Download axolotl paedomorphosis - UKnowledge

Document related concepts
no text concepts found
Transcript
University of Kentucky
UKnowledge
Theses and Dissertations--Biology
Biology
2013
AXOLOTL PAEDOMORPHOSIS: A
COMPARISON OF JUVENILE,
METAMORPHIC, AND PAEDOMORPHIC
AMBYSTOMA MEXICANUM BRAIN GENE
TRANSCRIPTION
Carlena Johnson
[email protected]
Recommended Citation
Johnson, Carlena, "AXOLOTL PAEDOMORPHOSIS: A COMPARISON OF JUVENILE, METAMORPHIC, AND
PAEDOMORPHIC AMBYSTOMA MEXICANUM BRAIN GENE TRANSCRIPTION" (2013). Theses and Dissertations--Biology.
Paper 13.
http://uknowledge.uky.edu/biology_etds/13
This Master's Thesis is brought to you for free and open access by the Biology at UKnowledge. It has been accepted for inclusion in Theses and
Dissertations--Biology by an authorized administrator of UKnowledge. For more information, please contact [email protected].
STUDENT AGREEMENT:
I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been
given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright
permissions. I have obtained and attached hereto needed written permission statements(s) from the
owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic
distribution (if such use is not permitted by the fair use doctrine).
I hereby grant to The University of Kentucky and its agents the non-exclusive license to archive and make
accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the
document mentioned above may be made available immediately for worldwide access unless a
preapproved embargo applies.
I retain all other ownership rights to the copyright of my work. I also retain the right to use in future
works (such as articles or books) all or part of my work. I understand that I am free to register the
copyright to my work.
REVIEW, APPROVAL AND ACCEPTANCE
The document mentioned above has been reviewed and accepted by the student’s advisor, on behalf of
the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we
verify that this is the final, approved version of the student’s dissertation including all changes required
by the advisory committee. The undersigned agree to abide by the statements above.
Carlena Johnson, Student
Dr. S. Randal Voss, Major Professor
Dr. Brian Rymond, Director of Graduate Studies
AXOLOTL PAEDOMORPHOSIS: A COMPARISON OF JUVENILE,
METAMORPHIC, AND PAEDOMORPHIC AMBYSTOMA MEXICANUM BRAIN
GENE TRANSCRIPTION
____________________________________
THESIS
____________________________________
A thesis submitted in partial fulfillment
of the requirements for the degree of Master of Science in the
College of Arts and Sciences
at the University of Kentucky
By
Carlena K. Johnson
Lexington, Kentucky
Director: Dr. S. Randal Voss, Professor of Biology
Lexington, Kentucky
2013
Copyright © Carlena K. Johnson 2013
ABSTRACT OF THESIS
AXOLOTL PAEDOMORPHOSIS: A COMPARISON OF JUVENILE,
METAMORPHIC, AND PAEDOMORPHIC AMBYSTOMA MEXICANUM BRAIN
GENE TRANSCRIPTION
Unlike many amphibians, the paedomorphic axolotl (Ambystoma mexicanum)
rarely undergoes external morphological changes indicative of metamorphosis. However,
internally, some axolotl tissues undergo cryptic metamorphic changes. A previous study
examined interspecific patterns of larval brain gene expression and found that these
species exhibited unique temporal expression patterns that were hypothesized to be
morph specific. This thesis tested this hypothesis by examining differences in brain gene
expression between juvenile (JUV), paedomorphic (PAED), and metamorphic (MET)
axolotls. I identified 828 genes that were expressed differently between JUV, PAED, and
MET. Expression estimates from JUV were compared to estimates from PAED and MET
brains to identify genes that changed significantly during development. Genes that
showed statistically equivalent expression changes across MET and PAED brains provide
a glimpse at aging and maturation in an amphibian. The genes that showed statistically
different expression estimates between metamorphic and paedomorphic brains provide
new functional insights into the maintenance and regulation of paedomorphosis. For
genes that were not commonly regulated due to aging, paedomorphs exhibited greater
transcriptional similarity to juvenile than metamorphs did to juvenile. Overall, gene
expression differences between metamorphic and paedomorphic development exhibit a
mosaic pattern of expression as a function of aging and metamorphosis in axolotls.
KEYWORDS:
expression
Ambystoma;
metamorphosis;
paedomorphosis;
microarray;
gene
Carlena K. Johnson
(July 1, 2013)
AXOLOTL PAEDOMORPHOSIS: A COMPARISON OF JUVENILE,
METAMORPHIC AND PAEDOMORPHIC AMBYSTOMA MEXICANUM BRAIN
GENE TRANSCRIPTION
By
Carlena K. Johnson
DR. S. RANDAL VOSS
DIRECTOR OF THESIS
DR. DAVID WESTNEAT
DIRECTOR OF
GRADUATE STUDIES
(JULY 1, 2013)
ACKNOWLEDGEMENTS
I thank the many people who have helped me with this project. I expressly thank Randal
Voss for all of his time and patience in teaching me about genomics and the writing
process. I am immensely grateful for your help and assistance during this project. I
thank Antony Athippozhy for his assistance and guidance on how to manipulate and
analyze microarray data. I also thank all the members of the Voss lab, especially, John
Walker, Kevin Kump, Sri Putta, and Tyler Frye for their support and encouragement. I
thank my committee members, Jeramiah Smith and Vinnie Cassone for their support in
this project. Thank you to the faculty and staff of the Biology Department. This project
was supported through grants received by Randal Voss from NSF and NIH. I would like
to thank my friends and family for their unwavering support through this project. Thank
you to my mom and fiancé, Tom, for their encouragement and reassurance.
iii TABLE OF CONTENTS
Acknowledgements.…………………………………………………………………… iii
List of Tables.………………………………………………………………………….. vi
List of Figures.………………………………………………………………………… vii
Chapter One: Literature Review.……………………………………………………….. 1
Why paedomorphosis and why salamanders?………………………………..… 1
Paedomorphosis and salamander life history variation…….…………………... 4
Hormonal basis of salamander metamorphosis………………………..……….. 6
Hormonal basis of paedomorphosis……………………………….……..…..… 8
Paedomorphosis in the Mexican Axolotl……………………….…….….…… 11
Transcriptional analysis of salamander metamorphosis
and paedomorphosis.………………………………………………………….. 13
Genetic analysis of paedomorphosis………………………………….….….... 15
The link between met1 and thyroid hormone regulation
of metamorphosis…………………………………………………………...… 17
What genes map to metamorphic timing QTL?………………………….…… 18
Synthesis: evolution of salamander paedomorphosis……………………..…... 19
Chapter Two: Microarray comparison of juvenile, paedomorphic
and metamorphic brain gene transcription……………………………..…….... 28
Introduction ………………………………………………………………....… 28
Methods……………………………………………………………..…………. 29
Animals and tissue sampling…………………………………...…….... 29
RNA isolation and microarray analysis………………………...…….... 29
Global analyses of microarray data……………………………...…….. 30
Identification of differently expressed genes (DEGs)
using microarray……………………………………………………….. 30
Identification of statistically enriched biological processes………....… 30
Bioinformatic validation of morph specific gene expression……...…... 31
iv Results ………………………………………………………………………….. 31
Global analysis of gene expression……………………………………... 31
Identification of differently expressed genes among
juvenile and adult brains………………………………………….…..… 32
Classification of adult gene expression patterns relative to juvenile…… 32
Identification of statistically enriched biological processes………….… 33
Bioinformatic analysis of morph specific
gene expression ……………………………………………………….... 35
Discussion………………………………………………………………………. 36
Aging and transcription in the A. mexicanum brain………………….…. 37
Genes associated with metamorphosis and paedomorphosis…………… 39
Aging, species, and morph specific differences in
hemoglobin expression…………………………………………………. 43
References…………………………………………………………………………...…. 61
v LIST OF TABLES
Table 1.1, General characteristics of salamander life history………………………..… 22
Table 2.1, List of genes that show greater than three standard
deviations from predicted values………….………………………..……………….…. 46
Table 2.2, List of over-represented ontology terms………………………….………… 47
Table 2.3, Pattern of gene expression comparison with DEGs
from Page et al. (2010)……………………………………………………………….… 51
vi LIST OF FIGURES
Figure 1.1, Representative salamander larva, adult metamorph (Ambystoma tigrinum)
and adult paedomorph (A. mexicanum)……………………………………………...… 23
Figure 1.2, Timeline showing timing of fore and hindlimb development
and the metamorphic period of laboratory reared A. t. tigrinum………………………. 24
Figure 1.3, Photo of Ambystoma dumerilii……………………………………………. 25
Figure 1.4, Gene expression profiles……………………...…………………………... 26
Figure 1.5, Model showing relationships among salamander life history
variation, physiology, and habitat stability…………………………………...……….. 27
Figure 2.1, Box plots of GeneChips……………………….………………………....... 54
Figure 2.2, Principle components analysis………………….……………………….… 55
Figure 2.3, Hierarchical clustering analysis………………………………….………... 56
Figure 2.4, Venn diagram of significant DEGs…………………………..……………. 57
Figure 2.5, Patterns of gene expression change for PAED and MET
relative to JUV..……………………………………………………………………….. 58
Figure 2.6, Bivariate plot of PAED and MET expression relative to JUV……….…… 59
Figure 2.7, Bar graph of expression for significant collagen genes…………………… 60
vii CHAPTER ONE: LITERATURE REVIEW
Salamanders are one of three primary groups of amphibians, the other two being
caecilians and anurans. Ancestrally, all three groups trace their origins to ancestors that
present biphasic life cycles with an aquatic larval phase and a more terrestrial adult
phase (Duellman & Trueb, 1986). However, alternate modes of development
subsequently evolved within all three groups. Interestingly, ancestral vestiges of
metamorphosis are observed during early development of direct developing anurans and
salamanders
(Callery & Elinson, 2000; Kerney et al., 2012), suggesting shared
evolutionary potential for radical and early shifts in the timing of metamorphosis.
Radical, later shifts in metamorphic timing are only observed in salamanders. In the most
extreme cases, metamorphosis has been abolished completely, yielding bizarre larvalform adults with completely aquatic life cycles (Gould, 1977; Shaffer & Voss, 1996).
These unique paedomorphic forms are ideal for investigating mechanisms of thyroid
hormone (TH) regulation that are associated with adaptive delays in metamorphic timing
and the evolution of novel life histories.
What is paedomorphosis and why salamanders?
Paedomorphosis is a heterochronic term that describes a specific pattern of
evolution, the retention of ancestral juvenile traits in the adult stage of a derived species
(Gould, 1977). In this meaning, paedomorphosis references an evolutionary change in
developmental timing between an ancestor and descendant species. However,
paedomorphosis is often used in a more general sense to describe the retention of larval
morphological traits in adult salamanders, irrespective of phylogeny. The evolution or
expression of paedomorphosis is clearly associated with TH, the primary metamorphic
hormone in anurans, salamanders, and some fish. In species that undergo a
metamorphosis, TH induces the regression of larval traits and the development of traits
typical of a more terrestrial adult (Figure 1.1). As discussed in more detail below, many
paedomorphic salamanders can be induced to undergo partial or complete
metamorphosis by simply placing them in a bath of TH. This suggests the possibility that
paedomorphosis evolves “simply” by blocking the synthesis, secretion, or reception of
TH in target cells (Page et al., 2010).
1 However, TH secretion is regulated by complex, neuroendocrine pathways that must first
develop during the larval period before these systems become competent to signal TH
release in response to intrinsic and extrinsic cues (Denver et al., 2002). Thus,
paedomorphosis could conceivably evolve by altering the development or function of a
number of different cells, tissues, and organs that regulate the release or reception of TH.
Given this complexity and the broad pleiotropic role that TH assumes in amphibian
development and physiology, it seems likely that paedomorph evolution requires
multiple changes across many genes.
The specific heterochronic process most commonly associated with salamander
paedomorphosis is neoteny (Gould, 1977) – relative to the ancestor or metamorphic
condition, somatic development is delayed but rate or timing of reproductive maturation
is the same. However, paedomorphosis in some species is associated with an earlier time
to first reproduction, achieved by accelerating the rate of gonadal development
(progenesis in Triturus alpestris; Denoel & Joly, 2000) or by initiating reproductive
maturation earlier in the life cycle (peramorphosis in Ambystoma talpoideum; Ryan &
Semlitsch, 1998). Thus, very different patterns of growth and development may be
associated with the expression of paedomorphosis within and between species
(Whiteman, 1994). Among species that occur in stable aquatic habitats, paedomorphosis
is a fixed trait with little or no incidence of metamorphosis. Within species that use less
permanent bodies of water for breeding, individuals may reproduce initially as
paedomorphs but undergo metamorphosis in subsequent years (Denoel et al., 2007). The
different patterns of growth and development observed among paedomorphs yield
different adult morphologies, and this suggests different mechanistic bases for the
evolution of paedomorphosis within and among salamander families. However, exactly
how patterns of development and morphology map onto mechanisms that block TH
signaling and metamorphosis, or affect the rate of gonadal development, is essentially
unknown. Without such mechanistic information, it is not possible to distinguish among
examples of paedomorphosis that evolve convergently among lineages, which clearly
has occurred during salamander phylogenesis (Shaffer & Voss, 1996; Weins &
Hoverman, 2008). Thus, all terms that describe the evolutionary origin or developmental
expression of paedomorphic salamanders are potentially useful metaphors, but if they are
2 not based upon mechanistic insight and genetically based models, they are not likely to
resolve outstanding questions concerning salamander life history variation and life cycle
evolution.
It is interesting to consider why paedomorphosis is unique to salamanders and
not other amphibians. In as much as paedomorphosis evolves as a prolongation of the
larval period, it is notable that salamanders with biphasic life cycles typically have
longer larval periods than caecilians or anurans (Duellman & Trueb, 1986). In biphasic
salamander species, metamorphosis may occur within the same season or year that eggs
are laid or larvae may overwinter in permanent habitats and metamorphose after one or
more additional years of development (e.g., Beachy, 1995; Castanet et al., 1996; Voss et
al., 2003). Although there are examples of anurans with multiyear larval periods,
salamander larval periods are typically longer. So, there are fundamental differences
between salamanders and other amphibians; salamanders have greater evolutionary
potential to protract (and contract in cases of direct development) the length of the larval
period. This greater evolutionary potential is associated with at least five aspects of
salamander development and life history. First, salamanders exhibit slow rates of growth
and development. Some anurans can complete metamorphosis in just a couple of weeks,
while the fastest developing salamander larvae require more than a month, and typically
several months (Duellman & Trueb, 1986). In general, salamander metamorphosis is
developmentally less radical and more protracted than that of anurans. For example,
salamander forelimb development does not coincide with other morphological
metamorphic changes, as it does in anurans (Figure 1.2). Instead, it occurs shortly after
hatching and before gills and tailfins are resorbed at metamorphosis. This suggests either
a decoupling of forelimb development from metamorphosis or an incredibly gradual
metamorphosis in salamanders that essentially spans the entirety of the larval period
(Duellman & Trueb, 1986). Second, aspects of TH signaling may differ between
salamanders and other amphibians. While it is clear that TH is required for salamander
metamorphosis, Ambystoma mexicanum tolerates levels of TH that cause abnormal
development and mortality in some anurans (Page et al., 2008). Also, several
orthologous genes, including thyroid hormone receptor beta (thrb), do not show the
same pattern of expression between TH-induced A. mexicanum and metamorphosing
3 anurans (Page et al., 2007, 2008). Third, salamanders often utilize low temperature
habitats, either because they occur at relatively high elevations or latitudes (Bizer, 1978;
Sexton & Bizer, 1978). Low temperature slows growth and development, depresses
activity of neurons and glands that regulate metamorphosis, and reduces the
responsiveness of tissues to TH (Uhlenhuth, 1919; Moriya, 1983a, 1983b). Fourth,
salamander larvae present a highly similar body design to adults. Thus, unlike anuran
tadpoles that must reorganize the body cavity at metamorphosis to allow space for
gonads to develop, salamanders can potentially allocate resources to developing gonads
and associated fat bodies during the larval period. Fifth, salamander larvae are
carnivorous like adults, and thus unlike herbivorous anuran tadpoles, do not rely upon
seasonal, primary productivity of ponds for food. Many anuran species have tadpoles
that are highly specialized for rapid development, utilizing seasonally abundant
resources in ephemeral ponds that select for finite larval periods (Wilbur, 1980).
Multiyear larval periods are simply not possible within the context of anuran life history
strategies that depend upon seasonally limiting resources. In contrast, salamander larvae
present greater developmental flexibility to utilize more permanent aquatic habitats by
extending the length of the larval period, indefinitely in the case of some paedomorphic
species.
Paedomorphosis and salamander life history variation
Salamander life histories can be broadly classified into three categories: biphasic,
paedomorphic, and facultative (Table 1.1). These are simply categories of convenience
and do not adequately treat the continuum of life histories represented by each class, and
the possibility of different genetic, developmental, and physiological bases within
classes. Species with biphasic life histories almost invariably undergo a metamorphosis.
This life history strategy is viewed as an adaptation to exploit transient opportunities for
larval growth and is often found where aquatic habitats are ephemeral (Wilbur & Collins,
1973; Wilbur, 1980). At the opposite extreme are salamanders that almost invariably
remain paedomorphic throughout their lives. Paedomorphic salamanders are associated
with isolated and stable aquatic habitats, such as large closed-basin lakes, spring-fed
lakes, caldera lakes, and river systems (Sprules, 1974; Shaffer, 1984). In facultative
4 species, metamorphs and paedomorphs are observed at varying frequencies within the
same population. In such species, it is thought that the ability to express paedomorphosis
and metamorphosis is advantageous when a landscape contains a mixture of ephemeral
and stable aquatic habitats (Wilbur & Collins, 1973; Sexton & Bizer, 1978; Wilbur,
1980; Semlitsch et al., 1990; Whiteman, 1994; Denoel et al., 2005;). Facultative life
histories are sometimes associated with aquatic habitats that may only be permanent for
a single year of reproduction; the fitness advantage must be considerable because it is not
uncommon to observe a paedomorphic brood in a cattle-watering tank on an arid
landscape with no permanent ponds (Randal Voss, personal observation). When
facultative species lay eggs in habitats that decline in quality during the larval period,
metamorphosis is initiated in all or some proportion of individuals in the population
(Semlitsch, 1987; Semlitsch et al., 1990).
