A Dissertation
entitled
Evolutionary Patterns and Occurrences of the fish Viral Hemorrhagic Septicemia Virus in
the Laurentian Great Lakes
by
Megan Denise Niner
Submitted to the Graduate Faculty as partial fulfillment of the requirements for the
Doctor of Philosophy Degree in Biology (Ecology)
______Dr. Carol A. Stepien, Committee Chair
______Dr. Douglas Leaman, Committee Member
______Dr. Daryl Moorhead, Committee Member
______Dr. Vikram Vakharia, Committee Member
______Dr. William Sigler, Committee Member
______Dr. Cyndee Gruden, Dean College of Graduate Studies
The University of Toledo
August 2019
Copyright © 2019, Megan Denise Niner
Chapters 1 and 4 of this document are copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of
The Evolution and Presence of Viral Hemorrhagic Septicemia Genogroup IVb in the Great Lakes
by
Megan Denise Niner
Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biology (Ecology)
The University of Toledo August 2019
Studying naturally occurring viruses outside of the lab can provide insight into the different evolutionary pathways between hosts and pathogens, aiding future predictions for their spread and levels of severity. One such virus is the focus of this dissertation:
Viral Hemorrhagic Septicemia virus (VHSV). This global pathogen has four major genogroups, I-IV, among which, our main focus, IVb emerged in the Laurentian Great
Lakes in the early 2000s, killing >30 fish species in large outbreaks from 2005-2010.
Few detections were reported between 2010 and 2016, however, two minor outbreaks occurred in 2017. Following recovery of 21 new isolates from sampling between 2015-
2017, we compared the genetic composition of these isolates with historical sequences using a combined approach of phylogenetic analysis with population genetics to examine spatial and temporal trends in IVb based on the sequence of the glycoprotein, or G-gene, along with a concatenated five-gene dataset using a reduced number of isolates. The 185
G-gene sequences revealed 36 unique haplotypes, whose relationships evolved from two apparent original haplotypes, designated “a” and “b”. Most haplotypes differed from those by single to a few nucleotides, with “a” and “b” diverging by just one. Prior to
iii 2011, “a” and “b” were prevalent, buy have gradually become less frequent over time.
The 14 haplotype descendants of “a” primarily occurred in the Upper and Central Great
Lakes, and the 22 “b” variants mostly were found in the Lower Great Lakes. Significant genetic divergences characterized the geographic distributions of haplotypes among the
Upper, Central, and Lower Great Lakes, and those from the later time period significantly differed from the early and middle time periods. Patterns of nucleotide substitutions corresponded to geographic divergences, whereas amino acid changes characterized temporal trends. We thus found that VHSV-IVb has continued to diversify in the Great
Lakes, following spatial and temporal with distinct spatial and temporal patterns. Using genomics, we conducted a phylogenetic comparison of 46 whole IVb genomes, along with comparisons to the North American and Asian VHSV-IVa, and the European I/III genogroups. A total of 253 nucleotide substitutions were found in IVb, with 85 resulting in amino acid changes. Most substititutions occurred in the non-coding region and Nv- genes, which also was true for the IVa and I/III genomes. IVa was estimated to have the fastest substitution rate (2.01x10-3), followed by IVb (6.64x10-5) and I/III (4.09x10-5).
Phylogenetic analysis of the novirhabdoviruses based on full genomes agreed with past results that were based on partial genes. We also investigated possible origins of the Nv- gene in VHS, whose results showed sister relationships but were homoplastic beyond that level. Cell culture experiments compared three isolates collected from fish hosts in 2016 to the original 2003 isolate for viral production and host immune response suppression.
The 2016 isolates displayed lower virulence and less suppression of the host antiviral response, which suggested that changes in the genome have phenotypic consequences.
iv Acknowledgements
I would like to thank all those who put their time and energy into this project. To
Dr. Stepien for providing me the opportunity and resources to work on this fascinating project. To Dr. Moorhead for handling the complicated logistics of my mostly-absent committee. To Dr. Leaman for letting me work with his lab and helping me understand the cell culture related portions of this project. To my other committee members, Dr.
Vakharia and Dr. Sigler, for agreeing to be involved with this project and their support.
To Dr. Krishnamurthy for allowing me a space in her lab. To Shelby Edwards, for countless hours in the field, lab and on the road. To my fiancé, parents, sister, coworkers at the LEC and the main campus for not letting me give up.
