Maternal environment and offspring physiology: the inheritance of
information across a generation
by
Nicholas 0. Burton
B.S., Biology
University of Wisconsin-Madison, 2009
Submitted to the Department of Biology in Partial Fulfillment of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
May 2017 ['June- Z017
2017 Nicholas 0. Burton. All rights reserved
The author hereby grants to MIT permission to reproduce and to distribute publicly
paper and electronic copies of this thesis document in whole or in part
Signature of
author
Certified
by
Accepted
by
Signature redacted
Nicholas 0. Burton
May 4, 2017
~ignafure redacted
U- H. Robert Horvitz
Professor of Biology
Thesis Supervisor
ignature redacted
MASSACHUSETTS INSTITUTE!
OF TECHNOLOGY.
MAY 2 3 2017
UIBRARIES
1
Amy Keating
Chair of the Graduate Committee
Professor of Biology
The author hereby grants to MIT permisslon t
reproduce and to distribute publicly paper and
electronic copies of this thesis document in
whole or in part in any medium now known orhereafter created.
Maternal environment and offspring physiology: the transmission of
information across a generation
by Nicholas 0. Burton
Submitted to the Department of Biology in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
Abstract
From the 4th century BC until the late 19th century AD philosophers and
biologists ranging from Hippocrates to Charles Darwin hypothesized that
information about the environment could be passed from parents to progeny.
However, in 1893, German biologist August Weismann tested these hypotheses and
based on his observations concluded that information about the environment could
not be transmitted from parents to progeny. Weismann's hypothesis became known
as the Weismann barrier and served as one of the founding pillars of modern
evolutionary synthesis, which postulates that genetic and phenotypic variability in
plant and animal populations are brought about by genetic recombination resulting
from sexual reproduction and random mutations.
Nonetheless, throughout the 20th century there have been several
observations of plants and animals where parental exposure to environmental
stress modified offspring physiology. These changes in progeny physiology
sometimes enhanced progeny survival in response to repeated environmental
stress, suggesting that information about the environment might be passed from
parent to progeny. The mechanisms by which parental environment can alter
progeny physiology to enhance survival remain unknown.
To explore such mechanisms I investigated how parental exposure of the
nematode C. elegans to osmotic stress affects its progeny's response to continued
osmotic stress. First, I found that C. elegans arrests its development during periods
of osmotic stress to enhance survival and that this developmental arrest is caused
by a loss of insulin-like signaling to the intestine. I then discovered that exposure of
parents to mild osmotic stress enhances progeny resistance to osmotic stress and
determined that this adaptation is the result of a loss of insulin-like signaling to the
maternal germline, which results in increased expression of the glycerol
biosynthetic enzyme GPDH-2 in embryos; the increased GPDH-2 expression results
in increased glycerol production, which in turn protects progeny from osmotic
stress. These results indicate that insulin can cross the Weismann barrier and
suggest that changes in maternal insulin signaling might be responsible for effects of
the maternal environment on human diseases that involve insulin signalling, such as
obesity and type-2 diabetes.
From a screen for mutants that fail to arrest development in response to
osmotic stress I identified the cytosolic sulfotransferase SSU-1 and found that SSU-1
functions in the ASJ sensory neurons to control development and insulin-sensitivity
in response to osmotic stress.
Thesis Advisor H. Robert Horvitz
Title: Professor of Biology
2
Acknowledgements
First and foremost, I would like to thank my thesis advisor, Bob Horvitz, for
allowing me the freedom to pursue any research topic I was interested in. I might
have spent a considerable amount of time pursuing "dead ends," but the freedom to
continue experimenting eventually led me to observations I truly find interesting
None of the observations in this thesis would have been possible without this
freedom and support. Bob also provided considerable guidance on how to present
and communicate my results as a scientific story. For these and many other
instances of help and guidance I am extremely grateful.
I would also like to thank my thesis committee members Dennis Kim and
Mary Gehring. Over the years they have provided considerable guidance both
scientifically and career-wise during the course of many committee meetings, which
often went well past the scheduled time frame. I very much appreciated your advice
and guidance. In addition, I would like to thank my outside committee member
Oliver Rando for reading my thesis and agreeing to be part of my thesis committee.
I would like to thank all the members of the Horvitz lab for their general
scientific and life advice and for generally making the Horvitz lab an amazing place
to spend the last five years. A special thanks to Na An for a monumental amount of
help in ordering, freezing, thawing, and maintaining the myriad of strains I ordered
or tested. This thesis would similarly have not been possible without your help as I
3
can only imagine what would have happened to all my various worm strains
without you; and an additional special thanks to Kirk Burkhart for all your editing
and scientific advice going back to UW-Madison and the Kennedy lab.
Thank you to my Mom and Dad, my brothers Joe and Jim, my sister-in-law
Kyrie, and (for the latter part of my time in graduate school) my nieces Adeline and
Azaleah. I know that my pursuit of a career as a researcher has continually resulted
in my moving further and further away, from Madison, to Boston, and now to the
UK, but your steadfast support and encouragement to do what I love (even if you're
not quite sure what that is) has been incredibly helpful and kept me a generally
cheerful person despite not being able to see you all as much as I might like. I would
similarly like to thank my wife's parents, Sarah and Mort Orlov, for being my family
here in Boston and the considerable amount of support (and meals) that have
helped power me through graduate school.
Finally, and most importantly, I would like to thank my wife, Lindsay, for
more help and support than I can possibly write about here (although if I tried it
might considerably help my page count). Your love and support has made my time
in grad school immensely more enjoyable and successful and your continual editing
of my writing has directly contributed to everything that I've done.
4
Table of Contents
Title Page 1
Abstract 2
Acknowledgments 3
Table of Contents 5
Chapter One: Introduction 9
Part 1: A brief history of the inheritance of information about the environment 10
Part 2: Defining information and the potential mechanisms by which information
about the environment can be inherited 12
Soma to soma inheritance of information 13
Soma to germline inheritance of information 14
Changes in resource allocation 14
Phenotypic effects that depend on parental environment but do not represent the
transfer of information 15
Part 3: Effects of parental environment on progeny phenotype in plants 16
Heritable effects caused by pathogen infection 16
Heritable effects caused by herbivory 17
Heritable effects caused by osmotic stress 18
Conclusions from plants 19
Part 4: Effects of parental environment on progeny phenotype in invertebrates 20
The inheritance of information in pea aphids 20
The inheritance of information in Bombyx mori 21
5
The inheritance of information in C. elegans 22
The inheritance of information in Daphnia 25
Other examples of the inheritance of information in invertebrates 26
Conclusions from invertebrates 26
Part 5: Effects of parental environment on progeny phenotype in vertebrates 28
The inheritance of information about population density in squirrels 28
The inheritance of information about predation in vertebrates 29
The inheritance of information about parental trauma in mice 29
The heritable deleterious effects of parental dietary stress in mice 30
Conclusions from vertebrates 31
Part 6: Effects of parental environment on progeny phenotype and its relationship
to human disease 32
Part 7: Concluding Remarks 35
Part 8: Outline of Thesis 37
Bibliography 39
Chapter 2: Insulin-like signalling to the maternal germline controls progeny
response to osmotic stress 46
Summary 47
Introduction 48
Results 48
Discussion 56
Methods 57
Acknowledgements 67
Bibliography 68
6
Figure 1: Insulin-like signalling to the intestine regulates developmental arrest in
response to osmotic stress. 71
Figure 2: Insulin-like signalling to the maternal germline regulates progeny
response to osmotic stress 73
Figure 3: Insulin-like signalling to the maternal germline modifies progeny response
to osmotic stress by regulating the RAS-ERK-like pathway 75
Figure 4: RAS-ERK signalling regulates C. elegans response to bacterial infection and
starvation 77
Figure Si: C. elegans arrests development in response to osmotic stress. 79
Figure S2: Comparison of gene expression in response to osmotic stress and
starvation. 82
Figure S3: Insulin-like signalling to the germline regulates developmental arrest in
response to osmotic stress. 84
Figure S4: Dense-core vesicle release from sensory neurons regulates MPK-1
activity in the germline. 86
Figure S5: Quanitication of dpMPK-1 in additional germlines of wild-type animals at
50 mM NaCl and 300 mM NaCl 88
Table Si: Most insulin peptides do not regulate development in response to osmotic
stress. 90
Chapter 3: SSU-1 functions in the ASJ sensory neurons to control metabolism
and development in C. elegans 93
Abstract 94
Results and Discussion 95
Materials and Methods 103
Acknowledgements 108
References 109
Figure 1: The cytosolic sulfotransferase SSU-1 functions in the ASJ sensory neurons
to regulate developmental arrest in response to osmotic stress. 111
7
Figure 2: SSU-1 and DAF-16 regulate the transcription of a common set of target
genes in response to osmotic stress 113
Figure 3: SSU-1 functions in parallel to insulin-like signaling to regulate
developmental arrest in response to osmotic stress 115
Figure 4: SSU-1 and DAF-16 regulate the metabolism of certain metabolites in C.
elegans 117
Figure Si: SSU-1 is not required for DAF-16 translocation into the nucleus in
response to osmotic stress 119
Chapter 4: Discussion and Future Directions 121
Future directions related to maternal environmental stress
and offspring physiology 122
Future directions related to SSU-1 123
Additional future directions 125
8
Chapter 1
Introduction
Nicholas Burton
9
Part 1: A brief history of the inheritance of information about the environment
The hypothesis that information about the environment can modify an
individual's physiology and that these changes in physiology can be inherited by
their offspring can be traced back as far as Hippocrates (Zirkle, 1935). By the late
18th century and early 19th century this hypothesis was widely accepted and
notably written about in 1794 in Zoonomia by English biologist Erasmus Darwin
(Charles Darwin's grandfather) and in 1809 in Philosophie Zoologique by French
biologist Jean-Baptiste Lamarck.
In his 1868 publication The Variation of Plants and Animals Under
Domestication, English biologist Charles Darwin proposed that "gemmules"
(imaginary particles of inheritance) might be shed from organs of the body and
collected in reproductive cells, transmitting some limited information about the
parental environment from parents to their offspring (Darwin, 1868). This
hypothesis was the first proposal of a potential mechanism underlying the
inheritance of information from parent to offspring and was proposed in part to
explain the hypotheses of Erasmus Darwin and Jean Baptiste Lamarck.
Francis Galton, Charles Darwin's cousin, tested Darwin's theory of gemmules
in a series of experiments from 1869-1871. Galton hypothesized that 1) the coat
color of rabbits was controlled by gemmules 2) gemmules were transmitted to
reproductive cells via the blood and 3) blood transfusions from one color rabbit to
another would alter the gemmules that are transmitted to reproductive cells and
10
therefore change the coat color of the resulting offspring. Galton found that blood
transfusions did not affect coat color in rabbits and concluded that gemmules likely
did not exist (Galton, 1870).
German biologist August Weismann later also tested Darwin's hypothesis of
gemmules and similarly concluded that gemmules did not exist (Weismann, 1893).
Specifically, Weismann tested whether information about the environment could be
transmitted to offspring by cutting the tails off of 68 mice repeatedly over 5
generations, observing that "901 young were produced by five generations of
artificially mutilated parents, and yet there was not a single example of a
rudimentary tail or of any other abnormality in this organ" (Weismann, 1889).
These data led Weismann to hypothesize that information about the environment
could not be transmitted from somatic cells to reproductive cells. Weismann's
hypothesis eventually became known as the "Weismann barrier," and the
widespread acceptance of the existence of the Weismann barrier led Darwin's
theory of gemmules to be largely forgotten.
The Weismann barrier went on to serve as one of the founding pillars of
modern evolutionary synthesis, which is a set of hypotheses that remains the
leading explanation for genetic and phenotypic variability among populations of
plants and animals. One of the basic conclusions of modern evolutionary synthesis is
"genetic and phenotypic variability in plant and animal populations is brought about
by genetic recombination resulting from sexual reproduction and random mutations
along the parent-offspring sequence" (Kutshera and Niklas, 2004). This hypothesis
11
excludes any inheritance of information or phenotypic variability dependent on
parental environment, such as the gemmules proposed by Charles Darwin.
Now, in the 21st century, several studies have demonstrated that parental
exposure to environmental stresses can lead to significantly altered offspring
development and physiology of several organisms, ranging from plants to humans
(Kyle and Pichard, 2006; Lumey et al. 2007; Storm and Lima, 2010; Sharma et al.
2016; Chen et al. 2016; Huypens et al. 2016). These observations raise the
possibility that the Weismann barrier is not impenetrable, allowing certain
environmental conditions to trigger changes in reproductive cells that ultimately
result in a DNA-independent shift in physiology from parent to offspring. However,
the mechanisms by which parental environmental stress modifies offspring
physiology, as well as whether any such mechanism represents either (1)
information crossing the Weismann barrier or (2) something resembling Charles
Darwin's gemmules, has remained unknown.
Part 2: Defining information and the potential mechanisms by which
information about the environment might be inherited
Classically, genetic information refers to the information in the genome --
encoded in the DNA sequence -- which represents the vast majority of all biological
information that is passed from one generation to the next. Nonetheless, over the
last several decades there have been a number of observations for which phenotypic
differences between individuals are controlled by parental environment and passed
on to their offspring independently of changes to DNA sequence. For the purposes of
12
this chapter I will define these observations as the inheritance of environmental
information. Similar to genetic information, the inheritance of environmental
information requires that information be passed from one generation to the next.
However, unlike genetic information, the phenotypic differences regulated by
parental environment are transient and in most cases last only for a single
generation. In addition, unlike genetic information, there are many possible
mechanisms by which environmental information might be passed from parent to
offspring. These possible mechanisms can be broken into three broad classes: the
inheritance of information from soma to soma, which does not require information
crossing the Weismann barrier, the inheritance of information from soma to
germline, which does require information crossing the Weismann barrier, and
changes in resource allocation.
Soma to soma inheritance of information
One possible mechanism for the transfer of information from soma to soma
across a generation is the transfer of signaling molecules from somatic cells of the
mother to somatic cells of the offspring. For example, thyroid hormones in humans
can be transported across the placenta and are thought to regulate fetal
development by directly activating receptors in the fetus (James et aL. 2007). These
types of observations represent the transfer of information from somatic cells of the
mother to somatic cells of the offspring and do not require information to cross the
Weismann barrier.
13
Soma to germline inheritance of information
Information about parental environment could also be transmitted from
parents to progeny via germ cells and there are several possible mechanisms by
which such information might be transmitted. Broadly speaking, the transfer of
information from somatic cells to germ cells requires signaling from parental
somatic cells to germ cells in such a way that modifies germ cells and results in
changes in the offspring's phenotype. For example, parental signaling could cause
changes in germ cell RNA composition, protein composition, nucleosome
positioning, histone modifications, or DNA methylation, all of which could affect
offspring development, physiology, or metabolism. These types of changes, where
signaling from parental somatic cells to germ cells causes changes to offspring
phenotype, would all require information to cross the proposed Weismann barrier.
Changes in resource allocation
Finally, maternal exposure to environmental stress could cause changes in
the amount, type, or quality of resources allocated to offspring. For example,
environmental stress might alter the amount or composition of yolk deposited into
the eggs of oviparous animals. Similarly, environmental stress might alter the
amount, type, or quality of resources deposited into the seeds of plants or
14
transported across the placenta in mammals. Changes in resource allocation do not
require information to cross the proposed Weismann barrier.
Phenotypic effects that depend on parental environment but do not represent the
transfer of information
One of the difficulties of studying the transfer of information from parent to
progeny is that parental environment can also affect progeny phenotype via
mechanisms that do not require the transfer of information. Differentiating between
these possible mechanisms remains challenging. For example, fetal alcohol
syndrome is caused, in part, by alcohol in the mother's blood crossing the placenta
and entering the fetus through the umbilical cord (Weinberg et al. 2007). In this case
the direct exposure of the fetus to the environmental stress of alcohol causes long-
term changes in offspring physiology. Such direct exposure of germ cells, the
embryo, or the fetus to environmental stress does not represent the transfer of
information from parent to progeny.
Similarly, in plants, such as Arabidopsis thaliana, germ cells are specified
from somatic cells (Scott and Spielman, 2006). The incomplete erasure of epigenetic
marks from the somatic cells upon transition to the reproductive cell fate might also
result in phenotypic changes to offspring. In such a case, it remains possible that
information about the environment might be maintained into the next generation.
However, in this example the environmental stress is still directly affecting the cells
that become the next generation of the species and thus does not require the
15
transfer of information, but rather the maintenance of information. Nonetheless,
such cases represent the blurring of the lines between somatic and germ cells.
Part 3: Effects of parental environment on progeny phenotype in plants
Heritable effects caused by pathogen infection
Studies of plants were among the first to reveal that, in some cases, parental
exposure to environmental stress could enhance progeny resistance to repeated
environmental stress. For example, in 1983, D. A. Roberts demonstrated that the
infection of tobacco plants (Nicotiana tabacum) by tobacco mosaic virus (TMV)
resulted in progeny that were resistant to repeated infection by TMV (Roberts,
1983), and follow-up studies found that parental infection of N. tabacum by TMV not
only enhanced progeny resistance to TMV, but also enhanced progeny resistance to
the bacterial pathogen Pseudomonas syringae and the pathogenic oomycete
Phytophthora nicotianae (Kathiria et al. 2010). These studies indicate that there is a
heritable link between parental exposure to pathogen and progeny response to
pathogen.
In 2012, similar studies of Arabidopsis thaliana found that parental exposure
to the bacterial pathogen Pseudomonas syringae resulted in progeny that were
resistant to both repeated infection by P. syringae and infection by the oomycete
pathogen Hyaloperonospora arabidopsidis (Luna et al. 2012; Slaughter et al. 2012).
However, unlike previous studies, these studies found that parental infection of A.
16
thaliana by P. syringae resulted in progeny that were more susceptible to the fungal
pathogen Alternaria brassicicola (Luna et al. 2012). The defense response of A.
thaliana to P. syringae and H. arabidopsidis is regulated by salicylic acid, while the
defense response of A. thaliana to the fungal pathogen A. brassicicola is regulated by
jasmonic acid. Based on these observations, Luna et al. (2012) proposed a model in
which parental infection of A. thaliana by P. syringae primed progeny to elicit a
stronger defense response to salicylic acid signaling and a weaker defense response
to jasmonic acid signaling. In addition, Luna et al. (2012) found increased histone 3
lysine 9 acetylation (H3K9ac) at the promoters of genes regulated by salicylic acid
and increased histone 3 lysine 27 trimethylation (H3K27me3) at the promoters of
genes regulated by jasmonic acid in the progeny of plants exposed to P. syringae
when compared to the progeny of plants that were not exposed to P. syringae (Luna
et al. 2012). Furthermore, these studies showed that both the adaptation to
infection and the observed changes in histone modifications depended on the
defense regulatory protein NPR1. Collectively, these data suggest that the infection
of A. thaliana by P. syringae results in specific changes in histone modifications at
the promoters of defense genes and that these changes in histone modifications can
be inherited by progeny, resulting in progeny that are resistant to future infection
by P. syringae. The mechanism by which this information in inherited and whether
this phenomena represents a change in signaling to the germline or a change in
maternal provisioning remains unknown.
