics on these communities and is probably analogous to dropping a nuclear bomb on h this ecosystem. you can end up with a depleted microbiota, some of the pathogens that maybe present or passing through can overgrow. there can be epithelial cell

damage which leaves open the lumen of the gi tract to gut associated immune system, you can get systemic kiss semination of pathogens and commensals and this can often drive a very robust inflammatory response. so i have been very interested in trying to figure out exactly

what happens in going from a healthy commune thety to a disrupted community, particularly interested in asking whether or not how easy it is to go in the other direction from a disrupted community back to an intact community.

the other thing i should mention when you end up with a disrupted community, it's not just the host and the microbiota that can play a role in the outcome of infectious disease, a number of pathogens in this case salmonella and enteric pathogen can use strategies to further

disrupt this system. so there's a lot going on. and i have been very fortunate to have an opportunity to be involved in a number of collaborative studies with stein and colleagues at vaccine development at maryland to look at this inner play between the

gut microbiota and enteric pathogens. i will tell you a couple of stories today. these are preliminary studies and we have some intriguing results and they raised far more questions than we have answers for but it's beginning to give

us some insight i think, to suggest this interaction between pathogens microbiome and host doesn't necessarily operate the same way in every context. this is work looking at shegella in vince young's lab in michigan. so we know with shigella, there

are for serotypes that can cause disease, shegella attacks tocolonnic epithelial cells, enters the cells and there's a well cascade of effects where shegella can get into the bloodstream, the cause shegella is endemic worldwide, it's responsible for 120 million

cases of distear each year. dysentery each year. that's live attenuate waited vaccines against shegel will bea in various phases of clinical trial. but one thing that's within a problem not only with development of shegella vaccines

but others is the fact that vaccines show different efficacy across various populations and oftentimes these vaccines when given to children in developing countries are less effective than they are when given to children in western society. and the reason for that is

really not entirely clear. so we set about to study the interaction of shegella infection and gut microbiome and vaccination in a cynamologous macaque model of shigalosis, first to characterize the gut microbiota in the macaques, this is the overall relative

abundance here. what we see is the same organisms at the phylum level present in the monkeys are what we see in humans but different relative abundance. this is just showing you the relative abundance of the gene that's level.

we also then went on to look at all the data that we generated and ask whether or not we saw evidence for the presence of enterotypes in sin moll goes macaque cynamologous gut microbiota and we did. we found evidence for four community types, relative

abundance of dominant members of these communities are shown they are clearly different. we also look at the shannon diversity index, a measure that takes into account the different numbers of organisms that are present and relative abundance and what we found was that there

was a very high diversity community type 2 very low diversity community type 3 and two other community types equivalent in diversity in between those two. so this was the study design that was followed. this was really looking at the

impact of two live attenuated vaccines against shegella developed at center of vaccine development comparing these to a pbs control in this macaque model. there was a first vaccination, second vaccination, at four weeks, and then after another

month there was a challenge with wild type shegella ideally studies are done in humans but it is not easy to get irb approval to do a wild type shegehla a challenge, thus the use of monkey model. the idea was that there would be a clear winner between these two

vaccines from this first set of experiments. there was a winner many this vaccine 1256 was further evaluated in a different set of animals under a different immunization regimen. larger number of doses over a shorter period of time, again,

followed by wild type challenge. seems simple enough in concept, my colleague marsella stein was doing immunology measurements and microbiome measurements to see whether immunization and challenge affected the gut these are the data from the monkeys in study one.

these are the two vaccines here in pbs. what we have color coded here are the four different community types that i refer to before. monkeys started with commune thety type 1 following each immunization there was a transient shift in the community

type 2 community type 3 lower diversity after the first immunization essentially all the monkeys recovered. after the second immunization that recovery with was slower again think about your ecology principles here, the concept of resilience.

and this red triangle showed the challenge with wild type shegella there was emergence of community type 4 that persisted in a number of monkeys. you can see these changes were relatively consistent across all the animals in this study. when we then look at clinical

symptoms in these animals, this is what just appeared now, green means normal stool. yellow to red represents die rera, different severity. essentially all the animals showed clinical symptoms even one of the vaccinated monkeys did but so there was evidence of

clinical disease in essentially all these monkeys. with we saw the same in study animals not vaccinate but challenged with wild type shegella contrast to second study looking at one of the vaccines, here now in these four doses, this case the majority of

the animals started out with a different community type, the high diversity community type 2 and this tended to be fairly resistant even following four doses of vaccine over a seven day period and more so after shegella. when you look at the second set

of bars looking at the clinical outcome with the exception of one incident of loose stool or mile diarrhea in one monkey, there were no clinical symptoms seen in these monkeys at all. my colleague asked what's going on, there must be a problem with this preparation of of shegella.