v Table of Contents
Abstract……………………………………………………………………………………iii
Acknowledgements………………………………………………………………………..v
Table of Contents………………………………………………………………………….vi
List of Tables………………………………………………………………………………x
List of Figures…………………………………………………………………………...xiii
Preface……………………………………………………………………………………xx
1. Introduction…………………………………………………………………………….1
2. Evolutionary trajectory of the fish Viral Hemorrhagic Septicemia virus across its
history in the Laurentian Great Lakes: Temporal and spatial patterns……………….10
2.1. Abstract…………………………………………………………………..10
2.2. Introduction………………………………………………………………11
2.2.1. VHS evolution, outbreaks, and hosts…………………………….13
2.2.2. Aim and objectives………………………………………………16
2.3. Materials and Methods…………………………………………………..16
2.3.1. Sampling…………………………………………………………16
2.3.2. RNA extraction and reverse transcription………………………..17
2.3.3. qPCR tests for VHSV-IVb and quantification…………………...18
2.3.4. Preparation of historic isolates in cell culture……………………19
2.3.5. Sequencing VHSV isolates………………………………………19
2.3.6. Genetic data analyses…………………………………………….20
2.4. Results ……………………………………………………………………23
2.4.1. VHSV-IVb detections between 2015-2017……………………...23
vi 2.4.2. Evolutionary patterns from the G-gene…………………………..24
2.4.3. Patterns from concatenated gene analyses……………………….26
2.4.4. Population genetic trends………………………………………...27
2.5. Discussion………………………………………………………………..29
2.5.1. VHS-IV occurrences and evolutionary trajectory………………..29
2.5.2. Evolutionary patterns across space and time.……………………30
2.5.3. Substitution rates, types, and patterns..…………………………..32
2.5.4. Gene specific variation…………………………………………..34
2.5.5. Host species generality, specificity, and infection……………….37
2.5.6. Selection and co-evolution……………………………………….41
2.5.7. Summary and conclusions……………………………………….42
2.6. Acknowledgements………………………………………………………42
3. Genomic and immunogenic changes across the evolutionary history of the VHSV-IVb
fish virus (Piscine novirhabdovirus) in the Laurentian Great Lakes………………….75
3.1. Abstract…………………………………………………………………..75
3.2. Introduction………………………………………………………………76
3.3. Materials and Methods…………………………………………………...81
3.3.1. Sample nomenclature…………………………………………….81
3.3.2. Virus isolation……………………………………………………81
3.3.3. Full genome sequencing…………………………………………82
3.3.4. Genetic analyses………………………………………………….83
3.3.5. Evaluating evolution and selection………………………………84
3.3.6. Cell lines and cell culture experiments…………………………..85
vii 3.4. Results……………………………………………………………………88
3.4.1. Genomic and genic changes……………………………………...88
3.4.2. Evolutionary relationships……………………………………….91
3.4.3. Differences in cytopathicity and immune response……………...94
3.4.4. IFN and virus production and expression of mRNAS…………...94
3.5. Discussion………………………………………………………………..95
3.5.1. Evolutionary trends………………………………………………95
3.5.2. Phylogenetic patterns: novirhabdoviruses……………………….98
3.5.3. Phylogenetic Patterns: VHSV-IVb………………………………99
3.5.4. Evolutionary Perspectives on the Nv-gene……………………..103
3.5.5. Differences in cytopathogenicity……………………………….103
3.5.6. Summary and conclusions……………………………………...105
3.6. Acknowledgments………………………………………………………105
4. Discussion…………………………………………………………………………...131
4.1. General Conclusions……………………………………………………131
4.1.1. Presence and distribution of VHSV-IVb 2015 to 2017………...132
4.1.2. Recent changes and population genetic trends of IVb………….133
4.1.3. Genomic trends of IVb, with comparisons to IVa and I/III…….135
4.1.4. Genomic characterization of novirhabdoviruses and the origin of
the Nv-gene……………………………………………………..136
4.1.5. In vitro comparison of 2016 isolates versus the original 2003
isolate…………………………………………………………...137
4.2. Future Research and Recommendations………………………………..138
viii 4.2.1. Continued surveillance with narrowed focus…………………...138
4.2.2. Resolving the relationship between genogroups Ia and III……..139
4.2.3. Genomics: sequencing of additional samples…………………..139
4.2.4. Pathogenicity of new isolates and changes in host immune
suppression……………………………………………………...140
References………………………………………………………………………………141
ix List of Tables
Table 2-1 VHSV positives from our collections in 2012, 2015, and 2016...... 44
Table 2-2 Primers for PCR and sequencing per gene region, with name, reference,
sequence, annealing temperature, and extension time ...... 46
Table 2-3 Haplotype numbers, diversity, and private haplotypes per time period (A)
and Great Lakes region (B) ...... 47
Table 2-4 Mean numbers of nucleotide (NT) and amino acid (AA) changes (+/-
standard errors) among VHS-IVb haplotypes for (A) time periods and (B)
geographic regions from the G-gene and concatenated genes ...... 48
Table 2-5 Pairwise genetic divergences of VHSV populations between (A) sampling
time periods, Early (2003-6), Middle (2007-10), and Later (2011-18), (B)
Great Lakes regions (Upper, Central, Lower), and (C) individual water
bodies (C), based on variation for the (1) G-gene and (2) concatenated gene
data sets, using exact tests (GENEPOP; above diagonal) and θST divergences
(ARLEQUIN; below diagonal). *=P0.05, **=remained significant (P<α)
following sequential Bonferroni correction, NS=P>0.05...... 49
Table 2-6 Relative distribution of genetic variation among VHSV-IVb isolates using
Analysis of Molecular Variance (AMOVA, Excoffier et al., 1992),
calculated from 185 partial G-gene sequences for A) nucleotide sequences,
and B) amino acid changes, using ARLEQUINv3.5.1.3 (Excoffier &
Lischer, 2010). *=significant ...... 50
Table 2-7 Tajima’s D test values (ARLEQUIN) for selection pressures on VHSV-IVb
evolution based on A) all samples, B) geographic region (Upper, Central,
x Lower Great Lakes), C) time period (Early (2003-6), Middle (2007-10),
Later (2011-18), D) individual water bodies, and E) host species.
