A recent analysis of
Essential Science IndicatorsSMfrom
Reuters data showed that the work of Dr. David
Relman had the highest percent
total citations in the field of Microbiology. His
record in this field includes 24 papers cited a total
of 1,467 times between January 1, 1999 and April 30,
Dr. Relman is the Thomas M. and Joan C. Merigan Professor in the
Department of Medicine, Division of Infectious Diseases and the Department
of Microbiology & Immunology at Stanford University School of Medicine.
He is also the Chief of Infectious Diseases for the VA Palo Alto Healthcare
In the interview
below, ScienceWatch.com correspondent Gary Taubes
talks with Dr. Relman about his highly cited research on
As a pioneer of the modern study of human
microflora, there must have been a moment when you first came on the
subject and found it worthy of interest. How did that happen?
I was finishing a postdoctoral fellowship at Stanford with Stanley Falkow
and my plan was to pursue a career in the research basis for microbial
pathogenesis and combine that with some clinical care. The only catch was
that I had gotten interested in what had started out to be a side project.
It had less to do with the molecular basis of bacterial disease and more to
do with pathogen discovery.
I was going to these weekly clinical case conferences, and a case was
presented on a disease called bacillary angiomatosis, which was a
relatively common problem in AIDS patients. The interesting feature was
that it clearly seemed to be associated with a bacterial organism. When you
looked at stains under the microscope you could see clusters of what were
obviously bacteria in places you don't expect to find them. And yet no one
had been able to identify these bacteria because they couldn't be
cultivated. Because cultures are the mainstay of bacterial diagnosis, it
meant there was no answer for this disease. Stan had been with me at the
case conference, and afterward he said this ought to be solvable with
molecular tools. He pointed me to the work of some leading environmental
microbiologists who at the time—this was now the late
1980s—were looking at DNA sequences obtained directly from
environments where there was obviously a great diversity of organisms. They
were using these sequences to say what in the bacterial world might be
"...one of the most important
ecosystems on the planet might be the human
So I simply borrowed the approach of these environmental microbiologists
and applied it to these clinical specimens and, lo and behold, a sequence
emerged that we could associate with the presence of these visible
organisms. It turned out to be a new sequence and an organism that to this
day is believed to be the cause of bacillary angiomatosis. It's also the
cause of cat-scratch disease.
To me the important lesson was that it may be possible to identify
previously unrecognized pathogens by taking this approach of trying to
discover telltale diagnostic sequence information directly from clinical
specimens. I started down that path as a parallel to what I was doing in
bacterial pathogenesis, looking in different disease settings where there
was a suspicion of a bacterial cause, yet nothing had been identified at
that point. I started looking in different kinds of human tissue. The
challenge was to find tissue that would potentially contain the pathogen,
but wouldn't contain any of the myriad of normally occurring bacterial
commensals, the bacteria that we live with all the time, in the human body.
So the initial problem was how to identify the pathogen
from the background of normal microflora?
That seemed to be the problem at the time. Using this molecular approach
meant that, in theory, we would pick up any and all bacteria. We couldn't
know upfront what the nature of the bacterial target or the pathogen might
be. So we were targeting conserved sequences believed to be found in nearly
all bacteria, which, of course, includes commensals. Along the way it
became clear to me that what I was considering the background problem was
actually an interesting topic unto itself.
Who else was working on commensals at the time, and how
were they approaching it?
Certainly this was nothing new in essence. Exploring commensal microbiota
was a long-standing interest of many folks, but they were taking the more
traditional approach—using cultivation and other kinds of classic
microbiological approaches. I was interested in using the same kind of
molecular survey tools that environmental microbiologists were using, so
instead of finding one or a discrete number of sequences from the pathogen,
I was finding hundreds or thousands of discrete sequences that would
correspond to our normal indigenous microbiota. So for me, what became a
new area of investigation was something that was previously, for me, a
distraction and an annoyance. Using these molecular techniques, I started
to deliberately sample the kinds of sites where I knew there would be a
complex microbial community.