it may not be fully virulent. after much experimentation he convinced himself the problem wasn't with shegella, there was something going on here in these animals. we saw these differences and were anxious to figure what the possible explanation might be

for the difference in clinical outcomes. and one of the things that we looked at was the geographic origin of these monkeys that had been ordered at different periods of time for these two studies. and it's been known in

cynamologous macaques there can be genetic diversity that maps with geographic origin. and this maybe related to differences many the mhc regions in these monkeys. the mhc haplotype in macaque is important in disease susceptibility particularly when

studying siv so we look at genetic diversity and brought non-mhc and mhc loci in population. 10 and 20 short repeats to determine genotypic diversity and relate it to geographic origin and seven microsatellite regions spanning the mhc region.

this is looking at the short 10 repeat data. what you see is a group of monkeys shown here in green that were all from separated from the rest of the monkeys that were from endochina or philippines. comparing these data to data that had been published

previously, suggested to us that the geographic origin suggested by the vendor was indeed correct based on the genotype data that we obtained. i'm going to tell you as you might suggest, it was these monkeys that were essentially used and studied too.

so we're starting to see they're different and they appear to be different in terms of their susceptibility to infection with wild type shegella. we also did the same analysis at mhc repertoire, class 1 and class 2 gene and when we construct a phylogenetic tree

based on that information we again see the maricious monkeys in green cluster and the rest of the monkeys in this study. so where with we ended with this initial project was that we clearly saw that there were differences in -- also differences in immune response

and i'll show you that in a second but differences in clinical outcome that seem to map with different genotype with different mhc composition, and with different microbiota competition, it was the community type 2, the high diversity community type that we

found in the moricious monkeys which e w saw no symptoms of clinical disease so we looked further for correlation between strength and type of immune response and microbiota. this is the figure that we ended one this network figure t. it's busy and hard to know

what's going on here but there seems to be a very complex relationship between all of the genre present in the monkey microbiota, the stool for clinical outcome and the antibody response. so we are trying to dig down deeper into this to figure what

might be going on. the first study suggested to us genotype may play a role shaping microbiota composition, that's not surprising there's a number of reports on that. it suggested to us that this high diversity community type might be protective against

shegella but also suggested to us that there maybe a need to characterize the gut microbiome in future vaccine studies particularly focused on enteric pathogen. so let me go on and tell you another story about another enteric pathogen that's the

subject of great interest at cbd, this is salmonella. salmonella is causative agent of typhoid fever, it's a human restack of pathogens that has severe impact on global public health, 27 million infections annually with 1% mortality. there are twolysis vaccines

currently available, again, salmonella and for -- p you go back to the '60s and '70s there were a number of human challenge studies that were done using wild type salmonella, they fell out of favor and coming back, they are important in assessing host response to

immunization, because you can imagine in the development of a vaccine against human pathogen the inability to do a challenge study, this is something we have seen most recently with ebola really is a limitation because we don't always know if we are identifying a good protective

correlate when looking at various community parameters. a lot has been done in deciphering the complex immune response to s type e in humans. i'm not going to get into any of this except to say that this immune response appears to involve both humoral and cell

mediated immunity and it involves a number of antigen presenting cells as well as various effector and regula aretory cells. so here is a first case study looking at the impact of ty 21a typhoid vaccine. this was a study to look at

effect of vaccine on local immune response and also on the gut microbiota. we were curious whether perturbations whether there were prob tenial perturbations of gut microbiota following vaccination and whether we could identify any associations between the

microbiota and immunogenicity that was elicited by this oral vaccine. this was the study design here. we looked at single dose or four doses of ty 21a vaccine compared to what we saw in group of unvaccinate subjects. this did not include a

challenge, this was just looking at immune response. as within outcome. this is a summary of microbiome data using the jenson shannon divergence index comparing across samples the size of the circle here represents differences, largest circle here

represent large differences small are circled smaller differences, these are vaccinees these are days followed, what you can see is vaccination single dose or four doses here, didn't create any consistent changes in the structure of the gut microbiota as assessed

basissteen rna studies. we then looked at cell mediated immune responses, this is one piece of the data, multiple parameters examined but these are the cd8 positive interferon production over time days post immunization in the vaccinees that received four doses of ty 2

#a vaccine, what you can see here is there are two different types of responses. a late response in two subjects doesn't peak until about day # 2. the other four subjects was a multi-fay sic response, increase in interferon production seen

starting relatively early after the immunization here. we look for correlations between two different response it is multi-fay sic response an late response. this is a heat map here, summarizing 16s data of the microbiota, we found indeed we

could separate the late responders from the multi-fay sic responders you can see here in late responders much higher level bacteria didibut there were a group of fermicutes here the clostridial microbes more abundant in the multi-fay sic responders.

