1 Relaxed selection in erythropoietic gene hemogen among high-latitude Antarctic notothenioids by Carmen M. Elenberger B.A. in Anthropology, University of Florida A thesis submitted to The Faculty of the College of Science of Northeastern University in partial fulfillment of the requirements for the degree of Master of Science December 12, 2018 Thesis directed by H. William Detrich Professor of Biochemistry and Marine Biology
73
Embed
hemogen among high-latitude Antarctic notothenioids by ...m044c681x/fulltext.pdfblooded and white-blooded notothenioid fish show reduced hematocrit, which is potentially an adaptive
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Relaxed selection in erythropoietic gene hemogen among high-latitude Antarctic notothenioids
by Carmen M. Elenberger
B.A. in Anthropology, University of Florida
A thesis submitted to
The Faculty of
the College of Science of
Northeastern University
in partial fulfillment of the requirements
for the degree of Master of Science
December 12, 2018
Thesis directed by
H. William Detrich
Professor of Biochemistry and Marine Biology
2
Copyright 2018
Carmen Elenberger
3
Acknowledgements
First and foremost, I would like to thank my advisor, Dr. H. William Detrich, for his guidance
and his support over the past four years. He challenged me to broaden my horizons and gave me
the opportunity to travel to the ends of the earth in order to do so. I would also like to thank Dr.
Thomas Desvignes, as well as Laura Goetz and Sierra Smith, for their assistance in conducting
field work for this project. I would like to extend further thanks to Dr. Jacob Daane for
permitting me to use his unpublished data to expand my analyses. Many thanks to Biology Open
for allowing me to reproduce their figure with permission [1].
I would like to thank my committee members, Dr. A. Randall Hughes and Dr. Steve
Vollmer, for their interest in my research and their advice in analyzing and framing the results of
my research. I would also like to thank my labmate, Dr. Michael Peters, and our lab manager,
Sandra Parker, for their advice, assistance, and encouragement over the years. Additionally, I
would like to thank the faculty and staff of the Marine Science Center, as well as the funding
sources for this research. Special thanks to the staff of Palmer Station and the crew of the
Laurence M. Gould for a productive and memorable field season. Finally, I would like to thank
my friends and family for their unwavering support and encouragement, now and always.
4
Abstract of Thesis
Antarctic icefish (Channichthyidae) are the only vertebrate taxon with an erythrocyte-null
phenotype, and present an interesting model for studying the evolution and regulation of
erythropoiesis. The gene hemogen has been identified to encode a protein which plays a role in
regulating erythropoietic processes in vertebrates. hemogen may have been potentially impacted
by the loss of globin-expression. I investigated possible relaxed selection at the hemogen locus
by looking for evolutionary change to the regulatory elements or segments encoding the
Hemogen protein, and assessed the evolutionary processes that drove hemogen variation among
Antarctic notothenioids. While regulatory mechanisms remain intact, icefish show a significant
90bp indel in exon 3 of hemogen that would disrupt conserved modules in the Hemogen protein
that are critical for erythropoiesis. Despite this, hemogen still remains expressed at low levels in
adult icefish and possesses a novel splice variant that encodes a truncated protein possibly
serving as a dominant negative for wild-type Hemogen. I conclude that while hemogen has
undergone relaxed selection and accumulated mutations that would impact erythropoietic
function in non-Antarctic fish, the observed mutations may be tolerated due to erythrocyte and
hematocrit modifications in notothenioid blood phenotypes. hemogen may have a decreased—
but still important—role to play in icefish, possibly functioning as a dominant negative for
hemogen’s role in erythropoiesis.