Heritable effects caused by herbivory
17
In addition to the observations that information about plant responses to
pathogens can be inherited from parent to progeny, several studies have found that
plant responses to herbivory can also be inherited from parent to progeny. Rasmann
et al. (2012) found that when both A. thaliana and the tomato plant Solanum
lycopersicum were exposed to the herbivore Helicoverpa zea their progeny were
more resistant to herbivory than the progeny of plants that were not exposed to H.
zea. In the case of A. thaliana this adaptation was regulated by jasmonic acid and the
production of small interfering RNAs (siRNAs) (Rasmann et al. 2012). Similar
studies of several other species of plants, including Raphanus raphanistrum
(Agrawal et al. 1999), Mimulusguttatus (Holeski, 2007), and Lotus wrangelianus
(terHorst and Lau, 2012) found that information about herbivory could be passed
from parent to progeny and that this information enhanced progeny resistance to
herbivory. Much like the observations of pathogen infection, the mechanism by
which parental exposure to herbivory can modify offspring response to herbiovry
remains unknown.
Heritable effects caused by osmotic stress
Finally, recent studies of A. thaliana demonstrated that information about
osmotic conditions can also be passed from parent to progeny to enhance progeny
resistance to hyperosmotic stress (Wibowo et al. 2016). Wibowo et al. (2016) found
that the progeny of plants that were osmotically stressed exhibited enhanced
18
survival and germination rates in response to osmotic stress when compared to
progeny of plants that were not osmotically stressed. In addition, this study found
that the adaptation to osmotic stress was regulated by DNA methylation and was
predominantly inherited via the mother.
Conclusions from plants
The precise mechanisms underlying each of these observations remain
unknown. However, in the cases of herbivory and pathogen infection, the
reproductive cells are not directly exposed to the herbivore or pathogen and the
effects appear to be mediated by hormonal signaling. Changes in hormonal signaling
in the parents could affect progeny resistance to stress by at least three possible
mechanisms. First, changes in hormonal signaling could alter the allocation of
resources into seeds, which in turn results in changes in offspring physiology.
Second, the hormones themselves could be deposited into seeds and subsequently
modify the somatic cells of the next generation. Finally, changes in hormonal
signaling to germ cells could program the next generation in such a way that
enhances their resistance to future stress. The lack of a mechanistic understanding
of how parental exposure to a variety of environmental stresses primes plant
offspring to respond to a similar stress means that it is unclear if any of these
example represent either information crossing the Weismann barrier or anything
resembling Darwin's gemmules.
19
Part 4: Effects of parental environment on progeny phenotype in
invertebrates
The inheritance of information in pea aphids
Among the first examples of the inheritance of information about the
environment in invertebrates was the finding that pea aphids could develop along
two distinct developmental trajectories, a winged morph and a wingless morph.
These two morphs differ not only in the presence or absence of wings, but also in
the presence or absence of flight muscles, the development of compound eyes,
antennae length, number of rhinaria, and the sclerotization of the head and neck
(Kalmus, 1945; Kring, 1977; Kawada, 1987; and Miyazaki, 1987). Despite these
significant differences, winged and wingless pea aphids are genetically identical and
in many of the approximately 5000 species of pea aphids, such as Acyrthosiphon
pisum, the decision to develop into either the winged or wingless morph depends on
their mother's environment. Under favorable maternal conditions, a majority of
progeny develop into wingless females, which reproduce more efficiently than the
winged morph. (A. pisum can reproduce either sexually or asexually by
parthenogenesis and therefore do not require males to reproduce.) By contrast,
under unfavorable maternal conditions, such as during high aphid density or the
presence of predators, a majority of progeny develop into winged males and
females. The molecular mechanism underlying how parental environment regulates
progeny development in pea aphids remains unknown.
20
The inheritance of information in Bombyx mori
One of the best-studied examples of the inheritance of information about the
environment is the regulation of diapause in silkworm (Bombyx mori) embryos,
which have been domesticated and cultured for over 5,000 years for the purpose of
silk production. B. mori embryos can enter diapause to overwinter just before
segmentation (approximately 12,000 cell-stage) (Watanabe, 1924, and Kitazawa et
aL. 1963). Whether or not embryos enter diapause is regulated by the temperature
of the environment during embryonic development of the mother (Wantanabe,
1924; Xu et aL. 1995; and Sato et aL. 2014). Specifically, when maternal embryos are
cultured at approximately 25'C they develop and lay diapause eggs (Wantanabe,
1924). By contrast, when maternal embryos are cultured at 15'C they develop and
lay non-diapause eggs (Wantanabe, 1924). Molecularly, this decision is regulated by
the release of a 24 amino acid peptide, diapause hormone, from the subesophageal
ganglion (Fukuda 1952; Hasegawa 1952), which signals to oocytes to increase the
transcription of the trehalase gene treh-2. This signaling to oocytes results in
increased glycogen production in embryos and diapause arrest of embryos at the
approximately 12,000 cell stage (Yamashita, 1996; Kitazawa et aL. 1963; Kamei et al.
2011). The molecular mechanisms regulating diapause entry and recovery from
diapause remain largely unknown. Nonetheless, these studies represent one of the
first observations of signaling from somatic cells to oocytes to regulate the
development and physiology of offspring. In addition, these results appear to
21
represent a crossing of the Weismann barrier by diapause hormone and the transfer
of environmental information from parent to offspring.
The inheritance of information in C. elegans
Another example of the inheritance of information about the environment
from parent to progeny in invertebrates was the discovery that gene silencing by
double-stranded RNA (dsRNA), also referred to as RNA interference (RNAi), could
be inherited in the nematode Caenorhabditis elegans (Fire et aL. 1998). Since this
discovery, considerable work has been done to determine the mechanism
underlying the inheritance of RNAi. Breifly, C. elegans can uptake dsRNA into their
intestine from their food source, E. coli (Timmons and Fire, 1998; Winston et aL.
2002). From the intestine dsRNA is thought to be transported throughout the
organism, including to the germline, via a process called systemic RNAi (Winston et
aL. 2002; Winston et aL. 2007), which requires the RNAi-spreading defective genes
rsd-1, rsd-2, rsd-4, rsd-6, and rsd-8 among others (Winston et aL. 2002; Feinberg and
Hunter 2003; Tijsterman et aL. 2004). Once RNAi silencing is established in the
germline, it is maintained throughout the life of the progeny by the nuclear RNAi
pathway, which is composed of the argonaute NRDE-3 and three proteins of
unknown function, NRDE-1, NRDE-2, and NRDE-4 (Burton et aL. 2011). Notably, the
grand-progeny of animals exposed to dsRNA do not inherit somatic RNAi silencing
(Burton et aL. 2011), suggesting that the mechanism underlying RNAi inheritance
functions in such a way to transmit information from somatic cells of the parents to
22
the germline and back to the somatic cells of the progeny for only a single
generation. However, when RNAi silencing is directed to germline-expressed genes,
such as oma-1, RNAi silencing can be inherited for 10 or more generations
(Vastenhouw et al. 2006; Ashe et al. 2012; Shirayama et al. 2012; Buckley et al.
2012). This germline-specific inheritance also requires NRDE-1, NRDE-2, and NRDE-
4, but is regulated by a different argonaute, HRDE-1 (Buckley et al. 2012).
The biological function of RNAi inheritance in germline tissue and somatic
tissues appears to be distinct. The inheritance of RNAi silencing specifically in germ
cells is required for maintaining germ cell immortality, and animals with mutations
in hrde-1, nrde-1, nrde-2, or nrde-4 go progressively sterile over many generations
(Buckley et al. 2012). By contrast, the likely function of somatic RNAi inheritance is
to pass on protection from viral infection from parent to progeny and consistent
with this hypothesis Gammon et aL (2017) found that parental exposure of C.
elegans to the negative-sense ssRNA vesicular stomatitis virus (VSV) results in
offspring that are more resistant to viral infection and that this adaptation depends
on RNAi silencing machinery.
In addition to RNAi inheritance, two recent studies of C. elegans found that
parental exposure to a variety of environmental stresses, including osmotic stress,
starvation, and heavy metal stress, can enhance progeny survival in response to
oxidative stress (Kishimoto et al. 2017) and that maternal dietary restriction can
enhance progeny resistance to starvation (Hibshman et al. 2016). In both studies it
was determined that parental activation of the FOXO transcription DAF-16 in
somatic cells was required for parental exposure to environmental stress to protect
23
progeny from future environmental stress (Hibshman et al. 2016; Kishimoto et al.
2017), indicating that parental DAF-16 activation might generally enhance progeny
resistance to environmental stress. However, Kishimoto et al. (2017) found that the
effects of parental exposure to environmental stress on progeny could be
transmitted via either the paternal or maternal germline, while Hibshman et aL
(2016) found that the effects of parental dietary restriction on progeny could be
transmitted only via the maternal germline. This difference in transmission
indicates that these two observations might be regulated by distinct mechanisms,
despite both effects requiring parental DAF-16 activation.
Kishimoto et al. (2017) found that an H3K4me3 regulatory complex member,
WDR-5, was required for parental exposure to a variety of environmental stresses to
enhance progeny resistance to oxidative stress. This observation suggests a possible
role for WDR-5 in the transfer of environmental information from parent to
offspring. Interestingly, WDR-5 was previously shown to function in the
transgenerational inheritance of longevity in C. elegans (Greer et al. 2011).
Specifically, Greer et al. (2011) observed that the loss of WDR-5 can enhance
longevity and that this extension of longevity can be inherited for up to four
generations. These observations suggest that WDR-5 activity in the germline might
generally regulate the inheritance of information across generations. Alternatively,
since longevity in C. elegans is also regulated by insulin-like signaling and DAF-16
activation (Martins et al. 2016), it is possible that WDR-5 activity in the germline
modifies insulin-like signaling and DAF-16 activation in progeny, which might
24
explain both the observed effects of WDR-5 on longevity (Greer et al. 2011) and
progeny response to stress (Kishimoto et al. 2017).
The inheritance of information in Daphnia
The development of helmets in the water flea Daphnia cucullata is regulated
by maternal exposure to predators. Specifically, parental exposure of the D.
cucullata to aquatic chemicals released by predators (kairomones), such as
Leptodora kindtii and Chaoborusflavicans, results in offspring with helmets that are
almost twice the size as offspring from mothers that were not exposed to
kairomones (Agrawal et al. 1999). Increased helmet size acts as a defense against
predators by lowering predator capture success and Daphnia with larger helmets
exhibited enhanced survival in response to both L. kindtii and C. flavicans (Agrawal
et al. 1999). In addition to helmet size, maternal exposure to kairomones can also
regulate progeny developmental rate of Daphnia ambigua (Walsh et al. 2014) and
neckteeth formation in Daphnia pulex (Tollrian, 1995; Miyakawa et al. 2010). These
studies found that the progeny of mothers exposed to kairomones develop faster in
the case of D. ambigua (Walsh et al. 2014) and develop more neckteeth in the case of
D. pulex (Tollrian 1995; and Miyakawa et al. 2010) than the progeny of mothers not
exposed to kairomones. These and other studies suggest that multiple phenotypes
in several species of Daphnia can be regulated by maternal exposure to predators.
Other examples of the inheritance of information in invertebrates
25
The inheritance of information about the environment from parent to
progeny has also been reported for several other species of invertebrates. In many
cases, such as those already described for pea aphids and water fleas, the exposure
of parents to predators alters progeny development to enhance progeny survival in
response to predators in the environment. Several studies have found that in
addition to modifying development, maternal exposure to predators can also alter
progeny behavior. For example, studies of the field cricket Gryllus pennsylvanicus
found that the offspring of gravid crickets exposed to the wolf spider Hogna helluo
spend more time as immobile animals and exhibit enhanced survival in
environments with wolf spiders than offspring of mothers not exposed to wolf
spiders (Storm and Lima, 2010). Similarly, studies of the predatory mite
Phytoselulus persimilis and the intraguild predator Amblyselus andersoni found that
P. persimilis mothers exposed to their predator A. andersoni produce progeny that
spend more time immobile than progeny from mothers not exposed to A. andersoni
(Seiter and Schausberger, 2015). These observations of both crickets and mites
suggest that maternal exposure to predators can alter the behavior of offspring.
However, any potential mechanism for the inheritance of behavioral changes
remains unknown.
Conclusionsfrom invertebrates
26
Similar to observations of plants, parental environment regulating progeny
physiology of both pea aphids and Daphnia appear to be regulated by hormonal
signaling, but whether they represent changes in the allocation of resources,
hormonal signaling from somatic cells of the mother to somatic cells of the offspring,
or hormonal signaling from somatic cells to germ cells remains unknown. Thus,
these observations do not indicate whether information can cross the Weismann
barrier.
By contrast, in the cases of Bombyx mori and of RNAi inheritance in C.
elegans, environmental information is transmitted from somatic cells of the parent
to germ cells, indicating that the Weismann barrier is not impermeable in all cases.
In addition, these studies indicate that multiple types of molecules (diapause
hormone and dsRNA) have evolved to cross the Weismann barrier, suggesting that
somatic to germ cell signaling might be more common than previously anticipated.
In addition to the Weismann barriers being permeable in these cases, it should be
noted that both diapause hormone and dsRNA resemble Charles Darwin's proposed
gemmules in that (1) they are released from somatic cells of the parent, (2) they are
transmitted to germ cells, and (3) they transmit some limited information that
modifies offspring physiology.
It remains unclear if similar mechanisms occur outside of invertebrates,
since diapause hormone does not appear to be conserved outside of invertebrates
and dsRNA is not known to be transmitted from somatic cells to germ cells in
mammals. These observations suggest that similar mechanisms of inheritance might
be limited to invertebrates.
27
Part 5: Effects of parental environment on progeny phenotype in vertebrates
The inheritance of information about population density in squirrels
North American red squirrels (Tamiasciurus hudsonicus) defend limited
territories, and squirrels that fail to acquire a territory before their first winter do
not survive (Dantzer et al. 2013). Acquisition of a territory depends largely on
population density. Dantzer et al. (2013) found that in years during which squirrel
density is high, mothers give birth to faster growing offspring and that these faster
growing offspring are more likely to acquire a territory and survive their first
winter. In addition, Dantzer et al. (2013) found that this effect on offspring growth
rate was caused by perceived squirrel density and could be reproduced even in low
squirrel density environments by either playing audio recordings of red squirrel
rattles or by adding cortisol to the squirrels' food source. These results suggest that
perceived high squirrel density is sensed audibly, which results in increased stress
levels and increased cortisol levels in pregnant mothers. These stressed mothers
then give birth to faster growing offspring. Interestingly, the authors found that
maternal cortisol levels affect postnatal growth rate rather than fetal growth rate,
suggesting that the effects of maternal stress persisted after birth. However, the
mechanisms by which maternal cortisol levels affect offspring postnatal growth
remain unclear and could be regulated by changes in provisioning or cortisol
crossing the placenta in utero.
28
The inheritance of information about predation in vertebrates
Similar to invertebrates, predation has also been found to affect offspring
development, growth rate, and behavior in both fish (Gasterosteus aculeatus and
Eretmodus cyanostictus) and birds (Turdus philomelos) (Segers and Taborskiy, 2011;
Coslovsky and Richner, 2011; Giesing et al. 2010; Mommer and Bell, 2014).
Specifically, in the case of the stickleback fish (G. aculeatus), when mothers are
exposed to predation they produced larger eggs with higher cortisol content. Upon
hatching, fry that develop from larger eggs with higher cortisol content exhibit
tighter shoaling behavior and antipredator defense (Giesing et al. 2010). These
results, similar to observations of squirrels, suggest that maternal cortisol can
influence offspring development and behavior and are likely the result of maternal
cortisol directly and stably programming somatic cells of the offspring. While such
observations do not represent a crossing of the Weismann barrier, the mechanisms
underlying how maternal cortisol levels can affect the postnatal physiology of
offspring might have important ramifications for human biology, as cortisol is a
conserved stress-signaling molecule in mammals.
The inheritance of information about parental trauma in mice
Dias and Ressler (2014) examined the inheritance of parental traumatic
experience in mice by pairing odorants with foot shock. They found that the
29
offspring of fear-conditioned males show increased sensitivity to the odorant
administered to the father but no change in sensitivity to an unpaired odorant. In
addition, Dias and Ressler found that this change in offspring behavior was paired
with an increased number of olfactory sensory neurons in the main olfactory
epithelium, which is known to sense the odorant tested (acetophenone) (Dias and
Ressler, 2014). Finally, the authors found that the changes in sensory neuron
number in offspring of fear-conditioned fathers were still present even if the
offspring were conceived by in vitro fertilization. These results suggest that changes
in behavior and neurodevelopment can be inherited via sperm. The authors suggest
that changes in DNA methylation in sperm might underlie this heritable effect.
However, the functional significance of the observed incremental changes in DNA
methylation remains unknown.
The heritable deleterious effects of parental dietary stress in mice
Recently, several studies have found that the effects of dietary stress on
parent mice can be transmitted to the next generation via either the maternal or
paternal germline. For example, with respect to the female germline, Huypens et al.
(2016) found that feeding female mice a high-fat diet results in offspring with
increased prevalence of obesity and glucose intolerance and that these effects can
be transmitted via oocytes (Huypens et al. 2016). These observations are consistent
with other observations of maternal diet and maternal diabetes regulating offspring
physiology and metabolism via oocytes (Sasson et al. 2015, Luzzo et al. 2012,
30
Jungheim et al. 2010, Wyman et al. 2008) and the many observations of maternal
dietary stress during pregnancy regulating offspring physiology and metabolism
(reviewed by Lakshmy (2013).
Much like observations of maternal dietary stress, paternal dietary stress can
also have adverse effects on offspring metabolism and physiology and that these
effects can be transferred via sperm (Anderson et aL. 2006; Ng et aL. 2010; Sharma et
aL. 2016; Chen et aL. 2016). In addition, Sharma et aL. (2016) and Chen et aL. (2016)
demonstrate that the effects of dietary stress on fathers can be transmitted to
offspring via changes in the population of tRNA fragments deposited into sperm by
epididymosomes (vesicles that fuse with sperm during epididymal transit) that are
shed from somatic tissue (Sharma et aL. 2016). Collectively, these results indicate
that dietary stress can modify both sperm and oocytes to result in obesity and
glucose intolerance in offspring. However, why such mechanisms exist remains
unclear, as it appears that such effects have an adverse effect on offspring health. It
is possible that under certain circumstances the effects of parental dietary stress
might enhance offspring survival and/or reproduction. Alternatively, the adverse
effects of parental dietary stress on progeny might simply be due to deleterious
effects of stress on oocyte or sperm maturation, including the proper provisioning of
tRNA fragments into sperm. Regardless of the biological "goal," it appears that the
effects of dietary stress on sperm and oocytes can cause common metabolic
disorders, such as type 2 diabetes, and thus further research is warranted.