-- phasic responders. when we look at phylogenetic diversities in the microbial communities with every measure we look at these communities in which we saw a multi-phasic response deemed to be a more robust immune response with associated with higher diversity

communities than what we saw in the individuals that exhibited just the late response. we also looked at correlations between the gut microbiota and humoral response to ty 2 #a. this is using a cut off of four fold anti-body titer in number of key antigens.

as a way to define responders versus non-respond p bars. these are bar charts that show you relative abundance of genre color co-ed by phylum. what's very apparent is there is indeed a difference when you look at responders versus non-responders using antibody

titers as a measure of the immune response. what we see is that in the responders these -- again, these tend to be much higher diversity communities so i hope you see there is a theme here high diversity in the go microbiota is associated with more robust

immune response or in the shegella study immune response against infection. we create correlation networks based on antibody data, there was no networks when we looked in the non-responders but in the responders we saw much more complex networks.

this is now looking at components of the microbiota and various immune responses. so we are now in the process of trying to delve into more detail but we have some clues now based on these kinds of data about what components of the gut microbiota might be associated

with a more robust ty 21a our challenge is going to be to identify the causal relationship. we have had the opportunity to partner with andy pollard's lab at oxford who is carrying out a series of vaccinations, trials with ty 21a and newer

anti-salmonella vaccine. what's interesting about this, they received approval to do a challenge in these vaccinees so we'll have an opportunity to look at what happens when human subjects challenged with wild type salmonella. we generated meta transcriptome

data. what that suggested to us in just our initial review of that data is that in the subjects who don't develop typhoid disease following challenge, the gut -- what we see in the gut microbiota is a very robust and sustained stress response

globally and subjects that end up with typhoid disease, that stress response seems to be completely missing so stay tuned , this will turn out to be very interesting. so where we are i think the we can be looking at maybe a well balanced high diversity

microbiota comparing to low diversity or dysfunctional microbiota and what we maybe seeing that there are indeed differences in overall vaccine responses or susceptibility to infection with enteric pathogens, i believe we are the first group to begin to start to

look at some of these questions i will be the first to admit these are tough studies to take forward to figure out cause and effect but we're willing to do that. if you will bare with me a couple of additional minutes, i want to leave you with another

closing thought and provide additional data. what's given me pause to think about is if the indeed the composition of the gut microbiota correlates with clinical outcomes when thinking about enteric pathogens, it certainly raises the question in

my mind whether or not we can intentionally manipulate the microbiota to achieve a beneficial outcome. perhaps prior to immunization. how might we do that? number of ways but involved in a series of studies to look at effect of probiotics on

that's certainly a viable potentially viable approach to begin to think about deliberately pa manipulating microbiota for outcomes so running through couple of slides with data we recently published on, this was a collaboration with pat hubbard at mass general

hospital work done by emily (inaudible) in my lab. this was part of a series of clinical trials toe look at effect of single organism probiotic lakotasbiosis, you can buy this as cultural capsules. and this was to -- the idea of the clinical trials was to look

at effect of lgt as potential adjuvant for mucosal immunology in elderly subjects receiving the list of vaccines. i was going to present data today from the overall label part -- open label part of clinical trial assessing safety in 12 volunteers.