Did you start with the human intestines, the subject of
your highly cited 2005 Science article (Eckburg PB, et
al., "Diversity of the human intestinal microbial flora,"
308: 1635-8, 10 June 2005)?
The first deliberate study I did looking at indigenous microbiota was in
the gum pocket, and it was actually my own gum pocket in this case. It was
published in 1999 (Kroes I, Lepp PW, Relman DA, "Bacterial diversity within
the human subgingival crevice," Proc. Nat. Acad. Sci. USA 96:
14547-52, 7 December 1999). I had this idea as I was getting ready to go to
my dentist. I went into work first, picked up some sterile collection
tubes, brought them with me to the dentist's office, and asked him, as he
was cleaning my teeth, would he mind putting this stuff into these tubes,
instead of throwing it out.
Part of the idea was to compare this molecular approach to the classic
microbiological culture-based approach on the same material. What we found
was, not surprisingly, that the majority of organisms could be found only
with the molecular approach, not with cultivation.
What led you to move from the gum pocket to the
Well, it has been known for a while that the distal gut, the large
intestine, is the site in the body most likely to contain the greatest
density of bacteria. In the early 2000s, I was interested in the idea of
comparing bacterial diversity in the setting of health vs. disease, as I
still am. I was approached by a clinician in Canada, Charles Bernstein, who
had collected an interesting set of colonic biopsy samples from patients
with Crohn's disease. He's one of the co-authors on that 2005
Science paper. He said, "I have this population-based study. It's
really well controlled; we have lots of clinical data, and we've also
recruited healthy family members of patients with Crohn's disease. Would
you be interested in studying these samples?" I said, "Sure." But I thought
we should start by examining the healthy samples.
With Charles's help, we came up with a sampling scheme, going for defined
regions of the large intestinal tract in each of the healthy controls,
moving right down the large intestine, starting with the cecum. So we had
these samples, and we also had the opportunity to sequence more deeply
through a collaboration with some folks at what was then The Institute for
Genomic Research and is now the J. Craig Venter Institute. They said that
at low cost they could sequence clone libraries for us at much greater
numbers than we could ever do on our own. So this seemed to be a good
opportunity to re-examine the patterns of diversity in health from this
well-studied site, but a site not well-studied with these modern molecular
techniques. That's how this study and paper came to be.
Why do you think that paper has garnered so many
citations in such a short time? What makes it so influential?
There are a couple of possible answers to that. It might be a
chicken-and-egg or horse-and-cart kind of issue. First, the paper
highlighted what seemed to be some pretty interesting features to this
ecosystem that hadn't been appreciated or realized—in particular, the
extent of the diversity in this microbiota and how that diversity varied
from person to person and site to site. It suggested that one of the most
important ecosystems on the planet might be the human body. All the
interest in the microbial world was focused everywhere in the "outside"
world, but less so within each of us, and what we learned was that this
particular ecosystem was as interesting and diverse as any of them. By
studying this ecosystem we might be learning about some fundamentally
important aspects of human health and disease.
"...I was interested in the idea of comparing
bacterial diversity in the setting of health vs.
The other point is that this paper may have just come along at the right
time. It highlighted the value of bringing together capabilities and
perspectives from different disciplines—clinical medicine,
traditional ecology, environmental microbiology, etc.—at a time when
people were just beginning to realize that this particular integration of
fields might hold a lot of promise. I think our paper influenced funding
agencies and other researchers to start thinking about focusing their
efforts on this particular sampling site—the human body.
The fact that NIH and other institutions in the European Union have all
decided to invest large amounts of money in what people are calling human
microbiome projects has now, if for no other reason, drawn a lot more
people into this area. Considering the economic challenges for the research
community right now, to have this kind of money directed towards this topic
has led to people setting aside what they were doing and turning to this
area of investigation. At the time that this paper was published there were
relatively few people deploying these kinds of perspectives and
technologies on the human microbiota.
How has research on human microflora evolved in the four
years since you published that paper?