Conclusions from vertebrates
31
Compared to plants and invertebrates, for vertebrates there are relatively
few examples for which parental exposure to environmental stress enhances
offspring survival. Nonetheless, examples involving squirrels, mice, stickleback fish,
and thrush song birds suggest that similar heritable effects do occur in vertebrates,
and observations of mice suggest that the heritable effects of parental dietary stress
might underlie certain human metabolic pathologies. Furthermore, observations by
Sharma et al. (2016) and Chen et aL (2016) demonstrate that tRNA fragments can be
transmitted from somatic cells of the parent to germ cells of the father to modify
offspring physiology. These observations indicate that the Weismann barrier might
also be permeable in mammals. In addition, similar to diapause hormone in B. mori
and dsRNA in C. elegans, tRNA fragments resemble Charles Darwin's proposed
gemmules in that they are shed by somatic tissue and transmitted to germ cells in
way that alters offspring physiology. However, the evolutionary function of these
tRNA fragments remains unclear, as the observed changes in tRNA populations in
response to dietary stress do not appear to enhance the survival of the offspring.
Part 6: Effects of parental environment on progeny phenotype in humans and
its relationship to human disease
The thrifty phenotype hypothesis speculates that the well-documented
epidemiological association between fetal growth and metabolic disorders such as
cardiovascular diseases and type 2 diabetes are the result of poor maternal
nutrition during pregnancy (Hales and Barker, 1991). Specifically, in a study of men
32
aged 59-70 years who were born and still living in Hertfordshire (UK), and whose
birth weights were below the median, Hales and Barker (1991) found that 26% had
impaired glucose tolerance and 15% had diabetes. By contrast, of men whose birth
weights were above the median, 5% had impaired glucose tolerance and 2% had
diabetes. These epidemiological associations of low birth weight and metabolic
disorders in later life have been reproduced in numerous studies and diverse
populations throughout the world (Hales and Barker, 2001).
The hypothesis that in utero malnutrition causes metabolic disorders in later
life was initially met with significant skepticism, as it was speculated that genetic
polymorphisms that underlie type 2 diabetes might also underlie low birth weight.
However, subsequent studies of the 45 known type 2 diabetes susceptibility loci
found that only two are associated with low birth weight (Vaag et al. 2012). In
addition, the 45 known genes associated with diabetes account for only
approximately 10-15% of the observed cases of diabetes (Ahlqvist et al. 2011).
These studies indicate that the environment might be the main cause of the
correlation between diabetes and low birth weight.
In addition to epidemiological studies of humans, laboratory studies of mice,
rats, and non-human primates have found that maternal dietary stress during
pregnancy can cause fetal growth restriction and result in obesity and impaired
glucose tolerance in adulthood (reviewed in Williams et aL 2015). Collectively, these
studies applied a range of dietary stresses, including low-fat diets, high-fat diets, and
low protein diets, and observed changes in a wide variety of physiological readouts.
33
Nonetheless, almost all studies confirmed that low birth weight caused by maternal
dietary stress correlated with increased metabolic defects in adult offspring.
Dietary stress during pregnancy is thought to restrict the allocation of
resources to offspring during in utero development to result in low-birth weight.
However, as previously mentioned, several studies have found that dietary stress of
mice can modify both oocytes and sperm to affect progeny physiology and
metabolism even if oocytes are fertilized in vitro and transferred to surrogate
mothers that were not exposed to dietary stress. These effects do not rule out
changes in the allocation of resources, as effects of stress on germ cells can modify
placental development in mammals, which in turn can affect the transfer of
resources to the fetus. However, these results demonstrate that dietary stress
before pregnancy can result in similar changes in progeny physiology and
metabolism as dietary stress during pregnancy.
These studies demonstrate that parental dietary stress in mammals,
including humans, can result in changes to offspring physiology and metabolism
that enhance progeny risk for common metabolic disorders such as type 2 diabetes.
In addition, the studies of Chen et al. (2016) and Sharma et al. (2016) indicate that,
at least in some cases, the effects of parental dietary stress on offspring physiology
and metabolism are not mediated by the direct effects of dietary stress on germ
cells, but rather by changes in the deposition of RNA populations into sperm. These
studies suggest that the transfer of information about the environment from somatic
cells to sperm cells, rather than the direct effects of dietary stress on germ cells,
underlies certain metabolic disorders such as type 2 diabetes. However, the
34
mechanisms by which maternal environment modifies oocytes to cause similar
metabolic changes, such as those reported by Huypens et al. (2016), remains
unknown.
Part 7: Concluding Remarks
Collectively, studies of diverse organisms suggest that parental environment
can have profound effects on offspring development, physiology, and metabolism
and that such effects can contribute to human disease. However, the molecular
mechanisms underlying how parental environment affects offspring remain largely
unknown. Perhaps the most compelling mechanistic studies to date have been of
invertebrates, specifically, studies of B. mori and C. elegans.
In the case of B. mori, it has been demonstrated that diapause hormone
signals to oocytes to regulate offspring entry into diapause. These observations
convincingly demonstrate that environmental information, in the form of diapause
hormone, can cross the Weismann barrier and signal to oocytes. However, how this
information is maintained though embryonic development to drive diapause arrest
is unknown. In addition, since diapause hormone is not conserved in vertebrates, it
is unclear if such a mechanism is restricted to B. mori or if similar mechanisms of
hormones signaling to oocytes exist in other organisms. Additional studies of how
the information about parental environment in maintained throughout embryonic
development in B. mori might reveal additional players in this heritable signaling
pathway and help determine if a similar mechanism could exist in other organisms.
35
Similar to B. mori, studies of C. elegans have convincingly demonstrated that
environmental information, in the form of dsRNA, can be transmitted from somatic
cells of the parent to germ cells and that this transfer of information results in the
maintenance of RNAi silencing across a generation. Furthermore, it has been
demonstrated that RNAi inheritance can protect offspring from viral infection
(Gammon et al. 2017). These studies of RNAi inheritance in C. elegans represent the
most detailed mechanistic description of how environmental information can pass
from parent to offspring and perhaps the most convincing evidence that the
Weismann barrier is permeable. However, it appears unlikely that similar
mechanisms exist in mammals since they are not known to transport dsRNA
between tissues. Thus, similar to the studies of B. mori, it is unclear if similar
mechanisms could exist in vertebrates.
Studies of plants and vertebrates have also observed cases where parental
environment modifies offspring development, metabolism, and physiology, but in
almost all of these cases either the molecular mechanism underlying the inheritance
remains unknown, as is the case for all of the examples in plants, or the biological
function remains unknown, as is the case of tRNA fragments in mice.
In conclusion, preliminary mechanistic studies indicate that parental
environment can affect offspring development and physiology by multiple different
mechanisms, including, in some cases, signaling from somatic cells of the parent to
germ cells. Future studies of how parental environment can modify progeny
development, physiology, and metabolism will likely shed significant light on
several outstanding questions in the field. Most notably, how common are parental
36
effects on offspring physiology? What are the mechanisms by which parental
environment can regulate offspring physiology? And what role does parental
exposure to environmental stress play in human disease?
Part 8: Outline of Thesis
In this thesis I describe how the nematode Caenorhabditis elegans responds
to osmotic stress and how maternal exposure to osmotic stress protects progeny
from future osmotic stress. Specifically, in Chapter Two, Tokiko Furuta, Amy
Webster, Rebecca Kaplan, Ryan Baugh, Swathi Arur, H. Robert Horvitz and I describe
how C. elegans arrests its development in response to osmotic stress and how this
developmental arrest is regulated by insulin-like signaling to the intestine. We
demonstrate that maternal exposure to mild osmotic stress can protect progeny
from the effects of strong osmotic stress and that this protection is mediated by
insulin-like signaling to the maternal germline. Finally, we show that reduced
insulin-like signaling to the maternal germline modifies the production of glycerol in
embryos to protect progeny from osmotic stress.
These observations demonstrate that, like diapause hormone in B. mori,
dsRNA in C. elegans, and tRNA fragments in mice, insulin-like peptides can signal
from somatic cells to germ cells in C. elegans. Given the strong conservation of the
insulin signaling pathway as a regulator of growth and metabolism throughout
metazoans, we propose that insulin signaling to the germline also regulates progeny
physiology in mammals and that abnormal insulin signaling to the germline might
37
underlie certain human diseases caused by abnormal insulin signaling, such as
intraruterine growth restriction (IUGR) and Type-2 diabetes.
In chapter Three Vivek Dwivedi, Kirk Burkhart, Rebecca Kaplan, Ryan Baugh,
H. Robert Horvitz and I discuss a mutagenesis screen for mutants that fail to arrest
development in response to osmotic stress, which we anticipated would identify
additional factors that function maternally to regulate offspring response to osmotic
stress. Unexpectedly, this screen revealed that the cytosolic sulfotranferase SSU-1 is
required in embryos to promote developmental arrest in response to osmotic stress.
Furthermore, we demonstrate that SSU-1 functions in a single pair of sensory
neurons, the ASJ neurons, to cell nonautonomously control metabolism, insulin
sensitivity, and development in C. elegans. Given that the decision to arrest
development in response to osmotic stress is bimodal, these results suggest that
maternal insulin-like signaling and signaling via SSU-1 might converge on a single
downstream target to control the decision to arrest development in response to
osmotic stress. Future studies of SSU-1 might elucidate such a downstream target
and shed light on the mechanisms regulating developmental arrest in response to
osmotic stress.
In Chapter Four I discuss possible future directions.
38
Bibliography
Agrawal, A. A., Laforsch, C. & Tollrian, R. Transgenerational induction of defences in
animals and plants. Nature 401, 60-63 (1999).
Ahlqvist, E., Ahluwalia, T. S. & Groop, L. Genetics of type 2 diabetes. Clinical
chemistry 57, 241-254 (2011).
Anderson, L.M., Riffle, L., Wilson, R., Travlos, G.S., Lubomirski, M.S., and Alvord, W.G.
(2006). Preconceptional fasting of fathers alters serum glucose in offspring of mice.
Nutrition 22, 327-331.
Ashe, A., Sapetschnig, A., Weick, E.-M., Mitchell, J., Bagijn, M.P., Cording, A.C., Doebley,
A.-L., Goldstein, L.D., Lehrbach, N.J., and Le Pen, J. (2012). piRNAs can trigger a
multigenerational epigenetic memory in the germline of C. elegans. Cell 150, 88-99.
Buckley, B.A., Burkhart, K.B., Gu, S.G., Spracklin, G., Kershner, A., Fritz, H., Kimble, J.,
Fire, A., and Kennedy, S. (2012). A nuclear Argonaute promotes multigenerational
epigenetic inheritance and germline immortality. Nature 489, 447-451.
Burton, N. 0., Burkhart, K. B. & Kennedy, S. Nuclear RNAi maintains heritable gene
silencing in Caenorhabditis elegans. Proceedings of the National Academy of Sciences
108, 19683-19688 (2011).
Chen, Q., Yan, M., Cao, Z., Li, X., Zhang, Y., Shi, J., Feng, G., Peng, H., Zhang, X., and
Zhang, Y. (2016). Sperm tsRNAs contribute to intergenerational inheritance of an
acquired metabolic disorder. Science 351, 397-400.
Coslovsky, M. & Richner, H. Predation risk affects offspring growth via maternal
effects. Functional Ecology 25, 878-888 (2011).
Dantzer, B., Newman, A.E., Boonstra, R., Palme, R., Boutin, S., Humphries, M.M., and
McAdam, A.G. (2013). Density triggers maternal hormones that increase adaptive
offspring growth in a wild mammal. Science 340, 1215-12 17.
Darwin, C. The variation of animals and plants under domestication. 2, (0. Judd,
1868).
Darwin, E., Darwin, R. W. & Dibner, B. Zoonomia; or, The laws of organic life / by
Erasmus Darwin. v.1, (J. Johnson, 1794).
Dias, B. G. & Ressler, K. J. Parental olfactory experience influences behavior and
neural structure in subsequent generations. Nature neuroscience 17, 89-96 (2014).
39
Feinberg, E. H. & Hunter, C. P. Transport of dsRNA into cells by the transmembrane
protein SID-1. Science 301, 1545-1547 (2003).
Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., and Mello, C.C. (1998).
Potent and specific genetic interference by double-stranded RNA in Caenorhabditis
elegans. Nature 391, 806-811.
Fukuda, S. Function of the pupal brain and suboesophageal ganglion in the
production of non-diapause and diapause eggs in the silkworm. Japan zoology
bulletin 25, 149-155 (1952).
Galton, F. Experiments in pangenesis, by breeding from rabbits of a pure variety,
into whose circulation blood taken from other varieties had previously been largely
transfused. Proceedings of the Royal Society of London 19, 393-410 (1870).
Gammon, D.B., Ishidate, T., Li, L., Gu, W., Silverman, N., and Mello, C.C. (2017). The
Antiviral RNA Interference Response Provides Resistance to Lethal Arbovirus
Infection and Vertical Transmission in Caenorhabditis elegans. Curr Biol 27, 795-
806.
Giesing, E. R., Suski, C. D., Warner, R. E. & Bell, A. M. Female sticklebacks transfer
information via eggs: effects of maternal experience with predators on offspring.
Proceedings of the Royal Society of London B: Biological Sciences rspb20101819
(2010).
Greer, E.L., Maures, T.J., Ucar, D., Hauswirth, A.G., Mancini, E., Lim, J.P., Benayoun,
B.A., Shi, Y., and Brunet, A. (2011). Transgenerational epigenetic inheritance of
longevity in Caenorhabditis elegans. Nature 479, 365-371.
Hales, C.N., Barker, D.J., Clark, P.M., Cox, L.J., Fall, C., Osmond, C., and Winter, P.D.
(1991). Fetal and infant growth and impaired glucose tolerance at age 64. Bmj 303,
1019-1022.
Hales, C. N. & Barker, D. J. The thrifty phenotype hypothesis. British medical bulletin
60, 5-20 (2001).
Hasegawa, K. Studies on the voltinism in the silkworm, Bombyx mori L., with special
reference to the organs concerning determination of voltinism. Tottori University
Dept. of Agriculture 1, 83-124 (1952).
Hibshman, J. D., Hung, A. & Baugh, L. R. Maternal Diet and Insulin-Like Signaling
Control Intergenerational Plasticity of Progeny Size and Starvation Resistance. PLoS
Genetics 12, e1006396 (2016).
Holeski, L. M. Within and between generation phenotypic plasticity in trichome
density of Mimulus guttatus. Journal of evolutionary biology 20, 2092-2100 (2007).
40
Huypens, P., Sass, S., Wu, M., Dyckhoff, D., Tsch6p, M., Theis, F., Marschall, S., de
Angelis, M.H., and Beckers, J. (2016). Epigenetic germline inheritance of diet-
induced obesity and insulin resistance. Nature Genetics 48(5): 497-9
James, S. R., Franklyn, J. A. & Kilby, M. D. Placental transport of thyroid hormone.
Best practice & research Clinical endocrinology & metabolism 21, 253-264 (2007).
Jungheim, E.S., Schoeller, E.L., Marquard, K.L., Louden, E.D., Schaffer, J.E., and Moley,
K.H. (2010). Diet-induced obesity model: abnormal oocytes and persistent growth
abnormalities in the offspring. Endocrinology 151, 4039-4046.
Kalmus, H. CORRELATIONS BETWEEN FLIGHT AND VISION, AND PARTICULARLY
BETWEEN WINGS AND OCELLI, IN JKSECTS. Proceedings of the Royal Entomological
Society of London. Series A, General Entomology 20, 84-96 (1945).
Kamei, Y., Hasegawa, Y., Niimi, T., Yamashita, 0. & Yaginuma, T. Trehalase-2 protein
contributes to trehalase activity enhanced by diapause hormone in developing
ovaries of the silkworm, Bombyx mori. Journal of insect physiology 57, 608-613
(2011).
Kawada, K. (1987) Polymorphism and morph determination. pp.299-314 in Minks,
A.K. & Harrewijn, P. (Eds) Aphids, their biology, natural enemies and control. Vol. A,
Amsterdam, Elsevier.
Kathiria, P., Sidler, C., Golubov, A., Kalischuk, M., Kawchuk, L.M., and Kovalchuk, I.
(2010). Tobacco mosaic virus infection results in an increase in recombination
frequency and resistance to viral, bacterial, and fungal pathogens in the progeny of
infected tobacco plants. Plant Physiology 153, 1859-1870.
Kishimoto, S., Uno, M., Okabe, E., Nono, M. & Nishida, E. Environmental stresses
induce transgenerationally inheritable survival advantages via germline-to-soma
communication in Caenorhabditis elegans. Nature Communications 8, 14031 (2017).
Kitazawa, T. & Takami, T. Staging of embryogenesis in Bombyx mori. II. On
hibernating and nonhibernating hybrids. Tech. Bull. Sericult. Exp. Sta 81, 1-4 (1963).
Kring, J. B. Structure of the eyes of the pea aphid, Acyrthosiphon pisum. Annals of the
Entomological Society ofAmerica 70, 855-860 (1977).
Kutschera, U. & Niklas, K. J. The modern theory of biological evolution: an expanded
synthesis. Naturwissenschaften 91, 255-276 (2004).
41
Kyle, U. G. & Pichard, C. The Dutch Famine of 1944-1945: a pathophysiological
model of long-term consequences of wasting disease. Current Opinion in Clinical
Nutrition & Metabolic Care 9, 388-394 (2006).
Lakshmy, R. Metabolic syndrome: role of maternal undernutrition and fetal
programming. Reviews in Endocrine and Metabolic Disorders 14, 2 29-240 (2013).
Lamarck, J. B. P. A. de M. de. Philosophie zoologique: Ou exposition; des considerations
relative a l'histoire naturelle des animaux. 1, (Cambridge University Press, 1809).
Lumey, L.H., Stein, A.D., Kahn, H.S., van der Pal-de, K.M., Blauw, G.J., Zybert, P.A., and
Susser, E.S. Cohort profile: the Dutch Hunger Winter families study. International
Journal of Epidemiology 36, 1196-1204. (2007).
Luna, E., Bruce, T. J., Roberts, M. R., Flors, V. & Ton, J. Next-generation systemic
acquired resistance. Plant physiology 158, 844-853 (2012).
Luzzo, K.M., Wang, Q., Purcell, S.H., Chi, M., Jimenez, P.T., Grindler, N., Schedl, T., and
Moley, K.H. High fat diet induced developmental defects in the mouse: oocyte
meiotic aneuploidy and fetal growth retardation/brain defects. PloS One 7, e49217.
(2012).
Martins, R., Lithgow, G. J. & Link, W. Long live FOXO: unraveling the role of FOXO
proteins in aging and longevity. Aging Cell 15, 196-207 (2016).
Miyakawa, H., Imai, M., Sugimoto, N., Ishikawa, Y., Ishikawa, A., Ishigaki, H., Okada, Y.,
Miyazaki, S., Koshikawa, S., and Cornette, R. Gene up-regulation in response to
predator kairomones in the water flea, Daphnia pulex. BMC Developmental Biology
10, 45. (2010).
Miyazaki, M. Forms and morphs of aphids. Aphids: their biology, natural enemies, and
control/edited by AK Minks and P. Harrewijn (1987).
Mommer, B. C. & Bell, A. M. Maternal experience with predation risk influences
genome-wide embryonic gene expression in threespined sticklebacks (Gasterosteus
aculeatus). PloS one 9, e98564 (2014).