Whereas metamorphic and facultative taxa are capable of dispersal among aquatic
habitats, obligatorily paedomorphic taxa are confined to isolated bodies of water. This
suggests that biphasic or facultative ancestors colonized aquatic habitats and ecological
conditions (e.g., permanent aquatic habitat and/or arid terrestrial conditions) selected for
paedomorphic life histories. The expression of paedomorphic life histories is known to
affect population structure and may affect the probability of speciation (Shaffer, 1984;
Shaffer & Breeden, 1989). The fitness benefits of paedomorphosis are an earlier time to
first reproduction, more than one breeding event per year, larger clutch size, and a higher
probability of mating success (Scott, 1993; Krenz & Server, 1995; Ryan & Semlitsch,
1998). It is important to note that paedomorphosis is not a pathology or default life
history strategy – a failure to undergo metamorphosis. Paedomorphic species are highly
adapted to their aquatic habitats. For example, Ambystoma dumerilii has secondarily
evolved webbed-feet and extra gill filaments that develop at the time metamorphosis
occurs in related metamorphic species (Figure 1.3). Thus, evolution of paedomorphosis
may release tissues from the constraint of TH remodeling, allowing new functions to
evolve that better adapt adults to aquatic habitats. The combination of selection and
genetic drift acting on isolated paedomorphic lineages may explain rapid divergence of
morphology and gene expression within some species groups (Shaffer & Voss, 1996;
Page et al., 2010).
5 Paedomorphic life history strategies are widespread among salamanders,
occurring in nine out of 10 families with a total of 57 species. Studies show that
paedomorphic salamanders have evolved multiple times in different lineages (Shaffer &
Voss, 1996; Denoel et al., 2005; Weins & Hoverman, 2008). Four salamander families
(Amphiumidae, Sirenidae, Proteidae, and Cryptobranchidae) are comprised entirely of
paedomorphic species, having no extant biphasic species. Some salamanders, such as
species in the families Cryptobranchidae and Amphiumidae, are said to undergo an
incomplete metamorphosis – meaning they undergo some anatomical changes, but not
others, and remain aquatic throughout their lives. For example, adult Hellbenders
(Cryptobranchus alleganiensis) lose their larval external gills, but do not develop eyelids
and retain a single gill-slit that appears as a circular opening on the neck (Larson et al.,
2003). The most species-rich salamander family, Plethodontidae, includes many biphasic
and direct developing species, with the Tribe Hemidactyliini (Subfamily Plethodontinae)
containing numerous paedomorphic species that are cave dwelling and exhibit reduced
pigmentation and vision (Larson et al., 2003). Perhaps the best-studied salamander
families with respect to life history variation are Salamandridae and Ambystomatidae.
Studies of paedomorphic and facultative species within these families have better
elucidated environmental and broad-sense genetic correlates of metamorph and
paedomorph expression, and these species continue to provide exceptional models for
investigating ecological aspects of life history evolution (Denoel et al., 2005; Doyle &
Whitemann, 2008; Takahashi & Parris, 2008; Denoel et al., 2009; Whiteman et al.,
2012).
Hormonal basis of salamander metamorphosis
Biphasic, paedomorphic, and facultative life histories begin the same way, with
embryonic and larval development occurring in an aquatic habitat. The primary
difference is when and if metamorphosis occurs. Although not as dramatic as
metamorphosis in anurans, multiple tissues undergo hormonally regulated changes in
cellular composition, biochemistry, and morphology during salamander metamorphosis
(Dodd & Dodd, 1976). These changes include degeneration of external gills and tailfins;
remodeling of skeletal elements and integument; shortening of the intestine;
6 modification of pigmentation; muscle degeneration and differentiation; and changes in
nitrogen excretion, urogenital function, and oxygen transport (Duellman & Trueb, 1986).
Also, new structures arise de novo, including eyelids and skin glands. All of these
changes adapt aquatic larvae for survival and reproduction as adults in more terrestrial
environments.
Although salamanders have a more protracted and gradual metamorphosis than
anurans, TH is the primary metamorphic hormone in both amphibian groups. Serum TH
in the form of tetraiodothyroine (T4) is low during embryonic and early larval
development, but as larvae approach metamorphosis, T4 levels increase (Norman et al.,
1987). This pattern of hormone secretion is not unlike that observed for other examples
of hormonally regulated, postembryonic development (e.g., insect and fish
metamorphosis; Gilbert et al., 1996; Power et al., 2001; Laudet, 2011). The rise in T4
during amphibian larval development is associated with maturation and stimulation of
brain regions that regulate thyroid activity, in response to extrinsic and intrinsic cues.
These brain regions include the hypothalamus and pituitary gland. Increasing levels of
T4 during larval development stimulate corticotropin-releasing hormone (CRH) release
from modified nerve endings within the median eminence. CRH travels through
capillaries to the anterior pituitary where it stimulates thyrotropes to release thyrotropin
(TSH) into the blood stream. TSH in turn stimulates the thyroid gland to release T4. The
TSH-releasing factor in amphibians is CRH, which also stimulates the release of
corticotropin (ACTH) and subsequently, glucocorticoids (Denver & Licht, 1989). The
potent, synergistic action of corticoids and TH in promoting anuran and salamander
metamorphosis clearly implicates the hypothalamic-pituitary- thyroid (HPT) and HPinterrenal
(HPI) axes in
the
physiological regulation of metamorphosis (Dodd &
Dodd, 1976; Buscaglia et al., 1985; Galton, 1991, 1992; Gancedo et al., 1992; Hayes et
al., 1993; Kuhn et al., 2005). Exactly how T4 levels initially increase at appropriate times
during amphibian development to trigger the onset of metamorphosis, and sustain T4 at
high levels during metamorphosis, is not fully understood (Manzon & Denver, 2004).
Peripherally, the availability and activity of TH is precisely regulated within
cells. In mammals, TH is transported into target cells via TH transporters, which actively
7 and preferentially import the hormone across the plasma membrane (Hennemann et al.,
2001). TH transporters have been studied in X. laevis (Ritchie et al., 2003; Connors et al.,
2010), but there have been no studies using salamanders. Deiodinases have received
considerable attention because they directly activate or inactivate intracellular forms of
TH in amphibians (Buscaglia et al., 1985; Galton, 1991; Kuhn et al., 2005). The low
activity T4 form of TH that is delivered to cells is converted intracellularly by
deiodinase-type 2 (DIO2) into highly active 3,5,3’-triiodothyronine (T3). A second
deiodinase, DIO3, inactivates intracellular T3. Although T4 may participate in cellular
signaling (Caria et al., 2008), T3 is the primary TH signaling molecule in anurans
(Denver et al., 2002) and presumably this is also true for salamanders, acting through
nuclear receptors (thra, thrb) that function as transcription factors. The higher activity of
T3 is associated with a higher affinity for TH receptor (TR) binding. The binding of T3
activates transcriptional regulatory networks in cell-specific manners (Shi, 2000). TH is
also known to cause cellular changes via nongenomic mechanisms, including membrane
receptors, transporter molecules, and cytoplasmic receptors that elicit changes in cells
through signaling pathways (Caria et al., 2008; Davis et al., 2008; Furuya et al., 2009;
Cheng et al., 2010). The effects of these nongenomic mechanisms of TH action and their
role in amphibian metamorphosis have yet to be studied in detail, however, Buchholz et
al. (2004) used transgenic analysis to show that TR is sufficient to mediate TH signaling
during metamorphosis in X. laevis. Regulation of deiodinases, T4/T3 concentrations, and
developmental expression of TRs help explain how a single hormone can coordinate
responses among different cell types, and more generally regulate the temporal sequence
of remodeling events during amphibian metamorphosis (Brown & Cai, 2007).
Hormonal basis of paedomorphosis
Only a year after, Gudernatsch (1912) established TH as the anuran metamorphic
hormone; TH was shown by Laufberger to induce metamorphosis in A. mexicanum
(Mexican axolotl) (reviewed by Huxley & Hogben, 1922). Over the next few decades,
the responsiveness of many different paedomorphic salamanders to TH was investigated.
Studies showed that morphologically similar paedomorphic salamanders (with external
gills or gill slits, no eyelids, larval pigmentation, etc.) responded differently to TH.
8 Members of the family Proteidae, such as the American mudpuppy (Necturus
maculosus), are known to be refractory to TH, such that treating them with exogenous or
injected TH does not induce morphological metamorphosis (Swingle, 1922). Larsen
(1968) proposed that peripheral tissues of N. maculosus had lost sensitivity to TH, given
that it exhibits a morphologically normal thyroid gland that had been tested for
functionality via trans- plant experiments (Swingle, 1922; Charipper, 1929; Grant, 1930).
This prompted investigations into the functionality of the animal’s deiodinase activity
and TR expression. Galton (1985) found that deiodinase activity capable of converting
T4 to T3 was present within N. maculosus, and Safi et al. (2006) showed that N.
maculosus expresses functional TRs that bind to DNA and activate transcription in
response to TH treatment, including transcriptional changes of classical TR target genes
like stomelysin 3 (mmp11) and thrb (Safi et al., 2006; Vlaeminck-Guillem et al., 2006).
Additionally, in situ hybridization has been used to show expression of thra and thrb in
central nervous system, epithelia of the digestive tract, myocardium, and skeletal muscle
(Vlaeminck-Guillem et al., 2006). These studies suggest that the paedomorphic
phenotype exhibited by N. maculosus may be due to the loss of key TH response genes
needed to induce morphological changes, even though juveniles experience a
postembryonic period of high TH sensitivity. Apparently, individuals undergo natural
transcriptional and bio- chemical changes during a “cryptic metamorphosis” that does
not result in gross morphological alterations (Vlaeminck-Guillem et al., 2006; Laudet,
2011). Species of the families Cryptobranchidae and Amphiumidae are responsive to
THs in some respects, but not others. They often reabsorb external gills and develop
adult characteristics of some features, but do not develop eyelids and retain gill slits as
mature adults, indicating that some metamorphic processes may have been decoupled
from TH regulation (Larson et al., 2003). When species of these families are treated with
high doses of TH, often the only result is skin shedding, a trait that is regulated by TH in
adults of metamorphic species, though otherwise the individual continues to exhibit
paedomorphic characteristics (Dent, 1968; Fox, 1984). It was originally thought that the
Texas blind salamander (Eurycea rathbuni, formerly assigned to the genus Typhlomolge)
was the only known vertebrate to lack a thyroid gland (Emerson, 1905). However,
Gorbman (1957) later showed the presence of a thyroid gland in E. rathbuni and Dundee
9 (1957) showed that an individual of this species underwent some morphological changes
(gill and tail reabsorption) in response to T4, but other larval features were resistant
(histological sections of skin revealed only larval characteristics). However, related
paedomorphic species (E. tynerensis and E. neotenes) complete metamorphosis after T4
treatment in less than 2 weeks (Kezer, 1952). Paedomorphic species of Ambystoma also
complete metamorphosis when treated with T4 but show species-specific variation in
metamorphic timing (Voss et al., 2012). Collectively, these studies establish that
paedomorphosis is associated with TH regulation and tissue responsiveness, and these
characteristics vary among closely and distantly related species.
Based on results of early TH induction experiments, it was proposed that
paedomorphic species evolve via several mechanisms, including disruption of peripheral
and central mechanisms of TH regulation (Kezer 1952; Dent, 1968). Species that were
shown to be refractory to TH but had active thyroid glands were presumed to be deficient
in peripheral mechanisms that receive and transduce the T4 signal, such as TRs. Species
that responded to high levels of TH but did not apparently metamorphose in nature were
assumed to be deficient in some aspect of central regulation within the hypothalamus or
pituitary. Such a deficiency would normally prevent T4 release but could artificially be
over-ridden in the lab by T4 treatment. It also seems likely that central regulatory
mechanisms have been altered in the evolution of facultative species from biphasic
ancestors. This is because the brain and pituitary regulate appropriate physiological and
developmental responses in response to environmental cues. In the case of facultative
paedomorphosis, the HPT axis is not sufficiently activated if cues favor a paedomorphic
life history, or environmental conditions suppress hormone activity, or if larval
development and maturation is delayed
(Whiteman, 1994). Several environmental
conditions that are associated with habitat quality affect the probability that an individual
will express paedomorphosis, including low density of conspecifics
(Harris, 1987;
Semlitsch, 1987), food availability (Semlitsch, 1987; Denoel & Poncin, 2001; Ryan &
Semlitsch, 2003), pond permanence (Semlitsch, 1987; Semlitsch et al., 1990), and low
temperature (Snyder, 1956; Sprules, 1974; Eagleson & McKeown, 1980).
10 Paedomorphosis in the Mexican Axolotl
The exemplar of all paedomorphic salamander is the Mexican axolotl (A.
mexicanum). The Mexican axolotl is native to the Xochimilco lake system, near presentday Mexico City. Whereas historical populations were sufficiently large to provide an
important food source to native people for thousands of years (Smith, 1989), the current
A. mexicanum population at Xochimilco is on the verge of extinction (Graue, 1998;
Zambrano et al., 2007). Smith (1989) provides an enjoyable historical account of the
original collection of A. mexicanum from Xochimilco in 1863. Six A. mexicanum were
transferred to Paris where Dumeril (1870) subsequently reported that individuals
reproduced in an aquatic, larval state and that some of the resulting offspring underwent
a metamorphosis. Smith (1989) suggests that some of the original stock that arrived in
Paris included closely related members of the Tiger salamander species complex, that
are capable of expressing a paedomorphic or metamorphic life history. However, it
seems likely that domestication altered the penetrance for expressing paedomorphosis
and the original axolotl stock was pure A. mexicanum that maintained a higher
propensity to express metamorphosis in nature (Voss & Shaffer, 2000). Even though
metamorphic forms have been culled from the Ambystoma Genetic Stock Center axolotl
collection for decades, the frequency of spontaneous metamorphosis is only 1–2%, with
a 10% frequency observed if A. mexicanum experience stressful conditions (Randal
Voss, unpublished data).
The simplest explanation for paedomorphosis in A. mexicanum is a faulty thyroid
gland. However, the thyroid gland is functional; it can synthesize T4 and release T4 after
TSH stimulation (Prahlad, 1968; Taurog et al., 1974). Numerous studies, including our
own (Page et al., 2007, 2008), have demonstrated that individuals can be stimulated to
undergo morphological metamorphosis by administering T4. This suggests that
intracellular receptors and deiodinase enzymes are functional. Putative T3 receptors have
been identified in red blood cells (Galton, 1991) and TR function has been shown in
vitro using mammalian cells (Safi et al., 2004). Also, the activity of deiodinases has been
demonstrated (Galton, 1991; Darras et al., 2002). Blount (1950) showed that
metamorphosis is inhibited if the pituitary of A. mexicanum is transplanted into the
11 metamorph A. tigrinum, while the reciprocal transplantation experiment resulted in
metamorphosis. Tassava (1969) pointed out that these grafts were probably not precise
enough to rule out a hypothalamic contribution. TSH is present in the A. mexicanum
pituitary (Taurog et al., 1974) at a quantity sufficient for inducing metamorphosis
(Darras & Kuhn, 1983). Thus, the pituitary is capable of producing sufficient functional
TSH for metamorphosis, but for some reason, it is not released. There are at least four
explanations for this observation in A. mexicanum: (1) The pituitary does not receive
appropriate stimulation from the hypothalamus, (2) the pituitary is insensitive to
hypothalamic stimulation, (3) the pituitary is defective in releasing TSH, and (4) the
pituitary is functional but stimulated to release TSH at the wrong time during ontogeny
(Rosenkilde & Ussing, 1996). Galton (1992) suggested that T4 levels might never
eclipse a threshold during development to stimulate brain regions to increase activity of
the thyroid gland. Kuhn et al. (2005) similarly proposed that brain T3 availability is the
limiting factor for metamorphic onset. They found that DIO2 activity in the brain, which
is required to generate T3, requires appropriate doses of corticoids and T4. Thus,
stimulation of both the HPA and HPI axes is required for metamorphosis, and in fact, it
is possible to induce metamorphosis in A. mexicanum if sub-metamorphic levels of T4
and corticoids are administered together. Because injection of amniote CRH stimulates
corticotropin but not thyrotropin release, Kuhn et al. (2005) implicated the thyrotroph
CRH receptor as a likely candidate gene for paedomorphosis in A. mexicanum.
However, researchers have demonstrated T4 increases after stimulating hypothalamic
regions in neotenic tiger salamanders (A. tigrinum) and A. mexicanum; this could
seemingly only occur if downstream thyrotrophin signaling pathways are functional
(Norris & Gern, 1976; Rosenkilde & Ussing, 1996).