it was a simple study design, 12 volunteers per day for a period of 28 days. there was then a 28 day wash out period and my lab characterized the microbiome in these subjects using 16 s ribosomal rna meta genome sequencing and meta transcription transscriptomic

approaches. this is a summary of the 16s data, there's a lot on the slide suffice it to say the administration of lgg for 28 days had absolutely no effect on the composition of the gut that was not a surprise to us, that had been reported previously and does not colonize

the lgg tract so we wouldn't have expected to see anything. where things got far more interesting was when we look at our meta transcriptomics data set and this was a figure showing what we did. we made use of the reference genomes generated as part of hmp

and mapped rna seq reads back to reference genomes as a means of doing tax nomic binning and once reads were assigned to a particular taxa went to do gene annotation. one thing we were not expecting but became very apparent is we were analyzing the rna seq data was emergence

of what appeared to be three trains scrip tome groups based on recruiting reads to various genomes. i think this is being analogous to what i described earlier about enterotype defined by 16s but this is now looking at different transcriptome groups

based on functional properties based on the genes that are actively being transcribed. these were the three groups you can see here. we characterize these a couple of ways, one was looking at the dominant taxa, group one high diversity type dominated by

multiple members of the fermicutes phylum. group 2 was distinct group where predominant transcripts derived from gut antigens an group 3 dominated by transcripts from a single species of pneumococcus. we look functionally at these groups and found that based on

mapping predicted genes to k categories that there seem to be some functional differences across the three groups. if we then went to look at the in vivo behavior of the probiotic lgg, what you see here is this is a schematic diagram of lgg genome, these are the 12 consent trache

circles represent data from each of 12 subjects, outer circles represent meta genome data but we mapped the reads in all the data sets back to the lgg genome and it was very comforting to see that we only saw evidence of lgg expression at day 28 at baseline and at the day 56 wash

out there was no evidence of lgg present. looking more closely at what's going on at day 28, there was strong correlation between transcript recovery and genome abundance from meta genome data. there was high level concordance between lgg expression and

transcriptome type. the four outer most circles which we see the highest level of lgg expression were all from members of high diversity transcriptome type that i the described a couple of slides ago. the highest level expression

within lgg across all subjects tended to be associated with a number of small non-coding rna and we're not sure what that means other than there maybe a lot of gene regulation going on. we asked what at global level whether the presence of lgg had impact on gene expression from

dominant members of the gut microbiota we're showing two strains of 16 back roy dees, five pneumococcus bacterium, for the most part, no, there was no differences over the time frame of this experiment suggesting lgg has an effect, it's not a global effect.

but when we look at differential gene expression associated with lgg something interesting emerged. what we saw day 28 compared to baseline or day 28 compared the day 56 was a select set of genes upregulated during the time probiotics was present.

these were genes involved in mow untilty and chemotaxis from the predominant receptors in the gut. a log fold change mapping these genes on to a schematic of the bacterial flagellula are and mapping to bacterial chemotaxis. so what is all this mean?

production of beauty rate is one of the most important metabolic functions carried by the gut and the two microbes in which we saw upregulation ophelia jig already genes are involved in carrying out complex conversion of plant polysaccharides into the production of buterate, it

plays an important role maintaining the epithelial cells in the gi tract described having potent anti-inflammatory effects. so one hypothesis with regard to potential benefit of lgg is that it may increase level of flagellar gene expression with

which then may serve to move these burerate producers into mucous layer closest to the epithelial cell layer and without changing the production at all we maybe shifting the buterate gradiant in the gut and may result in higher level of buterate level to the epithelial

cells. this is back to an interesting paper in 1984 by stanton and savage suggesting that flagellula in these modal bacteria is critically important to penetrate into the mucous layer. the other interesting

possibility is that something else entirely maybe going on. we know flagellan genes play a role interacting with the gut associated immune system, flagellan is a ligand for toll like receptor five. tlr-5 recognizes flagellans from a wide range of bacterial

species an exposure in a mouse model is shown to suppress asthma through generation or increase in number of of treg cells so it maybe this increase in the level of flagellin express in the buterat producers may have an effect on the host through its interaction with the

immune system. this is my slide before acknowledgments. i won't go into detail other than to say as someone who started in microbial genomics focusing almost exclusively on pathogens and believing that pathogens are most important in

directing outcome following infection with a particular disease agent, my thinking and the field have certainly changed considering this as interaction triad that includes not only pathogen but host and microbiome. so this is some of the most

exciting work i have done and i -- it's either good or bad, it will take a long period of time to sort out. acknowledgments with regard to studies i described on shegel healthcarea and salmonella, folk ms. my lab, emily and anna deserve credit, cellular

immunology group led by march sell stein and andy pollard and colleagues at university of oxford. with regard to the probiotics study this was also emily padrosh's work with patricia hubbard's group mass general with funding from ncamm and nsf.