Ng, S.-F., Lin, R.C., Laybutt, D.R., Barres, R., Owens, J.A., and Morris, M.J. Chronic high-
fat diet in fathers programs [bgr]-cell dysfunction in female rat offspring. Nature
467, 963-966. (2010).
Rasmann, S., De Vos, M., Casteel, C.L., Tian, D., Halitschke, R., Sun, J.Y., Agrawal, A.A.,
Felton, G.W., and Jander, G. Herbivory in the previous generation primes plants for
enhanced insect resistance. Plant Physiology 158, 854-863. (2012).
42
Roberts, D. A. Acquired resistance to tobacco mosaic virus transmitted to the
progeny of hypersensitive tobacco. Virology 124, 161-163 (1983).
Sasson, 1. E., Vitins, A. P., Mainigi, M. A., Moley, K. H. & Simmons, R. A. Pre-gestational
vs gestational exposure to maternal obesity differentially programs the offspring in
mice. Diabetologia 58, 615-624 (2015).
Sato, A., Sokabe, T., Kashio, M., Yasukochi, Y., Tominaga, M., and Shiomi, K.
Embryonic thermosensitive TRPA1 determines transgenerational diapause
phenotype of the silkworm, Bombyx mor. Proceedings of the National Academy of
Sciences 111, E1249-E1255. (2014).
Scott, R. J. & Spielman, M. Genomic imprinting in plants and mammals: how life
history constrains convergence. Cytogenetic andgenome research 113, 53-67
(2006).
Segers, F. H. & Taborsky, B. Juvenile exposure to predator cues induces a larger egg
size in fish. Proceedings of the Royal Society of London B: Biological Sciences
rspb20111290 (2011).
Seiter, M. & Schausberger, P. Maternal intraguild predation risk affects offspring
anti-predator behavior and learning in mites. Scientific reports 5, (2015).
Sharma, U., Conine, C.C., Shea, J.M., Boskovic, A., Derr, A.G., Bing, X.Y., Belleannee, C.,
Kucukural, A., Serra, R.W., and Sun, F. Biogenesis and function of tRNA fragments
during sperm maturation and fertilization in mammals. Science 351, 391-396.
(2016).
Slaughter, A., Daniel, X., Flors, V., Luna, E., Hohn, B., and Mauch-Mani, B. Descendants
of primed Arabidopsis plants exhibit resistance to biotic stress. Plant Physiology 158,
835-843. (2012).
Storm, J. J. & Lima, S. L. Mothers forewarn offspring about predators: a
transgenerational maternal effect on behavior. The American Naturalist 175, 382-
390(2010).
terHorst, C. P. & Lau, J. A. Direct and indirect transgenerational effects alter plant-
herbivore interactions. Evolutionary Ecology 26, 1469-1480 (2012).
Tijsterman, M., May, R. C., Simmer, F., Okihara, K. L. & Plasterk, R. H. Genes required
for systemic RNA interference in Caenorhabditis elegans. Current Biology 14, 111-
116 (2004).
Timmons, L. & Fire, A. Specific interference by ingested dsRNA. Nature 395, 854-
854 (1998).
43
Tollrian, R. Predator-induced morphological defenses: costs, life history shifts, and
maternal effects in Daphnia pulex. Ecology 76, 1691-1705 (1995).
Vaag, A. A., Grunnet, L. G., Arora, G. P. & Brons, C. The thrifty phenotype hypothesis
revisited. Diabetologia 55, 2085-2088 (2012).
Vastenhouw, N.L., Brunschwig, K., Okihara, K.L., Mdller, F., Tijsterman, M., and
Plasterk, R.H. Gene expression: long-term gene silencing by RNAi. Nature 442, 882-
882. (2006).
Walsh, M. R., Whittington, D. & Funkhouser, C. Thermal transgenerational plasticity
in natural populations of Daphnia. Integrative and comparative biology icu078
(2014).
Watanabe, K. Studies on the voltinism of the silkworm, Bombyx mori. Bull Sericult
Exp Sta (Tokyo) 6, 411-455 (1924).
Weinberg, J., Sliwowska, J. H., Lan, N. & Hellemans, K. G. C. Prenatal alcohol exposure:
Foetal programming, the hypothalamic-pituitary-adrenal axis and sex differences in
outcome.Journal of neuroendocrinology 20, 470-488 (2008).
Weismann, A. The Duration of Life, 1881. Essays Upon Heredity and Kindred
Biological Problems 1-66 (1889).
Weismann, A. The germ-plasm: a theory of heredity. (Scribner's, 1893).
Wibowo, A., Becker, C., Marconi, G., Durr, J., Price, J., Hagmann, J., Papareddy, R.,
Putra, H., Kageyama, J., and Becker, J. Hyperosmotic stress memory in Arabidopsis is
mediated by distinct epigenetically labile sites in the genome and is restricted in the
male germline by DNA glycosylase activity. Elife 5, e13546. (2016).
Williams, T. C. & Drake, A. J. What a general paediatrician needs to know about early
life programming. Arch Dis Child 100, 1058-1063 (2015).
Winston, W. M., Molodowitch, C. & Hunter, C. P. Systemic RNAi in C. elegans requires
the putative transmembrane protein SID-1. Science 295, 2456-2459 (2002).
Winston, W. M., Sutherlin, M., Wright, A. J., Feinberg, E. H. & Hunter, C. P.
Caenorhabditis elegans SID-2 is required for environmental RNA interference.
Proceedings of the National Academy ofSciences 104, 10565-10570 (2007).
Wyman, A., Pinto, A. B., Sheridan, R. & Moley, K. H. One-cell zygote transfer from
diabetic to nondiabetic mouse results in congenital malformations and growth
retardation in offspring. Endocrinology 149, 466-469 (2008).
44
Xu, W., Sato, Y., Ikeda, M. & Yamashita, 0. Stage-dependent and temperature-
controlled expression of the gene encoding the precursor protein of diapause
hormone and pheromone biosynthesis activating neuropeptide in the silkworm,
Bombyx mori. Journal of Biological Chemistry 270, 3804-3808 (1995).
Yamashita, 0. Diapause hormone of the silkworm, Bombyx mori: structure, gene
expression and function. Journal of Insect Physiology 42, 669-679 (1996).
Zirkle, C. The Inheritance of Acquired Characters and the Provisional Hypothesis of
Pangenesis. The American Naturalist 69, 417-445 (1935).
45
Chapter 2
Insulin-like signalling to the maternal germline controls progeny response to
osmotic stress
Nicholas 0. Burton, Tokiko Furuta, Amy K. Webster, Rebecca E. W. Kaplan, L. Ryan
Baugh, Swathi Arur, and H. Robert Horvitz
This chapter was originally published as: Burton, N. 0., Furuta T., Webster A.K., Kaplan R.E.W., Baugh
L.R., Arur S., Horvitz H.R. Insulin-like signalling to the maternal germline controls progeny response to
osmotic stress. Nat Cell Biol 19, 252-257 (2017).
I performed all experiments except Fig. 3c and Fig. S4a, and Fig. S5 (Tokiko Furuta and Swathi Arur) and
Fig. 4c and Fig. S2 (Amy Webster, Rebecca Kaplan, and Ryan Baugh).
46
Summary
In 1893 August Weismann proposed that information about the environment could
not pass from somatic cells to germ cells1, a hypothesis now known as the
Weismann barrier. However, recent studies have indicated that parental exposure
to environmental stress can modify progeny physiology2 7 and that parental stress
can contribute to progeny disorders 8.The mechanisms regulating these phenomena
are poorly understood. We report that the nematode C. elegans can protect itself
from osmotic stress by entering a state of arrested development and can protect its
progeny from osmotic stress by increasing the expression of the glycerol
biosynthetic enzyme GPDH-2 in progeny. Both of these protective mechanisms are
regulated by insulin-like signalling: insulin-like signalling to the intestine regulates
developmental arrest, while insulin-like signalling to the maternal germline
regulates glycerol metabolism in progeny. Thus, there is a heritable link between
insulin-like signalling to the maternal germline and progeny metabolism and gene
expression. We speculate that analogous modulation of insulin-like signalling to the
germline is responsible for effects of the maternal environment on human diseases
that involve insulin signalling, such as obesity and type-2 diabetes 8.
47
Introduction
Maternal exposure to a wide variety of environmental stresses alters progeny
growth, development, and physiology of diverse organisms 2-7 and is thought to be a
contributing factor to several human pathologies, including obesity and diabetes 8.
The mechanisms by which the maternal environment can modify progeny biology
are poorly understood. Parental exposure of the nematode C. elegans to mild
osmotic stress can protect progeny from the effects of strong osmotic stress9. This
finding and similar observations of other organisms 2 suggest that besides the
potentially deleterious effects of maternal environmental stress on progeny,
maternal exposure to environmental stress might epigenetically precondition
progeny and protect them from similar environmental insults in the future. How
maternal exposure to environmental stress can protect progeny from future
environmental stress remains largely unknown.
Results
To determine how parental exposure to mild osmotic stress (300 mM NaCl) protects
progeny from the effects of strong osmotic stress (500 mM NaCl) we first examined
the effects of 500 mM NaCl on C. elegans. Embryos placed at 500 mM NaCl
completed embryonic development and hatched but arrested development
immediately after hatching (Fig. la and Supplementary Fig. la). These arrested
animals were unable to move, feed, or respond to touch (Supplementary Fig. 1a).
48
However, when arrested animals were returned to normal osmotic conditions (50
mM NaCi), they regained mobility and resumed development (Fig. 1b). We made
similar observations when NaBr, KCl, or sucrose was used to cause osmotic stress
(Supplementary Fig. lb-d). We conclude that young C elegans larvae can enter a
state of immobile arrested development in response to osmotic stress.
C. elegans also arrests development in response to other environmental stresses,
such as starvation'0-11. These arrests are caused by the loss of insulin-like signalling
via the insulin receptor DAF-2 and the consequent activation of the FOXO
transcription factor DAF-1610-1 2.We tested whether the activation of DAF-16 is
required for developmental arrest in response to osmotic stress. Animals lacking
DAF-16 were less likely than the wild type to arrest development in response to
osmotic stress (Fig. la), whereas animals with increased DAF-16 activation, such as
daf-2 mutants 12, were more likely to arrest development in response to osmotic
stress (Fig. la). Genes in other stress response pathways, such as those that regulate
the responses of C. elegans to oxidative stress' 3 or infection' 4 , were not required for
developmental arrest in response to osmotic stress (Supplementary Fig. le-1f). In
addition, DAF-2 activity in the intestine was required for development at 300 mM
NaCl (Fig. 1c), and exposure to 500 mM NaCl caused the translocation of DAF-16
from the cytoplasm to the nucleus (Fig. 1d). Like the developmental arrest caused
by starvation, the arrest caused by osmotic stress was regulated by dense-core
vesicle release' 0 ; furthermore, we found that it is dense-core vesicle release from
sensory neurons that regulates developmental arrest in response to osmotic stress
49
(Supplementary Fig. 1g), and daf-1 6 mutants, which are resistant to developmental
arrest, showed an increased susceptibility to osmotic stress (Fig. 1b). These results
indicate that like starvation, osmotic stress causes developmental arrest by
inhibiting insulin-like signalling and that arrested development correlates with
enhanced survival.
In contrast to starvation induced developmental arrest, we found that
developmental arrest in response to osmotic stress is regulated by a different
insulin-like peptide, INS-3 (Supplementary Fig. lh-li and Supplementary Table 1),
than arrest in response to starvation (INS-4 and DAF-2815) and that animals that
arrest development in response to osmotic stress are immobile and unable to
respond to touch, unlike arrest in response to starvation in which animals remain
mobile10 . In addition, we found that a majority (78%) of genes the expression of
which reproducibly changed in response to osmotic stress were not affected by
starvation (Supplementary Fig. 2). These results suggest that these two arrest
phenotypes are controlled by partially overlapping but distinct pathways.
Frazier and Roth (2009) found that parental exposure of C. elegans to mild osmotic
stress protects progeny from the effects of strong osmotic stress and that this
protection required DAF-2 activation 9.These authors described an apparently
anomalous result: daf-2; daf-16 double mutants were significantly better at
adaptation to osmotic stress than both wild-type animals and either daf-2 or daf-16
single mutants 9. We hypothesized that this anomaly might be at least in part caused
50
by differing effects of maternal insulin-like signalling and progeny insulin-like
signalling with respect to progeny response to osmotic stress. Specifically, we
suspected that DAF-2 activation in embryos was required for adaption to osmotic
stress, consistent with both the observations of Frazier and Roth and our findings
(Fig. la). We also suspected that it was an inhibition rather than an activation of
parental insulin-like signalling that resulted in the protection of progeny from
osmotic stress, just as the inhibition of larval insulin-like signalling induced by
osmotic stress protects larvae from osmotic stress by causing developmental arrest.
To test this hypothesis, we crossed wild-type animals with daf-2 and daf-2; daf-1 6
double mutant animals in normal osmotic conditions (50 mM NaCl) and assayed the
response of their progeny to 500 mM NaCl. Approximately 60 percent of the
progeny from the cross of wild-type males with daf-2 mutant hermaphrodites
hatched and developed at 500 mM NaCl (Fig. 2a). By contrast, the reciprocal cross of
daf-2 mutant males with wild-type hermaphrodites and the cross of wild-type males
with daf-2; daf-16 double mutant hermaphrodites did not produce any progeny that
hatched and developed at 500 mM NaCl (Fig. 2a). These results demonstrate that
like parental exposure to mild osmotic stress, reduced maternal insulin-like
signalling can protect progeny from the effects of strong osmotic stress.
Importantly, these observations also indicate that there is a previously undescribed
link between maternal insulin-like signalling and progeny physiology.
Parental exposure to osmotic stress has been hypothesized to protect progeny from
the effects of osmotic stress by increasing the deposition of glycerol from mothers
51
into embryos, since embryos from parents exposed to 300 mM NaCl have more
glycerol than embryos from animals grown at normal osmotic conditions (50 mM
NaCl) 9 and glycerol is known to be protective against various environmental
stresses 16"1 7. We confirmed that exposure of parents to 300 mM NaCl resulted in
progeny that are resistant to the effects of 500 mM NaCl (Fig 2b-c and
Supplementary Table 3), and we discovered that the glycerol biosynthetic enzyme
GPDH-2 is required for parental exposure to 300 mM NaCl to protect progeny from
500 mM NaCl (Fig. 2b) but does not affect the response to osmotic stress of animals
with parents grown at 50 mM NaCl (Supplementary Fig. 3a). These observations are
consistent with the hypothesis that an increased level of glycerol is required for
adaptation to osmotic stress. However, it remained unclear how these observations
relate to our finding that reduced maternal insulin-like signalling can protect
progeny from strong osmotic stress (Fig. 2a), since previous studies indicated that
embryos from daf-2 mutant hermaphrodites contain the same amount of glycerol as
embryos from wild-type animals 9 and hence that daf-2 mutant mothers do not
deposit more glycerol into embryos. We hypothesized that reduced maternal
insulin-like signalling protects progeny from the effects of osmotic stress not by
increasing deposition of glycerol from mothers into embryos but rather by
increasing glycerol production in embryos via GPDH-2. To test this hypothesis, we
crossed gpdh-2 mutant males with daf-2; gpdh-2 double mutant hermaphrodites.
GPDH-2 was required for reduced maternal insulin-like signalling to protect
progeny from developmental arrest (Fig. 2d). To test if GPDH-2 functions
maternally, we crossed wild-type males with daf-2; gpdh-2 double mutant
52
hermaphrodites. GPDH-2 was not required in mothers to protect progeny from
developmental arrest in response to osmotic stress (Fig. 2d). We conclude that the
inhibition of maternal insulin-like signalling does not result in the increased
deposition of glycerol into embryos but rather results in increased glycerol
production in embryos. Importantly, these results reveal that there is a heritable
link between maternal insulin-like signalling and progeny metabolism.
We hypothesized that the heritable effects of maternal insulin-like signalling on
progeny might be mediated by insulin-like signalling to the germline. To test this
hypothesis, we expressed rescuing copies of the wild-type daf-2 or daf-1 6 genes
specifically in the germline. Germline-specific expression of either daf-2 or daf-1 6
was sufficient to rescue the effects of deficient maternal insulin-like signalling on
progeny response to osmotic stress (Fig. 2a). In addition, overexpression of daf-2 in
the germline blocked the protective effects of parental exposure to 300 mM NaCl on
progeny response to 500 mM NaCl (Supplementary Fig. 3b). These data suggest that
maternal exposure to 300 mM NaCl inhibits insulin-like signalling to the germline
and that this loss of insulin-like signalling to the germline protects progeny from the
effects of osmotic stress.
Insulin-like signalling to the C elegans germline both inhibits DAF-161 8 and activates
the RAS-ERK pathway' 9, which includes the Raf protein LIN-45, the Mek protein
MEK-2 and the Erk protein MPK-120. We found that partial loss-of-function mutants
in lin-45, mek-2, or mpk-1 (null mutants are lethal) did not arrest development at
53
500 mM NaCl (Fig. 3a and Supplementary Fig. 3c). In addition, (i) treatment of wild-
type animals with the MEK inhibitor U0126 2 or RNAi knockdown of mek-2
prevented developmental arrest in response to 500 mM NaCl (Supplementary Fig.
3d-3f), (ii) RAS-ERK signalling functioned maternally to regulate progeny response
to osmotic stress (Fig. 3b), (iii) GPDH-2 was required for reduced RAS-ERK
signalling to protect animals from developmental arrest (Fig. 3a) and (iv) maternal
exposure to osmotic stress inhibited MPK-1 activation in the germline (Fig. 3c and
Supplementary Fig. 5), possibly by inhibiting the release of insulin-like peptides
from sensory neurons (Supplementary Fig. 4a). We conclude that reduced insulin-
like signalling from the soma (likely from sensory neurons) to the maternal
germline protects progeny from the effects of osmotic stress by both activating DAF-
16 and inactivating MPK-1.
To further examine how reduced maternal insulin-like signalling via the RAS-ERK
pathway modifies progeny physiology we performed RNA-seq of wild-type and lin-
45 mutant embryos. We identified a total of 616 genes upregulated more than 2-fold
and 1,310 genes downregulated more than 2-fold in lin-45 mutant embryos (NCBI
GSE91073). Among the 616 upregulated genes was gpdh-2, and we confirmed that
gpdh-2 mRNA expression is upregulated approximately 3-fold in lin-45 mutant
embryos using qRT-PCR (Fig. 3d). To test if increased GPDH-2 expression results in
increased glycerol production, we compared the glycerol-to-glucose ratios in wild-
type and lin-45 mutant embryos by mass spectrometry. We observed an 82%
increase in the glycerol-to-glucose ratio in lin-45 mutant embryos compared to that
54
in wild-type embryos (Fig. 3e). These data are consistent with our hypothesis that
reduced insulin-like signalling to the maternal germline protects progeny from the
effects of osmotic stress by increasing the expression in embryos of the rate-limiting
glycerol biosynthetic enzyme GPD H-2.