It is also possible that a recently discovered TSH signaling pathway may be
implicated in salamander paedomorphosis. Studies of neuroendocrine regulation of
amniote reproductive cycles showed that TSH from the pars tuberalis (PT) of the anterior
pituitary gland is released to signal DIO2 transcription in the hypothalamus (Hanon et
al., 2008, Nakao et al., 2008). This increases local T3 concentrations and thus stimulates
canonical TH signaling pathways. Surprisingly, these studies show that the flow of
information may be bidirectional between the pituitary and hypothalamus. If a similar PT
12 signaling pathway is necessary for metamorphosis, this pathway would not be activated
in A. mexicanum because of a block in TSH release. This might also explain why A.
mexicanum does not show strong seasonality in breeding within lab populations and can
breed multiple times per year. Whether paedomorphosis involves a PT signaling
mechanism or a classical TSH signaling mechanism, the metamorphic block in A.
mexicanum is likely associated with hypothalamus and pituitary function (Prahlad &
DeLanney, 1965; Norris et al., 1973; Rosenkilde & Ussing, 1996; Kuhn et al., 2005;
Laudet, 2011).
Transcriptional analysis of salamander metamorphosis and paedomorphosis
Most mechanistic studies of metamorphosis and paedomorphosis have focused
on physiology and to a lesser extent biochemistry. Recently, microarrays developed for
A. mexicanum have been used to characterize global transcriptional responses of
metamorphic and paedomorphic salamanders. These studies have focused on the skin
and brain, the former because it is a peripheral target tissue of TH and much is known
about anatomical changes that occur during metamorphosis (Fahrmann, 1971a, 1971b,
1971c; Fox, 1983; Page et al., 2009). The brain has also been targeted because it is
where HPT and HPI axes are centrally regulated. Reviewed below are three important
findings from these studies.
T4 precisely activates metamorphic transcriptional programs in a concentrationdependent manner
Page et al. (2008) varied T4 concentration by an order of magnitude and
examined transcriptional patterns in A. mexicanum skin after 0, 2, 12, and 28 days of T4
treatment. Individuals exposed to 50 nM T4 initiated metamorphosis approximately 1
week prior to those in 5 nM, however, the sequence of transcriptional and morphological
changes during metamorphosis, and the length of the metamorphic period did not differ
(Figure 1.4A). Over 1000 genes were identified as differently expressed in both
treatments, showing the same directional trends, but the patterns were temporally shifted
by T4 concentration. The study showed that the timing of metamorphic onset is T4
concentration dependent, and surprisingly, high T4 concentration did not disassociate
13 subsequent transcriptional programs. It remains to be determined if the T4 concentration
effect in A. mexicanum is manifested at the level of target cells, the HPT axis, or possibly
both (Rosenkilde & Ussing, 1996).
Patterns of gene expression are complex, reflecting tissue remodeling, changes in
cellular metabolism, and loss and gain of larval and adult cell types
One of the advantages in working with A. mexicanum is that they present a
natural, hypothyroid condition throughout life (Huggins et al., 2012). Thus,
transcriptional programs can be activated precisely at juvenile or adult stages of the life
cycle. When T4 is administered, thousands of genes show significant changes in
transcript abundance. In a very general sense, these genes show four temporal gene
expression patterns: they increase or decrease in abundance, or they transiently increase
or decrease during metamorphosis (Figure 1.4B). Linear, quadratic, and piece-wise
regression models have been used to identify groups of genes that similarly increase or
decrease in abundance as a function of days of T4 treatment. The earliest gene
expression changes precede tissue-remodeling events and are associated with cells that
are specific to larvae or cells that differentiate to form adult tissues. For example, after 2
days of T4 treatment, genes that are specifically expressed in apical and Leydig cells of
the larval skin show decreases in transcript abundances (Page et al., 2007, 2008, 2009).
At this time, apical and Leydig cell numbers are normal for larval skin and there is no
histological evidence of remodeling. However, after 12 days of T4 treatment, apical cells
are no longer present in histological sections and Leydig cells are significantly reduced
in number (Page et al., 2009); presumably, both cell types undergo TH-induced,
programmed cell death, as has been demonstrated for apical cells in metamorphosing
anurans (Schreiber & Brown, 2002). As larval cells and their gene expression products
decrease in abundance during metamorphosis, TH induces the proliferation and
differentiation of adult progenitor cells. As a result, transcripts associated with adult cells
increase in abundance. The most dramatic changes are seen for basal keratinocytes that
proliferate and differentiate during metamorphosis and transcribe adult specific keratins
that generate a more cornified epithelial layer to limit desiccation during the adult phase.
Genes that transiently increase in abundance code for matrix-remodeling proteins (e.g.,
14 MMPs) and cellular processes like metabolism, transcription, and translation, while
genes that decrease in abundance include negative regulators of cellular processes (Page
et al., 2008). These gene expression patterns illuminate tissue remodeling and
macromolecular biosynthetic processes that occur during metamorphic climax.
Although some genes are expressed similarly between metamorphic and paedomorphic
individuals during larval development, the majority of the differentially expressed genes
are unique
For example, brain transcription has been compared between A. mexicanum and
A. t. tigrinum individuals during larval development and the period spanning the
metamorphic period of the later species (Page et al., 2010). As A. t. tigrinum individuals
near the onset of anatomical metamorphosis, more gene expression differences are
observed between larvae of these species. Presumably, many of these expression
differences trace to increasing levels of glucocorticoids and THs in A. t. tigrinum
individuals as they initiate early metamorphic changes. In support of this idea, a
glucocorticoid receptor (nr3c2) showed significantly higher expression in A. t. tigrinum.
Also, some genes that were more highly expressed during natural metamorphosis in A. t.
tigrinum were also similarly upregulated in A. mexicanum brains when individuals were
induced to metamorphose with T4 (Huggins et al., 2012). However, hundreds of genes
were expressed more highly in A. t. tigrinum than A. mexicanum throughout the larval
period. At this point, it is not clear if all of these differences are associated with
metamorphic and paedomorphic life histories; however, the observed systematic bias in
transcription suggests more than species-specific divergence of gene expression.
Genetic analysis of paedomorphosis
The idea that paedomorphic salamanders simply need a dose of TH to induce the
second half of the ancestral ontogeny influenced the way paedomorphic life histories
were thought to originate and evolve. For example, Goldschmidt (1940) cited A.
mexicanum as the “hopeful monster”, an example of evolution waiting around for the
right macromutation – simply block a single physiological step in TH regulation and
instantly a novel form is originated. The idea that paedomorphosis results from a simple
15 mechanistic change gained further support when genetic studies showed that
metamorphic and paedomorphic forms could be segregated according to Mendelian
expectations. In particular, Humphrey (1967) showed that when A. mexicanum/A. t.
tigrinum hybrids were backcrossed to A. mexicanum, there is 1:1 segregation of
phenotypes. Later, Tompkins (1978) and Gould (1981) argued from Humphrey’s single
cross that A. mexicanum was an example of single gene, macromutational evolution. To
further test the single gene hypothesis, Voss (1995) crossed domesticated A. mexicanum
and A. t. tigrinum to create F1 hybrids and like Humphrey (1967), backcrossed these to
A. mexicanum. The ratios of metamorphs and paedomorphs arising from two relatively
large backcrosses were consistent with single locus control, thus supporting the classical
idea of a single mutation underlying the evolution of paeodomorphosis. In these crosses,
the A. t. tirginum and A. mexicanum alleles were dominant and recessive in effect,
respectively. Subsequently, DNA was isolated from individuals of these crosses,
amplified fragment length polymorphisms (AFPLs) were typed, and a genetic linkage
map was constructed. This map was used to roughly locate the position of the major
gene (met1) for paedomorphosis (Voss & Shaffer, 1996). To determine if met1 arose
independently in the domesticated stock, the back-cross design was repeated using A.
mexicanum collected from Lake Xochimilco (Voss & Shaffer, 2000). The segregation of
phenotypes did not fit a single gene model, and the segregation of AFLPs marking the
met1 region exhibited segregation distortion, suggesting an epistatic effect between met1
and other loci on viability. Still, met1 genotypes did associate with metamorphic and
paedomorphic phenotypes, indicating a difference in genetic background between wildcaught A. mexicanum and the domesticated stock.
Voss and Smith (2005) made additional A. mexicanum/A. t. tigrinum backcrosses
and reared over 500 offspring to rigorously quantify the genetic effect of met1 from
wild-caught A. mexicanum. These crosses showed that met1 not only explained discrete
variation in the expression of metamorphosis and paedomorphosis but also continuous
variation in metamorphic timing. Approximately 20% of the individuals that inherited
two met1 alleles from wild A. mexicanum were paedomorphic. The remainder of these
homozygotes delayed metamorphic timing by an average of 36 days, relative to
heterozygotes. These results showed that met1 determines the expression of
16 metamorphosis and paedomorphosis by altering metamorphic timing. Within the context
of hybrid crosses, alleles from metamorphic and paedomorphic species decrease and
delay the time to metamorphosis, respectively. Variation in the number of paedomorphic
individuals generated from wild-caught versus domesticated A. mexicanum crosses
reflect differences in genetic background. Here, genetic background is defined as the
summative effect of loci whose independent effects are too small to identify and quantify
statistically. Within the genetic background of the domesticated stock, which has
undergone strong artificial selection for paedomorphosis, the penetrance of met1 is
greater, and this results in a higher proportion of paedomorphic individuals in hybrid
crosses (Voss & Smith, 2005).
The link between met1 and thyroid hormone regulation of metamorphosis
The link between met1 and TH was recently established by combining
quantitative trait locus (QTL) analysis with TH induction. Recognizing that
paedomorphic species show variation in time to complete metamorphosis when treated
with TH, Voss et al. (2012) crossed two paedomorphic species (A. mexicanum and A.
andersoni) and then performed a backcross to segregate TH response alleles among
second generation offspring. At 120 days postfertilization (dpf), backcross offspring were
treated with 2.5 nM T4 and then scored for time to complete metamorphosis. Essentially,
all offspring initiated and completed metamorphosis over an ~ 160 day interval of time.
A QTL screen identified the genomic position of met1 and two additional QTL (met2,
met3), each explaining approximately 10% of the variation in metamorphic timing; the
effects of all alleles were additive with A. andersoni alleles decreasing the time to
metamorphosis. On average, individuals that inherited met1-3 alleles from A. andersoni
metamorphosed earlier, just as did hybrid A. mexicanum that inherited met1 alleles from
metamorphic A. t. tigrinum (Voss & Smith, 2005). The decrease in metamorphic timing
was approximately 19 days for A. andersoni met1 versus 36 days for A. t. tigrinum met1.
These results demonstrate that variation in metamorphic timing is determined in part by
alleles that segregate at TH-responsive QTL, and alleles from paedomorphic species may
vary in effect.
17 In natural populations of amphibians, variation in metamorphic timing can affect
the values of other life history traits. Life history theories predict that individuals should
metamorphose early if conditions for larval growth are poor, but delay metamorphosis
and attain large larval body sizes if growth conditions are optimal (Gould, 1977; Wilbur
& Collins, 1973). This is because larger body sizes at metamorphosis translate into
increased reproductive success in the adult phase (Semlitsch et al., 1988). From a
physiological perspective, metamorphic timing and body size are expected to covary as a
function of resource allocation. From an evolutionary perspective, metamorphic timing
and body size are expected to covary and coevolve if they share a common genetic basis.
This prediction holds in the case of met2. In the A. andersoni/A. mexicanum backcross
study described above, met2 not only explained significant variation in metamorphic
timing but it also explained significant variation in adult body size (Voss et al., 2012).
Individuals that delayed metamorphosis significantly increased body length at 300 dpf
and weight at 400 dpf, and the effect was approximately equivalent for sexually
dimorphic males and females. Overall, the genetic results reviewed here show that THresponsive QTL account for variation in metamorphic timing, expression of
metamorphosis versus paedomorphosis, and adult fitness traits.
What genes map to metamorphic timing QTL?
The A. mexicanum genome is in early stages of genome sequencing and thus
comparative mapping has been used to more finely resolve the positions of met1-3 and
identify candidate genes. The most promising candidate that has been identified is
pou1f1, a transcription factor associated with combined pituitary hormone deficiency
(CPHD) in humans. Pou1f1 is predicted to map to the position of met3 on linkage group
(LG) 7. CPHD is associated with deficiencies in the secretion of pituitary hormones that
regulate growth and development, but not hormones associated with reproductive
physiology. Thus, pou1f1 is a candidate gene for evolutionary developmental changes
that are independent of reproductive maturation (neoteny), as is seen in the example of
salamander paedomorphosis.
In contrast to met3, identifying candidates for met1 and met2 has proven more
difficult. For example, the 2 cm interval defining the location of met1 on LG2 marks a
18 unique synteny disruption in Ambystoma and perhaps other salamanders (Voss et al.,
2011). This is a bit surprising and unlucky because the Ambystoma genome has
undergone relatively few chromosomal rearrangements relative to other vertebrates and
especially amniotes (Smith & Voss, 2006). Genes (e.g., rai1, shmt1, drg2, med9) from
the Smith–Magenis disease syndrome region in the human genome flank one side of
met1, while several genes associated with neural development and function (e.g., setd2,
ngfr, ccm2) flank the other side. In no other vertebrate genome are these groups of
flanking markers observed in synteny. Recently, microarray analysis was performed to
identify genes that are expressed differently during larval development as a function of
met1 genotype (Robert Page and Randal Voss, unpublished data). It was reasoned that
this combined genetic-genomic approach would identify differentially expressed genes
that map to the position of met1, potentially identifying candidate genes. As predicted,
the majority of differentially expressed genes from LG2 mapped within 10 cm of met1,
and six of the 14 genes that located within 2 cm of met1 were differentially expressed as
a function of genotype during larval development and/or metamorphosis. At this point, it
is not clear if the met1 genomic region is more enriched for differentially expressed
genes than other regions of the Ambystoma genome, and perhaps genes that are TH
responsive and associated with brain development. Regardless, it is clear from
microarray studies that many genes are differentially expressed between paedomorphic
A. mexicanum and metamorphic A. t. tigrinum, and met1 locates to a uniquely structured
genomic region with several candidate genes for paedomorphosis.
Synthesis: evolution of salamander paedomorphosis
Many times during salamander evolution, mechanisms that regulate TH induction
of metamorphosis were altered, resulting in paedomorphic salamanders that retain larval
traits into the adult, breeding phase. Over the past 100 years, researchers have treated
paedomorphic salamanders with TH to investigate the physiological and evolutionary
basis of this unique salamander trait. Studies show that tissues within some species are
entirely resistant to TH, while others show partial or complete metamorphosis. These
varied responses are not explained by phylogeny or morphology. Salamanders are a
morphologically conservative group and paedomorphosis has evolved independently
19 during salamander evolution. It is not likely that investigations of convergent
morphological differences among deeply divergent paedomorphic species will yield a
synthetic model for the origin of salamander paedomorphosis. Instead, an understanding
of the origins of paedomorphosis requires knowledge of physiological and genetic
mechanism, environmental context, and life history traits associated with fitness (Figure
1.5). In the case of ambystomatid salamanders, paedomorph expression is associated
with TH-responsive loci with pleiotropic affects on metamorphic timing and adult fitness
traits. According to our model, paedomorphosis and metamorphosis are mechanistically
connected by loci that segregate allelic variation for responsiveness to T4. Phenotypes
arising from hybrid ambystomatid crosses clearly show that it is possible to artificially
phenocopy the complete spectrum of metamorphic timing variation observed among
species in nature. Given variation in habitat stability, selection should favor low TH
response alleles in biphasic populations when habitats allow for extensions of the larval
period, as these alleles will delay metamorphic timing and increase adult body size.
Under directional selection for mutations that increase the length of the larval period,
central mechanisms of TH-release become progressively desensitized to intrinsic and
extrinsic signals that provide reliable cues for initiating metamorphosis within the
context of a biphasic life history. It is hypothesized that central mechanisms are
desensitized because TH-response loci become fixed for alleles that are less responsive
to positive TH feedback. As genetic background changes, the likelihood of paedomorph
expression increases. In support of this model, studies of facultative paedomorphic
populations (A. talpoideum) have demonstrated the efficacy of selecting for higher
paedomorph expression in relatively few generations (Semlitsch & Wilbur, 1989). If
directional selection occurs over many generations and within the context of large and
stable aquatic habitats, as has happened independently in the case of Tiger salamanders
species in Mexico (Shaffer, 1984; Shaffer & McKnight, 1996; Shaffer & Voss, 1996),
further changes in genetic background are predicted to yield individuals that are no
longer capable of initiating metamorphosis in nature.
Many studies have shown that environmental factors, and especially low
temperature, are often associated with paedomorphic populations (Snyder, 1956;
Sprules, 1974; Eagleson & McKeown, 1980). It is possible that some examples of
20 paedomorphosis may be pathological, reflecting extreme physiological conditions that
may be exacerbated by inbreeding depression within small, isolated populations.
However, Matsuda (1982; 1987) suggested that low temperature is pivotal in the
adaptive evolution of paedomorphic salamanders; low temperature initially causes the
expression of paedomorphosis in a population, and then through genetic assimilation
(Waddington, 1953) the paedomorphic phenotype becomes fixed. The model proposed
above is consistent with genetic assimilation because it also assumes directional
selection of TH-response alleles that increase an individual’s likelihood of expressing a
paedomorphic phenotype. The selection process and allelic effects depend upon
environmental context.
The experimental approaches and evolutionary model outlined above provide a
possible framework from which to initiate studies of the genetic basis of metamorphic
timing in natural amphibian populations. At this point, it is not clear how many THresponse genes contribute to adaptive differences in metamorphic timing within natural
populations and among species, or whether the same loci and physiological mechanisms
are implicated in independent examples of paedomorph evolution. Also, much remains
to be learned about facultative paedomorphosis at a mechanistic level. For example, does
facultative expression of paedomorphosis mechanistically present an intermediated
condition between metamorphosis and paedomorphosis, and how is it regulated? It is
exciting to think that these and other questions can now be resolved using genetic and
genomic methods that are applicable to nongenetic model organisms. The future looks
bright for investigating evolutionary, developmental, and physiological aspects of
salamander life history and life cycle evolution.