Well, in a couple of ways. First, an acceleration of science and technology
forces and trends started to become important around 2005. One of the most
obvious is the development of sequencing technologies, and the amazing,
almost exponential growth in capabilities in DNA sequencing. Part of the
attraction to that 2005 paper was just the number of sequences we were able
to report at the time. Today, for the same amount of money and time, the
number of sequences you can generate is astronomically larger.
As an example, we published a paper this past fall in PLoS Biology
(Dethlefsen L, et al., "The pervasive effects of an antibiotic on
the human gut microbiota, as revealed by deep 16S rRNA sequencing," 6:
2383-2400, November 2008), another look at microbial diversity in the
distal gut in a study of the effect of antibiotics on a number of healthy
individuals, and the number of sequences we were able to report was about
50-, almost 100-fold greater than in that 2005 paper, and for less money.
So just the power of next-generation sequencing is one of the big changes
and a dominant feature of the new scientific landscape.
The other way this research has evolved is in the increasing attraction of
this field for people who were trained and then practiced in other fields
of science. We now have some really prominent environmental microbiologists
becoming interested in the human body. We have some very prominent
microbial ecologists interested in the human body instead of grass lands or
river sediment. We have some very gifted mathematicians and statisticians
interested in data from the human body rather than data from economics or
meteorology. It all creates a convergence of interest and insight and
capability that's tremendously exciting.
What would you say is your single favorite research
project at the moment?
One thing I'm actively pursuing is the effect of deliberate perturbation of
these microbial communities. The tool we're using is antibiotics. We still
have a lot to learn about the nature of these communities and one way to do
it is to study how they respond and behave when they're perturbed. This
approach also has a lot of clinical relevance. Clinicians are interested in
what antibiotics are actually doing, and we already know they have untoward
effects. Even in healthy people who don't experience side effects, we now
know there are all kinds of major disturbances going on. So that's one area
I'm very interested in.
Another is how the diversity of this microbiota in the human body is
distributed in space, and in time. What's the biogeography of bacterial
diversity in the human body? Are there patterns of distribution that tell
us something about the physiology of the human body? That tell us something
about how communities might be involved in maintaining health and promoting
disease at certain sites and not others? Could it be related to why one
segment of the intestines becomes inflamed and the one next to it doesn't?
Or why one section of the intestinal tract develops a flare-up of Crohn's
disease and 10 centimeters away everything is fine? We're trying to look
more carefully at whether patterns of microbial diversity predict
subsequent disease, or whether subtleties in the patterns of diversity,
say, along the length of the bowel might tell us about why this disease
occurs in one place not another.
Are there particular diseases in which you think the
microflora play a causal role?
Yes, although at this point we're still guessing as much as anything. I do
think we'll find differences in the bacterial communities in a number of
diseases, although we'll then have to establish whether these differences
are responsible for the disease or simply a result of it. That's the
causation problem, and it's really not yet adequately addressed for a lot
of diseases. If these are going to be important associations, even
important causal associations, it will be in settings like inflammatory
bowel disease, or chronic gum disease or antibiotic-associated diarrhea.
There are number of other conditions where people have speculated that they
may be important—irritable bowel syndrome, for instance, which is
different from inflammatory bowel disease. There is some evidence for
differences in the composition of the microbiota. There's a long list of
disease states where some early work or partial data has suggested a link.
What message would you like to give lay readers about
I guess one message is that there is a great deal of microbial diversity in
the human body and these are almost entirely all organisms whose presence
is not something to fear or be worried about, but rather, something for
which to be thankful. To a large degree our microbial inhabitants are
contributing to our own health. They are a fundamental part of who we are
as healthy, compensated, adaptable living organisms.
The second message might be that the result of all this work—of this
whole emerging community of investigators—is that it may have real
impact on our ability to recognize early signs of disease and then perhaps
our ability to do something about it. Once we understand these bacterial
communities better, we'll be able to figure out how to maintain a healthy
community picture or restore it if it's already become disturbed. That's
also a positive message—that this research may indeed lead to options
for positive intervention.
David Relman, M.D.
Stanford University School of Medicine
Stanford, CA, USA
VA Palo Alto Healthcare System
Palo Alto, CA, USA