5
Table of Contents
Acknowledgements 3
Abstract of Thesis 4
Table of Contents 5
List of Tables 6
List of Figures 7
List of Abbreviations 9
Introduction 11
Methods 15
Results 21
Discussion 29
Tables and Figures 40
References 67
6
List of Tables
1 Primers used in PCR and qRT-PCR reactions to amplify hemogen gDNA and cDNA in
Antarctic notothenioids (pg 40)
2 Species sequenced and included in study of Antarctic notothenioid hemogen (pg 41)
3 Codon usage bias for hemogen (total coding sequence) among Antarctic notothenioids
(pg 42)
4 Mean pairwise dN/dS for within-family comparisons of Antarctic notothenioid families
(pg 43)
5 Mean pairwise dN/dS for between-family comparisons of Antarctic notothenioid families
(pg 44)
6 Results of codon-based site tests conducted in CodeML on the Antarctic radiation (pg 45)
7
List of Figures
1 Zebrafish Si:dkey-25o16.2 and human Hemogen are orthologous and encode related
proteins that differ in size (pg 46)
2 Icefish transcript variants for hemogen and their putative effects on translation illustrated
in representative species Champsocephalus gunnari (pg 48)
3 Maximum likelihood tree used to test for positive selection on the branch leading to the
Antarctic notothenioid clade (pg 50)
4 Maximum likelihood tree used in site-tests for positive/pervasive selection among
Antarctic notothenioids (pg 51)
5 RELAX tree shows relaxed selection on the branches contained Bathydraconidae and
Channichthyidae, demonstrating a trend of relaxed selection in hemogen on the way to
the erythrocyte-null phenotype (pg 53)
6 Gene structure and size remains conserved among red-blooded and white-blooded
notothenioids, including regulatory regions conserved among teleost fish (pg 54)
7 Conservation of conserved non-coding elements CNE1 and CNE2 in Antarctic
notothenioids relative to Gasterosteus aculeatus and Danio rerio (pg 56)
8 hemogen exon 3 deletions in representative species from Channichthyidae relative to a
red-blooded notothenioid, and their predicted effects on transcription and translation (pg
57)
9 Variant forms of hemogen “exon 3” deletion mapped onto the Channichthyidae species
tree (pg 59)
10 hemogen indels in Antarctic notothenioids mapped onto a maximum parsimony tree (pg
60)
11 Pairwise dN/dS comparisons plotting total dN/dS of whole Hemogen-encoding sequence
with the dN/dS values for the N-terminus and C-terminus of notothenioid Hemogen,
within families Nototheniidae (A & B) and Channichthyidae (C & D). (pg 62)
12 Pairwise dN/dS trends between families Nototheniidae and Channichthyidae, plotting
whole-Hemogen dN/dS vs the N-terminus (A) or C-terminus (B). (pg 63)
13 qPCR quantification of hemogen transcript variants in representative icefish species C.
aceratus and C. gunnari, comparing adult head kidney hemogen expression with N.
coriiceps adult head kidney for both hemgn-L and hemgn-s splice variants (pg 64)
8
14 Changes to the bipartite nuclear localization signal in icefish (Champsocephalus gunnari)
relative to red-blooded notothens (Notothenia coriiceps). (pg 66)
9
List of Abbreviations
aa amino acid
bp base pair
CAI Codon Adaptation Index
cDNA complementary DNA
CNE conserved non-coding element
dN nonsynonymous mutation rate
DNA deoxyribonucleic acid
dN/dS ratio of nonsynonymous to synonymous mutation rates
dS synonymous mutation rate
EDAG erythroid differentiation-associated gene
GATA1 GATA-binding protein 1
gDNA genomic deoxyribonucleic acid
HoxB4 homeobox B4
KLF4 Krueppel-like Factor 4
-lnL negative log likelihood
MMCT Middle Miocene Climate Transition
MRCA most recent common ancestor
Mya million years
Myb MYB Proto-Oncogene, Transcription Factor
NLS nuclear localization signal
p300 histone acetyltransferase p300
PCR polymerase chain reaction
10
qPCR quantitative polymerase chain reaction
RNA ribonucleic acid
Sox9 transcription factor SOX-9
UTR untranslated region
11
INTRODUCTION
Cold-driven evolution of the Antarctic notothenioid lineage began roughly 46 Mya [2]
concurrent with the emergence of the Drake Passage (55-41 Ma) [3] and the initial formation of
the Antarctic Circumpolar Current [4]. The development of antifreeze glycoproteins [5, 6]
permitted colonization and persistence in the Southern Ocean [7] and set the stage for further
diversification during successive cooling periods and accompanying geological events. The
radiation of the high latitude Antarctic notothenioids (Cryonotothenioidea) occurred during a
period of diversification driven by intensified cooling of the Southern Ocean during the Middle
Miocene Climate Transition (MMCT) [7, 8], with species diversification beginning ~14 Mya and
accelerating ~11 Mya during the Late Miocene [7, 9-11]. Cooling during the MMCT led to
contemporary Antarctic conditions (-2℃ to + 2℃) and resulted in the scouring of continental
shelves by ice [12, 13]. This opened ecological niches for potential colonization by removing
more temperate adapted competitors [14] and leading to rapid morphological and ecological
diversification [15]. Current day Antarctic notothenioids comprise 77% of Antarctic teleost
diversity and constitute a marine species flock [16] derived via adaptive radiation [17-19]. High
levels of morphological diversity and intense speciation make Antarctic notothenioids a useful
evolutionary model for studying cold adaptation.