We note that previous studies found that adaptation to osmotic stress resulted in up
to a 1,000% increase in glycerol in C. elegans9, significantly higher than the increase
we observed in lin-45 mutants. Our genetic data demonstrate that GPDH-2 is
required for reduced RAS-ERK signalling in the maternal germline to protect
progeny from developmental arrest (Fig. 3a). However, given the modest increase in
glycerol levels in lin-45 mutants it remains possible, and perhaps likely, that one or
more of the additional 615 genes upregulated in lin-45 mutants also contribute to
the resistance of these animals to developmental arrest in response to osmotic
stress.
We tested whether parental exposure to other environmental stresses, such as
bacterial infection or starvation, could similarly modify progeny response to
environmental stress via RAS-ERK signalling. Exposure of C. elegans to the
opportunistic pathogen P. aeruginosa PA14 slowed early larval development (Fig.
4a), parental exposure to PA14 enhanced this slowing of larval development (Fig.
4a), and this heritable slowing of larval development in response to bacterial
infection required RAS-ERK signalling (Fig. 4a). However, maternal exposure to
PA14 did not protect progeny from developmental arrest at 500 mM NaCl
55
(Supplementary Fig. 4b) but rather resulted in progeny that were more sensitive to
arrest in response to osmotic stress (Fig. 4b). In addition, we found that RAS-ERK
signalling was required for Li arrest in response to starvation (Fig. 4c). Collectively,
these results suggest that maternal exposure to environmental stress modifies
progeny physiology via RAS-ERK signalling, but that the effects of these stresses on
progeny are different for different environmental stresses.
In conclusion, we propose a model in which the inhibition of insulin-like signalling
to both the intestine and the germline can enhance C. elegans survival during
osmotic stress but in which the effects of inhibition of insulin-like signalling to these
two tissues are distinct. Specifically, the loss of insulin-like signalling to the intestine
enhances resistance to osmotic stress by promoting developmental arrest, whereas
the loss of insulin-like signalling to the maternal germline enhances progeny
resistance to osmotic stress by increasing glycerol synthesis in embryos (Fig. 4d). In
this model, information about the maternal environment is inherited via germ cells
to enhance progeny resistance to future environmental stress.
Discussion
The salt concentrations at which C. elegans arrests development in response to osmotic
stress are approximately those of seawater, 480 mM Na' and 559 mM C- 22 . We speculate
that both the state of immobile arrested development and the ability of parents to confer
progeny resistance to osmotic stress evolved to enhance organismal survival in response
56
to osmotic stress caused by seawater. The insulin signalling pathway is broadly
conserved among metazoa, and we postulate that insulin signalling to the germline plays
a role in several human developmental and metabolic abnormalities known to result from
abnormal insulin signalling, such as intrauterine growth restriction (IUGR), obesity, and
type-2 diabetes, all of which have been linked to maternal environmental stress8 , 23 .
Consistent with this hypothesis, Huypens et al. (2016) recently reported that feeding
parental mice a high-fat diet causes epigenetic changes to oocytes that results in progeny
that are more susceptible to both obesity and diabetes7. These observations of a maternal
high-fat diet modifying progeny physiology via oocytes in mice are similar to our
observations of maternal exposure to osmotic stress modifying progeny physiology via
oocytes in C. elegans. We propose that modified insulin-like signalling to the mouse
germline might be the mechanism underlying these epigenetic changes in mouse oocytes.
57
Methods:
Strains. All C. elegans strains were cultured as described previously 24 and
maintained at 20'C unless noted otherwise. The Bristol strain N2 was the wild-type
strain. Mutations used are:
LGI: daf-16(mu86), mek-2(ku114), ins-18(okl672), ins-18(ok2478), ins-26(tm1983),
ins-28(ok2722), ins-29(tm1922), ins-30(ok2343), ins-33(tm2988), ins-
36(tm6125), rrf-1(pk1417); gpdh-1 (ok1558)
LGII: ins-2(tm4467), ins-3(ok2488), ins-3(tm3608), ins-4(ok3534), ins-S(tm2560), ins-
6(tm2416), ins-11(tm1053), ins-12(tm2918), ins-13(tm4856), ins-14(tm4886),
ins-15(ok3444), ins-19(tmS155), ins-20(tm5634), ins-31(ok3543), ins-
32(tm6109), ins-37(tm6268), age-i(hx546), knuSi379 [Pm ex-5::daf-
16cDNA::GFP::nos-2 3' UTR,unc-1 19(+)]
LGIII: daf-2(e1370), mpk-1(n5639), ins-17(tm790), ins-21(tm5180), ins-22(tm4639);
gpdh-2(okl 733)
LGIV: lin-45(n2018), lin-45(n2506), ins-i(nj32), ins-7(tm1907), ins-8(tm4144), ins-
34(tm3095); unc-31 (fti); zIs356 (daf-1 6::GFP); pmk-1 (km25); skn-1 (zu67)
LGV: daf-28(tm2308), ins-10(tm3498), ins-27(ok2474), ins-35(ok3297), him-S(e1490)
LGX: ins-9(tm3618), pdk-i(sa709) nIs349[Pceh-28::4xNLS::mCherry; lin-15(+)];
unknown linkage: vizIs23 [pie-1p::GFP::daf-2(WT)::pie-1 3'UTR + unc-i 19(+)],
vizIs22 [pie-1 p::GFP::daf-2(WT)::pie-1 3'UTR + unc-i 19(+)], nIs343 [Pegl-
1::4xNLS::GFP; lini5(+)]; otIs39[unc-47p::GFP];
Extrachromosomal arrays: naEx187[pGC467 (ins-3(+)), pRF4 )rol-6(su1006)],
hpEx2906 [Prgef-1::daf-2; Pmyo-2::m Cherry], hpEx3369 [Pges-1::daf-2; Pmyo-
58
2::mCherry], hpEx2905 [Pmyo-3::daf-2; Pmyo-2::m Cherry] fx325 [Posm-6::unc-
31a; Pmyo-3::mCherry; unc-119(+)]
Assay for developmental arrest:
Approximately 200 developing eggs from mothers grown at 50 mM NaCl (unless
otherwise noted) were collected and placed on standard NGM plates containing
varying concentrations of NaCl at 25'C for 48 hrs. After 48 hrs, animals that
remained immobile and were not feeding were scored as arrested. Mobile animals
that were feeding were scored as developing. %L2+ is defined by the percent of
animals mobile and feeding and that have developed past the Li larval stage.
%Failing to arrest is defined by the percent of animals mobile and feeding (unlike
animals normally arrested in response to osmotic stress) but includes Li stage
larvae.
Assay for survival after arrest:
Approximately 100 developing eggs from mothers grown at 50 mM NaCl were
collected and placed on standard NGM plates containing varying concentrations of
NaCl at 20*C for 24 hrs. After 24 hrs arrested animals were picked onto plates
containing 50 mM NaCl and allowed to recover for 24 hrs. After 24 hrs the fraction
that regained mobility and resumed development were scored as surviving and the
fraction that failed to resume development and did not respond to touch were
assumed to be dead.
59
DAF-1 6::GFP localization:
Confocal microscopy was performed using a Zeiss LSM 800 instrument. The
resulting images were prepared using ImageJ software (National Institutes of
Health). Image acquisition settings were calibrated to minimize the number of
saturated pixels and were kept constant throughout the experiment.
mRNA expression analysis by RNAseq and qRT-PCR:
L4-stage wild-type and lin-45 animals were placed on standard NGM plates
containing either 50 mM or 300 mM NaCl for 24 hrs. Developing eggs from these
animals were collected and placed at either 50 mM or 500 mM NaCl for 6 hrs. After 6
hrs embryos were collected in M9, and RNA was extracted using TissueRuptor and
the RNeasy Mini kit (QIAGEN). For RNAseq, RNA integrity and concentration were
checked on a Fragment Analyzer (Advanced Analytical). The mRNA was purified by
polyA-tail enrichment, fragmented, and reverse transcribed into cDNA (Illumina
TruSeq). cDNA samples were then end-repaired and adaptor-ligated using the SPRI-
works Fragment Library System I (Beckman Coulter Genomics) and indexed during
amplification. Libraries were quantified using the Fragment Analyzer (Advanced
Analytical) and qPCR before being loaded for paired-end sequencing using the
Illumina NextSeq. For qRT-PCR reverse transcription was performed using
SuperScript III (Invitrogen), and quantitative PCR was performed using Applied
Biosystems Real-Time PCR Instruments. All results are normalized to the mRNA
levels of histone gene his-24.
60
Primers
gpdh-2
forward: tttgatccaaccgtccgtat reverse: cgaattgatgtggaacaacg
his-24
forward: atgatcaaggaggccatcaa reverse: tgagcattgatctggatgaca
Assay for wild-type adaptation to osmotic stress:
Wild-type embryos were placed on standard NGM plates containing 300 mM NaCl at
20*C for 60 hrs and allowed to develop to adulthood. Embryos from these adult
animals were extracted using a razor blade into M9 solution and pipetted onto
plates containing 500 mM NaCl. These animals were allowed to develop for 48 hrs at
20'C. % failing to arrest is defined by the fraction of animals that were mobile and
feeding.
Cross progeny analysis:
Approximately 20 L4 hermaphrodites were crossed with wild-type or daf-2(e1370)
males for 24 hrs at 25'C. After 24 hrs, embryos were dissected from hermaphrodites
using a razor blade and then placed on plates containing 500 mM NaCl at 20*C for
24 hrs. %L2+, the percentage of animals that developed past the Li-larval stage. All
males contained nIs343 (Pegl-1::4xN LS::mCherry); nIs349 (Pceh-
28::4xNLS::mCherry) for the identification of cross progeny.
Photographs:
61
Photographs of animals and embryos on standard NGM plates containing either 300
mM or 500 mM NaCl were obtained using an AxioCam MRm camera (Zeiss).
dpMPK-1 imaging:
Wild-type animals at the L4 stage were placed on either 50 mM or 300 mM NaCl
NGM plates seeded with OPSO at 20'C for 24 hrs, and germlines were extruded.
Dissections were performed as described previously. 25 -2 6 Briefly, dissections were
performed within 5 min of adding levamisole to achieve optimal diphosphorylated
MPK-1 (dpMPK-1) staining. The dissected germlines were then fixed in 3%
paraformaldehyde for 10 min, followed by a post-fix in 100% methanol at -200C. The
fixed germ lines were then processed for immunoflourescence staining as
described 27-29. anti-MAP Kinase was used at a dilution of 1:200 (Clone MAPK-YT,
Sigma, St. Louis, MO). Secondary antibodies were donkey anti-mouse Alexa Fluor
594 and used at a dilution of 1:400. Each gonad was photographed as a montage,
with each image taken as a 0.15 pt section and captured with overlapping cell
boundaries at 63x magnification. Images were taken using a Zeiss Axio Imager
upright microscope with AxioVs40 V4.8.2.0 microimaging software and an Axio
MRm camera (Zeiss). Montages were then assembled using Adobe Photoshop CS5.1,
and white levels were uniformly adjusted using Affinity Illustrator to reduce
background.
RNAseq prep and data analysis:
62
For supplemental tables 1 and 3 RNA integrity and concentration were checked on a
Fragment Analyzer (Advanced Analytical). The mRNA was purified by polyA-tail
enrichment, fragmented, and reverse transcribed into cDNA (Illumina TruSeq).
cDNA samples were then end-repaired and adaptor-ligated using the SPRI-works
Fragment Library System I (Beckman Coulter Genomics) and indexed during
amplification. Libraries were quantified using the Fragment Analyzer (Advanced
Analytical) and qPCR before being loaded for paired-end sequencing using the
Illumina NextSeq. For QC purposes, BEDTools 30 (version: 2.25.0) was used to count
the reads falling into genes, coding regions, intronic regions, 5' or 3' UTRs, flanking
3-kb genic regions and intergenic regions. Other basic statistics, including mapping
rate, ratio of sense vs. anti-sense reads and rRNA percentages were also collected
for each sample. The reads were first cleaned up by removing reads aligned to
rRNAs (BWA 0.6.131 and BEDTools 2.25.0 were used). RSEM 3 2 (version 1.2.15) was
used to estimate gene levels based on ce10 ensembl annotations downloaded from
UCSC genome table browser 33. The gene expression count table was then imported
into DESeq33 (version 1.10.1) for differential gene expression test. See README tab
in supplemental table 2 for detailed descriptions of RNA-seq and analysis for
supplemental table 2.
Pathogen exposure and development assay:
Wild-type embryos were placed onto either NGM plates seeded with E. coli OPSO or
slow-killing assay plates seeded with P. aeruginosa PA14 and allowed to grow at
63
25C for 72 hrs. After 72 hrs embryos were collected from adults and placed on new
plates seeded with either OP50 or PA14 and placed at 25C.
M-cell division in response to starvation:
M-cell division analysis was performed as described previously' 0 .
Statistics and Reproducibility
ANOVA analysis with post hoc p-value calculations were used for Fig. 2, Fig. 3a, Fig.
3b, Fig. 4a, Supplemental Fig. 5 and Supplemental Fig. 8. Unpaired two-tail students
t-test was used for Fig. 1, Fig. 3d, Fig. 3e, Fig. 4b, Fig. 4c, Supplemental Fig. 1,
Supplemental Fig. 3, Supplemental Fig. 6 and Supplemental Fig. 9 * = p < 0.05, ** = p
< 0.01, *** = p < 0.001, **** p < 0.0001. No statistical method was used to
predetermine sample size. The experiments were not randomized. The investigators
were not blinded to allocation during experiments and outcome assessment.
Metabolite preparation for quantification
Approximately 100 ul of concentrated embryos were collected by egg prep and
placed on normal NGM agar plates for 3 hrs to recover. After 3 hrs embryos were
collected in M9, pelleted, and frozen. Frozen embryos were resuspended in 400 ul
PBS and homogenized by douncing. Homogenized embryos were centrifuged at
3000 RPM for 2 min to remove undounced tissue and 200 ul of supernatant was
mixed with 800 ul of methanol and dried to extract polar metabolites. Dried samples
were stored at -80* C.
64
Metabolite profiling
Liquid chromatography and mass spectrometry were performed as described
previously33.
Assay for L2, L3 and L4 stage developmental arrest:
Approximately 50 L2, L3 or L4-stage animals were placed onto standard NGM Petri
plates containing 500 mM NaCl for 24 hrs. After 24 hrs animals that were immobile
and not developing or responding to touch were moved to plates containing 50 mM
NaCl. Percent developing, percent of animals that resumed development and
mobility after returning to normal growth conditions.
MEK inhibitor exposure
The MEK inhibitor U0126 (U120 Sigma-Aldrich) was resuspended in DMSO and
added to standard NGM Petri plates at a final concentration of 100 uM. These plates
were then seeded with Escherichia coli OPSO. Wild-type embryos were placed on
plates containing either DMSO alone or DMSO and the MEK inhibitor U0126.
Animals were allowed to grow for 72 hrs at 25'C. After 72 hrs, embryos from adult
animals were collected and placed onto Petri plates containing 500 mM NaCl at 25*C
for 24 hrs and the fraction of animals mobile and developing were counted.
mek-2 RNAi exoosure
65
Wild-type embryos fed Escherichia coli HT115 containing either the empty vector
L4440 or a mek-2 RNAi vector (Ahringer library - Source Biosciences) were grown
at 20'C for 72 hrs. After 72 hrs, embryos were collected and placed onto Petri plates
containing 500 mM NaCl for 24 hrs at 25*C and the fraction of animals mobile and
developing were counted.
Assay for daf-2 adaptation to osmotic stress:
L4 animals were placed on standard NGM plates containing 300 mM NaCl at 25'C
overnight. Embryos from resulting adult animals were extracted with a razor blade
into M9 solution and pipetted onto plates containing 500 mM NaCl. These animals
were allowed to develop for 48 hrs at 20*C. %L2+, the percentage of animals that
developed past the Li-larval stage.
Data availability
RNA-seq data that support the findings of this study have been deposited in the
Gene Expression Omnibus (GEO) under accession code GSE91073 and GSE91039.
Data supporting the findings of Fig. 1b, 1c, 1d, 2a, 2b, 2d, 3a, 3b, 3d, 3e, 4a, 4b, 4c and
Suppl. Fig. 1b, 1c, 1d, le, 1f, 1g, 3b, 3d, 3e, 3f, 4b, 4c, and 4d are provided in
Supplementary Table 6. All other relevant data are available from the authors on
reasonable request and/or are included with the manuscript.
66
Acknowledgements
We thank E. J. Hubbard, K. Ashrafi, S. Mitani and the Caenorhabditis Genetic Center,
which is funded by the NIH National Center for Research Resources (NCRR), for
providing strains. N. An for strain management; and K. Burkhart, S. Luo, A. Doi, N.
Paquin, and A. Corrionero for helpful discussions. HRH and NOB were supported by
NIH grant GM024663 and NSF grant 1122374. TF and SA were supported by NIH
grant GM98200 and ACS grant RSGO14-044-DDC. LRB, AKW and REWK were
supported by NIH grant GM117408.
67
Bibliography
1. Weismann A. The Germ-plasm: A Theory of Heredity. Charles Scribner's Sons
(1893)
2. Dantzer, B. et al. Density triggers maternal hormones that increase adaptive
offspring growth in a wild mammal. Science 340, 1215-1217 (2013).
3. Radford, E. J. et al. In utero undernourishment perturbs the adult sperm
methylome and intergenerational metabolism. Science 345, (2014).
4. Dias, B. G. & Ressler, K. J. Parental olfactory experience influences behavior and
neural structure in subsequent generations. Nature Neuroscience 17, 89-96 (2014).
5. Ost, A. et al. Paternal diet defines offspring chromatin state and intergenerational
obesity. Cell 159, 1352-1364 (2014).
6. Carone, B. R. et al. Paternally induced transgenerational environmental
reprogramming of metabolic gene expression in mammals. Cell 143, 1084-1096
(2010).
7. Huypens, P. et al. Epigenetic germline inheritance of diet-induced obesity and
insulin resistance. Nat Genet 48, 497-499 (2016).
8. Gallo L., Tran M., Master J., Moritz K., Wlodek M. Maternal adaptations and
inheritance in the transgenerational programming of adult disease. CellAnd Tissue
Research September 2012;349(3):863-880
9. Frazier III, H. N. & Roth, M. B. Adaptive sugar provisioning controls survival of C.
elegans embryos in adverse environments. Current Biology 19, 859-863 (2009).
10. Baugh, L. R. & Sternberg, P. W. DAF-16/FOXO Regulates Transcription of cki-
1/Cip/Kip and Repression of lin-4 during C. elegans Li Arrest. Current Biology 16,
780-785 (2006).
11. Hu, P. J. Dauer. WormBook 1-19 (2007). doi:10.1895/wormbook.1.144.1
12. Murphy, C. and Hu, P.J. Insulin/Insulin-like growth factor signaling in C elegans.
WormBook (2013) doi:10.1895/wormbook.1.164.1
13. An, J. H. & Blackwell, T. K. SKN-1 links C. elegans mesendodermal specification to
a conserved oxidative stress response. Genes Dev 17, 1882-1893 (2003).
68
14. Kim, D. H. et al. A conserved p38 MAP kinase pathway in Caenorhabditis elegans
innate immunity. Science 297, 623-626 (2002).