21 Table 1.1: General characteristics of salamander life history. This table uses common
salamander life history terminology and is meant to provide an overview of common
characteristics within these categories.
Metamorphic
•
•
•
•
•
Habitat often
contains ephemeral
ponds (1)
Almost invariably
metamorphose in
nature (1)
Generally reach
sexual maturity after
metamorphosis (2,3)
Thyroid hormone
causes
morphological
changes (e.g. gill
reduction, larval to
adult skin changes,
tail fin reduction,
etc.) (4)
Species may
respond to stressful
environmental
factors with
precocious
metamorphosis (5)
Paedomorphic
•
•
•
•
•
Aquatic habitat is
often stable (6)
Rarely/never
metamorphose in
nature (6)
Reach sexual
maturity while
remaining in larval
aquatic habitat (2,7)
May or may not
respond to thyroid
hormone with
morphological signs
of metamorphosis
(4)
In rare cases,
stressful
environmental
factors may cause
some species to
metamorphose (for
example, A.
mexicanum) (8,9)
Facultative
•
•
•
•
•
Habitat often contains a
mixture of temporary
and stable aquatic
features (3)
Populations of the same
species often vary in the
proportion of individual
that metamorphose and
the time to
metamorphosis (3)
Timing of sexual
maturity may vary
between metamorphic
and paedomorphic
individuals of the same
population (but not
always) (10,11,12)
Hormonal regulation has
not been studied in great
3
detail . Species are
thyroid hormone
responsive (13,14,15)
Environmental factors
affect the probability of
an individual to express
metamorphosis or
paedomorphosis
(6,16,17)
(1) Wilbur, 1980 (2) Duellman and Trueb, 1986 (3) Denoël et al., 2005 (4) Bruce, 2003
(5) Wilbur and Collins, 1973 (6) Sprules, 1974 (7) Armstrong et al., 1989 (8) Smith, 1969
(9) Voss, unpublished observations (10) Scott, 1993 (11) Ryan and Semlitsch, 1998 (12)
Denoël and Joly 2000 (13) Lynn and Wachowski, 1951 (14) Snyder, 1956 (15) Voss et
al., 2012, (16) Eagleson and McKeown, 1980 (17) Semlitsch, 1987
22 Figure 1.1: A representative salamander larva, adult metamorph (Ambystoma tigrinum),
and adult paedomorph (A. mexicanum). Critical levels of thyroid hormone induce
metamorphosis in salamanders. Environmental and genetic factors may cause thyroid
hormone levels to be low, resulting in paedomorphic salamanders. The pictures were
taken by Jeramiah Smith.
23 Figure 1.2: Timeline showing timing of fore and hindlimb development and the
metamorphic period of laboratory reared A. t. tigrinum. The dashed line indicates
variation among individuals in presenting early signatures of anatomical metamorphosis
(bulging eyes, reduced tail fins). In the anuran X. laevis, hindlimb and forelimb
development initiates 2 weeks after hatching. Hindlimbs complete development prior to
metamorphosis, while forelimbs emerge during metamorphosis.
24 Figure 1.3: Ambystoma dumerilii is a paedomorph that has secondarily evolved webbedfeet and extra filaments that develop on the dorsal surfaces of each gill. The bottom
panel is a magnified image of the dorsal gill surface. The arrow indicates dorsal
filaments (<1.0 mm in length) that are sprouting on the gill surface. These extra fibers
form late in the larval period, at the time metamorphosis normally occurs in related
metamorphic species. The A. dumerilii picture was taken by Brad Shaffer.
25 Figure 1.4: (A) Representative expression profile for a gene that shows a linear increase
in mRNA abundance after TH induction of A. mexicanum using 5 or 50 nM T4. (B) Four
general types of gene expression response observed after TH induction of
metamorphosis in A. mexicanum. See text for details.
26 Figure 1.5: Model showing relationships among salamander life history variation,
physiology, and habitat stability. Paedomorphosis is associated with low thyroid hormone
responsiveness, permanent aquatic habitats, delayed metamorphic timing, and large body
size. Metamorphosis is associated with contrary values for these variables.
27 CHAPTER TWO: MICROARRAY COMPARISON OF JUVENILE,
PAEDOMORPHIC, AND METAMORPHIC BRAIN GENE TRANSCRIPTION
INTRODUCTION
Mexican axolotls (Ambystoma mexicanum) have long intrigued scientists because
they present a paedomorphic mode of development (Gould, 1977). Unlike closely related
species with aquatic and terrestrial phases of development, axolotls do not initiate
metamorphosis and as a result, retain larval characteristics throughout life (Johnson and
Voss, 2013). Thus, in comparison to species that undergo metamorphosis, axolotls
seemingly do not mature several of their external anatomical features, including tail fins,
gills, eyes, and epidermis. However, internally, gonads and brain mature similarly to
metamorphic species. For example, axolotls and terrestrial A. tigrinum share many
temporal patterns of brain gene expression at the time A. tigrinum larvae initiate
metamorphosis (Page et al., 2010). Conservation of gene expression among metamorphic
and paedomorphic salamanders suggests axolotls retain cryptic maturational changes of
the metamorphic ancestor. Additionally, Page et al. (2010) showed that axolotls also
present unique, temporal gene expression patterns that differ from A. tigrinum; this
suggests that these differences may correlate with paedomorphic development. However,
the above speculations are tenuous because an interspecific comparative approach cannot
differentiate paedomorphic and metamorphic characteristics from characteristics that
arise independently among species during evolution.
In this study, I used an intraspecific comparative approach to identify gene
expression changes that correlate with metamorphic and paedomorphic modes of
development. Although axolotl almost always exhibit paedomorphosis, some individuals
spontaneously undergo metamorphosis, the frequency of which is < 1-2% in the
Ambystoma Genetic Stock Center (AGSC) (Johnson and Voss, 2013). I obtained juvenile,
paedomorphic, and metamorphic A. mexicanum brains, isolated RNA, and performed
microarray analysis. Using estimates of gene expression, I performed statistical contrasts
to accomplish two objectives: (1) To determine how transcription differs between
juvenile and adult brains, and (2) To identify genes and associated ontologies that change
significantly as a function of paedomorphic and metamorphic development.
28 METHODS
Animals and tissue sampling
Three adult male axolotls (four years old) that had spontaneously metamorphosed
and three paedomorphic axolotls of the same age (two males and one female) were
obtained from the Ambystoma Genetic Stock Center (AGSC) at the University of
Kentucky. Additionally, three juvenile axolotl siblings were obtained from the AGSC and
reared under the same laboratory conditions until 130 days post hatching (dph). The
juveniles were immature with respect to gonad maturation, so it was not possible to
determine genetic sex. All individuals were anesthetized in 0.02% benzocaine, and the
brains and pituitaries were removed, flash frozen in liquid nitrogen, and stored
individually at -80 C until RNA isolation was performed. Samples were divided into
three groups: juvenile (JUV), adult paedomorph (PAED), and adult metamorph (MET).
Animal care and use was carried out under University of Kentucky IACUC protocol
#1087L2006.
RNA isolation and microarray analysis
Total RNA was isolated independently from whole brain, including pituitary, of
each of the three individuals in the JUV, PAED, and MET groups using TRIzol
(Invitrogen, Carlsbad, CA). Samples were further purified using Qiagen RNeasy minicolumns following manufacture’s guidelines. All RNA samples were quantified using a
NanoDrop ND-1000 and an Agilent BioAnalyzer. RNA expression profiling was
conducted using custom Amby_002 microarrays (Huggins et al. 2012). The nine isolated
RNA samples were labeled and hybridized to independent microarray GeneChips and
scanned by the University of Kentucky Microarray Core Facility according to standard
Affymetrix protocols. A total of nine GeneChips were used in the experiment – one chip
per individual, three individuals per group, and three groups (JUV, MET, PAED).
Background correction, normalization, and expression summaries were accomplished
using the robust multi-array average (RMA) algorithm (Irizarry et al. 2003)
29 Global analyses of microarray data
The distribution of raw hybridization intensities for each array was evaluated
using box plots to identify any abnormal arrays. Microarray chips were then processed
using RMA and grouped using principal components and hierarchical cluster analysis in
JMP genomics version 5.0 (SAS Institute, Inc.) For the principal components analysis,
arrays were mean centered and scaled to a variance of one. For the cluster analysis, “Fast
Ward’s” method was used (Ward, 1963).
Identification of differently expressed genes (DEGs) using microarray
Only probesets that measured signal intensities > 6.25 (the average value of the
midpoint of the saddle region from a signal intensity histogram for each individual array)
(Buechel et al., 2011) on at least two chips per group, for at least one group, were
retained (N = 12,785). The retained probesets were log2 transformed and analyzed using
a one-way ANOVA with three groups. A fold-change cutoff of 1.5 and an FDR of q =
0.05 (the p-value, corresponding to a q-value of 0.05, is 0.009) were used to create a list
of differently expressed genes (DEGs) (N = 828). Then, t-tests were applied to the list of
DEGs, and a protected Fisher’s least significant difference (LSD) was used to detect
significantly different changes in transcript abundance among the three pairwise
contrasts: JUV and PAED, JUV and MET, and PAED and MET. DEGs identified from ttests were classified into groups to investigate how gene expression changed between
juvenile and adult brains. MET and PAED expression values were evaluated relative to
JUV because it is likely that metamorphic and paedomorphic axolotls exhibit similar
patterns of gene expression as juveniles, but diverge after paedomorphic adults undergo
spontaneous metamorphosis. Operationally, this grouped genes according to statistical
significance and direction of change.
Identification of statistically enriched biological processes
To identify biological processes that were statistically enriched in the lists of
DEGs, overrepresentation analyses were conducted using PANTHER (Thomas et al.
2003). To generate a reference list, probesets were filtered to identify non-redundant
human NCBI gene symbols. Of the 12,785 probesets that exceeded the low expression
30 cutoff, 10,054 probesets annotated to human NCBI gene symbols; the list was further
reduced by removing redundant human NCBI gene symbols (N = 8,088). DEGs were
analyzed according to statistical significance and direction of change (see above). A
Bonferroni correction of α = 0.05 was used to identify overrepresented ontology terms.
Bioinformatic analysis of morph specific gene expression
DEGs identified from the JUV vs. MET and JUV vs. PAED contrasts were
compared to genes identified by Page et al. (2010). In that study, genes were identified as
commonly and differently expressed between A. mexicanum and A. tigrinum during
larval development (42-84 dph). Because the Page et al. (2010) study used a different
Affy GeneChip (Amby_001), a BLAST search was performed to identify probesets
between Amby_001 and Amby_002 that were designed to target the same transcript, and
thus could be reliably compared between the studies.
RESULTS
Global analysis of gene expression
Gene expression variation at the GeneChip level was examined using box plots,
principal components analysis (PCA), and hierarchical clustering. Box-plots were
generated to examine the overall quality of the data. These plots showed the distribution
of raw intensity values to be consistent among all GeneChips (Figure 2.1); thus, all
GeneChips were retained for RMA normalization. The normalized gene expression
values were then subjected to PCA (Figure 2.2) and hierarchical clustering (Figure 2.3).
Both of these analyses yielded the same result: replicate GeneChips within experimental
treatments grouped together, forming JUV, PAED, and MET groups. The first principal
component explained 98.6% of the variation in gene expression and the first and second
principal components resolved distinct groupings for JUV, PAED, and MET replicate
GeneChips. There was no evidence that female and males of the PAED group yielded
variable gene expression estimates. These results show that juveniles, metamorphs, and
paedomorphs are distinct in terms of overall gene expression.
31 Identification of differently expressed genes among juvenile and adult brains
A total of 828 probesets (i.e. DEGs) yielded statistically significant transcript
abundance estimates among the three brain types, and exceeded a 1.5 fold-change
threshold between any of the three pairwise contrasts (Figure 2.4). Approximately twice
as many DEGs were differently expressed between JUV and MET (N = 722) than
between JUV and PAED (N = 364). This indicates that metamorphosis is associated with
more unique brain transcriptional changes than paedomorphosis. There was considerable
overlap of DEGs between the JUV vs. MET and JUV vs. PAED contrasts; 271 of the 364
DEGs identified between JUV and PAED were also differently expressed between JUV
and MET. Thus, 74% of the differences observed between JUV and PAED can be
attributable to aging because they were also identified as significant between JUV and
MET. These results show that transcriptional patterns within the brains of adult axolotls
are associated with metamorphosis, paedomorphosis, and aging.
Classification of adult gene expression patterns relative to juvenile
To more finely resolve expression patterns, DEGs were classified into groups
(Figure 2.5) according to statistical significance and direction of change relative to JUV.
No DEGs exhibited significantly lower expression for MET and significantly higher
expression for PAED (LMHP) or significantly higher expression for MET and
significantly lower expression for PAED (HMLP); in other words, no genes showed
opposite directional changes in transcript abundance between MET and PAED. More of
the DEGs unique to MET decreased in transcript abundance (LM, N = 240; HM, N =
193). This pattern was also observed for genes that were differently expressed within
MET and PAED brains (LMP, N = 191; HMP, N = 98) and DEGs unique to PAED (LP,
N = 43; HP, N=32). Thus, overall, DEGs tended to show lower transcript abundances in
adults than juveniles, indicating that lower transcript abundance correlates with aging and
maturation.
To determine if PAED and MET expression showed the same pattern and
magnitude of divergence from JUV expression values, the 797 DEGs that were
significant relative to JUV were graphed as a bivariate plot and fitted using least squares
regression (Figure 2.6). Under a null hypothesis that PAED and MET expression is
32 equivalent for all DEGs, a best-fitting line through these data would have a slope of 1.0.
However, the line with the best fit has a positive slope, the value of which is not included
within the 95% confidence interval of the null model. The positive slope indicates that on
average, MET expression estimates deviated more from JUV expression estimates, than
did PAED. Indeed, 12 of 14 DEGs that deviated by three or more standard deviations
from predicted values of the best fitting line were classified as LM (Table 2.1). Thus, for
genes that were differently expressed between juveniles and adults, transcriptional
patterns were more similar between juveniles and paedomorphs, than between juveniles
and metamorphs.
Identification of statistically enriched biological processes
Overrepresentation analyses were conducted using PANTHER to identify
statistically enriched gene ontologies within groups listed in Figure 2.5. Pathway, protein
class, cellular component, molecular function, and biological process gene ontologies
were enriched for LM and LMP (Table 2.2). Only a few ontologies were identified as
enriched for HM (molecular function, biological process) and HMP (protein class,
biological process), and no significantly over-represented ontologies were identified for
LP and HP. The number of significantly enriched ontologies identified within groups
correlated with group size. This suggests that the over-representation analyses may have
been too statistically underpowered to identified DEGs within LP and HP groups. Thus,
for these groups, genes were screened by eye to identify candidates with functions
potentially associated with metamorphosis and paedomorphosis.
Over-representation of ontologies for LMP and HMP
To identify similarities in gene expression between paedomorphic and
metamorphic axolotls, the gene lists for LMP (lower expression) and HMP (higher
expression) were analyzed. Over-representation analysis for the LMP group identified
ontologies associated with DNA replication, cell cycle, and mitosis as significantly
enriched (Table 2.2). The genes that enriched these terms encode histones (e.g. cenpa,
hist1h2bj, hist2h2be) and other chromosome associated proteins (e.g. hmgn2, incep,
ncapd2, ncapd3, ncapg, smc2, tox3), cell cycle regulators (e.g. bub1, bub1b, ccna2,
33 ccnb1, ccnb3, ccni, cdc20), kinesins (e.g. aurka, aurkb, cdc2, cdkn1c, cks2, kif4a, kif11,
kif15, kif22, kifc1, plk1), DNA replication associated genes (e.g. mcm2, mcm3, mcm4,
mcm5, mcm7, pcna, pola), genes associated with cell fate (e.g. notch1, notch2), and
transcription factors (e.g hes6). Six extracellular matrix (ECM) genes (col2a1, col4a1,
col4a6, col5a1, col11a1, fn1) enriched ontologies associated with antibacterial response
and two embryonic hemoglobin genes (hbe1, hbz) enriched the blood circulation
ontology. Thus in comparison to juveniles, metamorphic and paedomorphic salamanders
expressed relatively fewer transcripts for genes associated with cell division, extracellular
matrix, and oxygen transport. Genes that were similarly up regulated in metamorphs and
paedomorphs (HMP) yielded lipid metabolism, hydroxylase, and oxygenase terms that
were enriched for cytochrome p450 genes (cyp2a6, cyp2a7, cyp2a13, cyp2f1, cyp7b1,
cyp39a1). These patterns suggest that aging in A. mexicanum is associated with
moderation of cell division, extracellular matrix restructuring, hemoglobin switching, and
lipid metabolism.
Over-representation of ontologies for LM and HM
The ontologies identified for genes that uniquely decreased in metamorphs (LM)
partially overlapped the list described above for LMP. Genes with extracellular functions,
and especially collagens (col1a1, col1a2, col3a1, col4a2, col4a5, col5a2, col6a1,
col6a2), enriched several gene ontologies in LM. Although this group of collagens was
not identified as significant between JUV and PAED, transcript abundances were
generally lower in PAED and thus showed the same direction of change as MET (Figure
2.6). Ontology terms associated with blood circulation, immune function, and cellular
and developmental processes were identified as enriched in LM. The blood circulation
term identified a fetal hemoglobin paralog (hbg2) as enriched within this group. Only two
ontologies were identified as enriched for genes that uniquely increased in metamorphic
brains (HM): system process and enzyme inhibitor activity. The system processes
included voltage gated potassium (kcnc2, kcng1) and calcium (cacna1d) ion channels,
neurotransmitter binding (gabra4, npy1r), synapse remodeling (nlgn4x), and two
somatostatin receptors involved in neurotransmission and endocrine signaling (sstr1,
sstr5). The enzyme inhibitor term was enriched by a group of protease inhibitors (cst6,
34 pkib, ppp1r14a, spint1), including a regulator of extracellular matrix remodeling (timp2).