Antarctic notothenioids possess a number of remarkable changes to erythropoiesis and
the oxygen-transport system at large that resulted in the evolution of the only known vertebrate
clade devoid of erythrocytes—the family Channichthyidae, characterized by a “white-blooded”
phenotype [20]. It has been hypothesized that the high oxygen concentration in polar seawater
could lead to potential relaxed selection on erythrocytes and other oxygen-binding pigments, as
hypoxic stress becomes less of a relevant factor with oxygen in such high abundance [21].
12
Evidence for such relaxed selection can be seen in changes to blood content: a study of “red-
blooded” Antarctic species from McMurdo Sound showed decreased numbers of erythrocytes,
lowered hematocrit, and lowered hemoglobin concentrations when compared with temperate fish
[22]. General trends throughout the radiation show that the more derived the family, the fewer
erythrocytes present in circulating blood and the lower the hemoglobin content [21]. Both red-
blooded and white-blooded notothenioid fish show reduced hematocrit, which is potentially an
adaptive feature to contend with the increased viscosity of blood under low temperatures [23,
24]. Hemoglobin multiplicity is reduced among notothenioids relative to temperate fish [25-27]
and cold anemia responses became genetically assimilated [28-30]. At some point notothenioid
dependence on hemoglobin for respiration became so reduced even red-blooded fish could
continue to effectively absorb and utilize oxygen even in the presence of carbon monoxide [31],
suggesting that the stage had well been set for hemoglobin loss before it disappeared.
Channichthyidae are characterized by loss of the vertebrate oxygen-transport molecules
the α2β2 hemoglobin tetramer carried within erythrocytes. This occured in the most recent
common ancestor (MRCA) of all icefish via large genomic lesions within the respective loci [25,
32-36]. Furthermore, there have been multiple, independent losses of myoglobin during
diversification [36]. Icefish possess few erythroblasts, and their blood contains mostly leukocytes
and plasma [35]. The evolution of the white-blooded phenotype is unique among vertebrates and
has far reaching consequences for the cardiovascular system and key globin partners. As a result
of hemoglobin loss, we would anticipate changes to the genetic machinery involved in red blood
cell production and maintenance, as selective constraints on this may relax in the absence of key
globin partners. It is also possible that this began somewhere within the red-blooded families, as
oxygen transport molecules became less necessary for survival. Relaxed selection in the
13
regulatory regions of globin has been detected among dragonfish, prior to the emergence of a
white-blooded phenotype [37].
The gene hemogen has been identified as an interesting candidate for further study in
notothenioid fish, given evidence based on subtraction libraries that expression may be impaired
or entirely absent in icefish. The hemogen gene encodes the transcription factor Hemogen (Fig
1), which acts as a regulator in hematopoietic development by stimulating the differentiation of
hematopoietic cells into both the erythroid and megakaryocytic lineages [38-43]. In teleost fish,
Hemogen is encoded by four exons and contains domains similar to those predicted in the human
ortholog: a coiled-coil domain, a bipartite nuclear localization, a series of tandem repeats and an
acidic domain (Fig 1) [1, 38]. It is promoted via two conserved non-coding elements, one
proximal and one distal, both critical for promoting primitive erythropoiesis (Fig 1) [1].
Hemogen also plays a role in cell apoptosis [39] and has been implicated in the regulation of
tumor cells in acute myeloid leukemia [44]. Other possible roles include spermatogenesis [45],
sex-determination [46], and osteoblast recruitment and bone calcification [47-49]. Research
show Hemogen’s role in hematopoiesis takes place via interactions with a number of key
proteins involved in erythropoiesis and development, including GATA1 and p300. GATA1 is
critical for erythroid differentiation [50-52] and functions in both primitive and definitive
hematopoiesis [53]. Nonsense mutations in GATA1 lead to a “bloodless” phenotype [54].
