Archives of Medical Research - (2018) -

REVIEW ARTICLE
Evolving Ecosystems: Inheritance and Selection
in the Light of the Microbiome
Santiago Sandoval-Motta,a,b,c Maximino Aldana,a,d and Alejandro Franka,e,f
a
Centro de Ciencias de la Complejidad, Biolog
ıa de Sistemas, Universidad Nacional Aut
onoma de M
exico, Ciudad de M
exico, M
exico
b
Instituto Nacional de Medicina Gen
omica, Gen
omica Computacional, Ciudad de M
exico, M
exico
c
Consejo Nacional De Ciencia Y Tecnolog
ıa, C
atedras CONACyT, Ciudad de M
exico, M
exico
d
Instituto de Ciencias F
ısicas, Fen
omenos no lineales y complejidad, Universidad Nacional Aut
onoma de M
exico, Morelos, M
exico
e
Instituto de Ciencias Nucleares, Estructura de la Materia, Universidad Nacional Aut
onoma de M
exico, Ciudad de M
exico, M
exico
f
Colegio Nacional, Ciudad de M
exico, M
exico
Received for publication October 21, 2017; accepted January 12, 2018 (ARCMED-D-17-00605).

The importance of microorganisms in human biology is undeniable. The amount of

research that supports that microbes have a fundamental role in animal and plant physi-
ology is substantial and increasing every year. Even though we are only beginning to
comprehend the broadness and complexity of microbial communities, evolutionary the-
ories need to be recast in the light of such discoveries to fully understand and incorporate
the role of microbes in our evolution. Fundamental evolutionary concepts such as diver-
sity, heredity, selection, speciation, etc., which constitute the modern synthesis, are now
being challenged, or at least expanded, by the emerging notion of the holobiont, which
defines the genetic and metabolic networks of the host and its microbes as a single evolu-
tionary unit. Several concepts originally developed to study ecosystems, can be used to
understand the physiology and evolution of such complex systems that constitute ‘‘indi-
viduals.’’ In this review, we discuss these ecological concepts and also provide examples
that range from squids, insects and koalas to other mammals and humans, suggesting that
microorganisms have a fundamental role not only in physiology but also in evolution.
Current evolutionary theories need to take into account the dynamics and interconnected-
ness of the host-microbiome network, as animals and plants not only owe their symbio-
genetic origin to microbes, but also share a long evolutionary history together. Ó 2018
IMSS. Published by Elsevier Inc.
Key Words: Microbiome, Holobiont evolution, Missing heritability, Adaptability, Allostasis.

Ecological Evolution
The Object of Selection
Natural selection has been a controversial subject since
Darwinian times. With the discovery of genes and the
formulation of the modern synthesis some scientists
thought that selection should act upon genes, since they
were the subject of mutation and thus were the raw material
of evolution (1). However, evolutionary biologists such as
Address reprint requests to: Santiago Sandoval-Motta, Associate Pro-
fessor, Instituto Nacional de Medicina Gen
omica, Gen
omica Computacio-
nal, Perif
erico Sur No. 4809, Col. Arenal Tepepan, Delegaci
on Tlalpan,
Ciudad de M
exico 14610, Mexico City, Mexico; Phone: (þ52) (55)
33899339; E-mail: santiago.sandoval.m@gmail.com
Ernst Mayr and Sewall Wright argued that selective pres-
sures do not only act on single genes. Based on the obser-
vation that many genes’ functionalities depend on the
presence of other genes, Mayr and Wright proposed that
the whole genome of an organism was the target of selec-
tion (2,3). This idea prevails, along with Mayr’s definition
of species, and both are among the few lasting modifica-
tions to the basic Darwinian theory of evolution. Neverthe-
less, we now recognize a vast number of ways in which
selection acts apart from genetics. For example, selection
can act upon the morphology of an organism (which may
not depend entirely on genetics), on its behavior, on its
metabolic capabilities, and over all of these (and many
more) factors at the same time (Multilevel Selection The-
ory). It seems now that the object of selection is the

0188-4409/$ - see front matter. Copyright Ó 2018 IMSS. Published by Elsevier Inc.
https://doi.org/10.1016/j.arcmed.2018.01.002
2 Sandoval-Motta et al./ Archives of Medical Research
individual as a whole and therefore, in addition to its ge-
netics, we must also take into account the organism’s ca-
pacity to modify its environment to its advantage, its
heritage, and ultimately the interactions with individuals
of the same species or from other species that change its
survival and reproduction capabilities. Hence, if we
consider this wholeness for each organism, we will be
delving into the complex subject of ecosystem evolution.
An ecosystem can be described as a set of communities
of different species of organisms living in the same place, at
the same time, with a constant interchange of matter and
energy. This constant flux of matter and energy opens up
the possibility to determine the boundaries of an ecosystem
mathematically, as fluxes inside an ecosystem must be
greater within the community than with external compo-
nents. It is important to mention that this interchange of
matter and energy occurs between the biotic and abiotic
components altogether, and the mechanisms by which the
interchange occurs can also modify the environment itself.
A change in environment can consequently change the dy-
namics and composition of the living parts of an ecosystem,
creating a feedback loop where change is the only constant.
Always depending on the intensity, frequency and type of
changes, also known as perturbations, ecosystems can often
withstand them. Sometimes the biotic composition of the
ecosystem will remain unchanged, but often it will be
forced into other equilibrium states in which the abun-
dances, interactions and chemical reactions of the organ-
isms that compose it, need to adapt to the new imposed
challenges (4).