Overall, the gene functions identified for LM suggest relatively lower expression of
extracellular matrix components and immune related genes in metamorphic brains
relative to juveniles and paedomorphs. While collagens generally decreased in abundance
as a function of aging, a greater diversity of collagens decreased significantly in
metamorphic brains. The genes that showed significantly higher abundances in
metamorphic brains (HM) suggest active inhibition of extracellular matrix remodeling
relative to juveniles and paedomorphs, and fundamental differences in neural and
endocrine signaling.
Genes that were significantly differently expressed in paedomorphs
The DEGs that showed significantly higher or lower expression in paedomorphs
(HP and LP) were screened by eye to identify functions associated with paedomorphic
development, as these gene lists did not yield significantly enriched terms. The LP list
contained genes associated with regulation of gene transcription (znf207, dot1l, and
nr2e1), nuclear transport (nup107), mRNA processing (hnrnpk), and ribosome biogenesis
(urb2). The HP list included genes associated with blood circulation and iron uptake
(fth1, hba2), transcriptional regulation (crem), and homocysteine metabolism (bhmt,
cubn). Additionally, genes associated with human disease were identified, including
Fanconi anemia (fancl), type II diabetes (iapp), neurodegenerative diseases - Alzheimer’s
disease (aplp2, pcmt1) and Sandhoff disease (hexb).
Bioinformatic analysis of morph specific gene expression
A previous study (Page et al., 2010) examined brain gene expression between
larval axolotls and a closely related metamorphic species, Ambystoma tigrinum. Genes
that were commonly and differently expressed between the species were identified during
the larval period, using regression modeling. Although Page et al. (2010) considered the
differently expressed genes to be associated with metamorphic and paedomorphic
development, they could not rule out the possibility that expression differences were
species-specific and not morph-specific. In contrast, the DEGs identified in this study are
more likely to be associated with metamorphic and paedomorphic development because
35 gene expression was examined within a common genetic background. Considering 59
probesets that could be reliably compared between the studies, 59% of these were
classified as paedomorphic, metamorphic, or expressed similarly between metamorphs
and paedomorphs. This included 21 genes that showed the same pattern of expression
between metamorphs and paedomorphs, 13 that were metamorph specific, and one that
was paedomorph specific. These results validate 14 genes as exhibiting morph specific
expression patterns that are detectable in larvae and adults (Table 4.3). The metamorph
specific genes have functions relating to alcohol metabolism (adh1b), chromatin
remodeling (anp32e), DNA repair (apex1), extracellular matrix (col1a2, nid2), RNA
binding (msi1), lipid and cholesterol metabolism (pltp, psap), ribosomal binding (rrbp1),
sulfatase (sulf1), T-cell lymphoma (tiam1), and astrocytes (gfap). The paedomorph
specific gene is a homocystine methyltransferase (bhmt). Of the probeset pairs that did
not agree, 18 of 24 were metamorph specific in Page et al. (2010), but expressed similarly
among metamorphs and paedomorphs in this study. These genes showed functions
related to protease inhibition (a2m), cell cycle regulation and mitosis (aurka, aurkb,
ccnb3, kif11, kif22, nusap1, pbk. zwint), basement membrane (col4a1), nuclear transport
(kpna2), DNA replication (rfc2), ribonuclease activity (rnaseh2a), and chromatin
structure (smarca5).
DISCUSSION
In this study, a custom microarray was used to compare gene expression among
the brains of juvenile, paedomorphic and metamorphic A. mexicanum. I compared
expression estimates from juveniles to estimates from paedomorphs and metamorphs to
identify genes that changed significantly during development. The genes that showed
statistically equivalent expression changes across metamorphic and paedomorphic brains
provide the first glimpse of aging in an amphibian. The genes that showed statistically
different expression estimates between metamorphic and paedomorphic brains provide
new functional insights about the maintenance and regulation of these different modes of
development. Below, I discuss the significance of the results, highlighting specific DEGs
and gene functions associated with aging, metamorphosis, and paedomorphosis.
36 Aging and transcription in the A. mexicanum brain.
Pairwise contrasts among the three groups revealed metamorphic and
paedomorphic axolotls to express many genes in common. These genes are discussed
within the context of aging below, though genes that were expressed in common between
metamorphic and paedomorphic axolotls relative to juvenile could be related to
maturational changes within the brain. For example, gonadotropin-releasing hormone
(gnrh1) showed a similar increase in expression in both adults relative to juveniles,
suggesting equivalence of reproductive physiology between metamorphs and
paedomorphs. If so, this supports the idea that axolotl paedomorphosis is an example of
neoteny and not progenesis (Gould, 1976; Voss and Shaffer, 1996).
Genes that showed increased expression in metamorphs and paedomorphs relative
to juveniles were enriched for cytochrome p450 genes involved in lipid and steroid
metabolism. It may be that the increase in cytochrome p450 genes in adult axolotls is
related to age-related changes in lipid metabolism, as has been described for mammals.
Changes in lipid concentration and distribution in the brain has been linked to
neurodegenerative diseases in humans (Ledesma, et al., 2012; van Vilet, 2012) and
increased expression of genes associated with lipid metabolism in the mouse brain are
associated with aging (Zahn et al., 2007). Additionally, cyp7b1, a cytochrome p450 gene
involved in neurosteroid metabolism within mammalian brain shows decreased
enzymatic activity with aging; this gene’s product decreases levels of active
glucocorticoids in vitro, though the in vivo effects are unknown (Yau & Seckl, 2012).
The neurosteroids produced by enzymatic activity of cyp7b1 and another cytochrome
p450 gene involved in neurosteroid metabolism – cyp39a1 – are involved in the
regulation of the immune response, cell proliferation and apoptosis in mouse and rat
(Lathe, 2002). Thus, changes in lipid metabolism and neurosteroid synthesis, relating to
specific cytochrome p450 transcript abundance, may be related to changes in the immune
system, as well as in cell growth and cell death.
Genes that showed lower expression in metamorphs and paedomorphs relative to
juveniles were associated with DNA metabolism, cell cycle regulation, and mitosis. In
humans, aging is associated with a global decrease in gene expression across all regions
37 of the brain and an increase in expression of genes related to cell death (Berchtold et al.,
2008); aging in mouse brain tissues is associated with an increase in genes involved in
cell cycle arrest (Zahn et al., 2007). Though metamorphs and paedomorphs did not show
a similar increase in genes directly related to cell death or cell cycle arrest, a general
decrease in genes related to cell proliferation was observed.
Berchtold et al. (2008) showed an increase in expression of genes associated with
angiogenesis, oxygen transport, and immune function due to aging in humans, and Zahn
et al. (2007) noted an increase in expression of genes associated with iron ion transport
and inflammatory responses associated with aging in mouse brain tissues. While blood
circulation and immune function were enriched in axolotls, the genes that enriched these
categories showed lower transcript abundances in adults. However, a B-cell associated
receptor (bcap29) showed higher expression in both morphs and a T-cell associated gene
(tiam1) showed higher expression in metamorphs only. Thus, while immune related
genes were generally lower in axolotl adults, morph specificity in expression varied for
specific genes. Morph specificity of gene expression related to immune function may be
related to differences in pathogens that a terrestrial metamorph would be likely to
encounter compared to an aquatic paedomorph. Additionally, paedomorphs showed
higher expression of ferritin (fth1), which is involved in iron transport and metamorphs
showed lower expression of collagen genes associated with angiogenesis, indicating that
the aging related gene expression noted in mammals may vary among axolotl morphs.
Such variation may correlate with specializations necessary for paedomorphs to live in
aquatic habitats and metamorphs to transition to more terrestrial habitats, in the adult
phase of life (Dodd & Dodd, 1976; Duellman & Trueb, 1986).
The immune related genes that showed lower expression in metamorphs and
paedomorphs included five collagen genes (col2a1, col4a1, col4a6, col5a1, col11a1) and
an additional extracellular matrix protein (fn1). This group included fibrillar (col2a1,
col5a1, col11a1) and basement membrane (col4a1, col4a6) collagens. Additionally, one
microfibrillar collagen (col6a3) also showed the same pattern of expression. Collagens
are one of the most abundant components of the extracellular matrix (ECM) and are
commonly known to aid in cell attachment, cell migration, and structural integrity of a
38 tissue, equipping tissues with specific mechanical and biochemical properties (Bateman
et al., 1996). However, cell-matrix interactions can be mediated by cell receptors on
ECM molecules and can also stimulate cell differentiation and regulate gene expression
levels (reviewed by Gelse et al., 2003; Heino, 2007). Collagens can bind some growth
factors and cytokines, and in particular, type I collagens have been shown to indirectly
block TGF-beta-action (Yamaguchi et al., 1990) and to bind integrins (Tulla et al., 2001).
Additionally, changes in collagen expression within brain vasculature are associated with
aging in humans (del Zoppo, 2012). Neurodegeneration and gene expression changes due
to aging have mostly been considered in humans and other mammals, and have not been
well studied in non-mammalian vertebrates (Jiang et al., 2001; Berchtold et al., 2008;
Cookson, 2012). My findings suggest that some collagen and extracellular matrix
proteins are similarly regulated in salamanders as a function of aging.
Genes associated with metamorphosis and paedomorphosis
Prior to my work, paedomorphic and metamorphic brain gene expression was
compared between two species – A. mexicanum and A. tigrinum. That study examined
changes in gene expression over time during larval development and metamorphosis in A.
tigrinum. It identified many significant DEGs that were unique to larval A. mexicanum
and larval A. tigrinum, as well as many commonly expressed genes between the two
species (Page et al., 2010). The results suggested axolotl development to be a mosaic of
ancestral and derived gene expression patterns, where some genes were expressed
similarly with A. tigrinum, though with respect to other genes, there were morph-specific
differences. The Page et al. (2010) study, however, could not differentiate genes that
showed unique expression due to species-specific differences from those that showed
unique expression due to metamorphic and paedomorphic development. My study has
better resolved candidate genes for metamorphic and paedomorphic development because
it related the early gene expression changes noted by Page et al (2010) between
metamorphic A. tigrinum and axolotl larvae, to metamorphic and paedomorphic axolotls.
The majority of the comparable genes between Page et al. (2010) and my study showed
agreement, with the same pattern of expression observed in both studies.
39 Eighteen genes that Page et al. (2010) classified as metamorph specific during the
larval period were expressed similarly among metamorphic and paedomorphic axolotls in
my study. Many of these genes are related to cell cycle regulation and mitosis (aurka,
aurkb, ccnb3, kif11, kif22, nusap1, pbk. zwint), chromatin structure and transcriptional
regulation (smarca5), DNA replication (rfc2), and ribonuclease activity (rnaseh2a).
These results suggest that some genes with similar functions in paedomorphs and
metamorphs are transiently, differently expressed during metamorphosis, when processes
associated with cell proliferation, transcription, and protein synthesis are necessary for
remodeling events (Page et al., 2009, 2010; Huggins et al., 2012).
Genes with extracellular matrix functions, and especially collagens (col1a1,
col1a2, col3a1, col4a2, col4a5, col5a2, col6a1, col6a2), were more lowly expressed in
metamorphs than juveniles and adults (Figure 4.7). Similar to the collagen genes that
showed significantly lower expression in both metamorphs and paedomorphs, these
collagens are a mix of fibrillar (col1a1, col1a2, col3a1, col5a2), basement membrane
(col4a2, col4a5), and microfibrillar (col6a1, col6a2) collagens. Relative to juvenile, both
morphs exhibit lower expression of collagen genes, but the deviation was greater for
metamorphs. Two extracellular matrix (ECM) proteins that were classified as metamorph
specific in Page et al. (2010) and my study are collagen type 1 alpha 2 (col1a2) and a
basement membrane protein, nidogen (nid2), suggesting that these two ECM proteins are
similarly regulated as a function of metamorphosis. Interestingly, metamorphs also
showed a higher expression of TIMP metallopeptidase inhibitor 2 (timp2), a peptidase
involved in the degradation of the extracellular matrix. These differences in gene
expression suggest changes in ECM structure and regulation as functions of both aging
and morph specific development.
Hormone receptor genes showed significantly higher expression in metamorphs
but not paedomorphs; these included somatostatin receptor 1 (sstr1), somatostatin
receptor 5 (sstr5), and nuclear receptor corepressor 1 (ncor1). Somatostatin receptors are
known to inhibit the release of hormones and other secretory proteins. Ncor1 is a ligandindependent repressor of TR and has been shown in mouse and rat to be involved in CRH
and TSH secretion in the pituitary, as well as the recognition of the HPT set point (van
40 der Laan et al., 2008; Astapova et al., 2011; Costa-e-Sousa et al., 2012). Additionally, a
metamorph specific gene validated between both Page et al. (2010) and my study,
ribosomal binding protein 1 (rrbp1), also showed significantly higher expression in
metamorphs only. This gene functions in the terminal differentiation of secretory cells
(Benyamini et al., 2009) and is also known to enhance in vitro collagen biosynthesis
(Ueno et al., 2010). The higher expression of these neuroendocrine related receptors and
a gene involved in secretory cell differentiation appear to be correlated with
metamorphosis and may influence differences in circulating hormone levels between
paedomorphs and metamorphs.
As previously mentioned, cytochrome p450 genes associated with lipid
metabolism appear to be similarly regulated due to aging, however, other genes
associated with lipid metabolism appear to be regulated differently between metamorphs
and paedomorphs. For example, two genes that are morph specific in both Page et al.
(2010) and my study are a phospholipid transfer protein (pltp) and a homocystine
methyltransferase (bhmt). Pltp is metamorph specific and shows higher expression in
metamorphs, but not paedomorphs. It is found in blood plasma and involved in
cholesterol metabolism (Albers et al., 2012). Bhmt is paedomorph specific and shows
higher expression in paedomorphs relative to juvenile. It has been shown to be involved
in lipid metabolism (Teng et al., 2011, 2012). Additionally, apolipoprotein H (apoh) is
expressed significantly lower in metamorphs relative to juvenile and is known to be
involved in cholesterol and lipid metabolism. These results indicate that some genes
involved in brain lipid metabolism appear to be regulated differently as a function of
aging and metamorphosis. Interestingly, Hervant et al. (2001) found that the obligate
cave dwelling, paedomorphic species Proteus anguinus exhibits a consistently lower
standing metabolic rate (SMR) over a 24-hour cycle than the facultative cave dwelling,
metamorphic species Euproctus asper. Additionally, two paedomorphic salamanders that
are not cave dwelling, Necturus maculosus and Amphium means (Light & Lowcock,
1991), exhibit similar SMR to P. anguinus, though the ability to compare these species
was called into question by Hervant et al. (2001). This suggests that there may be
differences in SMR between paedomorphic and metamorphic salamanders, though
differences between facultative and obligate cave dwelling salamanders may also play a
41 role. Therefore, an investigation of the possible differences between juvenile,
metamorphic, and paedomorphic axolotls, as well as between axolotl and tiger
salamanders, may help to elucidate differences in metabolism that are associated with
metamorphosis.
Metamorphs and paedomorphs also appear to regulate some genes related to
neurological processes differently as a function of metamorphosis. Recently, apoh has
been implicated in neurodegenerative processes, including Alzheimer’s disease, and low
levels of apoh have been correlated with cognitive decline in humans (Thambisetty et al.,
2010; Katzav et al., 2011; Song et al., 2012). Other neuronal associated genes that
showed significantly higher expression in metamorphs but not paedomorphs, included
GABA receptor (gabra4), neuropeptide Y receptor (npy1r), and neuroligin 4 (nlgn4x),
which is involved in nervous system remodeling. A nuclear phosphoprotein (anp32e) was
classified as metamorph specific in Page et al. (2010) and my study. It showed
significantly higher expression in metamorphs and is commonly involved in chromatin
remodeling, However, anp32e is thought to be involved in protein trafficking during
synaptogenesis through its regulatory effect on protein phosphatase 2A activity (Costanzo
et al., 2006). Metamorphs also show significantly lower expression of glial fibrillary
acidic protein (gfap), which is an intermediate filament protein of mature astrocytes and
may indicate that metamorphs have fewer astrocytes than similarly aged paedomorphs.
Paedomorphs exhibited significantly higher expression of genes associated with
neurodegenerative disorders in humans as well, including aplp2, hexb, iapp, and pcmt1.
Aplp2 is an amyloid precursor protein that is implicated in the pathogenesis of
Alzheimer’s disease and pcmt1 has been shown to negatively regulate amyloid beta
formation and may play a protective role in the pathogenesis of Alzheimer’s disease (Bae
et al., 2011). Additionaly, iapp encodes an amyloid-like protein. Hexb is associated with
Sandhoff disease and degrades GM2 ganglioside, an overabundance of which can lead to
neurodegeneration. Metamorphs appear to show higher expression of genes relating to
neural signaling and nervous system remodeling, while both metamorphs and
paedomorphs show morph specific regulation of genes related to neurodegeneration in
humans, possibly due to differences in the maintenance of neurological function related
to aging and metamorphosis. These expression results indicate that some genes associated
42 with neurological function and neurodegenerative disorders are not regulated similarly
due to aging only, but instead show expression that is both age and morph specific. These
expression differences between morphs may be related to physiological remodeling of
neural tissues as a result of metamorphosis.