GATA1 recruitment is crucial for hemogen function and downregulation of hemogen expression
inhibits GATA1 activity [40, 43], while GATA1 recruits hemogen to the beta-globin locus [55].
p300 is crucial for cell differentiation [56, 57] and inhibition of p300 binding to Hemogen causes
decreased production of erythroid cells. Hemogen facilitates the interaction between GATA1 and
p300, making it a critical part of the erythroid differentiation process [55].
14
Decreased hemogen expression in white-blooded fish may indicate functional loss. Given
the decreasing importance of red blood cells to the notothenioid lineage, selective constraints on
known regulators of erythrocyte production may have relaxed prior to complete globin loss.
Hemogen interacts with Beta-globin and regulates erythroid production, raising the possibility
that erythropoietic features may be aberrant in icefish. However, hemogen demonstrates
pleiotropy, as described in the previous paragraph, and lists of potential partners implicate it in a
number of important cellular processes beyond erythropoiesis. Therefore, at least some features
must remain conserved in order to carry out non-erythropoietic roles.
In this thesis, I characterize hemogen genes in both red-blooded and white-blooded
Antarctic notothenioids and compare them with sub-Antarctic perciform outgroups to establish
hemogen’s history within this clade. I hypothesize that the hemogen locus is undergoing relaxed
selection among the icefish, and that relaxation of selective constraints began prior to the
emergence of Channichthyidae. I investigated partial conservation of the hemogen gene,
hypothesizing that pleiotropy would protect against total pseudogenization of hemogen. Features
under relaxed selection would be implicated in erythropoietic function and could be considered
targets for further study of hemogen in erythropoiesis. I hypothesize some level of differential
expression between white-blooded and red-blooded fish; if not complete loss of expression, than
loss in certain tissues or of certain key isoforms in Channichthyidae.
My results show a strong trend towards relaxed selection in high-latitude Antarctic
notothenioids relative to Sub-Antarctic relatives, with icefish showing intensified relaxation.
Confirmation of relaxed selection among-red-blooded fish supports the theory that the decreased
dependence on erythrocytes in notothenioid fish also correlates with larger-scale changes in the
erythropoietic paradigm on the genomic level. Three out of four key functional domains show
15
some form of degradation, either via nonsynonymous mutation or through the
transcriptional/translational impacts of indels on icefish hemogen. Three key evolutionary events
took place in the MRCA of all extant icefish: the degradation of the bipartite NLS, a 30aa loss in
a proline-rich region of tandem repeats, and the development of a novel splice form, hemgn-s,
which excludes all functional domains encoded by exon 3 and 4 and theoretically results in a
frameshifted and truncated hemogen protein. However, key promoter regions remain conserved
in icefish, and while expression is down-regulated in adult tissues relative to red-blooded
species, hemogen is still expressed in adult tissues of some icefish. This suggests that while the
decreased importance of erythropoietic functions may have significantly relaxed pressure on
hemogen and resulted in mutations impacting domains critical for erythropoietic-function, it is
not necessarily non-functional and may still be playing a decreased but critical role in other
processes.
METHODS
Sample collection & sequencing of notothenioid hemogen gDNA
The primary source of genomic material came from tissues obtained by the Detrich Lab
during the 2012, 2014 and 2016 winter fishing cruises conducted by the Research Vessel
Laurence M. Gould near Palmer Station, Antarctica. Tissues were flash-frozen in liquid nitrogen
and then stored at -80℃. I generated sequences from between 1-5 individual fish per species.
Molecular methods for gDNA extraction from tissues are as specified in the Quick-gDNA
miniprep kit (Zymo Research, D3024). Full notothenioid hemogen—from start codon to the 3’
UTR—was amplified by PCR from gDNA samples using 1 µM primers (Table 1) designed from
previously obtained Notothenia coriiceps sequences. The amplification protocol was as
16
follows—35 cycles of 98°C for 10 s, 59°C for 10 s, 72°C for 1 min. PCR products were cloned
into the pGEM T-easy vector (Promega, A1360), and recombinant clones were transformed into
DH5α competent cells (New England Biolabs, C2987H). Recombinant plasmids were identified
using blue/white screening, purified via the Wizard Plus Miniprep DNA Purification System
(Promega, A7500), and sequenced by GeneWiz. I obtained full genomic sequences for 18
notothenioid species (Table 2, Figure S1).