There are many spatial scales in which ecosystems can
be defined. It is still a matter of debate where to draw the
line in size (and in that sense the comparative measure-
ments of fluxes mentioned above could be helpful), but
research groups have defined and studied ecosystems as
small as a piece of cheese (5), or as vast as the Biosphere
(6). Nevertheless, a new kind of ecosystem is at hand:
The holobiont, which is defined by the genetic and meta-
bolic networks of the host and its microbes as a single
evolutionary unit. The main reason the holobiont defies
the standard concept of an ecosystem is because the ‘‘envi-
ronment’’ in which the species that comprise it thrive, is
another living organism. This is not a trivial factor. The
fluxes between biotic and abiotic components in traditional
ecosystems depend almost entirely on the abilities of the bi-
otic to extract, use or transform the abiotic. However, in
holobionts this exchange is highly regulated by the host,
for example through its immune system (7). In traditional
ecosystems, perturbations often come either from a cata-
strophic natural event (disruptions such as earthquakes,
fires, floods, etc.) and/or are caused by the biotic compo-
nents themselves. In holobionts this may not always be
the case, as the host can quickly introduce or remove car-
bon sources and other metabolites through a change in diet,
or even increase or decrease the diversity of the ecosystem
- (2018) -
through ingestion of pro/anti biotics. The holobiont concept
defines the object of evolution as the fusion of the nuclear
genome of the host, its organelles, and its microbiome. As
an evolutionary unit, selective pressures applied to the nu-
clear genome prevent deleterious mutations and promote
favorable ones. Selection over the microbiome encourage
the development of beneficial microbes (mainly in diges-
tive, immune and sexual processes) (8), while simulta-
neously suppress pathogenic organisms. It is also
important to understand that the role of specific microbes
in host biology is context dependent. This means than even
though a strain is considered neutral under normal circum-
stances, its role can change in a different condition. These
organisms, known as amphibionts or pathobionts, are
known to become pathogenic in response to inflammation,
antibiotics, infections and other types of stressful condi-
tions (9). It has been shown, for example, that the highly
antibiotic resistant bacterium Clostridium difficile is a com-
mon dweller of adult intestines, but it is only under certain,
still unclear, conditions that it becomes pathogenic (10). In
summary, if selection applies at the level of the holobiont,
then dysfunctionalities (host or microbiome related) will
impact the survival and reproduction of the individual, un-
dergoing natural selection and thus evolving as a unified
entity.
On the Origins of Complexity
By reconciling Mendelian genetics and Darwinian Natural
Selection, the modern synthesis theory of evolution ex-
plains: a) the diversity between individuals from a genetic
point of view, b) how natural selection can filter such ge-
netic variations, and c) how new species originate through
gradual and cumulative genetic mutations and recombina-
tion. However, it does not explain how complex organisms
arise. For example, the modern synthesis cannot explain
how eukaryotic cells were formed by the successive merg-
ing of prokaryotic cells (11). These limitations in the
Darwinian Theory of evolution are understandable, as at
the time of its formulation microbes were prominently
catalogued as agents of disease. Additionally, the technol-
ogy available to look into the genetics of such microbes
was far from being developed. However, the origin of the
eukaryotic cell is now widely accepted as being the result
of endosymbiosis, and now we have a less biased view of
microbial communities as well as many tools to study
them.
The term ‘‘holobiont’’ was first proposed by Lynn
Margulis (12) and derives from the Greek word holos,
meaning whole. This term was originally proposed
to describe symbiotic associations between organisms
throughout most of their lifetimes, and now it refers to
the collection of microbes associated with a host plus the
host itself. In 1967 Lynn Margulis also demonstrated the
endosymbiotic emergence of eukaryotic cells, proposed
 Evolution in the Light of the Microbiome
around 1910 by the botanist K. Mereschkowski. In this
work Margulis proved that the functional complexity and
organization of eukaryotic cells was produced by a well-
defined succession of mergers between protist organisms
(11). This theory showed that the nuclear genome and the
organelles of a eukaryotic cell work as a unity to perform
many of the complex tasks required for the cell to survive
and duplicate. It also shows that endosymbiosis, the fusion
of independent and already formed organisms, is another
mechanism by which life gains complexity and produces
new species. Sexual reproduction, along with genetic
recombination and mutation, ensure that the newly formed
offspring closely resemble their parents. However, endo-
symbiosis is capable of generating organisms that can
significantly differ from their predecessors, positioning it-
self as a mechanism that may produce a great evolutionary
change. The evolutionary pathways of a set of organisms is
usually represented as independent branches of a phyloge-
netic tree. Once a branch arises, it either continues its path
indefinitely or disappears, but it never merges with other
branches as what is thought to have happened at the origin
of eukaryotic cells (Figure 1). This merging of independent
organisms occurs very often (13), and although symbiotic
organisms do not necessarily integrate in the same way as
the eukaryotic cell (14,15) they can and must be studied
as an evolutionary unit.
The merging of two or more independent individuals
into a coordinated polygenic organism can yield many ad-
vantages. For instance, all of the metabolic capabilities of
former cells are joined allowing the new organism to utilize
the resources that were harnessed by them. Any specializa-
tion present in any of the former individuals may also be
preserved in the new being. It has been shown that commu-
nities composed of several species, tend to develop distinc-
tive characteristics to avoid competition over the same
nutrients (16). In these experiments, Turcotte et al. showed
Figure 1. (A) Representation of a traditional phylogenetic tree of independent species. (B) Representation of a phylogenetic tree where the evolutionary unit
is considered a metabolic and genetic junction of more than one organism.
3
that when a species is forced to grow separated from others
it often develops faster reproduction rates but does not
develop mechanisms for alternate resource usage. However,
the opposite happens when these species are grown
together. In such a case, three-quarters of the species that
were grown together developed mechanisms to consume
either new environmental resources or waste products pro-
duced by the other species, thus avoiding direct competi-
tion. It was also observed that only one species developed
a dramatically faster reproduction rate whereas the other
ones decreased their own duplication rates. In the light of
these results, and because processes such as autophagy
and endocytosis are induced in a low nutrient medium
(17), it is possible that cell merging arose from an incom-
plete digestion caused by starvation. Indeed, when colonies
of microorganisms are subjected to low levels of nutrients,
cells start ingesting each other to improve their survivabil-
ity (18). An incomplete phagocytosis can leave parts of the
ingested cell untouched, leading to a possible merging of
functions.
Coming back to the holobiont, it can be described as a
multilayer network in which encompassed networks can
have different properties but are all interconnected. In a
multi-species community, one species can alter the environ-
ment of another, which may facilitate its survival or impose
a new selective pressure (19). Evidence suggests that when
two species coevolve, specific adaptations exhibited by
either one of them would not exist in the absence of the
other, and emerging co-adaptations can feedback to influ-
ence ecosystem functioning (16). This process is known
as Diffuse Coevolution and as discussed below, is of the
utmost importance for the evolution of holobionts and their
correct functioning. Understanding the ecology of holo-
bionts requires considering not only changes in the abun-
dances of each individual species of a microbiome, but
also evolutionary co-adaptations that modify the metabolic
4 Sandoval-Motta et al./ Archives of Medical Research
capabilities of those species. These studies, in turn, will
require transgenerational tracking.
Heredity and Heritability; the Information within the
Microbiome
With respect to human biology, the holobiont concept has
deep consequences that go beyond lessening our sense of
independence. Several reviews on the influence of the mi-
crobiome on our biology are already available (20e23).
These include studies that demonstrate the influence of
microbes in weight gain, development of the immune sys-
tem, cancer, and even neurological disorders. However, it
is not our intention to further reinforce the importance of
the microbiome on our physiology but to focus on the
importance of the microbiome in the transmission of phe-
notypes from one human to another. In fact, we stress the
idea that microbiome-related phenotypes can be transmitted
between phylogenetically close holobionts (humans or not)
as well as transgenerationally.