Uromodulin (umod) and surfactant protein D (sftpd) both showed significantly
lower expression in metamorphs. Page et al. (2009) found that expression of umod
decreased in axolotl head skin during T4 induced metamorphosis. Surfactant genes were
also enriched among genes specific to metamorphs, and in particular, sftpd, showed down
regulation in the present study and the Page et al. (2010) study. Sftpd is expressed in
human brain, though it is primarily expressed in lung tissue, and may play a role in the
activated immune system (Madsen et al., 2000; Davé et al., 2004). Additionally, the sftpd
promoter can be activated by thyroid transcription factor-1 (ttf1) and abnormal regulation
of surfactant proteins has been implicated in brain-lung-thyroid syndrome, which
involves congenital hypothyroidism, neurological symptoms and respiratory distress in
humans (Guillot et al., 2010). These genes both appear to be good indicators of
metamorphosis – umod shows decreased expression in both brain and head skin of
metamorphic axolotls, and sftpd shows significant down regulation that correlates with
metamorphosis across axolotls and A. tigrinum. Aging, species, and morph specific differences in hemoglobin expression
Mammals and other vertebrates exhibit altered expression of hemoglobin subunits
at different stages of development (Palis, 2008; Dzierzak & Philipsen, 2013). Likewise,
axolotls are known to undergo a change in hemoglobin expression between 100-150 dph,
which is thought to be due to a cryptic metamorphosis (Tompkins & Townsend, 1977).
This shift in hemoglobin expression can also occur prematurely with T4 treatment that is
too low to result in morphological metamorphosis. The transcriptional analysis resulted in
the identification of morph specific hemoglobin expression changes that are most likely
due to gene expression changes within erythrocytes that were isolated from vasculature
within the brain tissue. Page et al. (2010) also identified differences in expression of
hemoglobin genes during larval and metamorphic development of A. tigrinum compared
to A. mexicanum, though species and morph specific differences in expression could not
43 be disentangled. My findings help to resolve differences in hemoglobin expression that
are associated with morph and species-specific variation.
In the Page et al. (2010) study, hemoglobin zeta (hbz) showed similar regulation
between A. tigrinum and axolotl larvae, and in my study hbz was regulated similarly
among metamorphs and paedomorphs. Thus, hbz expression is conserved among species
and morphs. Hemoglobin gamma (hbg) exhibited the same pattern of expression over
time in A. tigrinum compared to A. mexicanum (Page et al., 2010), though its expression
was consistently higher in A. tigrinum. Similarly, my study found hbg2 to exhibit lower
expression in both morphs relative to juveniles, though the difference in expression was
significant for metamorphs only. This indicates that lower expression of hbg2 is
associated with aging, but metamorphs and paedomorphs show differences in the
magnitude of expression during development.
Paralogs of hemoglobin alpha (hba)
showed both increasing and decreasing patterns of expression in A. tigrinum, but in larval
A. mexicanum, these genes showed increasing or flat patterns of expression (Page et al.,
2010). I also observed different patterns of expression between paralogs. One hba2
paralog showed significantly higher expression in paedomorphs but not metamorphs
(compared to juveniles), while the other hba2 paralog showed significantly decreased
expression in both morphs relative to juveniles. Thus, one paralog showed aging related
changes and the other showed changes attributable to aging, species, and morph.
Hemoglobin epsilon (hbe) showed higher levels of expression in A. mexicanum
throughout larval development compared to A. tigrinum (Page et al., 2010), but hbe1 was
similarly down regulated relative to juveniles in both metamorphic and paedomorphic
individuals in my study. The expression difference in this case seemingly implicates
species-specific variation in the timing of hemoglobin switching events that are
independent of metamorphosis. Overall, the complex regulation of hemoglobin loci
during development can only be deciphered using longitudinal experiments that are
capable of parsing variation attributable to species and morph. These differences in
hemoglobin expression, as well as enrichment of genes associated with angiogenesis that
showed lower expression in metamorphs relative to juvenile, may indicate that brain
vasculature differs with age and metamorphosis. An investigation of differences between
morph and age specific vascularization could indicate if this is indeed the case.
44 In summary, my study shows that metamorphosis and aging strongly affect the
transcriptional program of the brain in A. mexicanum. My study provides an initial
glimpse of gene expression related to aging in an amphibian. Biological processes
associated with aging were identified, including changes in lipid and neurosteroid
metabolism, cell proliferation, blood circulation, and immune function. My study also
identified morph specific gene expression patterns that are detectable during early
development and continue into adulthood, including differences in expression of
collagen, hemoglobin, hormone pathways and neurodegenerative associated gene
regulation. More significant DEGs exhibited higher or lower expression that were unique
to metamorphs (compared to juvenile), than those that were unique to paedomorphs. This
suggests that paedomorph brain gene expression is more similar to juvenile than
metamorph expression for genes are not commonly regulated due to aging. Future studies
should address possible differences in brain vasculature and metabolism between morphs,
as well as include longitudinal studies that would better clarify transient transcriptional
changes during metamorphosis across species from transcriptional changes due to aging
and maturation.
45 Table 2.1: Genes that deviated by three or more standard deviations (residual absolute value > 1.72) from values predicted by a linear
model fit to a bivariate plot of Paed and Met DEGs. LM = Lower Met, HM = Higher Met, LMP = Lower Met and Paed.
Probe Set ID
axo20166-r_s_at
axo20348-f_s_at
axo15318-f_at
axo22229-f_at
axo05346-f_at
axo20347-f_s_at
axo10397-f_at
axo27355-f_at
46 axo08766-f_at
axo02853-f_at
axo08393-r_at
axo10212-f_at
axo02258-f_at
axo06982-f_at
Gene
Symbol
HBG2
COL1A1
UMOD
--APOH
COL1A1
COL3A1
---
Gene Name
G-gamma globin [Homo sapiens]
alpha 1 type I collagen preproprotein [Homo sapiens]
uromodulin precursor [Homo sapiens]
apolipoprotein H precursor [Homo sapiens]
alpha 1 type I collagen preproprotein [Homo sapiens]
collagen type III alpha 1 preproprotein [Homo sapiens]
pulmonary surfactant-associated protein D precursor [Homo
SFTPD
sapiens]
ZNF605
zinc finger protein 605 [Homo sapiens]
serine (or cysteine) proteinase inhibitor, clade B (ovalbumin),
SERPINB5 member 5 [Homo sapiens]
COL1A2
alpha 2 type I collagen [Homo sapiens]
KRTCAP2 keratinocyte associated protein 2 [Homo sapiens]
baculoviral IAP repeat-containing protein 5 isoform 1 [Homo
BIRC5
sapiens]
Group
LM
LM
LM
LM
LM
LM
LM
LM
LM
LM
LM
LM
HM
LMP
47 Table 2.2: List of over-represented ontology terms, including the number of genes present for each term, and the associated p-value
for each term.
Significant
Ontology
Significant
Group
Over-represented Term
Gene List
Category
Gene List
(P-value)
cell cycle
48
6.10E-14
cellular process
76
4.99E-10
mitosis
25
5.93E-09
DNA replication
16
1.62E-07
chromosome segregation
14
5.96E-07
DNA metabolic process
17
5.20E-04
blood circulation
8
8.23E-04
cellular component organization
25
1.88E-03
regulation of liquid surface tension
5
2.19E-03
Biological
cytokinesis
9
6.94E-03
LMP
Process
developmental process
35
7.66E-03
ectoderm development
22
1.94E-02
defense response to bacterium
6
2.46E-02
nucleobase, nucleoside, nucleotide and nucleic acid metabolic
45
2.73E-02
process
establishment or maintenance of chromatin architecture
10
2.74E-02
asymmetric protein localization
5
2.81E-02
protein localization
5
2.81E-02
system development
25
3.18E-02
meiosis
8
3.89E-02
extracellular matrix
12
3.33E-03
Cellular
LMP
extracellular region
12
3.49E-03
Component
ribonucleoprotein complex
8
5.78E-03
Table 2.2 Continued
LMP
Molecular
Function
LMP
Pathway
LMP
Protein Class
48 LM
Biological
Process
extracellular matrix structural constituent
microtubule motor activity
structural molecule activity
DNA helicase activity
motor activity
Integrin signalling pathway
DNA binding protein
extracellular matrix structural protein
microtubule binding motor protein
surfactant
DNA helicase
extracellular matrix protein
antibacterial response protein
ribonucleoprotein
8
6
26
6
6
8
22
8
6
5
6
12
6
8
1.27E-04
1.62E-03
2.95E-03
9.14E-03
1.92E-02
2.84E-02
5.09E-05
1.54E-04
1.96E-03
2.31E-03
8.39E-03
1.67E-02
2.59E-02
2.89E-02
regulation of liquid surface tension
blood circulation
defense response to bacterium
cell-cell adhesion
cell adhesion
macrophage activation
homeostatic process
asymmetric protein localization
protein localization
ectoderm development
localization
mesoderm development
11
14
12
25
33
13
12
8
8
33
11
28
1.73E-10
2.26E-08
4.03E-07
1.59E-06
2.30E-06
3.91E-06
6.51E-05
1.58E-04
1.58E-04
2.23E-04
7.41E-04
8.64E-04
Table 2.2 Continued
49 LM
Biological
Process (Con't)
LM
Cellular
Component
LM
Molecular
Function
LM
Pathway
LM
Protein Class
HM
HM
Biological
Process
Molecular
Function
skeletal system development
receptor-mediated endocytosis
immune system process
developmental process
angiogenesis
response to stimulus
endocytosis
system process
system development
cellular process
extracellular matrix
extracellular region
extracellular matrix structural constituent
structural molecule activity
receptor activity
Integrin signalling pathway
surfactant
defense/immunity protein
cell adhesion molecule
antibacterial response protein
extracellular matrix protein
extracellular matrix structural protein
receptor
12
10
38
46
10
27
16
37
33
79
19
19
9
32
27
11
11
23
24
12
19
9
27
2.63E-03
4.10E-03
7.90E-03
9.35E-03
1.31E-02
1.67E-02
1.83E-02
2.21E-02
2.45E-02
3.59E-02
9.57E-06
1.03E-05
2.29E-04
1.20E-02
1.65E-02
3.03E-03
1.82E-10
1.93E-09
9.47E-08
4.24E-07
4.79E-05
2.77E-04
1.94E-02
system process
30
1.20E-02
enzyme inhibitor activity
10
3.77E-02
Table 2.2 Continued
Biological
HMP
Process
HMP
50 Protein Class
lipid metabolic process
oxygenase
hydroxylase
14
4.81E-03
8
5
1.76E-06
4.11E-04
Table 2.3: Comparison of DEGs from this study and Page et al. (2010). M = metamorph specific, P = paedomorph specific, MP =
shared between metamorph and paedomorph.
Expression Expression
Official Gene
Amby_001
Amby_002
pattern in
Pattern in
Sal-site V0.3 Contig
Symbol
Probeset
Probeset
Page et al.
current
(2010)
study
51 C6orf115
CDC2
CDC20
CENPV
COL2A1
FAAH
FXYD3
GDA
KIAA0101
KIFC1
LAMA1
MCM2
MCM3
PLK1
POLR2L
RCC1
RCC2
RRM1
RRM2
SEPT8
SFRP2
TMEM176B
SRV_05942_at
SRV_01180_at
SRV_00806_a_at
SRV_08376_at
SRV_01205_at
SRV_00949_at
SRV_02740_a_at
SRV_02137_at
SRV_03588_at
SRV_06485_a_at
SRV_02580_a_at
SRV_09459_at
SRV_01414_at
SRV_02420_at
SRV_04514_at
SRV_00812_a_at
SRV_04287_at
SRV_00703_a_at
SRV_00706_a_at
SRV_04195_at
SRV_09416_at
SRV_03419_at
axo18117-r_at
axo07587-r_at
axo07048-f_at
axo02633-f_at
axo07638-f_at
axo07273-r_at
axo06274-f_at
axo09923-r_at
axo21059-f_at
axo08033-f_at
axo11181-f_at
axo10169-r_at
axo08153-f_at
axo10701-f_at
axo18021-f_at
axo04428-f_at
axo06321-f_at
axo00677-r_at
axo00701-r_at
axo05007-r_at
axo08757-f_at
axo05154-f_at
contig79311
contig24691
contig07123
contig21590
contig81592
contig45884
contig32491
contig79334
Mex_NM_014736_Contig_2
contig08742
contig61801
contig03405
contig07339
contig64493
contig23009
contig32974
contig84321
contig04610
contig86260
contig97712
contig94308
contig44794
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
P
MP
MP
MP
MP
MP
MP
MP
MP
M
MP
MP
MP
M
MP
MP
52 Table 2.3 Continued
TMPO
SRV_01795_at
TYROBP
SRV_01821_at
UHRF1
SRV_03362_at
A2M
SRV_00116_at
ADH1B
SRV_00457_at
ANP32E
SRV_04811_a_at
APEX1
SRV_05165_a_at
AURKA
SRV_01907_at
AURKB
SRV_02115_at
CCNB3
SRV_05031_at
COL1A2
SRV_09970_at
COL4A1
SRV_01207_s_at
CRYAB
SRV_01225_at
DLGAP5
SRV_03593_at
GFAP
SRV_01304_at
KIF11
SRV_02235_at
KIF22
SRV_03220_at
KPNA2
SRV_01354_a_at
MSI1
SRV_10183_at
NID2
SRV_03227_at
NUP107
SRV_04423_a_at
NUSAP1
SRV_04253_a_at
PBK
SRV_04270_at
PLTP
SRV_07538_at
PSAP
SRV_01558_at
RFC2
SRV_05574_at
RNASEH2A
SRV_02906_a_at
axo01770-f_at
axo02980-f_at
axo13318-f_at
axo00006-f_at
axo07742-r_at
axo19462-f_at
axo00384-f_at
axo03050-f_at
axo09856-f_at
axo20258-f_at
axo10212-f_at
axo07643-f_at
axo07684-f_at
axo06865-f_at
axo07863-r_at
axo10165-f_at
axo12862-f_at
axo08037-f_at
axo08202-f_at
axo12902-f_at
axo17535-f_at
axo16931-r_at
axo16972-r_at
axo11843-r_at
axo04317-f_at
axo02595-r_at
axo12035-f_at
contig40054
contig03910
contig63025
contig97275
contig08448
contig77943
contig40552
contig06048
contig88180
contig86908
contig81602
contig84121
contig02235
contig08508
contig21475
contig84891
contig02837
contig93141
contig04157
contig82626
contig11105
contig05609
contig61944
contig08790
contig64611
contig05592
contig28679
MP
MP
MP
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
MP
MP
MP
M
M
M
MP
MP
MP
M
MP
MP
MP
M
MP
MP
MP
M
M
P
MP
MP
M
M
MP
MP
Table 2.3 Continued
RRBP1
SRV_02269_a_at
SFTPD
SRV_07632_at
SMARCA5
SRV_01909_at
SULF1
SRV_03675_at
TIAM1
SRV_01791_at
TM6SF1
SRV_04650_a_at
TPX2
SRV_03256_at
ZWINT
SRV_05024_at
BHMT
SRV_01149_at
CAPNS1
SRV_01164_a_at
53 axo04353-f_at
axo08766-f_at
axo09291-f_at
axo05681-f_at
axo08977-r_at
axo18613-r_at
axo12960-f_at
axo12663-r_at
axo07527-f_at
axo07565-r_at
contig03012
contig45505
contig20796
contig38682
contig28568
contig07019
contig13550
contig09898
contig38970
contig74383
M
M
M
M
M
M
M
M
P
P
M
M
MP
M
M
MP
MP
MP
P
MP
Figure 2.1: Box plots, generated prior to RMA correction, of log2 microarray signal
intensity. Juvenile microarray chips: 454-11, 454-12, 454-13. Adult Metamorph
microarray chips: 454-1, 454-2, 454-3. Adult Paedomorph microarray chips: 454-6
(female), 454-7 (male), 454-8 (male).
54 C11
C13
C12
B6
B7
B8
A3
A2
A1
Figure 2.2: Results from principle components analysis. Juvenile microarray chips are
blue (C11 = 454.11, C12 = 454.12, C13 = 454.13), Metamorph microarray chips are red
(A1 = 454.1, A2 = 454.2, A3 = 454.3), and Paedomorph microarray chips are green (B6
= 454.6 (female), B7 = 454.7 (male), B8 = 454.8 (male)).
55 $
$
5HSRUW8QWLWOHG
$
%
+LHUDUFKLFDO&OXVWHULQJ
%
0HWKRG
% )DVW:DUG
3
'HQGURJUDP
&
&
&
$
$
$
%
%
%
&OXVWHULQJ+LVWRU\
1XPEHURI
&OXVWHUV
'LVWDQFH
/HDGHU
%
&
%
&
$
$
$
&
-RLQHU
%
&
%
&
$
$
%
$
&OXVWHULQJ+LVWRU\
Figure 2.3:
Results of hierarchical clustering analysis. Microarray chips cluster according
1XPEHURI
to their sample
groups. Juvenile
microarray/HDGHU
chips are
C11 = 454.11, C12 = 454.12, C13 =
&OXVWHUV
'LVWDQFH
-RLQHU
454.13. Metamorph
microarray
chips are
% A1 =%454.1, A2 = 454.2, A3 = 454.3.
Paedomorph microarray
chips
are
B6
=
454.6
& (female),
& B7 = 454.7 (male), B8 = 454.8
(male). The clustering
history states the distance
between
the leader and the joiner (i.e.
%
%
between two GeneChips)
where distance&
is represented
as a function of the squared
&
$
$
Euclidian distance.
$
$
&
56 $
%
$
Figure 2.4: Venn diagram showing overlap of the 828 DEGs between the three different
pair-wise comparisons – JUV and MET, JUV and PAED, and PAED and MET.
57 %#""
&'"(")*+,""
%!""
&'"("*-,""
!"
$"
#"
$"
#"
!"