Cloning and sequencing of notothenioid cDNAs
I isolated total RNA from flash-frozen tissues of adult N. coriiceps and C. aceratus using
the RNEasy Mini Kit (Qiagen, 74104). Several potential hemogen transcripts had been
previously identified by other Detrich Lab members (Figure 2). To expand upon these results,
RNA samples were prepared from ten tissues: liver, spleen, head kidney, trunk kidney, white
muscle, pectoral red muscle, testes, brain, heart ventricle and gill. Total cDNA was produced
from the mRNA using M-MuLV reverse transcriptase and an oligo(dT)23 primer according to the
protocol outlined in the Protoscript II First Strand cDNA Synthesis kit (NEB, E6560S). cDNA
was amplified via PCR using the same primers as gDNA PCR (Table 1) according to the
following protocol: 35 cycles of 98°C for 10 s, 59°C for 10 s, and 72°C for 45 s. cDNA was then
cloned into pGEM T-easy vector and subsequently transformed and purified as outlined for
gDNA sequences.
Construction of genomic, coding and protein alignments for gene characterization,
phylogenies and evolutionary analysis
17
Nucleic acid alignments were constructed using MUSCLE [58] as implemented in
MEGA7 [59], with a gap opening penalty of 15 and gap extension penalty of 6.66. Alignments
were subsequently inspected and adjusted by eye in BioEdit [60]. Construction of gene trees and
evolutionary analysis relied primarily on three alignments: a gDNA alignment with all exons +
introns; a coding alignment based on cDNA sequences, transcriptome data and concatenated
exome data; and a protein alignment, translated from the coding sequences in MEGA7 [59].
The cDNA sequences that I generated were supplemented with hemogen cDNAs from
transcriptomic analyses of Pseudochaenichthys georgianus [unpublished data from Detrich lab]
and Parachaenichthys charcoti [unpublished data from Detrich lab], and aligned with my
genomic sequences to generate coding sequences for other notothenioids. Additionally, cDNA
and transcriptome sequences also served as a basis for alignment and quality control for
sequences obtained via an exome-capture analysis [unpublished] conducted by Dr. Jacob Daane
of the Detrich lab. A full list of all species included and the sequence sources can be found in
Table 2. In total 43 species representing Antarctic notothenioids from all high-latitude families
(Artedidraconidae, Bathydraconidae, Channichthyidae, Harpagiferidae, Nototheniidae) as well
as 7 Sub-Antarctic outgroups were included in evolutionary analyses.
Analysis of positive, pervasive and relaxed selection on Antarctic notothenioid hemogen
All trees were constructed in RAxML [61, 62] using nucleotide substitution model
GTRGAMMA to conduct an initial tree search of 20 trees and select the best tree from this pool.
No outgroups were specified. Branch tests were conducted using the CodeML module included
in PAML 4.0 [63, 64].
18
Two tests for positive selection were run. The first was conducted using branch-site
models [65, 66] on a subset of coding sequences (SFigure 2) to search for possible positive
adaptation in the Antarctic clade relative to several notothenioid outgroups. The branch leading
to the representative Antarctic notothenioids was specified a priori as the foreground branch
(Figure 3). The null model set NSsites = 2, fix_omega = 1, and omega = 1. This assumes two
categories of sites (purifying and neutral selection) and looks for a difference in proportions of
sites undergoing neutral selection on the foreground branch relative to the background. The
positive/alternative model set NSsites = 2, fix_omega = 0, and omega = 1, which allows for three
categories of sites (purifying, neutral, and positive selection) and looks to identify sites
undergoing positive selection on the foreground relative to the background branch. If the
alternative model is accepted over the null, this indicates a site has undergone episodic positive
selection (changed once, then retained in the clade)
The second test relied on codon-substitution site models [67, 68] to detect pervasive
positive selection among Antarctic notothenioids using the coding sequences (SFigure 2). This
would identify any possible sites which changed repeatedly throughout diversification of the
clade, possibly as a result of differing adaptive challenges related to the modification of the
hematic system. Models M0, M1a, M2a, M3, M7, and M8 were run by setting NSsites = 0 1 2 3
7 8 (respectively), fix_omega = 0, and omega = 1. Model M8a set NSsites = 8 but set fix_omega
= 1 and omega = 1. The submitted gene tree for the site tests can be found in Figure 4.