The distinctiveness between host associated microbes
and free living bacterial communities (24), along with the
similarity between phylogenies of microbial communities
and phylogenies of their hosts (25), are strong indicators
that many animals and plants have coevolved with their
associated microbes since ancient times. Powerful interde-
pendencies have arisen. For instance, it has been proposed
that the vertebrate immune system, given its potential to in-
fluence microbial development, has evolved as a means to
manage more complex microbial communities (26). How-
ever, the extent to which this coevolution has influenced
both the host and its microbial genomes is only beginning
to be revealed and could be more profound than we have
imagined. A recent study on mice has shown that when
two subspecies of mice are crossed, bacterial communities
of the hybrids were significantly different, and transgressive
phenotypes were encountered (extremely different from
parental features). Hybrid mice also displayed aberrant im-
mune systems and intestinal problems, both associated with
altered microbial communities. On the one hand, these re-
sults restate the importance of stabilizing selection (favor-
ing intermediate phenotypic values) and show that
hologenome variation correlates with phenotypic variability
upon which selection can operate. On the other hand, this
study demonstrates the consequences of evolutionary diver-
gences in holobionts, which as we discuss in a later section,
are an important factor of reproductive isolation, which in
turn may lead to speciation.
So far we have discussed two different levels of interde-
pendence in the holobiont. One is the dependence of mi-
crobes among each other (27) and the second is the
interdependence between the microbiome’s functionality
and the genetics of the host (28,29). This hologenome
(the collective genomes of the holobiont), can be modeled
as a multilayered genetic network where a complex
- (2018) -
plethora of factors can influence its composition. Diet, ge-
ography, physical activity and stress have been correlated
with changes in abundance of certain microbial species as
well as their behavior (21). These multiple interactions
create such a vast amount of variability across individuals,
particularly in humans, that it has not been possible to
define a ‘‘core microbiome,’’ that is a well-defined set of
species that are unambiguously present in any individual
(30). Nevertheless, recent studies have shown that it is
possible to identify certain conserved metabolic functions
present in the microbiome, allowing research groups to
define a ‘‘core functionality’’ of the human microbiome
(30). Regarding of the ecology of the holobiont, these re-
sults suggest that there are basic functions of the micro-
biome that are needed for maintaining an ecological
equilibrium within the species of the microbiome, as well
as between them and their host.
Horizontal and Vertical Transmission of Microbes
Endosymbionts (organisms that live within others with
others with mutual benefit) often carry out important func-
tions for their hosts (31). These organisms can be trans-
mitted directly from their parents (vertical transmission)
from other non-related individuals (horizontal transmis-
sion) or the environment. Although all types of transmis-
sion are widely distributed across phyla, it has been
suggested that most endosymbionts in vertebrates are envi-
ronmentally acquired (32). However, this environmental
acquisition is often mediated by the parents. It has been
shown that infant horses and iguanas consume the feces
of their parents thus colonizing their own intestines with
appropriate microbial communities (33). Mother koalas
are known to directly feed feces to their offspring providing
them with the necessary endosymbionts that metabolize
secondary toxic compounds of their food (eucalyptus)
(34). In many other species, parental care and interaction
with individuals of the same species is often enough to
ensure a correct colonization of the infant’s gut (35). Other
non-coprophagic mechanisms exist, particularly in aquatic
environments where it has been shown that the bacterial
symbiont Vibrio fischeri of the squid Euprymna scolopes
is transmitted to youngsters through increasing the concen-
tration of the bacteria around their birthplace. E. scolopes
blows high concentrations of Vibrio into the water (mecha-
nism know as winnowing), significantly increasing its envi-
ronmental concentration and thus helping the colonization
of the next generation of squids.
Vertical transmission has also been widely reported in
insects. Diverse associations with viruses, protists and bac-
teria have been studied in these arthropods. Many of their
associated microbes live exclusively within specialized or-
gans of their hosts, with some of them, for instance,
residing in the female reproductive organs. These
‘‘endemic’’ endosymbionts are essential for the survival
 Evolution in the Light of the Microbiome
and reproduction of their host (36,37). By supplying
missing dietary compounds, the complementary metabolic
networks of these symbionts allow the host to exploit food
sources that would be otherwise nutritionally insufficient
(e.g. tree sap) (38,39). Contrary to environmental acquisi-
tion, these indispensable endosymbionts are transferred
maternally to offspring, as in the case of some ants, cock-
roaches and termites (36,40). This means that the concept
of heredity, the passing of genetic information from one
generation to another, can be extended beyond sexual or
asexual reproduction. Heredity can also happen through
vertical and horizontal transmission of microbes (32).
The previous examples reflect a direct influence of par-
ents on the microbial colonization of their offspring. How-
ever, it is important to note that there are cases in which a
group of individuals of the same species does not only pass
this information onto youngsters but also create a ‘‘safe-
guard’’ for their microbial partners. This is the case of
the ant Lasius niger, which exhibits an interesting group
behavior that is possibly related to transgenerational inher-
itance of microbes. It has been shown that these ants store
their feces into specialized compartments inside their nests,
despite the fact that the hive’s waste products (e.g. corpses,
or food waste) are always disposed of (41). The authors
suggest that this behavior helps the ants acquire specific
odors that let other ants recognize them as part of the col-
ony. However, since essential digesting microbes reside in
the guts of these ants (a phenomenon also observed in ter-
mites), it is possible that these compartments function as a
safeguard for important microbial endosymbionts that are
to be transmitted to future generations.
Microbial Transmissions in Humans
In humans, several examples of transgenerational microbial
transmission have been reported (42). They consist essen-
tially of the mother transmitting her microbes to the
newborn. This microbial inoculation of human babies has
several steps that will be described in chronological order.
However, further studies are needed to determine the
importance of each step for the correct development of
the human microbiome.
Starting with the embryo, the placental barrier is known
to filter undesirable pathogens from newborns throughout
pregnancy. Nevertheless, experimental evidence suggests
that the placenta may not be as sterile as previously thought
and those specific microbes can inoculate the child’s body
before delivery (43). This hypothesis is still controversial,
as it has been demonstrated that intrauterine infections
are strongly correlated with preterm deliveries (occurring
in less than 30 weeks) (44). Nevertheless, when the first
stool of an infant (meconium) is tested for microbial pres-
ence, Enterococcus, Escherichia and other microbial mem-
bers present in adults’ intestinal tract can be found (45).