15. Chen, Y. & Baugh, L. R. ins-4 and daf-28 function redundantly to regulate C.
elegans Li arrest. Developmental Biology 394, 314-326 (2014).
16. Lamitina, S. T., Morrison, R., Moeckel, G. W. & Strange, K. Adaptation of the
nematode Caenorhabditis elegans to extreme osmotic stress. American Journal of
Physiology - Cell Physiology 286, C785-C791 (2004).
17. Lamitina, T., Huang, C. G. & Strange, K. Genome-wide RNAi screening identifies
protein damage as a regulator of osmoprotective gene expression. Proceedings of the
National Academy ofSciences 103, 12173-12178 (2006).
18. Michaelson, D., Korta, D. Z., Capua, Y. & Hubbard, E. J. A. Insulin signaling
promotes germline proliferation in C. elegans. Development (Cambridge, England)
137, 671-680 (2010).
19. Lopez III, A. L. et al. DAF-2 and ERK couple nutrient availability to meiotic
progression during Caenorhabditis elegans oogenesis. Developmental Cell 27, 227-
240(2013).
20. Sundaram, M. V. RTK/Ras/MAPK signaling (February 11, 2006), WormBook, ed.
The C. elegans Research Community, WormBook, doi/10.1895/wormbook.1.80.1
21. Favata, M. F. et al. Identification of a Novel Inhibitor of Mitogen-activated Protein
Kinase Kinase. Journal of Biological Chemistry 273, 18623-18632 (1998).
22. Pilson, E. Q. M. An introduction to the chemistry of the sea. (second edition). Pg.
67. Cambridge University Press 2013
23. Prada, J. A. & Tsang, R. C. Biological mechanisms of environmentally induced
causes of IUGR. EurJ Clin Nutr 52 Suppl 1, S21-7-28 (1998).
24. Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).
25. Arur, S. et al. MPK-1 ERK controls membrane organization in C. elegans
oogenesis via a sex-determination module. Developmental Cell 20, 677-688 (2011).
26. Arur, S. et al. Multiple ERK substrates execute single biological processes in
Caenorhabditis elegans germ-line development. Proceedings of the National Academy
ofSciences 106, 4776-4781 (2009).
27. Lee, M.-H. et al. Multiple functions and dynamic activation of MPK-1 extracellular
signal-regulated kinase signalling in Caenorhabditis elegans germline Development.
Genetics 177, 2039-2062 (2007).
69
28. Quinlan, A.R., BEDTools: The Swiss-Army Tool for Genome Feature Analysis.
Curr Protoc Bioinformatics, 2014. 47: p. 11 12 1-11 12 34.
29. Li, H. and R. Durbin, Fast and accurate short read alignment with Burrows-
Wheeler transform. Bioinformatics, 2009. 25(14): p. 1754-60.
30. Li, B. and C.N. Dewey, RSEM: accurate transcript quantification from RNA-Seq
data with or without a reference genome. BMC Bioinformatics, 2011. 12: p. 323.
31. Karolchik, D., et al. The UCSC Table Browser data retrieval tool. Nucleic Acids Res,
2004. 32(Database issue): p. D493-6.
32. Anders, S. and W. Huber, Differential expression analysis for sequence count
data. Genome Biol, 2010. 11(10): p. R106.
33. Birsoy, K. et al. An Essential Role of the Mitochondrial Electron Transport Chain
in Cell Proliferation Is to Enable Aspartate Synthesis. Cell 162, 540-551 (2015).
70
Figures and Figure legends
Figure 1. Insulin-like signalling to the intestine regulates developmental arrest in
response to osmotic stress. (a) Percent of wild-type, daf-2(e1370), age-1 (hx546),
pdk-1 (sa 709), daf-1 6(mu86) and daf-2(e1370); daf-1 6(mu86) animals developing
past the Li larval stage after 48 hrs. Error bars, s.d. n = 3 experiments of >100
animals P < 0.01 for all genotypes (two-tailed t-test) (b) Percent of wild-type, daf-
2(e1370), daf-16(mu86) and daf-2(e1370); daf-16(mu86) animals that resume
development after 24 hours of exposure to osmotic stress. Error bars, s.d. n = 3 P <
0.01 for all genotypes (two-tailed t-test) (c) Percent of wild-type and daf-2(e1370)
animals developing past the Li larval stage at 300 mM NaCl after 48 hrs. Neuron-
specific expression was driven by Prgef-1; intestine-specific expression was driven
by Pges-1; muscle-specific expression was driven by Pmyo-3. Error bars, s.d. n = 3
experiments of >100 animals (d) Confocal images of DAF-16::GFP after 6 hrs of
exposure to 50 mM and 500 mM NaCl. Scale bar 10 urm. 3 experimental replicates of
>100 animals. The quantified results are presented as mean s.d. using two-tailed
t-test ****P < 0.0001 were considered significant. n.s., not significant.
71
-a- wild-type
-- daf-2
-a- age-I
+ pdk-I
- daf-16
+def-2; def-16
bo
0S
0 200 400 600
mM NaCl
I
300 mM NaC
Hll&
d
100
0
eo 700 80 900 1000
mM Naed
50 mM NaCI
wild-type daf-2 daf-2 daf-2 dof-2
- - neurons muscle intestine
72
a
-J
100-
60-
-a-wild-type
-e-daf-2
--- daf-16
-a-daf-2; daf-16
C
+I 100
U.
0
Genotype
daf-2 transgene
expression
- - -.- * -
Figure 2. Insulin-like signalling to the maternal germline regulates progeny response
to osmotic stress (a) Percent of wild-type, daf-2(e1370) and daf-2(e1370); daf-
16(mu86) cross progeny failing to arrest development after 48 hrs at 500 mM NaCl.
Males contained (Pegl-1::4xNLS::GFP); him-S(e1490); nIs349 (Pceh-
28::4xNLS::mCherry) for the identification of cross progeny. The pie-1 promoter was
used to drive germline specific expression of DAF-2 and the mex-5 promoter was
used to drive germline specific expression of DAF-16. Error bars, s.d. n = 7, 3, 6, 3, 3,
3, and 3 see Supplementary Table 6. (b) Percent of wild-type,gpdh-1(ok1558), and
gpdh-2(okl 733) animals failing to arrest development at 500 mM NaCl after 48 hrs.
Error bars, s.d. n = 3 experiments of >100 animals (c) Average fold change of 2
replicates of the 25 most upregulated genes in embryos in response to osmotic
stress after 6 hrs. (d) Percent of wild-type, daf-2(e13 70) and gpdh-2(oki 733) cross
progeny failing to arrest development after 48 hrs at 500 mM NaCl. Males contained
otIs39 (Punc-47::GFP); him-S(e1490) for the identification of cross progeny. Error
bars, s.d. n = 3 experiments of >20 animals. The quantified results are presented as
mean s.d. using ANOVA. *P < 0.05, ***P < 0.001, **** P < 0.0001 were considered
significant. n.s., not significant.
73
Parental environment - 50 mM NaCl
Progeny environment - 500 mM NaCl
b
100-
01 ... 80-
t
0- 40-
nos LL 2 i0 -
Maternal genotype WT
Paternal genotype WT
Maternal transgene -
expression DAF-2
Maternal transgene -
expression DAF-I6
daf-2 daf-2 WT daf-2; daf-2 dat-2;daf-10 dat-I16
daf-2 WT daf-2 WT WT WT
- .
- -
germline -
Genotype: mild-type wild-type gpdh-1 gpdh-2 gpdh-1;gpdh-2
Parental Environment 60 mM 300 mM 300 mM 300 mM 300 mM
Progeny Environment 500 mM 500 mM 500 mM 500 mM 500 mM
gernline
d
900
Parental environment - 50 mM NaCl
Progeny environment - 500 mM NaCl
* Wlld-type 50 mM NaCl a 100-
700 - Wld-type 500 mM NaCl ( 1
- Wlld-type, 500 mM MaCI -SO
E 600 Parents preconditioned at 300 mM NaCl 801
500 -60
400 40. n.s.
300 20
200 
1
100 Maternal genotype WT dat-2 daf2 ; daf-2 1ef2
Paternal genotype WT wr gpdh-2 WT gpdh-2
74
a
8
* I
A4
LL0
-V
-F
C
Figure 3. Insulin-like signalling to the maternal germline modifies progeny response
to osmotic stress by regulating the RAS-ERK-like pathway. (a) Percent of wild-type
and lin-45(n2018), mek-2(kul 14), mpk-1 (n5639) and lin-45(n2018); gpdh-2(okl 733)
animals failing to arrest development at 500 mM NaCl after 48 hrs. Error bars, s.d. n
= 3 experiments of >100 animals (b) Percent of wild-type and lin-45(n2018) cross
progeny failing to arrest development at 500 mM NaCl after 48 hrs. Males contained
otIs39 (Punc-47::GFP); him-5(e1490) for the identification of cross progeny. Error
bars, s.d. n = 3 experiments of >20 animals. (c) Representative germlines dissected
from wild-type animals exposed to either 50 mM NaCl or 300 mM NaCl and stained
for DNA (DAPI, white) and diphosphorylated MPK-1 (dpMPK-1) (red) Each
condition was replicated 16 times. Scale bar 50 ptm (d) Relative expression of gpdh-2
mRNA in wild-type and lin-45(n2018) embryos measured by qRT-PCR and
normalized to the expression of the histone his-24. Error bars, s.d. n = 3
experiments from pellets of >1000 embryos (e) Glycerol-to-glucose ratio in wild-
type and lin-45(n2018) mutant embryos. Error bars, s.d. n = 3 experiments from
pellets of >1000 embryos. The quantified results are presented as mean s.d. using
ANOVA (a, b) and two-tailed t-test (d, e). *P < 0.05, **P < 0.01, ***P < 0.001 were
considered significant. n.s., not significant.
75
aParental environment - 50 mM NaCl
Progeny environment - 500 mM NaCl
100-
80-
.60-
C
- 40-
206 
n.s.
* 
. 2 0 
-
6- 11
wild-type Iin-45 mek-2 mpk-1 Iin-45,
gpdh-2
b
Parental environment - 50 mM NaCl
Progeny environment - 500 mM NaCl
(**
e 20-
S
S10
UL
U-O 
~ l
Maternal genotype
Paternal genotype
Progeny genotype
wild-type
wild-type
wild-type
C
Relative Fluorescence
wild-type 50 mM #1
wild-type 50 mM #2
wild-type 300 mM #1
wild-type 300 mM #2
d
gpdh-2 mRNA
expression at
50 mM NaCl
6-
4--
Fold 3E
Change
2-
widtIp
wild-type
e
Embryos -50 mM NaCI
3.50
3.00
2.50
GlycerolGlucose .
Ratio 20
1.50
1.00
0.500 d-y
wild-type
MIn-45
76
lin-45 Iin-45 wild-type
Iin-45 wild-type Iin-45
fin-45 Iin-451+ Iin-45/+
lin-45
n.s.
I
Figure 4. RAS-ERK signalling regulates C. elegans response to bacterial infection and
starvation (a) Percent of animals expressing lin-4::YFP after 24 hrs of exposure to
either E. coli OPSO or P. aeruginosa PA14. Error bars, s.d. n = 3 experiments of >100
animals (b) Percent of animals failing to arrest development after 24 hrs at 350 mM
NaCl. Error bars, s.d. n = 3 experiments of >100 animals (c) Percent of wild-type, Un-
45(n2018) and daf-16(mu86) mutants with a divided M-cell after 7 days without
food in S-basal at 20'C Error bars, s.e.m. n = 4 experiments of >100 animals (d)
Model for how maternal exposure to osmotic stress inhibits DAF-2 activity in the
germline and affects progeny response to osmotic stress. See text for details. Color
code: Red, embryo; green, germline; purple, intestine. The quantified results are
presented as mean s.d. (a, b) and s.e.m (c) using ANOVA (a) and two-tailed t-test
(b,c). *P < 0.05, **P < 0.01, ***P < 0.00 1, ****P < 0.0001 were considered significant. n.s.,
not significant.
77
a 100-
+
N
-J 50-
wild-type wild-type wild-type wild-type fin-45 lin-45
Environment OP50 OP50 PA14 PA14 OP50
Environment OP50 PA14 PA14 OP50 OP50
c
350 mM NaCi
04
100.
801
21
Parental Environment OP50
20%
15%-
0%
PA14
E
0
Starvation
wild-type Iin-45 daf-16
d
Embryo LI stage larva
deczzE
"-.
APITR
78
--1-
Genotype
Parental
Progeny
b
Iin-45
OP50
PA14
lin-45
PA14
PA14
PA14
OP50
Mother
Progeny
phenotype
Developmental
arrest
Continued
development
Figure S1 C. elegans arrests development in response to osmotic stress. (a)
Representative images of wild-type animals after 72 hrs of exposure to 50 mM or
500 mM NaCl. Experiment replicated 12 times Scale bar 200 ptm (b) Percent of wild-
type animals developing past the Li larval stage after 24 hrs of exposure to 50 mM
NaCl and 450 mM KCl . error bars, s.d. n = 3 experiments of >100 animals (c) Percent
of wild-type animals developing past the Li larval stage after 24 hrs of exposure to
50 mM NaCl and 450 mM NaBr. error bars, s.d. n = 3 experiments of >100 animals
(d) Percent of wild-type animals developing past the Li larval stage after 24 hrs of
exposure to 50 mM NaCl and 900 mM sucrose. Sucrose was added at twice the
concentration of salts to compensate for the difference in osmolarity. error bars, s.d.
n = 3 experiments of >100 animals (e) Percent of wild-type, pmk-1(km25), or skn-
1(zu67) animals developing past the Li larval stage after 48 hrs of exposure to 300
mM NaCl. error bars, s.d. n = 3 experiments of >100 animals (f) Percent of wild-type,
pmk-1(km25), or skn-1 (zu67) animals failing to arrest development at 500 mM NaCl
after 48 hrs. error bars, s.d. n = 3 experiments of >100 animals (g) Percent of wild-
type and unc-31 (ft) animals failing to arrest development after 48 hrs. The osm-6
promoter was used to drive the expression of UNC-31 specifically in sensory
neurons. error bars, s.d. n = 3 experiments of >100 animals (h) Percent of wild-type
and ins-3(tm3608) mutant animals failing to arrest development at 400 mM NaCl
after 48 hrs. error bars, s.d. n = 3 experiments of >100 animals (i) Percent of wild-
type, ins-3(ok2488), ins-3(ok2488);daf-2(e1370) and rescue animals failing to arrest
development at 400 mM NaCl 48 hrs post-hatching. error bars, s.d. n = 3
experiments of >100 animals The variation between the wild type in panels (a) and
79
(b) is likely to be a result of variations in evaporation between different batches of
Petri plates. The quantified results are presented as mean s.d. using two-tailed t-
test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 were considered significant. n.s.,
not significant.
80
wild4ype - 72 hrs at wild-type .72 hrs at
60 mM NaCI 800 mM NaCI
ed Sucrose - 24 hrs of exposure
+!UFM90m rmvlo
b KCI - 24 hrs of exposure
300 mM NaC
1001
o-
wild-type pmk-1 sakn-1
C NaBr - 24 hrs of exposure
500 mM NaCI
100-
0k
WIT p.at-I k.-I
unc-31
100- *
0o 1
h 0-
40-
20
0 d
wild-ty
400 mM NaCI
**
s-(3)
a Ins-3(tm3608)
400 mM NaCI
01S
40-
0.
GeOtype wild-type IMw-3(ok248) Ins-3(Ak2488) dAt-2;Ina-3(ok2486)
81
a
g
ln 3
rescue
Figure S2 Comparison of gene expression in response to osmotic stress and
starvation. (a) Venn-diagram of genes that exhibit statistically significant changes in
gene expression in response to osmotic stress and starvation. (b) Scatter plot of
gene expression changes for 1605 genes whose expression is regulated by both
osmotic stress and starvation.
82
a
Low Salt Food vs. High Salt Food
Low Sat Food
vs. Low Salt No
Food
5641 1605 612
p =1.57e-171
b
Fold changes of 1605 genes differentially expressed in both starvation and osmotic arrest
Genes per quadrant 1.
1v- 732
11.21
111.790
IV 62
E
-4 -
S- -*
Significant fold changes from starvation
83
Figure S3 Insulin-like signalling to the germline regulates developmental arrest in
response to osmotic stress. (a) Percent of wild-type and gpdh-1 (ok1558); gpdh-
2(okl 733) animals developing past the Li larval stage after 48 hrs. error bars, s.d. n
= 3 experiments of >100 animals (b) Percent of wild-type and daf-2(e1370) animals
failing to arrest development at 500 mM NaCl. The pie-1 promoter was used to drive
DAF-2 expression specifically in the germline. Germline #1 and germline #2
represent two independently integrated transgenes. error bars, s.d. n = 3
experiments of >100 animals (c) Percent of wild-type and lin-45(n2018) animals
developing past the Li larval stage after 48 hrs. error bars s.d. n = 3 experiments of
>100 animals (d) Percent of progeny from animals exposed to either DMSO or the
MEK inhibitor U0126 dissolved in dimethyl sulfoxide (DMSO) developing past the
Li larval stage after 48 hrs (500 mM NaCl). error bars s.d. n = 3 experiments of >100
animals (e) Percent of progeny from animals exposed to either L4440 empty vector
or mek-2 RNAi that developed past the Li larval stage after 48 hrs (500 mM NaCl).
error bars s.d. n = 3 experiments of >100 animals (f) Percent of rrf-1(pk1417)
progeny, which exhibit reduced RNAi silencing in somatic tissue but retain germline
RNAi silencing1 , from animals exposed to either L4440 empty vector or mek-2 RNAi
that developed past the Li larval stage after 48 hrs (500 mM NaCl). error bars s.d. n
= 3 experiments of >100 animals. The quantified results are presented as mean
s.d. using ANOVA (b) and two-tailed t-test (d,e,f). **P < 0.01, ***P < 0.00 1, were
considered significant. n.s., not significant.
84
b
100
ata
"o" Genotype
100 200 300 400 000 600PmnalE
oM NaCI Tranogeneof DAF-2
C d
0 .iod-W.
a Un-4s
mM NaCI
500 mM NaCI
261*
20-
+ 1j
15t
0
empty vector mek-2 RNAI
500 mM NaCl
15-
n.s. ns.
liMd-type daf-2 wild-type daf-2 def-2 d.f-2
vironment 50 mM 30 mM 300 mM 300 mM 300 mM 300 mM
txpression - - - - germline# 1 germllne 52
500 mM NaCI
10
4 6-
4-
2
0
**
Parental treatment DMSO MEK inhibitorU0126
f
100-
go-
+ 60-
40-
20-
rrf-1 600 mM NaCI
empty vector mek-2 RNAI
85
a
e
Figure S4 Dense-core vesicle release from sensory neurons regulates MPK-1 activity
in the germline. (a) Representative germlines dissected from wild-type or unc-
31(ftl) animals, which exhibit reduced dense-core vesicle release 2, at 50 mM NaCl
and stained for DNA (DAPI, white) and diphosphorylated MPK-1 (dpMPK-1) (red).