i went very late, i have a few minutes left, i'm willing to take whatever questions you may have. thank you very much. >> we do indeed have time for questions. microphones are in the aisle, please approach the microphone

if you have a question so people watching the video can hear. let me ask you, when you proposed this theme which seems to pop up in several enstances the diversity of microbiome is good for you, in terms of better opportunity for immune response to vaccine for instance, you

have to wonder i'm sure whether that diverse the city a surrogate for something for subtle going on, there maybe is a minor species in there which is not necessarily attracting your attention but needs to be there at certain level and if you have less diversity there's

less of it and problems arise. how do you take apart the hypotheses when there's a complicated system under you? >> that's a great question and critically important question. and to be quite true, i'm not sure how we go disentangling this.

i can tell you specifically with cynamologous macaque studies, with shegilla one thing we look to do now is see whether or not we can transfer this presumed protective effect we see with high diversity community into a monkey with a low diversity there are a number of caveats

and challenges that come from doing that study if we're successful great, if not we can probably find lots of -- come u up with reasons why that is not going to work. if we can do that then we can i think start to systematically create fractionate the

communities and try and identify is this something abundant, something less abundant. i don't know whether we can see this in other animal models but this is where we're going to start and we thought about this for a long time and i'm not sure how well we can do this.

i think the most important next step is to question nonstrait this is a property -- demonstrate this is a property of the microbiota that can be transferredded. >> i i have a question about the -- microbiome to the immune so how do we untangle the

correlation that may exist between our immune system of the individual and genetic makeup of individuals? and microbiome that might be correlated with that versus the -- in our role of this microbiome in response to the how do we associate causality

and from that correlation? >> again, another question and i think it follows on the question that you raised. we have made a number of observations. there are genotypic difference mhc differences in monkey study, microbiome differences and

differences in susceptibility to infection with shegella, it will be tough to figure out what comes first. i think we can perhaps get some insights from fecal transplantation studies look to try -- unfortunately when looking at shegella and

salmonella, we don't are the benefit of a lot in mouse models. these are pathogensna -- that are specific to non-human, human primates so that limits what we can study. perhaps we can find a simpler system like citrobacter

infection in mice for instance just to try and demonstrate these principles. is it the host genetics, is it a component of the microbiome, is it both? i think we'll beat our head against the wall a long time trying to get to these answers

with these studies. we have to find a way to to this in a simpler system. but what struck me over and over again is the fact that these high diversity communities that seem to be associated with the more robust outcome. i can't tell you what's driving

that but i can't let it go. there maybe something there. >> hi dr. fraser, thank you for the interesting talk. do you know in cases the humanbiome is changed how long does the human body try to recorrect a changed microbiome back to original template?

>> good question and i would say most cases the answer is yes. if we look at the studies that have been done there's various perturbations, indict looked at o diet looked at in a number of studies, if you switch from omnivorous diet to vegetation diet you see a reversion in the

-- the one exception is with antibiotic andry pooted exposure to antibiotics. some individuals you will see this reservation of community composition following administration of antibuy contribution. in other individuals, after the

path advantage of period of time the composition appears different and's suggested, and i would pose a question to the audience how many can remember how many times you have taken over the course of the lifetime, i can't and that's probably true for most of you.

so it maybe that again, in some of us, with these communities that are less resilient, those kinds of exposures over time start to change the composition and presumably the function of the microbayou that. if you're interested in this marty placer has done phenomenal work at

therapeutic levels, the impact of therapeutic levels of antibiotic as well as impact of subtherapeutic levels were exposed in our food supply to chronic low levels of antibiotics that seem? mouse models to be having a long term effect.

q. in those case it isbiome did change does human body adapt to change or is it -- >> that's a fantastic question. that's not going to be so easy to figure out but you can postulate it's a microbiome change and properties change it mace change our susceptibility

or ability to respond to a different perturbation. but anxious those questions require long term studies and why would great to do those, not sure we can afford. >> make get there, we can probably take one more question. >> thank you so much for great

talk. my question is about mental health, do you think microbiome influences mental health diseases such as depression or bipolar schizophrenia? >> there have been intriguing results from animal studies to suggest the answer to that

question is indeed yes. by extrapolation it wouldn't be implausible to think the same things could be in play in humans. but with these we tend to be talking about very complex very complex biology and we overlay on top of that the potential

role of very complex microbial i think it's important to always remind ourselves that if we really want to get at fundamental mechanisms we have to be doing that by onlying at the problem from a different direction. >> those of you interested in

continuing the conversation are warmly invited to refreshments at the nih library around the corner. meanwhile, let's thank our speaker one more time.