Despite the fact that the precise mechanisms by which
5
microbes enter the uterus are not well known, it is highly
likely that a human first inoculation of beneficial microbes
happens in the womb (42,46). Mode of delivery is the next
step by which human babies (as well as other mammals)
are inoculated by parental microbes. Passing through the
vaginal duct at birth is now recognized as a significant in-
fluence on the development of microbial communities
across the baby’s body, including skin, oral cavity and
gut. Cesarean section (C-section) babies are not exposed
to the same microbes. Instead, their microbial communities
resemble human skin, with no significant distinction be-
tween that of the mother or the medical personnel (47).
These different compositions (Vaginal or C-section) have
significant health consequences as C-section babies are
more prone to develop metabolic, allergic and immune
related diseases (48e51). In fact, some hospitals are start-
ing to ‘‘swab’’ C-section babies with maternal vaginal
fluids (52), however in more than 50% of the world C-sec-
tion rates are more than 15% (53), making C-section an
important health issue. Finally, breastfeeding represents
the final step of direct maternal microbial transmission
(non-parental transmissions are discussed in a later sec-
tion). As with the womb, breast milk was thought to be
sterile. However it is now known to contain hundreds of
microbial species correlated with the correct development
of the immune system (54). Furthermore, bacterial compo-
sition of the milk changes throughout the breastfeeding
process, probably to prime the digestive system to forth-
coming diet changes. Milk from six-month mothers shows
an increased abundance of species of genera such as Prevo-
tella and Leptotrichia while displaying less lactic acid bac-
teria of genera such as Streptococcus and Lactococcus (55).
In a very similar way as with C-section babies, formula-fed
infants are more prone to disease (54). As previously stated
by Funkhouser and Bordenstein: ‘‘[.] it seems that mod-
ern medicine designed to prevent infant mortality (such
as emergency Cesarean sections and formula feeding)
has likely contributed to the rise in immune-mediated dis-
eases in developed countries due to the inherent lack of
exposure to maternal microbes associated with these
practices.’’ (42).
Extinctions and Famines
It has been well established that many of our functions as
holobionts depend on our microbial partners (20,56). Thus,
any situation interfering with their structure and composi-
tion can alter our functioning as a reproducing ecosystem.
For instance, starvation is known to change the functioning
and structure of microbial communities (57). This may be
relevant to understanding a surprising case originated at
the end of World War II, when the population of the
German-occupied part of the Netherlands was subjected
to one of the most studied human famines in history. From
the end of 1944 to mid-1945 the Dutch population struggled
6 Sandoval-Motta et al./ Archives of Medical Research
to survive with only 30% of their normal dietary intake.
When this dreadful situation finally ended, survivors were
recruited by the Netherlands healthcare system to record
their recovery and to study any possible long term effects
of the famine. Unsurprisingly, mothers that were pregnant
during the majority of the famine, often gave birth to small
children. However, these children were never able to attain
normal weights (58). The surprising part of the story is that
the malnourishment effects remained present even in the
grandchildren of the famine-affected families. The most
accepted hypotheses for this transgenerational effect are
epigenetic-related, with no mechanistic description capable
of explaining such effects. However, it has been shown that
malnutrition and stress can affect the composition of the gut
microbiome, even causing the loss of certain species (57).
Since microbial transmissions occur from mothers to
offspring at birth, it is entirely possible that these transge-
nerational effects are due to an incomplete parental micro-
biome being transmitted to offspring causing an improper
development of their own microbiome.
Extinctions and Antibiotics: Microbial Transfers as a
Solution
Along with Cesarean section and feeding formula to in-
fants, antibiotic prescription is another common practice
in modern human societies that severely affects the struc-
ture and functions of the microbiome. Since the introduc-
tion of antibiotics as antimicrobial agents the minimum
inhibitory concentration baselines for antibiotics have
significantly increased (e.g. up to 250 fold for erythro-
mycin) (59). Apart from the widely accepted effects of an-
tibiotics in the development of super resistant organisms,
another issue is the permanent eradication of important mi-
crobial functionalities from the human microbial reservoir.
It has been demonstrated that modern societies have a
significantly decreased microbial diversity when compared
to the diversity in more traditional societies (57,60). Antibi-
otics are widely used and are often prescribed unnecessarily
(61). Antibiotic usage can have long lasting effects in the
composition of the microbiome which can last up to several
months (62). Furthermore, if important species are extin-
guished from the microbiome, reacquiring them may not
be a simple task. First, because even if probiotics are
administered, stable colonization is not always ensured.
Second, a probiotic for the specific extinct species may
not be available. Third, due to the interdependence of spe-
cies in the microbiome, the eradication of one species may
cause a cascade effect that ends up eradicating several other
species. This in turn can create an all-or-non phenomenon,
as newly extinct microbes could be necessary for re-
introducing the originals. Although antibiotics are still crit-
ical to counter bacterial infections, antibiotics function as a
double-edged sword that induces dysbiotic processes that
have been associated with diseases such as obesity, diabetes
- (2018) -
and many others (63e66). However, microbial transplants
have emerged as an alternative treatment for microbial in-
fections. Controlled microbial transplants were first re-
ported around the year 2010 and are often used to fight
highly antibiotic-resistant pathogens (67). This technique
aims to restore the microbial ecosystem of a patient by
introducing whole microbial samples from healthy donors.
Fecal samples are the most common, but oral and mucosal
samples have also been used (68). This practice is now
beginning to be used more often as it prompts rapid recov-
ery of the patients and shows restoration of the structure
and function of the microbiome without any of the side ef-
fects of antibiotics. All of the results discussed in this sec-
tion demonstrate that transmission of microbes can directly
affect holobiont phenotypes.
Missing Heritability Problem
Relevant problems in human genetics can be turned around
through the holobiont perspective. As a community, our mi-
crobiome has more than 2000 times the amount of protein-
coding genes than our nuclear genome (69,70). This fact
has significant consequences. For instance, it can provide
some answers to the missing heritability problem
(71e73). This particular challenge refers to the huge
discrepancy observed between the heritability calculated
in familial studies vs. the heritability calculated in Genome
Wide Association Studies (GWAS). Heritability is a mea-
sure of the phenotypic variance that can be explained by
the genotypic variance of a population (74). GWAS studies
only take into account the genetic variation in human cells
and do not consider all the effects on the phenotype caused
by the microbiome. Therefore, it is not surprising that
GWAS studies can only explain a very limited part of the
phenotypic variability for many human traits (75). Further-
more, when phenotype variability is measured in families
and genetic identity is inferred by kinship, the heritability
of several human phenotypes is significantly higher than
that resulting from GWAS studies. This is so because in
the familial studies what is observed is the net phenotype,
which includes any effect the microbiota can have on it
(in addition to several other important effects). In a previ-
ous paper, our group has proposed that considering the vari-
ability of our microbial companions can significantly
reduce the gap between these measurements as it increases
the genetic variability associated to any microbiota-related
phenotype (71).