Test for relaxed selection in the branch leading to Channichthyidae was conducted using
RELAX [69] as part of the HyPhy suite of hypothesis testing software [70]. RELAX conducts a
comparative test of whether an a priori specified branch or subset of branches has undergone
relaxed or diversifying selection relative to the rest of the tree. This makes it useful for
19
identifying trends and/or shifts in the stringency of natural selection on a given gene, provided
one has an idea of where that should occur. The branches ending in Channichthyidae and
Bathydraconidae were selected as the test branches, with all others used as reference branches.
The reference tree used was the putative species tree of Daane [unpublished]. The test was run
on the Datamonkey server [71, 72].
Bioinformatic comparison of notothenioid hemogen promoters and coding domains
Regulatory regions from Eleginops maclovinus, N. coriiceps and Chaenocephalus
aceratus were sequenced based on the annotations for the N. coriiceps genome (NCBI
Accession: PRJNA66471, ID: 66471) [73]. gDNA sequences were aligned to N. coriiceps and C.
aceratus scaffolds via BLAST in Geneious (v. 10.0.5) [74] to determine whether notothenioid
species possess conserved synteny around the hemogen locus as observed in other vertebrate
species [1]. Scaffold sequences were confirmed by sequencing from the upstream (anp32b) and
downstream (TRMO) genes towards hemogen. Promoter alignments for hemogen were obtained
using the whole genome alignments for D. rerio and Gasterosteus aculeatus (ENSEMBL v94)
[75]. Transcription factor binding sites were predicted with ConTra v2 using a similarity matrix
of 0.75 [76]. Protein domains were identified based on annotations from human [38] and
zebrafish hemogen [1].
Parsimony gene tree and deletion mapping
A hemogen gene tree was built using coding sequences and maximum parsimony method
[77] in Mega7. Gaps were treated as partial deletions with site coverage set for 90%. This
allowed for the inclusion of sites where a majority of species possessed sequence data but one
20
species (or genus) possessed a phylogenetically informative indel. The tree included 1st, 2nd, and
3rd codon positions and was computed using the Subtree-Pruning-Regrafting method, beginning
with 10 trees and retaining 100 trees. Following 1000 bootstrap iterations the best tree was
selected based on comparison with known species phylogenies. The phylogeny was edited to
include indel information using ggtree in R [78] and the Interactive Tree of Life (iTOL) v3 [79].
I ran the tree topology in CodeML [63, 64] using the M0 model (model = 0, NSsites = 0) [67] to
obtain the number of nucleotide substitutions per codon (dN+dS) as well as dN, dS, and dN/dS
for the whole tree.
An icefish species tree was constructed based on the species tree built from the exome
data of Daane [unpublished] with modifications derived from available Channichthyidae
phylogenies [80, 81].
Pairwise dN/dS comparisons
Pairwise dN/dS values were generated using a subset of the coding alignment (SFigure 2)
and were ran in PAML4 using yn00 [64]. yn00 calculates rates based on the method outlined in
Nielsen & Yang 2000 [82] and allows for codon usage bias as well as transition-transversion rate
differences. To assess codon usage bias in notothenioid hemogen, I used DnaSP v5 [83, 84] to
measure codon usage bias via the codon adaptation index (CAI/CBI) [85, 86]. Values for CAI
are shown in Table 3; the values fall within a range of 0.3-0.4 for all species, which represents
moderate codon usage bias (low bias < 0.3 and high > 0.5).
All forty-three high-latitude notothenioid species were included in this analysis (Table 2).
I examined two kinds of evolutionary relationships: within-family comparisons (ex: icefish vs
icefish) and between-family pairwise comparisons (ex: Channichthyidae vs Nototheniidae). For
21
each kind of comparison I ran the data with three partitions: the total protein coding sequence,
the coding sequence for the N-terminus only (1-79 aa, which represents the end of the bipartite
NLS), and the coding sequence for the C-terminus only (80 aa—end). This allowed for a more
nuanced analysis of the selective forces at work on different parts of the gene as well as within
different clades and is derived from work done parsing geographic effects on cichlids and
positive selection in notothenioids [87, 88].
qPCR
Previous qualitative PCR I conducted on C. aceratus cDNA established general
presence/absence of hemogen expression in several adult tissues—liver, head kidney, trunk
kidney, spleen, and brain—and isolated the predominant isoforms of hemogen expression in