SN::UNC-31(+) animals that expressed a rescuing copy of unc-31 specifically in
sensory neurons using the osm-6 promoter. Experiment replicated 4 times. Scale bar
100 prm (b) Percent of animals mobile and feeding after 48 hrs at 500 mM NaCl.
error bars s.d. n = 3 experiments of >100 animals. We note that the observation that
less wild-type animals adapted to 500 mM NaCl when parents were exposed to 300
mM NaCl than in Fig. 2a is likely due to the fact that in this experiment embryos had
to be collected by bleaching to remove any contaminating PA14 and this likely
added an extra stressor to the embryos.
86
a 50 mM NaCI
b
500 mM NaCI
267
U)
I 20-1M1
0
I.
6-
Parental Environment OP50 PA14 300 mM NaCI
87
Figure S5 Quantification of dpMPK-1 in additional germlines of wild-type animals at
50 mM NaCI and 300 mM NaCl. See Figure S3 of Chapter 2.
88
dpMPK-1 at 50 mM NaCI
2m0
1000 77
dismnce 100 200 30400 500 6001
dpMPK-1 at 300 mM NaCI
3000
2500 - -. aG-001
2000 -
15.00004
10000m
050N
distance 100 200 300 400 S00 600
89
Table S1 Most insulin peptides do not regulate development in response to osmotic
stress. Mutant animals that each lacked one of 35 insulin-like peptides were assayed
for the percent of animals that developed after exposure to 300 mM NaCl or 500 mM
NaCl. WT at 300 mM NaCl: fewer than 1% of animals arresting in response to 300
mM NaCl. WT at 500 mM NaCl: 0% of animals mobile and feeding at 500 mM NaCl.
WT, wild-type. The quantified results are presented as mean s.d. using two-tailed
t-test ***P < 0.001 was considered significant. n.s., not significant.
90
Gene Chr. Allele 300 mM 500 mM
arrested arrested
daf-28 V tm2308 and sal91 0 100
ins-1 IV nj32 0 100
ins-2 II tm4467 0 100
ins-3 II ok2488 and see Supp. see Supp. Fig.
tm3608 Fig. 9 9
ins-4 II ok3534 0 100
ins-5 II tm2560 0 100
ins-6 II tm2416 0 100
ins-7 IV tm1907 0 100
ins-8 IV tm4144 0 100
ins-9 X tm3618 0 100
ins-10 V tm3498 0 100
ins-1i II tm1053 0 100
ins-12 II tm2918 0 100
ins-13 II tm4856 0 100
ins-14 II tm4886 0 100
ins-15 I I ok3444 0 100
ins-1 6 III Not tested N/A N/A
ins-1 7 III tm790 0 100
ins-18 I ok1672 and 0 100
ok2478
ins-19 II tm5155 0 100
ins-20 II tm5634 0 100
ins-21 III tm5180 0 100
ins-22 III tm4639 0 100
ins-23 
- Not tested N/A N/A
ins-24 
- Not tested N/A N/A
ins-25 
- Not tested N/A N/A
ins-26 I tm1983 0 100
ins-27 V ok2474 0 100
ins-28 I ok2722 0 100
ins-29 I tm1922 0 100
ins-30 I ok2343 0 100
ins-31 II ok3543 0 100
ins-32 II tm6109 0 100
ins-33 I tm2988 0 100
ins-34 IV tm3095 0 100
ins-35 V ok3297 0 100
ins-36 I tm6125 0 100
ins-37 II tm6268 0 100
91
ins-38 Not tested N/A N/A
ins-39 Not tested N/A N/A
92
Chapter 3
The cytosolic sulfotransferase SSU-1 functions in sensory neurons to control
insulin sensitivity and development of C. elegans
Nicholas 0. Burton, Vivek K. Dwivedi, Kirk B. Burkhart, Rebecca E. W. Kaplan, L.
Ryan Baugh, and H. Robert Horvitz
I performed all experiments except Fig. id (Rebecca Kaplan and Ryan Baugh). Fig. 2b and Fig. 3c
were completed with assistance from Kirk Burkhart. Fig. 1c and Fig. 2c were completed with
assistance from Vivek Dwivedi.
93
Abstract
Many organisms, ranging from worms to humans, have evolved mechanisms to slow
or arrest development during periods of environmental stress, such as starvation,
bacterial infection, and osmotic stress (1-5). Failure to properly slow or arrest
development can lead to long-term metabolic disorders, developmental defects, and
death (6-9). The mechanisms by which multicellular organisms coordinately
regulate development during periods of environmental stress remain unclear. Here
we report that the C elegans cytosolic sulfotransferase SSU-1, which is predicted to
regulate hormone signaling, functions in a single pair of sensory neurons to cell-
nonautonomously control insulin sensitivity and larval development during osmotic
stress. We demonstrate that SSU-1 and insulin-like signaling antagonistically
regulate development during periods of osmotic stress and that the loss of SSU-1
activity can suppress the effects of reduced insulin-like signaling. Finally, we show
that the FOXO transcription factor DAF-16, which is a downstream effector of
insulin-like signaling, enters the nucleus in response to osmotic stress, but is unable
to activate gene expression in response to osmotic stress in the absence of SSU-1.
We propose that cytosolic sulfotransferases similarly regulate insulin sensitivity and
developmental rate during periods of environmental stress in other organisms and
that abnormal activity of cytosolic sulfotransferases might contribute to human
developmental rate pathologies, such as intrauterine growth restriction.
94
Results and Discussion:
To complete development efficiently and successfully, many multicellular organisms
must be able to adapt to variations in their environment. This adaptation is
particularly important during periods of environmental stress, such as starvation,
when resources are limited and organisms must choose between allocating
resources towards growth and development or towards responding to the
environment. To regulate this decision, a variety of mechanisms have evolved to
slow metabolism and development in response to environmental stress. For
example, the human fetus can adapt to maternal undernutrition by altering its
metabolism, increasing catabolism, and slowing its growth rate (2). Failure to
respond appropriately to environmental stress during fetal development can result
in developmental defects, long-term metabolic disorders, and fetal death (6-9). The
mechanisms by which multicellular organisms coordinately regulate growth,
metabolism, and development during periods of environmental stress remain
unclear.
In response to osmotic stress C. elegans arrests its development immediately after
hatching (4). To determine the mechanisms that regulate this developmental arrest,
we screened for mutants that failed to arrest development in response to osmotic
stress (500 mM NaCl). We identified a nonsense allele of the cytosolic
sulfotransferase ssu-1 (W2840pal) that caused approximately 40% of animals
failing to arrest development in response to 500 mM NaCl (Fig. la and 1b). Similarly,
95
we found that six additional mutations in ssu-1 (fc73, tm1117,gk266317, gk747222,
gk876992, and gk31 9712) caused animals to fail to arrest development in response
to osmotic stress (Fig. 1b). We conclude that SSU-1 regulates developmental arrest
in response to osmotic stress. SSU-1 is the only predicted cytosolic sulfotransferase
encoded in the C. elegans genome (10), and is most similar to the SULT1 family of
human cytosolic sulfotransferases, which regulate a variety of signaling molecules
including dopamine (SULT1D1) and estrogen (SULTiEl) (11). We propose that
SSU-1 similarly regulates neuromodular and hormone signaling in C elegans.
SSU-1 is expressed in a single pair of sensory neurons, the ASJ neurons (10). To
determine if SSU-1 functions specifically in the ASJ sensory neurons to regulate
developmental arrest in response to osmotic stress we expressed a rescuing
transgene of ssu-1 under the control of the ASJ-specific promoter trx-1 (12).
Expression of SSU-1 in the ASJ sensory neurons was sufficient to restore
developmental arrest in 100% of animals in response to osmotic stress (Fig. 1c),
indicating that SSU-1 functions in the ASJ sensory neurons to regulate
developmental arrest in response to osmotic stress.
C. elegans also arrests development in response to other environmental stresses,
such as starvation, and this arrest requires the FOXO transcription factor DAF-16. Li
larval stage arrest in response to starvation results in the inhibition of cell divisions,
including that of the mesodermal M cell (5). To determine if SSU-1 is also required
for developmental arrest in response to starvation or if SSU-1 specifically regulates
96
developmental arrest in response to osmotic stress we starved wild-type, daf-1 6 and
ssu-1 mutant animals for one week and assayed the percentage of animals with a
divided M cell. We found that 100% of M cells in wild-type animals and ssu-1
mutants arrested cell division (Fig. 1d). By contrast, 8% of M cells in daf-16 mutants
failed to arrest cell division in response to starvation (Fig. 1d), consistent with
previous observations (5). These results suggest that SSU-1 specifically regulates
developmental arrest in response to osmotic stress and does not regulate
developmental arrest in response to starvation.
In humans, cytosolic sulfotransferases can sulfonate hormones such as estrogen and
dehydroepiandrosterone (DHEA) (11). These hormones in turn regulate gene
expression by activating nuclear hormone receptors, such as the estrogen receptor
(11). We hypothesized that SSU-1 might similarly regulate the sulfonation of a
hormone that in turn regulates developmental arrest in response to osmotic stress
by controlling the transcriptional response to osmotic stress. To test if SSU-1
regulates gene expression in response to osmotic stress we exposed wild-type and
ssu-1 mutant embryos to either 50 mM or 500 mM NaCl for 3 hours and quantified
mRNA expression by RNA-seq. We found that the expression of 434 genes was
upregulated greater than 2-fold in response to osmotic stress and that the
expression of 106 of these genes was dependent on SSU-1, including 20 of the 25
genes that exhibited a greater than 10-fold increase in expression in response to
osmotic stress (Fig. 2a). For example, the superoxide dismutase sod-5 and the
osmotic stress resistance protein lea-1 (15) exhibited a greater than 25-fold
97
increase in expression in response to osmotic stress and their expression in
response to osmotic stress required SSU-1 (Fig. 2a). We confirmed that SSU-1 was
required for the expression of SOD-5 in response to osmotic stress using a GFP
reporter (Fig. 2b and 2c). Specifically, GFP was expressed broadly throughout the
animal in response to osmotic stress and this broad increase in expression required
SSU-1 function in the ASJ sensory neurons (Fig. 2c). These observations are
consistent with SSU-1 regulating the activity of a hormone that functions cell non-
autonomously to control the transcriptional response to osmotic stress in C. elegans.
Osmotic stress causes developmental arrest of C. elegans by activating the FOXO
transcription factor DAF-16 (4). To determine if DAF-16 and SSU-1 regulate the
expression of the same or different downstream target genes we exposed wild-type
and daf-16 mutant embryos to 50 mM or 500 mM NaCl and quantified mRNA
expression by RNA-seq. We found that 161 genes exhibited a greater than 2-fold
increase in expression in response to osmotic stress and were dependent on DAF-16
for their induction (Fig. 2d and 2e). Of the 161 genes regulated by DAF-16 in
response to osmotic stress, 64 were also regulated by SSU-1, including 20 of the 25
the genes previously identified to exhibit the largest increase in expression in
response to osmotic stress (Fig. 2d and 2e). These results indicate that a significant
fraction of the changes in gene expression in response to osmotic stress requires
both SSU-1 and DAF-16. We note that for several genes, such as sod-5, we observed
an approximately 4-fold larger increase in expression in wild-type animals in
response to osmotic stress (Fig. 2c) than our previous RNA-seq experiment (Fig. 2a).
98
These results are likely due to the fact that these genes are normally not expressed
or are lowly expressed under normal osmotic conditions and thus the differences in
fold-change between wild-type animals in these two experiments likely reflects
small differences in background expression in wild-type animals under normal
osmotic conditions rather than significantly increased expression of these genes at
500 mM NaCl.
Mutations in the insulin-like receptor daf-2 result in animals that are more likely to
arrest development in response to osmotic stress than wild-type animals (4). We
hypothesized that SSU-1 might modify insulin-like signaling via DAF-2 to control the
activation of DAF-16 in response to osmotic stress. To test this hypothesis, we
constructed double mutant animals harboring mutations in both daf-2 and ssu-1,
and exposed wild-type, daf-2, ssu-1, and daf-2; ssu-1 double mutant embryos to mild
osmotic stress (300 mM NaCl) and strong osmotic stress (500 mM NaCl). Consistent
with our previous findings, we found that nearly 100% of daf-2 mutant embryos
arrested development at 300 mM NaCl (Fig. 3a). However, none of the wild-type,
ssu-1, or daf-2; ssu-1 double mutant animals arrested development at 300 mM NaCl
(Fig. 3a), while approximately 30% of ssu-1 and daf-2; ssu-1 double mutant animals
failed to arrest development at 500 mM NaCl (Fig. 3b). These results suggest that
the loss of SSU-1 activity can suppress the effects of reduced insulin-like signaling
during osmotic stress.
99
Our data indicate that SSU-1 functions in the ASJ sensory neurons to regulate arrest
(Fig. 1c), while we previously reported that insulin-like signaling functions in the
intestine to regulate arrest (4). Insulin-like signaling to the intestine inhibits the
activation of the FOXO transcription factor DAF-16 by sequestering it in the
cytoplasm (4, 16), and the loss of insulin-like signaling to the intestine causes
developmental arrest because DAF-16 is no longer sequestered in the cytoplasm
and can translocate into the nucleus to activate a transcriptional response to stress
(4, 16). Our observation that both DAF-16 and SSU-1 are required for the expression
of a common set of osmotic stress response genes suggests that SSU-1 either
functions in the ASJ sensory neurons to regulate the translocation of DAF-16 into the
intestine in parallel to DAF-2 or that SSU-1 and DAF-16 function in parallel to
regulate the transcriptional activation of a common set of genes in response to
osmotic stress.
To test if SSU-1 regulates DAF-16 translocation into the nucleus we examined wild-
type and ssu-1 mutant animals that expressed a GFP-tagged copy of DAF-16 and
assayed DAF-16 translocation to the nucleus in response to osmotic stress. We
found that SSU-1 was not required for DAF-16 translocation into the nucleus in
response to osmotic stress (Fig. 3c and Fig. Si). We conclude that SSU-1 likely
functions in parallel to insulin-like signaling and DAF-16 translocation into the
nucleus in C. elegans to regulate development in response to osmotic stress. In
addition, our results suggest that animals are no longer sensitive to the effects of
100
reduced insulin-like signaling in response to osmotic stress in the absence of SSU-1,
indicating that SSU-1 is a modifier of insulin sensitivity in C elegans.
In humans, FOXO transcription factors interact with several nuclear hormone
receptors to regulate transcription (19) and cytosolic sulfotransferases function to
modify both the activity and metabolism of the ligands that bind nuclear hormone
receptors (11). Based on these observations, we propose that SSU-1 regulates the
metabolism and/or activity of a yet-to-be identified ligand for an also as yet-to-be
identified nuclear hormone receptor that functions with DAF-16 to activate the
transcriptional response to osmotic stress.
We previously found that increased levels of glycerol, an osmolyte that protects
animals from the effects of osmotic stress, can protect C. elegans from entering
developmental arrest in response to osmotic stress (4). We hypothesized that the
loss of SSU-1 or DAF-16 might similarly regulate the production of either glycerol or
some other osmolyte, such as betaine, to protect animals from entering
developmental arrest in response to osmotic stress. To test this hypothesis we
profiled 137 polar metabolites and 1,069 lipid metabolites in wild-type, ssu-1, and
daf-16 mutant embryos by mass spectrometry (data not shown). We found that the
levels of 8 polar metabolites and 18 lipid metabolites were changed (p <0.01) in ssu-
1 mutant embryos and that the levels of 11 polar metabolites and 58 lipid
metabolites were changed (p <0.01) in daf-16 mutant embryos (data not shown). Of
the metabolites regulated by SSU-1 and DAF-16, 4 polar molecules and 4 lipid
101
molecules were changed in both ssu-1 and daf-16 mutant embryos (Fig. 4a and 4b).
However, we did not observe an increase in the levels of any known osmolyte,
including glycerol and betaine. These results suggest that the loss of SSU-1 and DAF-
16 do not protect animals from developmental arrest in response to osmotic stress
by increasing the production of an osmolyte.
In short, our findings reveal a neuronal signaling pathway that regulates C. elegans
development and physiology in response to environmental stress. We propose that
a yet-to-be identified nuclear hormone receptor functions downstream of SSU-1 and
antagonistically to insulin-like signaling to regulate the transcriptional response to
osmotic stress. We speculate that the regulation of growth and development and the
modification of insulin sensitivity by nuclear hormone receptors are conserved
processes and that studies of how hormone signaling modifies development and
insulin sensitivity will provide important insights into disorders that involve insulin
signaling, including intrauterine growth restriction, obesity, and type-2 diabetes.
102
Materials and Methods:
Strains. All C. elegans strains were cultured as described previously (20) and
maintained at 20'C unless noted otherwise. The Bristol strain N2 was the wild-type
strain. Mutations used are:
LGI: daf-16(m26); daf-16(mu86)
LGIII: daf-2(e1370)
LGIV: ssu-1 (fc73); ssu-1(n5883); ssu-1(tm1117); ssu-1(gk266317); ssu-1(gk747222);
ssu-1 (gk876992); ssu-1 (gk319712); zIs356 [Pdaf-1 6::daf-1 6a/b-gfp; rol-6]
unknown linkage: wuIsS7 [sod-5p::GFP, rol-6(sul 006)]
Extrachromosomal arrays: nEx2685 [Ptrx-1::SSU-1::mCherry::unc-54 3'UTR; Punc-
122::GFP]
M cell division in response to starvation:
M cell division analysis was performed as described previously (5).
DAF-1 6::GFP localization:
Embryos were placed onto plates containing 500 mM NaCl for 5 hours. Confocal
microscopy was performed using a Zeiss LSM 800 instrument. The resulting images
were prepared using ImageJ software (National Institutes of Health). Image
acquisition settings were calibrated to minimize the number of saturated pixels and
were kept constant throughout the experiment.
103
Assay for developmental arrest:
Approximately 200 developing eggs from mothers grown at 50 mM NaCl (unless
otherwise noted) were collected and placed on standard NGM plates containing
varying concentrations of NaCl for 48 hrs. After 48 hrs, animals that remained
immobile and were not feeding were scored as arrested. Mobile animals that were
feeding were scored as developing. Percent failing to arrest is defined by the percent
of animals mobile and feeding (unlike animals normally arrested in response to
osmotic stress) and includes Li stage larvae.
Mutagenesis screen for animals that-fail to arrest development
Approximately 20,000 L4 stage wild-type animals were incubated with 20 uL of
ethyl methanesulfonate (Sigma) in 4 mL of M9 for 4 hrs at 20C. F3 generation
embryos were placed onto plates containing 500 mM NaCl and screened for mutants
that hatched and were mobile.
Metabolite preparation and quantification
Approximately 100 ul of concentrated embryos were collected by egg prep and
placed on normal NGM agar plates for 3 hrs to recover. After 3 hrs embryos were
collected in M9, pelleted, and frozen. Frozen embryos were resuspended in 600 ul
methanol and lysed using the BeadBug microtube homogenizer (Sigma) and 0.5 mm
Zirconium beads. After lysis, 300 ul of water and 400 ul of cholorform (containing
internal standards) was added to each sample and samples were then vortexed for 1
min at 4'C and centrifuged for 10 min at 15000 g at 4C. After centrifugation the
104
polar and lipid layers were separated and dried using a speedvac. Liquid
chromatography and mass spectrometry were performed as described previously
(21).