Robustness and Adaptability in Evolution, a General
Hypothesis for the Holobiont
Evolution of living systems is often considered to happen at
two separate scales. Microevolution is related to evolution
below the species level. It involves processes such as gene
flow, genetic drift, mutation, recombination, selection and
ecological interactions between species. Macroevolution,
 Evolution in the Light of the Microbiome
in contrast, includes phenomena involved in much larger
taxonomical groups. Geographical changes, mass extinc-
tions and long term diversification processes involved in
the generation of evolutionary novelties are examples of
macroevolution (76). Whether or not macroevolution can
be explained by microevolutionary processes is an ongoing
debate. However, metabolic and genetic features that led to
major evolutionary shifts did not occur through an accumu-
lation of small changes, but instead through a fast and large
acquisition of metabolic capabilities. That is, endosymbi-
osis. Endosymbiosis is the process where two or more inde-
pendent organisms join to become one. Lynn Margulis
argued that processes such as endosymbiosis are likely to
happen quickly and to bring saltational evolutionary events,
such as the acquisition of the energy generation capabilities
of the mitochondria (77). Even though it is still a matter of
controversy, present-day neo-Darwinian evolutionary the-
ory states that DNA mutations, gene duplications, and rear-
rangements act as the only source of novelty. Despite that
endosymbiosis is an important possibility for an organism
to outsource or include new metabolic processes, it is not
the only one. The metabolic capabilities of a holobiont
can expand or shrink in a matter of days either by incorpo-
rating or eradicating microbial species from its microbiome
(60,62,78). We now know that the metabolic repertoire of
holobionts is distributed among the host and their microbial
companions. Thus, if we consider the evolutionary impor-
tance of the microbiome to be mainly metabolic and not on-
ly genetic, as has been proposed for endosymbiosis (76), its
role would become much clearer. It is important to note that
from a theoretical point of view, this acquisition of micro-
bial metabolic capabilities cannot be easily treated as a
large mutational change, as has been proposed for endo-
symbiosis. In this metabolic context, endosymbiosis deals
with biochemical networks instead of particular genes,
phenotypic interactions rather than lineal inheritance, com-
munities rather than lineages, and focuses on the major di-
versifications of life. We believe this view is not exclusive
for endosymbiosis, as the more encompassing view of the
holobiont shares the same principles.
Holobionts can (and should) be treated as evolutionary
units, but the timescales at which their constituting parts
change and adapt to the environment are not equal.
Although major evolutionary changes may not come from
microevolutionary processes, microevolution can produce
significant amounts of genetic diversity, in turn proportional
to the duplication rate of the organism. In this sense the mi-
crobial part of the holobiont should adapt at a faster pace
than the host due to its faster duplication rates and higher
susceptibility to its surroundings (78e80). For instance, hu-
man Paneth cells in the intestine’s epithelium are replaced
on average every 20 d, whereas common microbial inhabi-
tants of the intestine such as Escherichia Coli have a 17 min
average generation time (81,82), representing a factor of
1,694. This means that even if we assume an equal mutation
7
rate in both (in fact it is much higher in bacteria), taking
into account the population size of E. coli in the intestines
and the higher division rate, it is clear that in the same time
lapse significantly more DNA mutations are produced by
the microbes. Thus, the rapid mutation rates and fast divi-
sion times of the microbiota provide the holobiont with
excellent mechanisms to adapt to sudden changes in their
environment. Nevertheless, biological systems need also
to be robust in order to survive in a constantly changing
environment (83). The host cells, with their slower division
rates and their vast repertoire of metabolic and genetic reg-
ulatory networks, must function as a stabilizing force for
the holobiont. We are currently testing these ideas in
models by using systems of coupled networks with
different reproductive and mutation rates.
Biological systems need to integrate the ability to with-
stand change with the ability to innovate and adapt, i.e.,
create new phenotypes. When computer simulated biolog-
ical networks are left to evolve and duplicate under such
constraints (to produce new phenotypeseattractors without
losing the original ones), the networks that survive usually
end up with a unique set of characteristics. For instance,
such networks tend to acquire the properties of ‘‘critical
systems,’’ in the sense that they respond to perturbations
in a collective way. Critical systems are those poised at
the brink of a phase transition between two different
dynamical regimes, for example, ordered and chaotic. In
the ordered regime the system dynamics is frozen, so that
the system does not respond to perturbations and hence
cannot evolve. Contrary to this, in the chaotic regime the
system is extremely sensitive to perturbations and can
completely change its entire state under even the smallest
perturbation. At the critical point the system is neither
frozen nor extremely sensitive to perturbations. Critical sys-
tems produce responses at many different scales character-
ized by power-laws. This means that the response of the
system to a perturbation can either be confined to a small
subset of its constituent elements, or it can travel across
the entire system (like a domino effect). In the context of
gene regulatory networks this means that if the organism
faces a new condition, or the state of some of its compo-
nents is externally modified, a subset of the regulatory
network will change its state in order to withstand these al-
terations without changing the genetic functions the organ-
isms have already acquired through evolution. It is the
delicate balance between phenotypic robustness
(conserving the old phenotypes) and phenotypic innovation
(generating new phenotypes to contend with new environ-
mental challenges) that drives the genetic network of the
evolving organism to operate at the critical point. Several
papers that deal with this topic in depth are available
(84e86). Here it suffices to say that many biological sys-
tems exhibit behaviors that would be difficult to explain
without this ‘‘critical dynamics’’ characteristic. The bal-
ance between robustness and innovation that generates
8 Sandoval-Motta et al./ Archives of Medical Research
critical dynamics may also be valid for the holobiont, but to
our knowledge there is yet no experimental evidence of this
hypothesis.
Conclusions and Perspectives
In this review we have covered a wide range of topics
related to the evolution of holobionts. We have shown that
the holobiont concept, in addition to being consistent with
experimental observations, also provides a new perspective
to reformulate fundamental evolutionary questions. Since
hosts and microorganisms have coevolved through millions
of years, it is reasonable to expect the latter to be well
suited to hold information about past and present environ-
mental conditions (together with genetic and perhaps epige-
netic mechanisms in the host). Furthermore, the holobiont’s
microbial information can be transferred to offspring, func-
tioning as a generator of phenotypic-memory for selection
to act upon.