.#""
&'"("/0-,""
.!"
&'"("-),""
#"
$"
!"
$"
!"
.#!"
&'"("01,"
#"
%#!"
&'"("/0/,""
!"
#"
$"
#"
!"
$"
%#.!"
&'"("+,"" !"
.#%!"
&'"("+,"" #"
$"
$"
#"
!"
Figure 2.5: Patterns of gene expression changes for PAED and MET relative to JUV
expression. LM = significantly lower expression for Met, but not Paed; LP = significantly
lower expression for Paed, but not Met; HM = significantly higher expression for Met,
but not Paed; HP = significantly higher expression for Paed, but not Met; HMP =
significantly higher expression for Met and Paed; LMP = significantly lower expression
for Met and Paed; LMHP = significantly lower expression for Met and significantly
higher expression for Paed; HMLP = significantly higher expression for Met and
significantly lower expression for Paed.
58 59 Figure 2.6: Bivariate plot of the mean PAED and mean MET expression relative to JUV for the 797 DEGs significant compared to
JUV. The green line is the expected best fitting line (y = x) and the red line is the best-fit line (y = 1.2640142x – 0.086525) for the
data. The colored data points correspond to the seven groups as follows, purple is LMP; dark blue is LM; light blue is LP; yellow is
HP; orange is HM; and red is HMP. Data points represented by r deviate by > three standard deviations from the best-fit line for the
data (red line). The fourteen corresponding probesets are listed in Table 4.1.
1
60 Log2 Fold Change (relative to Juvenile)
0
-1
-2
-3
juv
paed
-4
met
-5
-6
Gene
Figure 2.7: Bar graph of significant collagen gene expression relative to JUV for seventeen significant probsets that annotate to
fourteen collagen genes. Probesets that annotate to the same gene show the same pattern of expression between JUV, PAED, and
MET. LM = significantly lower in MET only; LMP = significantly lower in MET and PAED.
REFERENCES
Albers, J. J., Vuletic, S., Cheung, M. C. (2012). Role of plasma phospholipid transfer
protein in lipid and lipoprotein metabolism. Biochimica et Biophysica Acta, 1821,
345-357.
Arbeitman, M. N., Furlong, E. E. M., Imam, F., Johnson, E., Null, B. H., Baker, B. S., et
al. (2002). Gene expression during the life cycle of Drosophila melanogaster.
Science, 297, 2270-2275
Astapova, I., Vella, K. R., Ramadoss, P., Holtz, K. A., Rodwin, B. A. Liao, X. H., et al.
(2011). The nuclear receptor corepressor (NCoR) controls thyroid hormone
sensitivity and the set point of the hypothalamic-pituitary-thyroid axis. Molecular
Endocrinology, 25, 212-224.
Armstrong, J. B., Duhon, S. T., & Malacinski, G. M. (1989). Raising the axolotl in
captivity. In J. B. Armstrong & G. M. Malacinski (Eds.), Developmental biology
of the axolotl (pp. 220–227). New York: Oxford University Press.
Bae, N., Byeon, S. E., Song, J., Lee, S. J., Kwon, M., Mook-Jung, I., et al. (2011).
Knock-down of protein L-isoaspartyl O-methyltransferase increases β–amyloid
production by decreasing ADAM10 and ADAM17 levels. Acta Pharmacologica
Sinica. 32, 288-294.
Bateman, J. F., Lamande, S. R., Ramshaw, J. A. M. (1996). Collagen superfamily. In W.
D. Comper (Ed.), Extracellular Matrix (pp. 22-67). Melbourne: Harwood
Academic Press.
Benyamini, P., Webster, P., Meyer, D. I. (2009). Knockdown of p180 eliminates the
terminal differentiation of a secretory cell line. Molecular Biology of the Cell, 20,
732-744.
Berchtold, N. C., Cribbs, D. H., Coleman, P. D., Rogers, J., Head, E. Kim, R. et al.
(2008). Gene expression changes in the course of normal brain aging are sexually
dimorphic. Proceedings of the National Academy of Science of the United States
of America, 105, 15605-15610.
Blount, R. F. (1950). The effects of heteroplastic hypophyseal grafts upon the axolotl,
Ambystoma mexicanum. The Journal of Experimental Zoology, 113, 717–739.
Brown, D. D., & Cai, L. (2007). Amphibian metamorphosis. Developmental Biology,
20–33.
Bruce, R. C. (2003). Life histories. In D. M. Sever (Ed.), Reproductive biology and
phylogeny of Urodela (pp. 477–525). Enfield, NH: Science Publishers, Inc
61 Buchholz, D. R., Tomita, A., Fu, L., Paul, B. D., & Shi, Y. B. (2004). Transgenic analysis
reveals that thyroid hormone receptor is sufficient to mediate the thyroid hormone
signal in frog metamorphosis. Molecular and Cellular Biology, 24, 9026–9037.
Buchholz, D. R., Paul, B. D., Fu, L. Z., Shi, Y.B. (2006). Molecular and developmental
analyses of thyroid hormone receptor function in Xenopus laevis, the African
clawed frog. General and Comparative Endocrinology, 145, 1-19.
Buechel, H.M., Popovic, J., Searcy, J.L., Porter, N.M., Thibault, O., et al. (2011). Deep
Sleep and pariental cortex gene expression changes are related to cognitive
deficits with age. PLoS ONE 6, e18387.
Buscaglia, M., Leloup, J., & De Luze, A. (1985). The role and regulation of
monodeiodination of thyroxine to 3,5,3’-triiodothyronine during amphibian
metamorphosis. In M. Balls & M. Bownes (Eds.) Metamorphosis (pp. 273–293).
Oxford, UK: Oxford University Press.
Caria, M. A., Dratman, M. B., Kow, L. M., Mameli, O., & Pavlides, C. J. (2008). Thyroid
hormone action: Non-genomic modulation of neuronal excitability in the
hippocampus. Journal of Neuroendocrinology, 21, 98–107.
Charipper, H. A. (1929). Studies on amphibian endocrines. V. Accelerated
metamorphosis of Rana clamitans larvae by means of implants of fresh anterior
pituitary from adult Necturus maculosus. The Anatomical Record, 44, 117–141.
Cheng, S. Y., Leonard, J. L., & Davis, P. J. (2010). Molecular aspects of thyroid hormone
actions. Endocrine Reviews, 31, 139–170.
Connors, K. A., Korte, J. J., Anderson, G. W., & Degitz, S. J. (2010). Characterization of
thyroid hormone transporter expression during tissue-specific metamorphic events
in Xenopus tropicalis. General and Comparative Endocrinology, 168, 149–159.
Cookson, M. R. (2012). Evolution of neurodegeneration. Current Biology, 22, R753-761.
Costa-e-Sousa, R. H., Astapova, I., Ye, F., Wondisford, F. E., Hollenberf, A. N. (2012).
The thyroid axis is regulated by NCoR1 via its actions in the pituitary.
Endocrinology, 153, 5049-5057.
Costanzo, R. V., Vila-Ortiz, G. J., Perandones, C., Carminatti, H., Mantilla, A.,
Randrizzani, M. (2006). Anp32e/Cpd1 regulates protein phosphatase 2A activity
at synapses during synaptogenesis. European Journal of Neuroscience, 32, 309324.
Darras, V. M., & Kuhn, E. R. (1983). Effects of TRH, bovine TSH, and pituitary extracts
on thyroidal T4 release in Ambystoma mexicanum. General and Comparative
Endocrinology, 51, 286–291.
62 Darras, V. M., Van der Geyten, S., Cox, C., Segers, I. B., De Groef, B., & Ku¨ hn, E. R.
(2002). Effects of dexamethasone treatment on iodothyronine deiodinase
activities and on metamorphosis-related morphological changes in the axolotl
(Ambystoma mexicanum). General and Comparative Endocrinology, 127, 157–
164.
Davé, V., Childs, T., Whitsett, J. A. (2004). Nuclear factor of activated T cells regulates
transcription of the surfactant protein D gene (sftpd) via direct interaction with
thyroid transcription factor-1 in lung epithelial cells. Journal of Biological
Chemisty, 279, 34578-34588.
Davis, P. J., Leonard, J. L., & Davis, F. B. (2008). Mechanisms of nongenomic actions of
thyroid hormone. Frontiers in Neuroendocrinology, 29, 211–218.
Del Zoppo, G. J. (2012). Aging and the neurovascular unit. Annals of the New York
Academy of Sciences, 1269, 127-133.
Denoel, M., Ivanovic, A., Dzukic, G., & Kalezic, M. L. (2009). Sexual size dimorphism
in the evolutionary context of facultative paedomorphosis: Insights from
European newts. BMC Evolutionary Biology, 9, 278.
Denoel, M., & Joly, P. (2000). Neoteny and progenesis as two heterochronic processes
involved in paedomorphosis in Triturus alpestris (Amphibia: Caudata).
Proceedings of the Royal Society of London, Series B: Biological Sciences, 267,
1481–1485.
Denoel, M., Joly, P., & Whiteman, H. H. (2005). Evolutionary ecology of facultative paedomorphosis in newts and salamanders. Biological Reviews, 80, 663–671.
Denoel, M., & Poncin, P. (2001). The effect of food on growth and metamorphosis of
paedomorphs in Triturus alpestris apuanus. Archiv fu¨r Hydrobiologie, 152, 661–
670.
Dent, J. N. (1968). Survey of amphibian metamorphosis. In W. Etkin & L. I. Gilbert
(Eds.), Metamorphosis: A problem in developmental biology (pp. 271–311). New
York: Appleton-Century-Crofts.
Denver, R. J., Glennemeier, K. A., & Boorse, G. C. (2002). Endocrinology of complex
life cycles: Amphibians. In D. W. Pfaff, A. P. Arnld, A. M. Etgen, S. E. Fahrbach
& R. T. Rubin (Eds.), Hormones, brain and behavior (pp. 469–513). San Diego:
Elsevier.
Denver, R. J., & Licht, P. (1989). Neuropeptide stimulation of thyrotropin secretion in the
larval bullfrog: Evidence for a common neuroregulator of thyroid and interrenal
activity in metamorphosis. The Journal of Experimental Zoology, 252, 101–104.
Dodd, M. H. I., & Dodd, J. M. (1976). The biology of metamorphosis. In B. Lofts (Ed.),
Physiology of amphibia, Vol. 3, (pp. 467–599). New York: Academic Press.
63 Doyle, J. M., & Whitemann, H. H. (2008). Paedomorphosis in Ambystoma talpoideum:
Effects of initial body size variation and density. Oecologia, 156, 87–94.
Duellman, W. E., & Trueb, L. (1986). Biology of amphibians. New York: McGraw-Hill.
Dumeril, A. (1870). Quatrieme notice sur la menagerie des reptiles du Museum
d’Histoire Naturelle. Nouvelles Archives du Museum D’Histoire Naturelle De
Paris, 5, 47–60.
Dundee, H. A. (1957). Partial metamorphosis induced in Typhlomolge rathbuni. Copeia,
1957, 52–53.
Eagleson, G. W., & McKeown, B. A. (1980). T4-induced metamorphosis in Ambystoma
gracile (Baird) from two populations: The effects of aging and temperature.
Biochemistry and Experimental Biology, 16, 111–126.
Emerson, E. T. (1905). General anatomy of Typhlomolge rathbuni. Proceedings of the
Boston Society of Natural History, 32, 43–76.
Fox, H. (1984). Amphibian morphogenesis. Clifton, NJ: The Humana Press, Inc.
Fujiwara, K., Jindatip, D., Kikuchi, M., Yashiro, T. (2010). In situ hybridization reveals
that type I and III collagens are produced by pericytes in the anterior pituitary
gland of rats. Cell and Tissue Research, 342, 491-495.
Furuya, F., Lu, C., Guigon, C. J., & Cheng, S. Y. (2009). Nongenomic activation of phosphatidylinositol 3-kinase signaling by thyroid hormone receptors. Steroids, 74,
628–634.
Galton, V. A. (1985). 3,5,3’-triiodothyronine receptors and thyroxine 5’ monodeiodinating activity in thyroid hormone-insensitive Amphibia. General and
Comparative Endocrinology, 57, 465–471.
Galton, V. A. (1991). Thyroid hormone receptors and iodothyronine deiodinases in the
developing Mexican axolotl, Ambystoma mexicanum. General and Comparative
Endocrinology, 85, 62–70.
Galton, V. A. (1992). The role of thyroid hormone in amphibian metamorphosis. Trends
in Endocrinology and Metabolism, 3, 96–100.
Gancedo, B., Corpas, I., Alonso-Gomez, A. L., Delgado, M. J., Morreale de Excobar, G.,
& Alonso-Bedate, M. (1992). Corticotropin-releasing factor stimulates
metamorphosis and increases thyroid hormone concentration in prometamorphic
Rana perezi larvae. General and Comparative Endocrinology, 87, 6–13.
Gelse, K., Pöschl, E., Aigner, T. (2003). Collagens–structure, function, and biosynthesis.
Advanced Drug Delivery Review, 55, 1531–1546.
64 Gilbert, L. I., Tata, J. R., & Atkinson, B. G. (Eds.), (1996). Metamorphosis:
Postembryonic repro- gramming of gene expression in amphibians and insect
cells. San Diego: Academic Press.
Goplen, D., Bougnaud, S., Rajcevic, U., Boe, S. O., Skaftnesmo, K. O., Voges, J., et al.
(2010). αB-crystallin is elevated in highly infiltrative apoptosis-resistant
glioblastoma cells. The American Journal of Pathology, 177, 1618-1628.
Gorbman, A. (1957). The thyroid gland of Typhlomolge rathbuni. Copeia, 1957, 41–43.
Gould, S. J. (1977). Ontogeny and phylogeny. Cambridge, UK: Belkap Press.
Gould, S. J. (1981). Change in developmental timing as a mechanism of macroevolution.
In J. T. Bonner (Ed.), Evolution and development (pp. 333–346). Berlin,
Germany: Springer- Verlag.
Grant, M. P. (1930). The release of follicular colloid from the thyroid of Necturus
maculosus following heteroplastic anterior-pituitary implants. The Anatomical
Record, 46, 205–221.
Graue, W. V. (1998). Estudio gene´tico y demogra´fico de la poblacion del amfibio
Ambystoma mexicanum (Caudata: Ambystomaidae) del Lago de Xochimilco.
Mexico: Instituto de Ciencias del Mar y Limnologia, UNAM.
Gudernatsch, J. F. (1912). Feeding experiments in tadpoles: I. The influence of specific
organs given as food on growth and differentiations: A contribution to the
knowledge of organs with internal secretion. Archiv fu¨r Entwicklungsmechanik
der Organismen, 35, 457–483.
Guillot, L., Carre, A., Szinnai, G., Castanet, M., Tron, E., Jaubert, F., et al. (2010).
NKX2-1 mutations leading to surfactant protein promoter dysregulation cause
interstitial lung disease in “Brain-Lung-Thyroid Syndrome”. Human Mutations,
31, E1146-1162.
Hagemann, T. L., Boelens, W. C., Wawrousek, E. F., Messing, A. (2009). Suppression of
GFAP toxicity by alphaB-crystallin in mouse models of Alexander disease.
Human Molecular Genetics, 18, 1190-1199.
Hanon, E. A., Lincoln, G. A., Fustin, J.-M., Dardente, H., Masson-Pevet, M., Morgan, P.
J., et al. (2008). Ancestral TSH mechanism signals summer in a photoperiodic
mammal. Current Biology, 18, 1147–1152.
Harris, R. N. (1987). Density-dependent paedomorphosis in the salamander
Notophthalamus viridescens dorsalis. Ecology, 68, 705–712.
Hayes, T., Chan, R., & Licht, P. (1993). Interactions of temperature and steroids on larval
growth, development, and metamorphosis in a toad (Bufo boreas). The Journal of
Experimental Zoology, 266, 206–215.
65 Heino, J. (2007). The collagen family members as cell adhesion proteins. Bioessays, 29,
1001–1010.
Hennemann, G., Docter, R., Friesema, E. C., De Jong, M., Krenning, E. P., & Visser, T.
J. (2001). Plasma membrane transport of thyroid hormones and its role in thyroid
hormone metabolism and bioavailability. Endocrine Reviews, 22, 451–476.
Hervant, F., Mathieu, J., Durand, J. P. (2001). Circadian rhythmicity, respiration and
behavior in hypogean and epigean salamanders. Natura Croatica, 10, 141-152.
Huggins, P., Johnson, C. K., Schoergendorfer, A., Putta, S., Bathke, A. C., Stromberg, A.
J., et al. (2012). Identification of differentially expressed thyroid hormone
responsive genes from the brain of the Mexican Axolotl (Ambystoma
mexicanum). Comparative Biochemistry and Physiology Part C: Toxicology, 155,
128–135.
Huxley, J. S., & Hogben, L. T. (1922). Experiments on amphibian metamorphosis and
pigment response in relation to internal secretions. Proceedings of the Royal
Society B, 93, 36–53.
Irizarry, R. A., Hobbs, B., Collin, F., Beazer-Barclay, Y. D., Antonellis, K. J., Scherf, U.
et al. (2003). Exploration, normalization, and summaries of high density
oligonucleotide array probe level data. Biostatistics, 4, 249-264.
Jiang, C. H., Tsien, J. Z., Schultz, P. G., Hu, Y. (2001). The effects of aging on gene
expression in the hypothalamus and cortex of mice. Proceedings of the National
Academy of Science of the United States of America, 98, 1930-1934.
Katzav, A., Faust-Socher, A., Kvapil, F., Michaelson, D. M., Blank, M., Pick, C. G., et al.
(2011). Antiphospholipid syndrome induction exacerbates a transgenic Alzheimer
disease model on a female background. Neurobiology of Aging, 32, 272–279.