Preparation of RNA
Wild-type, ssu-1(fc 73), and daf-16(m26) embryos were placed on standard NGM
plates for 3 hrs. After 3 hrs embryos were collected in M9, lysed using BeadBug
microtube homogenizer (Sigma) and 0.5 mm Zirconium beads (Sigma), and RNA
was extracted using the RNeasy Mini kit (QIAGEN).
RNA-seq preparation
For RNAseq, RNA integrity and concentration were checked on a Fragment Analyzer
(Advanced Analytical) and libraries were prepared using the Illumina NeoPrep
RNAseq kit. Libraries were quantified using the Fragment Analyzer (Advanced
Analytical) and qPCR before being loaded for paired-end sequencing using the
Illumina NextSeq5OO.
RNA-seq analysis
75-nucleotide paired-end sequencing reads were mapped against the C. elegans
genome assembly ce1O using RSEM v. 1.2.15 , with bowtie v. 1.0.1 for read
alignment (flags --paired-end -p 6 --bowtie-chunkmbs 1024 --forward-prob 0, for
strand-specific libraries). Expected read counts per gene were retrieved and used to
perform differential gene expression with DESeq2 in the R v. 3.2.3 statistical
105
environment. Briefly, after rounding counts, library size factors and dispersions
were estimated, followed by fitting of negative binomial distribution in a GLM
framework and calculation of differential expression significance using a Wald test
using the DESeq() function. log2-transformed fold-change tables were generated in
the presence and absence of independent filtering and Cook's filtering of outliers.
Data obtained with independent and Cook's filtering turned off were used for
downstream analysis of the N2 vs. daf 16 experiment (these filters did not impact
the results for the N2 vs. ssu-1 comparisons).
sod-5p::GFP pictures
Embryos were placed onto plates containing either 50 mM or 500 mM NaCl for 24
hours. Confocal microscopy was performed using a Zeiss LSM 800 instrument. The
resulting images were prepared using ImageJ software (National Institutes of
Health). Image acquisition settings were calibrated to minimize the number of
saturated pixels and were kept constant throughout the experiment.
Cloning of Ptrx-1::SSU-1::m Cherry
ssu-1 cDNA fragment with synthetic introns replacing endogenous introns was
obtained from Integrated DNA Technologies using their custom gene synthesis
service. The ssu-1 fragment was amplified using primers 5'-
ctttgagcaattgatcatgaccccgaagaccccaaag-3' and 5'-gagaccatgaccggtgcctca
gcgaaagtggacaaatc-3'. The pJDM169 vector containing 1.1 kb of the trx-1 promoter
upstream of mCherry and unc-54 3'UTR was obtained from Joshua Meisel (Meisel
106
and Kim, 2014). pJDM169 was linearized by inverse PCR using the primers 5'-
gcaccggtcatggtctcaaagggtgaagaa-3' and 5'-gatcaattgctcaaagtcac-3'. The amplified
ssu-1 fragment was cloned into linearized pJDM169 by Infusion HD cloning kit as
per manufacturers instructions.
107
Acknowledgments: We thank David Gems and the Caenorhabditis Genetic Center,
which is funded by the NIH National Center for Research Resources (NCRR), for
providing strains, N. An for strain management, and A. Doi, and A. Corrionero for
helpful discussions. HRH, VD, KB, and NOB were supported by NIH grant GM024663
and NOB was supported by NSF grant 1122374. LRB and REWK were supported by
NIH grant GM117408.
108
References:
1. Hales, C.N., and Barker, D.J. (2001). The thrifty phenotype hypothesis. Br Med Bull
60, 5-20.
2. Harding JE, Johnston BM. Nutrition and fetal growth. Reprod Fertil Dev 1995; 7:
539-47
3. David J P Barker; The malnourished baby and infant: Relationship with Type 2
diabetes. Br Med Bull 2001; 60 (1): 69-88
4. Burton, N.O., Furuta, T., Webster, A.K., Kaplan, R.E.W., Baugh, L.R., Arur, S., and
Horvitz, H.R. (2017). Insulin-like signalling to the maternal germline controls
progeny response to osmotic stress. Nat Cell Biol 19, 252-257.
5. Baugh, L.R., and Sternberg, P.W. (2006). DAF-16/FOXO regulates transcription of
cki-1/Cip/Kip and repression of lin-4 during C. elegans Li arrest. Curr Biol 16, 780-
785.
6. Barker DJP, Gluckman PD, Godfrey KM, Harding JE, Owens JA, Robinson JS. Fetal
nutrition and cardiovascular disease in adult life. Lancet 1993; 341: 938-41
7. Forsen T, Eriksson JG, Tuomilehto J, Teramo K, Osmond C, Barker DJP. Mother's
weight in pregnancy and coronary heart disease in a cohort of Finnish men: follow
up study. BMJ 1997; 315: 837-40
8. Villar J, Smeriglio V, Martorell R, Brown CH, Klein RE. Heterogeneous growth and
mental development of intrauterine growth-retarded infants during the first 3 years
of life. Pediatrics 1984; 74: 783-91
9. Haas JD, Balcazar H, Caulfield L. Variation in early neonatal mortality for different
types of fetal growth retardation. Am J Phys Anthropol 1987; 73: 467-73
10. Carroll, B.T., Dubyak, G.R., Sedensky, M.M., and Morgan, P.G. (2006). Sulfated
signal from ASJ sensory neurons modulates stomatin-dependent coordination in
Caenorhabditis elegans. J Biol Chem 281, 35989-35996.
11. Coughtrie, M.W.H. (2016). Function and organization of the human cytosolic
sulfotransferase (SULT) family. Chem Biol Interact 259, 2-7.
12. Miranda-Vizuete, A., Fierro Gonzalez, J.C., Gahmon, G., Burghoorn, J., Navas, P.,
and Swoboda, P. (2006). Lifespan decrease in a Caenorhabditis elegans mutant
lacking TRX-1, a thioredoxin expressed in ASJ sensory neurons. FEBS Lett 580, 484-
490.
109
13. Lamitina, S.T., Morrison, R., Moeckel, G.W., and Strange, K. (2004). Adaptation of
the nematode Caenorhabditis elegans to extreme osmotic stress. Am J Physiol Cell
Physiol 286, C785-791.
14. Frazier, H.N. 3rd, and Roth, M.B. (2009). Adaptive sugar provisioning controls
survival of C. elegans embryos in adverse environments. Curr Biol 19, 859-863.
15. Gal, T. Z., Glazer, I. & Koltai, H. An LEA group 3 family member is involved in
survival of C. elegans during exposure to stress. (2004) FEBS Lett 577, 21-26
16. Murphy, C.T., and Hu, P.J. (2013). Insulin/insulin-like growth factor signaling in
C. elegans. WormBook 1-43.
17. Adeva-Andany M, L6pez-Oj6n M, Funcasta-Calder6n R, Ameneiros-Rodriguez E,
Donapetry-Garcia C, Vila-Altesor M, Rodriguez-Seijas J (2014) Comprehensive
review on lactate metabolism in human health. Mitochondrion 17C:76-100.
18. Chen AT-Y, Guo C, Itani OA, et al. Longevity Genes Revealed by Integrative
Analysis of Isoform-Specific daf-1 6/FoxO Mutants of Caenorhabditis elegans.
Genetics. 2015;201(2):613-629. doi:10.1534/genetics.115.177998.
19. van der Vos, K.E., and Coffer, P.J. (2008). FOXO-binding partners: it takes two to
tango. Oncogene 27,2289-2299.
20. Brenner, S. The genetics of Caenorhabditis elegans. (1974). Genetics 77, 71-94
21. Birsoy, K. et al. An Essential Role of the Mitochondrial Electron Transport Chain
in Cell Proliferation Is to Enable Aspartate Synthesis. (2015) Cell 162, 540-551
110
Figure Legends
Fig. 1. The cytosolic sulfotransferase SSU-1 functions in the ASJ sensory neurons to
regulate developmental arrest in response to osmotic stress. (a) Schematic of
mutations in ssu-1 that cause defects in developmental arrest in response to osmotic
stress. (b) Percent of ssu-1 mutants failing to arrest development in response to
osmotic stress. n = 3 experiments of greater than 100 animals. error bars s.d. (c)
Number of wild-type and ssu-1(fc73) animals failing to arrest development in
response to osmotic stress. The trx-1 promoter was used to drive ASJ cell specific
expression of SSU-1. Animals also contained wuIs57 [pPD95.77 sod-5p::GFP, rol-
6(su1006)] to confirm the response to osmotic stress. n = 100 (d) Percent of wild-
type, daf-16(mu86), and ssu-1(fc73) animals with a divided M cell after 1 week of
starvation. n > 200 animals error bars s.e.m.
111
Q108Amber W2840pal
gk319712 gkfc73
SSu-1 R470pal W195Amber gk268317
tmI117
sulfotransferase domain
600 mM NaCl
so-
g o.
40-
Oi,
'b
Number
genotype* failing to
arrest
wild-type 0
ssu-1 42
ssu-1; 0
ASJp::SSU-1
*strains n=100
contained
sod-5p::GFP
Starvation - 1 week
10-
& i 
-
wed-type dat-It 84u-I
112
a
b
Fig. 2. SSU-1 and DAF-16 regulate the transcription of a common set of target genes
in response to osmotic stress. (a) Average fold change gene expression measured by
RNA-seq in wild-type and ssu-ltfc73) mutant animals at 500 mM NaCI when
compared to 50 mM NaCl. The 25 genes indicated represent those 25 genes
exhibiting the highest increase in expression in wild-type animals at 500 mM NaCl
when compared to wild-type animals at 50 mM NaCl. n = 3 error bars s.d. (b)
Confocal and DIC images of sod-5p::GFP expression in wild-type and ssu-1(fc73)
mutants exposed to 500 mM NaCI for 24 hours. Scale bars 100 um (c) Confocal and
DIC images of sod-5p::GFP expression in wild-type and ssu-1(fc73) mutants exposed
to 500 mM NaCl for 24 hours. The trx-1 promoter was used to drive ASJ cell-specific
expression of SSU-1. Scale bars 100 um. (d) Average fold change gene expression
measured by RNA-seq in wild-type and daf-16(m26) mutant animals at 500 mM
NaCl when compared to 50 mM NaCl. The 25 genes indicated represent the 25 genes
exhibiting the highest increase in expression in wild-type animals at 500 mM NaCi
when compared to wild-type animals at 50 mM NaCl. n = 3 error bars s.d. (e) Venn
diagram representing the number of genes exhibiting a greater than 2-fold increase
in mRNA expression in wild-type animals exposed to 500 mM NaCl when compared
to wild-type animals exposed to 50 mM NaCl and dependent on either SSU-1 (blue)
and/or DAF-16 (yellow).
113
m wild-type
a SSu-1
I I I [ I
c
sod-5::GFP DIC
500 mM NaCl
And-An -- 3PP DIC
wild-type 60 mM NaCI
wild-type 600 mM NaCI
ssu-I(-) 50 mM NaCI
asu-1(-) 600 mM NaCI
e
" wild-type
* daf-16
LiIS LL
SSU-1 targets DAF-16 targets
42
u 64 97
114
a
70
40 I301Lul
10c
U-U
b
wild-type
ssu-1(-)
ssu-I-);pASJ.:SS U-I
d
E I
Fig. 3 SSU-1 functions in parallel to insulin-like signaling to regulate developmental
arrest in response to osmotic stress. (a) Percent of wild-type, daf-2(e1370) and ssu-
1(fc73) mutants failing to arrest development in response to 300 mM NaCl. n >100
animals error bars s.d. (b) Percent of wild-type, daf-2(e1370) and ssu-1(fc73)
mutants failing to arrest development in response to 500 mM NaCl. n >100 animals
error bars s.d. (c) Representative confocal images of DAF-16::GFP localization after
5 hours of exposure to 500 mM NaCl in wild-type and ssu-1(fc73) mutants. Scale
bars 10 um.
115
a
300 mM NaCl
100-
U 0 L11
b
500 mM NaCl
8O-
Go-
20
16-i
C
116
Fig. 4 SSU-1 and DAF-16 regulate the production of certain metabolites in C elegans
(a) Relative levels of polar metabolites exhibiting a statistically significant (p < 0.01)
change in levels in ssu-1 (fc 73) and daf-16(m26) mutant embryos. Metabolites were
normalized to the levels of histidine. n = 3 error bars s.d. (c) Relative levels of lipid
metabolites exhibiting a statistically significant (p < 0.01) change in levels in both
ssu-1(fc73) and daf-16(m26) mutant embryos. Metabolites were normalized to the
levels of total lipid. n = 3 error bars s.d. PC - phosphatidylcholine, AcCa -
Acylcarnitine, PE - phosphatidylethanolamine, TG - triglyceride (c) Model for how C.
elegans regulates its metabolism and development in response to osmotic stress.
117
Polar metabolites
C I C&NI0
ASJ
b
M wild-ty"
M ssu-1
11111 daf-16
Lipid metabolites
2.0-
I'M vwild-type
is- M ssu-1
~ 1 0-M daf-16
0 1.0
0.-5- -Iiji i
'N
Sensory neurons
/ INS-3
DAF-2
Developmental arrest
118
a
C
Madde
WOWS
I
Figure Si SSU-1 is not required for DAF-16 translocation into the nucleus in
response to osmotic stress. Confocal images of DAF-16::GFP localization after 5
hours of exposure to 500 mM NaCl in wild-type and ssu-1(fc73) mutants. Scale bars
10 um.
119
daf-16-:gfp
ssu-1-)
daf-16-:gfp
120
5.6
Chapter 4
Discussion and Future Directions
Nicholas Burton
121
Future directions related to maternal environmental stress and offspring physiology
Our results indicate that in the case of osmotic stress in C. elegans environmental
information can cross the Weismann barrier. The mechanism by which insulin-like
signaling to the germline modifies offspring physiology remains unclear. One
possible model is that the loss of insulin-like signaling to the germline results in the
activation of DAF-16 and inactivation of MPK-1 and that these activated and
inactivated versions of DAF-16 and MPK-1 are transmitted to the single-cell embryo
where they modify embryonic development and metabolism. Alternatively, it is
possible that the loss of insulin-like signaling to the germline modifies the uptake of
resources by the oocyte resultin'g in changes in maternal provisioning of embryos
such that there is increased glycerol production in embryos. A third possibility is
that the loss of insulin-like signaling in the germline epigenetically modifies oocytes
to alter embryonic metabolism. Differentiating among these models is
experimentally challenging. Nonetheless, determining the mechanism by which
reduced insulin-like signaling to the maternal germline modifies offspring
metabolism and physiology in C. elegans will be critical for determining the
generality of such maternal effects and whether or not similar maternal effects
might underlie human pathologies that involve insulin-like signaling, such as type 2
diabetes.
One possible approach to differentiate among these mechanisms is to screen for
mutants that function maternally to regulate offspring developmental arrest in
122
response to osmotic stress, similar to lin-45 mutants. If this screen identified genes
specifically involved in resource uptake or epigenetic reprogramming it might
provide evidence for a particular model. Another potential approach is to use
metabolomics to profile the metabolites present in embryos from parents exposed
to mild osmotic stress. Such profiling might identify additional metabolites
deposited into embryos when parents are stressed. I could then test if these
metabolites can be transmitted from mother to offspring and if these metabolites
can protect offspring from osmotic stress by supplementing increased amounts of
such metabolites into the mother's diet. As a third approach, I could profile the
relative amounts of common chromatin modifications, such as histone methylation
and histone acetylation, present at genes regulated by osmotic stress with and
without parental exposure to mild osmotic stress. These experiments might identify
changes in histone modifications at genes regulated by osmotic stress. If so, I could
then identify the enzymes that regulate these histone modifications by candidate
testing of mutants and RNAi silencing to determine whether the loss of these
enzymes affects C. elegans adaptation to osmotic stress.
Future directions related to SSU-1
From a chemical mutagenesis screen for genes required for developmental arrest in
response to osmotic stress I identified SSU-1. I found that SSU-1 functions in parallel
to insulin-like signaling to regulate developmental arrest in response to osmotic
stress. However, it remains unclear if or how SSU-1 function is impacted by changes
123
in maternal environment. Additionally, it remains unclear if insulin-like signaling to
the maternal germline and SSU-1 activity in the embryo function in parallel
pathways to regulate developmental arrest in response to osmotic stress or if these
two pathways converge on a common downstream target to regulate developmental
arrest in response to osmotic stress. Finally, it remains unclear how SSU-1 function
in the ASJ sensory neurons can modify metabolism, insulin sensitivity, and
developmental arrest in response to osmotic stress. Based on the function of
cytosolic sulfotransferases in humans, we propose that SSU-1 likely sulfonates an as
yet unidentified hormone and that this hormone regulates the activity of an also as
yet unidentified nuclear hormone receptor to control transcription, metabolism, and
developmental arrest in response to osmotic stress. It is likely that the identification
of such a hormone and hormone receptor functioning downstream of SSU-1 would
result in significant insight into how SSU-1 controls metabolism, insulin sensitively,
and development and how maternal insulin-like signaling to the germline integrates
with SSU-1 activity to control developmental arrest in response to osmotic stress.
To identify the hormone sulfonated by SSU-1 I could perform untargeted
metabolomics of small molecules looking for a sulfate group. Since SSU-1 is the only
cytosolic sulfotransferase in C. elegans it is likely that any sulfonated small molecule
would need to be sulfonated by SSU-1. However, any such molecule is likely present
in small amounts as SSU-1 is only expressed in a single pair of sensory neurons, thus
such an approach might not be effective.
124
To identify the hormone receptor that functions downstream of SSU-1 I could screen
for suppressors of SSU-1 using the SOD-5::GFP reporter. Such a screen might
identify genes that function downstream of SSU-1 to regulate gene expression in
response to osmotic stress, including a potential nuclear hormone receptor.
Alternatively, I could continue to screen for mutants that fail to arrest development
in response to osmotic stress and fail to activate SOD-5::GFP expression in response
to osmotic stress, similar to ssu-1 mutants. This screen could similarly identify a
potential nuclear hormone receptor that functions downstream of SSU-1. Finally, I
could use RNAi to specifically knockdown all of the predicted nuclear hormone
receptors in C. elegans and test if they affect developmental arrest in response to
osmotic stress or the expression of SOD-5::GFP in response to osmotic stress. If any
of these screens identified the predicted nuclear hormone receptor that functions
downstream of SSU-1, then I might be able to purify the receptor and use mass
spectrometry to identify any potential ligand bound to the nuclear hormone
receptor which might be the target ligand of SSU-1.
Additionalfuture directions
In addition to SSU-1 I identified several other mutations from this screen that did
not affect the expression of sod-5p::GFP in response to osmotic stress. These results
suggest that the predicted transcriptional response to osmotic stress might not be
sufficient to drive developmental arrest. I propose that future characterization of
these mutants will shed further insight into both the mechanisms by which maternal
125
environment can modify offspring physiology and the mechanisms by which
organisms can slow or arrest development in response to environmental stress.
126