In the light of recent developments about the relation-
ship between the host and its microbes, one can postulate
that:
a) The object of natural selection is the holobiont. The
phenotypes of the host are strongly influenced by its
microbiome.
b) Although the Darwinian theory cannot explain the
‘‘complexification’’ of organisms through evolution,
this complexity can arise from the union of distinct
phylogenetic branches. The holobiont is a prominent
example of the emergence of complexity at different
scales.
c) The concept of heritability needs reformulation in the
light of the microbiome, as parents systematically
transfer (inherit) microorganisms to their progeny.
d) A great variety of microorganisms play an essential
role in multiple metabolic and immunological func-
tions of the holobiont. The holobiont should be
considered as an ecosystem rather than as an
individual.
e) The microbiome provides a vast source of genetic
variability, which if not taken into account when
calculating heritability of phenotypes, may underesti-
mate its value.
f) An equilibrium between robustness and evolvability is
achieved in holobionts by the interplay of host and mi-
crobial genetics, basically due to their very different
reproduction and mutation timescale.
It is clear that many questions remain unanswered. For
instance, does the holobiont concept change the definition
of ‘‘species’’? Ernst Mayr defined a species as a group of
interbreeding organisms that are reproductively isolated
from other groups (87,88). However, microbiologists are
well aware that whenever microbes are involved, all
- (2018) -
traditional taxonomical classification methods including
sequence based similarity, exhibit subjective problems
(such as defining the similarity threshold to consider two in-
dividuals as members of the same species). In ecology, a
species is defined as a set of organisms adapted to a partic-
ular set of environmental resources (niche). Many organ-
isms fit both definitions and match standard taxonomical
classifications. However, complex microbial communities
where many genetically diverse organisms coexist and
share resources and genetic material, are hard to accommo-
date in either of these definitions. This issue is yet to be
resolved, but it is important to note that more complex
problems can arise when we deal with holobionts. It turns
out that that there are examples where the presence of spe-
cific microbes alters the reproductive capabilities of a set of
genetically identical macro organisms. For instance, it has
been reported that bacterial communities in hyenas are
capable of broadcasting scented information about gender
and the fertilization status of females. When these microbi-
al communities are compared between spotted and striped
hyenas, they prove to be compositionally different. More
importantly, these differences bias partner selection for
mating, preventing gene flow between spotted and striped
individuals (89). Thus, although it is perfectly possible
for spotted and striped hyenas to mate and exchange genetic
material, they do not do because they carry different sets of
microorganisms. A similar example occurs in Drosophila.
When two populations of fruit flies where given different
diets for 25 generations (starch vs. molasses), females were
more likely to mate with males that had the same diet as
they did (90). This effect was shown to be triggered by a
commensal bacterium living inside the flies. As stated in
(91), the role of microbial symbionts in behavior and speci-
ation can yield a broader understanding of the origin of spe-
cies. The hyenas and fruit flies examples show that the
microbiome can act as a non-physical barrier between
two genetically identical populations and thus promote an
allopatric-like speciation process. An even more dramatic
example is observed in Nasonia wasps. By a process known
as cytoplasmic incompatibility, it has been shown that due
to the presence of different bacteria, two species of the
jewel wasp (Nasonia giraulti and Nasonia vitripennis) were
unable to interbreed. Nevertheless, if individuals were sub-
jected to antibiotic treatments, hybrids were much more
likely to survive (91).
The holobiont concept is changing the way we think about
‘‘individuals’’ and their evolution. The evidence and ideas
presented in this article are based on specific cases, but they
cannot yet be considered as generalized rules for evolution.
However, we believe that it is already clear that when we
study the development and evolution of organisms we should
take into account their microbiome, as it can strongly influ-
ence variability, reproduction, and behavior. Perhaps an
appropriate summary of these ideas is that we should think
of ourselves more as a forest than as a tree.
 Evolution in the Light of the Microbiome
Acknowledgments
The authors thank Prof. Michele Gold Morgan and Prof. Teresa
Bosques for their critical reading of the manuscript and their many
valuable suggestions. AF thanks El Colegio Nacional for its
generous support. SSM thanks Catedras CONACYT 2017 pro-
gram for its support.
References
1. Dawkins R. The selfish gene, Oxford University Press, United
Kingdom, pp.464.
2. Wright S. Evolution and the genetics of populations. In: The theory
of gene frequencies, 2. United States: University of Chicago Press;
1969.
3. Mayr E. The objects of selection. Proc Natl Acad Sci U S A 1997;94:
2091e2094.
4. Elmqvist T, Folke C, Nystr€om M, et al. Response diversity, ecosystem
change, and resilience. Front Ecol Environ 2003;1:488e494.
5. Wolfe BE, Dutton RJ. Fermented foods as experimentally tractable
microbial ecosystems. Cell 2015;161:49e55.
6. Lovelock JE, Margulis L. Atmospheric homeostasis by and for the
biosphere: the gaia hypothesis. Tellus 1974;26(1e2):2e10.
7. Hooper LV, Littman DR, Macpherson AJ. Interactions between the
microbiota and the immune system. Science 2012;336:1268e1273.
8. Bordenstein SR, Theis KR, Woese C, et al. Host biology in light of
the microbiome: ten principles of holobionts and hologenomes. Wal-
dor MK, editor. PLoS Biol 2015;13:e1002226.
9. Kamada N, Chen GY, Inohara N, et al. Control of pathogens and
pathobionts by the gut microbiota. Nat Immunol 2013;14:685e690.
10. Riggs MM, Sethi AK, Zabarsky TF, et al. Asymptomatic carriers are
a potential source for transmission of epidemic and nonepidemic
clostridium difficile strains among long-term care facility residents.
Clin Infect Dis 2007;45:992e998.
11. Margulis L. Serial endosymbiotic theory (SET) and composite indi-
viduality. Microbiology Today 2004;31:172e174.
12. Margulis L. Symbiogenesis and symbionticism. In: Margulis L,
Fester R, eds. Symbiosis as a source of evolutionary innovation:
speciation and morphogenesis. United States: MIT press; 1991. pp.
49e92.
13. Guerrero R, Margulis L, Berlanga M, et al. Symbiogenesis: the hol-
obiont as a unit of evolution. Int Microbiol 2013;16:133e143.
14. Zilber-Rosenberg I, Rosenberg E. Role of microorganisms in the evo-
lution of animals and plants: the hologenome theory of evolution.
FEMS Microbiol Rev 2008;32:723e735.
15. Rosenberg E, Sharon G, Zilber-Rosenberg I. The hologenome theory
of evolution contains Lamarckian aspects within a Darwinian frame-
work. Environ Microbiol 2009;11:2959e2962.
16. Turcotte MM, Corrin MSC, Johnson MTJ. Adaptive evolution in
ecological communities. PLoS Biol 2012;10:e1001332.