Kezer, J. (1952). Thyroxine-induced metamorphosis of the neotenic salamanders Eurycea
tynerensis and Eurycea neotenes. Copeia, 1952, 234–237.
Krenz, J. D., & Server, D. M. (1995). Mating and ovoposition in paedomorphic
Ambystoma talpoideum preceeds the arrival of terrestrial males. Herpetologica,
51, 387–393.
Kuchenbauer, F., Hopfner, U., Stalla, J., Arzt, E., Stalla, G. K., Paez-Pereda, M. (2001).
Extracellular matrix components regulate ACTH production and proliferation in
corticotroph tumor cells. Molecular and Cellular Endocrinology, 175, 141-148.
Kuhn, E. R., De Groef, B., Van der Geyten, S., & Darras, V. M. (2005). Corticotropinreleasing hormone-mediated metamorphosis in the neotenic axolotl Ambystoma
mexicanum: Synergistic involvement of thyroxine and corticoids on brain type II
deiodinase. General and Comparative Endocrinology, 143, 75–81.
66 Larsen, J. H. (1968). Ultrastructure of thyroid follicle cells of three salamanders
(Ambystoma, Amphiuma, and Necturus) exhibiting varying degrees of neoteny.
Journal of Ultrastructure Research, 24, 190–209.
Larson, A., Weisrock, D. W., & Kozak, K. H. (2003). Phylogenetic systematics of
salamanders (Amphibia:Urodela), a review. In D. M. Sever (Ed.), Reproductive
biology and phylogeny of Urodela (pp. 31–108). Enfield, NH: Science
Publishers, Inc.
Lathe, R. (2002). Steroid and sterol 7-hydroxylation: ancient pathways. Steroids, 67, 967977.
Laudet, V. (2011). The origins and evolution of vertebrate metamorphosis. Current
Biology, 21, R726–R737.
Ledesma, M. D., Martin, M. G., Dotti, C. G. (2012). Lipid changes in the aged brain:
effect on synaptic function and neuronal survival. Progress in Lipid Research, 51,
23-25.
Light, L. E., & Lowcock, L. A. (1991). Genome size and metabolic rate in salamanders.
Comparative Biochemistry and Physiology, 100B, 83-92.
Lynn, W. G., & Wachowski, H. E. (1951). The thyroid gland and its functions in coldblooded vertebrates. The Quarterly Review of Biology, 26, 123-168.
Madsen, J., Kliem, A., Tornoe, I., Skjodt, K., Koch, C., Holmskov, U. (2000).
Localization of lunch surfactant protein D on mucosal surfaces in human tissues.
Journal of Immunology, 164, 5866-5870.
Manzon, R. G., & Denver, R. J. (2004). Regulation of pituitary thyrotropin gene
expression during Xenopus metamorphosis: Negative feedback is functional
throughout metamorphosis. The Journal of Endocrinology, 182, 273–285.
Melcangi, R. C., Magnaghi, V., Cavarretta, I., Riva, M. A., Piva, F., Martini, L. (1998).
Effects of steroid hormones on gene expression of glial markers in the central and
peripheral nervous system: variations induced by aging. Experimental
Gerontology, 33, 827-836.
Nakao, N., Ono, H., Yamamura, T., Anraku, T., Takagi, T., Higashi, K., et al. (2008).
Thyrotrophin in the pars tuberalis triggers photoperiodic response. Nature, 452,
317–322.
Norman, M. F., Carr, J. A., & Norris, D. O. (1987). Adenohypophysial-thyroid activity of
the tiger salamander, Ambystoma tigrinum, as a function of metamorphosis and
captivity. The Journal of Experimental Zoology, 242, 55–66.
Norris, D. O., & Gern, W. A. (1976). Thyroxine-induced activation of hypothalamohypophysial axis in neotenic salamander larvae. Science, 194, 525–527.
67 Norris, D. O., & Platt, J. E. (1973). Effects of pituitary hormones, melatonin, and
thyroidal inhibitors on radioiodide uptake by the thyroid glands of larval and adult
tiger salamanders, Ambystoma tigrinum (Amphibia: Caudata). General and
Comparative Endocrinology, 21, 368–376.
Nugent, B. M., Tobet, S. A., Lara, H. E., Lucion, A. B., Wilson, M. E., Recabarren, S. E.,
et al. (2012). Hormonal programming across the lifespan. Hormone Metabolism
Research, 44, 577-586.
Page, R. B., & Voss, S. R. (2009). Induction of metamorphosis in axolotls (Ambystoma
mexicanum). Cold Spring Harbor Protocols.
Page, R. B., Boley, M. A., Smith, J. J., Putta, S., & Voss, S. R. (2010). Microarray
analysis of a salamander hopeful monster reveals transcriptional signatures of
paedomorphic brain development. BMC Evolutionary Biology, 10, 199.
Page, R. B., Monaghan, J. R., Samuels, A. K., Smith, J. J., Beachy, C. K., & Voss, S. R.
(2007). Microarray analysis identifies keratin loci as sensitive biomarkers for
thyroid hormone disruption in salamanders (Ambystoma).
Comparative
Biochemistry and Physiology Part C: Toxicology and Pharmacology, 145, 15–27.
Page, R. B., Monaghan, J. R., Walker, J. A., & Voss, S. R. (2009). A model of
transcriptional and morphological changes during thyroid hormone-induced
metamorphosis of the axolotl. General and Comparative Endocrinology, 162,
219–232.
Page, R. B., Voss, S. R., Samuels, A. K., Smith, J. J., Putta, S., & Beachy, C. K. (2008).
Effect of thyroid hormone concentration on the transcriptional response
underlying induced metamorphosis in the Mexican axolotl (Ambystoma). BMC
Genomics, 9, 78.
Power, D. M., Llewellyn, L., Faustino, M., Nowell, M. A., Bjornsson, B. T., Einarsdottir,
I. E., et al. (2001). Thyroid hormones in growth and development of fish.
Comparative
Biochemistry and Physiology. Part C: Toxicology and
Pharmacology, 130, 447–459.
Prahlad, K. V. (1968). Induced metamorphosis: Rectification of a genetic disability by
thy- roid hormone in the Mexican axolotl, Siredon mexicanum. General and
Comparative Endocrinology, 11, 21–30.
Prahlad, K. V., & DeLanney, L. E. (1965). A study of induced metamorphosis in the
axolotl. The Journal of Experimental Zoology, 160, 137–145.
Riddiford, L. M. (1996). Molecular aspects of juvenile hormone action in insect
metamorphosis. In Gilbert, L. I., Tata, J. R., Atkinson, B. G., (Eds).
Metamorphosis: postembryonic reprogramming of gene expression in amphibian
and insect cells. (pp. 223-251) San Diego: Academic Press.
68 Ritchie, J. W., Shi, Y. B., Hayashi, Y., Baird, F. E., Muchekehu, R. W., Christie, G. R., et
al. (2003). A role for thyroid hormone transporters in transcriptional regulation by
thyroid hormone receptors. Molecular Endocrinology, 17, 653–661.
Rosenkilde, P., & Ussing, A. P. (1996). What mechanisms regulate neoteny and induced
metamorphosis in urodeles? The International
Journal of Developmental
Biology, 40, 665–673.
Ryan, T. J., & Semlitsch, R. D. (1998). Intraspecific heterochrony and life history
evolution: Decoupling somatic and sexual development in a facultatively
paedomorphic salaman- der. Proceedings of the National Academy of Sciences
of the United States of America, 95, 5643–5648.
Ryan, T. J., & Semlitsch, R. D. (2003). Growth and the expression of alternative life
cycles in the salamander Ambystoma talpoideum (Caudata: Ambystomatidae).
Biological Journal of the Linnean Society (London), 80, 639–646.
Safi, R., Bertrand, S., Marchand, O., Duffraisse, M., de Luze, A., Vanacker, J. M., et al.
(2004). The axolotl (Ambystoma mexicanum), a neotenic amphibian, expresses
functional thyroid hormone receptors. Endocrinology, 145, 760–772.
Safi, R., Vlaeminck-Guillem, V., Duffraisse, M., Seugnet, I., Plateroti, M., Margotat, A.,
et al. (2006). Paedomorphosis revisited: Thyroid hormone receptors are functional
in Necturus maculosus. Evolution and Development, 8, 284–292.
Scott, D. E. (1993). Timing of reproduction in paedomorphic and metamorphic
Ambystoma talpoideum. The American Midland Naturalist, 129, 397–402.
Semlitsch, R. D. (1987). Paedomorphosis in Ambystoma talpoideum: Effects of density,
food, and pond drying. Ecology, 68, 994–1002.
Semlitsch, R. D., Harris, R. D., & Wilbur, H. M. (1990). Paedomorphosis in Ambystoma
talpoideum: Maintenance of population variation and alternative life history
pathways. Evolution, 44, 1604–1613.
Sexton, O. J., & Bizer, J. R. (1978). Life history patterns of Ambystoma tigrinum in
montane Colorado. The American Midland Naturalist, 99, 101–118.
Shaffer, H. B. (1984). Evolution in a paedomorphic lineage. I. An electrophoretic
analysis of the Mexican ambystomatid salamanders. Evolution, 38, 1194–1206.
Shaffer, H. B., & Breeden, F. (1989). The relationship between allozyme variation and
life history: Non-transforming salamanders are less variable. Copeia, 1989,
1016–1023.
Shaffer, H. B., & Voss, S. R. (1996). Phylogenetic and mechanistic analysis of a
developmentally integrated character complex: Alternate life history modes in
ambystomatid salamanders. American Zoologist, 36, 24–35.
69 Shi, Y. B. (2000). Amphibian metamorphosis: From morphology to molecular biology.
New York: Wiley.
Smith, H. M. (1969). The Mexican axolotl: Some misconceptions and problems.
BioScience, 19, 593–597.
Smith, H. M. (1989). Discovery of the axolotl and its early history in biological research.
In G. M. Malacinski & J. B. Armstrong (Eds.), Developmental biology of the
axolotl. New York: Oxford University Press.
Smith, J. J., Putta, S., Walker, J. A., Kump, D. K., Samuels, A. K., Monaghan, J. R., et al.
(2005). Sal-Site: integrating new and existing ambystomatid salamander research
and informational resources. BMC Genomics, 6.
Snyder, R. C. (1956). Comparative features of the life histories of Ambystoma gracile
(Baird) from populations at low and high altitudes. Copeia, 1956, 41–50.
Song, F., Poljak, A., Crawford, J., Kochan, N. A., Wen, W., Cameron, B., et al. (2012).
Plasma apolipoprotein levels are associated with cognitive status and decline in a
community cohort of older individuals. PLoS One, 7, e34078.
Sprules, W. G. (1974). The adaptive significance of paedogenesis in North American
species of Ambystoma (Amphibia: Caudata): An hypothesis. Canadian Journal of
Zoology, 52, 393–400.
Swingle, W. W. (1922). Experiments on the metamorphosis of neotenous amphibians.
The Journal of Experimental Biology, 36, 397–421.
Takahashi, M. K., & Parris, M. J. (2008). Life cycle polyphenism as a factor affecting
ecological divergence within Notophthalmus viridescens. Oecologia, 158, 23–34.
Tassava, R. A. (1969). Survival and limb regeneration of hypophysectomized newts with
pi- tuitary xenografts from larval axolotls, Ambystoma mexicanum. The Journal of
Experimental Zoology, 171, 451–458.
Taurog, A. C., Oliver, R., Eskay, L., Porter, J. C., & Mckenzie, J. M. (1974). The role of
TRH in the neoteny of the Mexican axolotl (Ambystoma mexicanum). General
and Comparative Endocrinology, 24, 267-279.
Teng, Y. W., Mehedint, M. G., Garrow, T. A., Zeisel, S. H. (2011). Deletion of betainehomocysteine S-methyltranferase in mice perturbs choline and 1-carbon
metabolism, resulting in fatty liver and hepatocellular carcinomas.
Teng, Y. W., Ellis, J. M., Coleman, R. A., Zeisel, S. H. (2012). Mouse betainehomocysteine S-methyltransferase deficiency reduces body fat via increasing
energy expenditure and impairing lipid synthesis and enhancing glucose oxidation
in white adipose tissue. Joural of Biological Chemistry, 287, 16187-16198.
70 Thambisetty, M., Simmons, A., Velayudhan, L., Hye, A., Campbell, J., Zhang, Y., et al.
(2010). Association of plasma clusterin concentration with severity, pathology,
and progression in Alzheimer disease. Archives of General Psychiatry, 67, 739–
748.
Thomas, P.D., Campbell, M.J., Kejariwal, A., Mi, H., Karlak, B., Daverman, R., Diemer,
K., Muruganujan, A., Narechania, A. (2003). PANTHER: a library of protein
families and subfamilies indexed by function. Genome Res. 13, 2129–2141.
Tompkins, R., & Townsend, J. K. (1977). Control of metamorphic events in a neotenous
urodele Ambystoma mexicanum. Journal of Experimental Zoology, 200, 191-196.
Tulla, M., Pentikäinen, O., Viitasalo, T., Käpylä, J., Impola, U., Nykvist, P., et al. (2001).
Selective binding of collagen subtypes by integrin α1I, α2I, and α10I domains.
Journal of Biological Chemisty, 276, 48206–48212.
Ueno, T., Tanaka, K., Kaneko, K., Taga, Y., Sata, T., Irie, S., et al. (2010). Enhancement
of procollagen biosynthesis by p180 through augmented ribosome association on
the endoplasmic reticulum in response to stimulated secretion. Journal of
Biological Chemisty. 285, 29941-29950.
Van der Laan, S., Lachize, S. B., Vreugdenhil, E., de Kloet, E. R., Meijer, O. C. (2008).
Nuclear receptor coregulators differentially modulate induction and
glucocorticoid receptor-mediated repression of the corticotropin-releasing
hormone gene. Endocrinology, 149, 725-732.
Van Vliet, P. (2012). Cholesterol and late-life cognitive decline. Journal of Alzheimers
Disease, 30, S147-S162.
Vlaeminck-Guillem, V., Safi, R., Guillem, P., Leteurtre, E., Duterque-Coquillaud, M., &
Laudet, V. (2006). Thyroid hormone receptor expression in the obligatory
paedomrophic salamander Necturus maculosus. The International Journal of
Developmental Biology, 50, 553–560.
Voss, S. R., Kump, K. D., Walker, J. A., Shaffer, H. B., & Voss, G. J. (2012). Thyroid
hormone responsive QTL and the evolution of paedomorphic salamanders.
Heredity, 109, 293–298.
Voss, S. R., & Shaffer, H. B. (2000). Evolutionary genetics of metamorphic failure using
wild-caught versus laboratory axolotls (Ambystoma mexicanum). Molecular
Ecology, 9, 1401–1408.
Ward, J. H. (1963). Hierarchical grouping to optimize an objective function. Journal of
the American Statistical Association, 58, 236-244.
Weins, J. J., & Hoverman, J. T. (2008). Digit reduction, body size, and paedomorphosis
in salamanders. Evolutionary Development, 10, 449-463.
71 Wells, K. D. (2007). The ecology and behavior of amphibians. Chicago: University of
Chicago Press.
Whiteman, H. H. (1994). Evolution of facultative paedomorphosis. The Quarterly
Review of Biology, 69, 205–221.
Whiteman, H. H., Wissinger, S. A., Denoel, M., Mecklin, C. J., Gerlanc, N. M., &
Gutrich, J. J. (2012). Larval growth in polyphenic salamanders: Making the best
of a bad lot. Oecologia, 168, 109–118.
Wilbur, H. M. (1980). Complex life cycles. Annual Review of Ecology and Systematics,
11, 67–93.
Wilbur, H. M., & Collins, J. P. (1973). Ecological aspects of amphibian metamorphosis.
Science, 182, 1305–1314.
Yamaguchi, Y., Mann, D. M., Ruoslathi, E. (1990). Negative regulation of transforming
growth factor-h by the proteoglycan decorin. Nature 346, 281 – 284.
Yau, J. L., & Seckl, J. R. (2012). Local amplification of glucocorticoids in the aging
brain and impaired spatial memory. Frontiers in Aging Neuroscience, 4, 24.
Zahn, J. M., Poosala, S., Owen, A. B., Ingram, D. K., Lustig, A., Carter, A., et al. (2007).
AGEMAP: a gene expression database for aging in mice. PLoS Genetics, 3, 2326
– 2337.
Zambrano, L., Vega, E. E., Herrera, M., Prado, E., & Reynoso, V. H. (2007). A
population matrix model and population viability analysis to predict the fate of
endangered species in highly managed water systems. Animal Conservation, 10,
297–303.
72 VITA
Carlena K. Johnson
PLACE OF BIRTH:
Unadilla, New York
EDUCATION:
Bachelor of Art, Biology, 2007
Hartwick College, Oneonta, NY
PROFESSIONAL POSITIONS HELD:
Research Assistant, June 2005-July 2005
Dept. of Biology, Hartwick College
Research Assistant, June 2006-August 2006
Dept. of Biology, Hartwick College
Wetland/Turtle Intern, April 2008-December 2008
Hudsonia Ltd., Red Hook, NY
Research Assistant, August 2010-July 2012
Dept. of Biology, University of Kentucky
Research Assistant, August 2012-December 2012
Kentucky EPSCoR, University of Kentucky
Teaching Assistant, January 2013-April 2013
Dept. of Biology, University of Kentucky
SCHOLASTIC AND PROFESSIONAL HONORS:
Frank G. Brooks Award, April 2007
BBB, Hartwick College
Daniel R. Reedy Quality Achievement Fellowship, August 2010-April 2012
Dept. of Biology, University of Kentucky
Academic year Fellowship, August 2012-December 2012
Dept. of Biology, University of Kentucky
73