17. Redfield RJ. Do bacteria have sex? Nat Rev Genet 2001;2:634e639.
18. Nandy SK, Bapat PM, Venkatesh KV. Sporulating bacteria prefers pre-
dation to cannibalism in mixed cultures. FEBS Lett 2007;581:151e156.
19. Stinchcombe JR, Rausher MD. Diffuse selection on resistance to deer
herbivory in the ivyleaf morning glory, Ipomoea hederacea. Am Nat
2001;158:376e388.
20. Cho I, Blaser MJ. The human microbiome: at the interface of health
and disease. Nat Rev Genet 2012;13:260e270.
21. Yatsunenko T, Rey FE, Manary MJ, et al. Human gut microbiome
viewed across age and geography. Nature 2012;486:222e227.
22. Morgan XC, Segata N, Huttenhower C. Biodiversity and functional
genomics in the human microbiome. Trends Genet 2013;29:51e58.
23. Dave M, Higgins PD, Middha S, et al. The human gut microbiome:
current knowledge, challenges, and future directions. Transl Res
2012;160:246e257.
9
24. Ley R, Hamady M, Lozupone C, et al. Evolution of mammals and
their gut microbes. Science 2008;320:1647e1651.
25. Ochman H, Worobey M, Kuo C-H, et al. Evolutionary relationships
of wild hominids recapitulated by gut microbial communities. Acht-
man M, editor. PLoS Biol 2010;8:e1000546.
26. McFall-Ngai M. Adaptive immunity: care for the community. Nature
2007;445:153.
27. Rakoff-Nahoum S, Foster KR, Comstock LE. The evolution of coop-
eration within the gut microbiota. Nature 2016;533:255e259.
28. Goodrich JJK, Waters JJL, Poole AAC, et al. Human genetics shape
the gut microbiome. PubMed Commons. Cell 2014;159:789e799.
29. Blekhman R, Goodrich JK, Huang K, et al. Host genetic variation im-
pacts microbiome composition across human body sites. Genome
Biol 2015;16(1):191.
30. Turnbaugh PJ, Hamady M, Yatsunenko T, et al. A core gut micro-
biome in obese and lean twins. Nature 2009;457:480e484.
31. McFall-Ngai MJ. Unseen forces: the influence of bacteria on animal
development. Dev Biol 2002;242:1e14.
32. Nyholm SV, McFall-Ngai M. The winnowing: establishing the squi-
devibrio symbiosis. Nat Rev Microbiol 2004;2:632e642.
33. Crowell-Davis SL, Caudle AB. Coprophagy by foals: recognition of
maternal feces. Appl Anim Behav Sci 1989;24:267e272.
34. Osawa R, Blanshard W, Ocallaghan P. Microbiological studies of the
intestinal microflora of the koala, Phascolarctos-Cinereus .2. Pap, a
special maternal feces consumed by juvenile koalas. Aust J Zool
1993;41:611.
35. Lombardo MP. Access to mutualistic endosymbiotic microbes: an
underappreciated benefit of group living. Behav Ecol Sociobiol
2008;62:479e497.
36. McFall-Ngai M, Hadfield MG, Bosch TCG, et al. Animals in a bac-
terial world, a new imperative for the life sciences. Proc Natl Acad
Sci U S A 2013;110:3229e3236.
37. Moran NA, Wernegreen JJ. Lifestyle evolution in symbiotic bacteria:
insights from genomics. Trends Ecol Evol 2000;15:321e326.
38. Baumann P, Moran NA, Baumann LC. Bacteriocyte-associated endo-
symbionts of insects. In: The prokaryotes. Berlin, Heidelberg:
Springer Berlin Heidelberg; 2013. pp. 465e496.
39. Akman Gunduz E, Douglas A. Symbiotic bacteria enable insect to
use a nutritionally inadequate diet. Proc Biol Sci 2009;276:987e991.
40. Abe T, Bignell DE, Higashi M. Termites: evolution, sociality, symbi-
oses, ecology. Springer, Netherlands; 2000. p. 466.
41. Czaczkes TJ, Heinze J, Ruther J, et al. Nest etiquette, where ants go
when nature calls. d’Ettorre P, editor. PLoS One 2015;10:e0118376.
42. Funkhouser LJ, Bordenstein SR. Mom knows best: the universality of
maternal microbial transmission. PLoS Biol 2013;11:1e9.
43. DiGiulio DB. Diversity of microbes in amniotic fluid. Semin Fetal
Neonatal Med 2012;17:2e11.
44. DiGiulio DB, Romero R, Amogan HP, et al. Microbial prevalence,
diversity and abundance in amniotic fluid during preterm labor: a mo-
lecular and culture-based investigation. Fisk NM, editor. PLoS One
2008;3:e3056.
45. Jimenez E, Mar
ın ML, Martin R, et al. Is meconium from healthy
newborns actually sterile? Res Microbiol 2008;159:187e193.
46. Arora N, Sadovsky Y, Dermody TS, et al. Microbial vertical transmis-
sion during human pregnancy. Cell Host Microbe 2017;21:561e567.
47. Dominguez-Bello MG, Costello EK, Contreras M, et al. Delivery
mode shapes the acquisition and structure of the initial microbiota
across multiple body habitats in newborns. Proc Natl Acad Sci
2010;107:11971e11975.
48. Thavagnanam S, Fleming J, Bromley A, et al. A meta-analysis of the
association between Caesarean section and childhood asthma. Clin
Exp Allergy 2008;38:629e633.
49. Neu J, Rushing J. Cesarean versus vaginal delivery: long-term infant
outcomes and the hygiene hypothesis. Clin Perinatol 2011;38:
321e331.
10 Sandoval-Motta et al./ Archives of Medical Research
50. Darmasseelane K, Hyde MJ, Santhakumaran S, et al. Mode of deliv-
ery and offspring body mass index, overweight and obesity in adult
life: a systematic review and meta-analysis. Dewan A, editor. PLoS
One 2014;9:e87896.
51. Roduit C, Scholtens S, de Jongste JC, et al. Asthma at 8 years of age
in children born by caesarean section. Thorax 2009;64:107e113.
52. Dominguez-Bello MG, De Jesus-Laboy KM, Shen N, et al. Partial
restoration of the microbiota of cesarean-born infants via vaginal mi-
crobial transfer. Nat Med 2016;22:250e253.
53. Gibbons L, Beliz
an JM, Lauer JA, et al. The global numbers and
costs of additionally needed and unnecessary Caesarean sections per-
formed per year: overuse as a barrier to universal coverage, World
Health Report, Background Paper 30; 2010;1e31.
54. Fern
andez L, Langa S, Mart
ın V, et al. The human milk microbiota:
origin and potential roles in health and disease. Pharmacol Res 2013;
69:1e10.
55. Cabrera-Rubio R, Collado MC, Laitinen K, et al. The human milk
microbiome changes over lactation and is shaped by maternal weight
and mode of delivery. Am J Clin Nutr 2012;96:544e551.
56. Clemente JC, Ursell LK, Parfrey LW, et al. The impact of the gut micro-
biota on human health: an integrative view. Cell 2012;148:1258e1270.
57. Sonnenburg ED, Smits SA, Tikhonov M, et al. Diet-induced extinc-
tions in the gut microbiota compound over generations. Nature 2016;
529:212e215.
58. Emanuel I, Filakti H, Alberman E, et al. Intergenerational studies of
human birthweight from the 1958 birth cohort. 1. Evidence for a
multigenerational effect. Br J Obstet Gynaecol 1992;99:67e74.
59. Fern
andez L, Breidenstein EBMM, et al. Creeping baselines and
adaptive resistance to antibiotics. Drug Resist Updat 2011;14:1e21.
60. Sonnenburg ED, Sonnenburg JL. Starving our microbial self: the
deleterious consequences of a diet deficient in microbiota-
accessible carbohydrates. Cell Metab 2014;20:779e786.
61. CDC. Get smart for healthcare in hospitals and long-term carejGet
smart for healthcarejCDC. [Internet]. Available from: https://www.
cdc.gov/getsmart/healthcare/. Accessed December 28, 2017.
62. Dethlefsen L, Huse S, Sogin ML, et al. The pervasive effects of an
antibiotic on the human gut microbiota, as revealed by deep 16s
rRNA sequencing. Eisen JA, editor. PLoS Biol 2008;6:2383e2400.
63. Jayasinghe TN, Chiavaroli V, Holland DJ, et al. The new era of treat-
ment for obesity and metabolic disorders: evidence and expectations
for gut microbiome transplantation. Front Cell Infect Microbiol
2016;6:15.
64. Yoshimoto S, Loo TM, Atarashi K, et al. Obesity-induced gut micro-
bial metabolite promotes liver cancer through senescence secretome.
Nature 2013;499:97e101.
65. Cryan JF, Dinan TG. Mind-altering microorganisms: the impact of
the gut microbiota on brain and behaviour. Nat Rev Neurosci
2012;13:701e712.
66. Larsen N, Vogensen FK, Van Den Berg FWJ, et al. Gut microbiota in
human adults with type 2 diabetes differs from non-diabetic adults.
PLoS One 2010;5:e9085.
67. Khoruts A, Dicksved J, Jansson JK, et al. Changes in the composition
of the human fecal microbiome after bacteriotherapy for recurrent
clostridium difficile-associated diarrhea. J Clin Gastroenterol 2010;
44:354e360.
68. Pozhitkov AE, Leroux BG, Randolph TW, et al. Towards micro-
biome transplant as a therapy for periodontitis: an exploratory study
- (2018) -
of periodontitis microbial signature contrasted by oral health, caries
and edentulism. BMC Oral Health 2015;15:125.
69. Proctor LM. The human microbiome project in 2011 and beyond.
Cell Host Microbe 2011;10:287e291.
70. Pertea M, Salzberg SL. Between a chicken and a grape: estimating
the number of human genes. Genome Biol 2010;11:206.
71. Sandoval-Motta S, Aldana M, Mart
ınez-Romero E, et al. The human
microbiome and the missing heritability problem. Front Genet 2017;
8:80.
72. Marian AJ. Elements of ‘‘missing heritability’’. Curr Opin Cardiol
2012;27:197e201.
73. Llewellyn CH, Trzaskowski M, Plomin R, et al. Finding the missing
heritability in pediatric obesity: the contribution of genome-wide
complex trait analysis. Int J Obes (Lond) 2013;37:1506e1509.
74. Zaitlen N, Kraft P. Heritability in the genome-wide association era.
Hum Genet 2012;131:1655e1664.
75. Maher B. The case of the missing heritability. Nature 2008;456:
18e21.
76. O’Malley MA. Endosymbiosis and its implications for evolutionary
theory. Proc Natl Acad Sci U S A 2015;112:10270e10277.
77. Sapp J. Saltational symbiosis. Theory Biosci 2010;129:125e133.
78. Dethlefsen L, Huse S, Sogin ML, et al. The pervasive effects of an
antibiotic on the human gut microbiota, as revealed by deep 16s
rRNA sequencing. PLoS Biol; 2008;2383e2400.
79. Turnbaugh PJ, Ridaura VK, Faith JJ, et al. The effect of diet on the
human gut microbiome: a metagenomic analysis in humanized
gnotobiotic mice. Sci Transl Med; 2009;6e14.
80. Sommer MOA, Dantas G. Antibiotics and the resistant microbiome.
Curr Opin Microbiol 2011;14:556e563.
81. Cheng H, Leblond CP. Origin, differentiation and renewal of the four
main epithelial cell types in the mouse small intestine V. Unitarian
theory of the origin of the four epithelial cell types. Am J Anat
1974;141:537e561.
82. Reshes G, Vanounou S, Fishov I, et al. Timing the start of division in
E. coli: a single-cell study. Phys Biol 2008;5:46001.
83. Torres-Sosa C, Huang S, Aldana M, et al. Criticality is an emergent
property of genetic networks that exhibit evolvability. Rzhetsky A,
editor. PLoS Comput Biol 2012;8:e1002669.
84. Mora T, Bialek W. Are biological systems poised at criticality? J Stat
Phys 2011;144:268e302.
85. Shmulevich I, Kauffman SA, Aldana M. Eukaryotic cells are dynam-
ically ordered or critical but not chaotic. Proc Natl Acad Sci U S A
2005;102:13439e13444.
86. Masel J, Trotter MV. Robustness and evolvability. Trends Genet
2010;26:406e414.
87. De Queiroz K. Ernst Mayr and the modern concept of species. Proc
Natl Acad Sci U S A 2005;102(Suppl 1):6600e6607.
88. Mayr E. Animal species and evolution. Eugen Rev 1964;55:
226e228.
89. Theis KR, Venkataraman A, Dycus JA, et al. Symbiotic bacteria
appear to mediate hyena social odors. Proc Natl Acad Sci U S A
2013;110:19832e19837.
90. Sharon G, Segal D, Ringo JM, et al. Commensal bacteria play a role
in mating preference of Drosophila melanogaster. Proc Natl Acad Sci
U S A 2010;107:20051e20056.
91. Shropshire JD, Bordenstein SR. Speciation by symbiosis: the micro-
biome and behavior. MBio 2016;7:e01785.