*=p0.05 ...... 51
Table 2-A VHSV-IVb samples used for our analyses. Isolate names, years, location
information, host species, geographic coordinates, haplotypes, GenBank
Accession numbers, sources and analysis group for the (1) Early, (2)
Middle, and (3) Later time periods ...... 62
Table 2-B Substitutions in the concatenated gene sequence haplotypes (N, P, M, G, and
Nv-genes). Numbered from the start of the full haplotype “a” genome
(GQ385941), with corresponding haplotypes. *=non-synonymous ...... 70
Table 2-C Compositions of substitutions in concatenated gene sequence regions (N, P,
M, and Nv-genes). Proportions of substitutions for each region sequence are
in parenthesis. dN/dS = proportion of non-synonymous versus synonymous
changes ...... 73
Table 3-1 VHSV isolates for full genome analysis. IVb isolates were analyzed by
Niner and Stepien (in review) and their corresponding G-gene and combined
gene haplotypes are included below ...... 107
Table 3-2 Additional Rhabdovirus sequences used in phylogenetic trees ...... 110
Table 3-3 Single nucleotide polymorphisms (SNPs) from each gene’s coding region
and the combined non-coding regions (NCDS) in IVb sequence variants for
(A) VHSV-IVb, (B) IVa, and (C) Ia, Ib, & III combined. The number of
nucleotides (NT) is reported in front of amino acid (AA) changes. The
proportion of nonsynonymous (dN) to synonymous (dS) changes is reported
xi along with transversions (Tv) and the proportion of transversions to
transitions (Ts). Totals are shown in the final row ...... 111
Table 3-4 Positive (diversifying) or negative (purifying) selection pressures on
individual codons determined by FUBAR (fast, unconstrained Bayesian
approximation) and MEME (mixed effects model of evolution) analyses
(Murrell et al. 2012, 2013) for (A) IVb, (B) IVa, and (C) I/III. None of the
codons found in IVa or I/III matched the codons under selection for IVb.
Codons for tests with more than six codons found are not displayed ...... 112
Table 3-5 Pairwise genetic divergences between VHSV populations: (A) sampling
time periods, Early (2003-6), Middle (2007-11), and Later (2012-16) and
(B) Great Lakes regions (Lake Michigan and Budd Lake, Lake St. Clair,
Lake Erie and Lake Ontario) using exact tests (GENEPOP; above diagonal)
and θST divergences (ARLEQUIN; below diagonal). *=p0.05,
**=remained significant (p<α) following sequential Bonferroni correction,
NS=p>0.05...... 113
Table 3-A Single nucleotide polymorphisms (SNPs) and nonsynonymous changes per
individual isolates. *=Group, includes C06NP, C06RB, C06SR, C06YP,
C06FD, M08AM, C08LEa, C08LEb, and C09MU ...... 115
xii List of Figures
Figure 1-1 Rhabdovirus phylogenetic tree, based on full genome sequences as
determined by Bayesian analyses. Bayesian posterior probabilities are on
each branch. Tree is rooted to the Viral Hemorrhagic Septicemia isolate
MI03GL (VHSV, GenBank: GQ385941)...... 9
Figure 2-1 Maps showing locations (circles, colored by year) of VHSV-IVb isolates
analyzed here, per time period (A) Early (2003-6), (B) Middle (2007-10),
and (C) Later (2011-18) ...... 54
Figure 2-2 Concentrations of VHSV-IVb (+/- standard error) in wild-caught fish
tissues, compared to results from experimental laboratory haplotype “a”
challenged muskellunge, determined with our qPCR assay using internal
standards (Pierce et al., 2013b). Laboratory samples (squares) are named by
the number of days (6-42D) after VHSV-IVb inoculation, H=high virus
dosage (1x105 pfu/mL), L=low dosage (100 pfu/mL) (data from Pierce,
2013). Haplotype of each sample is listed above its standard error bars.
*above the haplotype indicates that the sample was sequences for both the
G-gene and concatenated gene analyses. Solid line denotes the experimental
symptom threshold and dashed line the cell culture detection threshold
(Pierce et al., 2013b). Wild caught samples (circles) are designated by
abbreviated common name, followed by collection year and sample number
(Table 2-1). Names: MUS=muskellunge, LMB= largemouth bass,
FRD=freshwater drum, WPE=white perch, ROG=round goby, GIZ=gizzard
shad, PUM=pumpkinseed, EMS=emerald shiner, ALE=alewife ...... 54
xiii Figure 2-3 VHSV-IVb G-gene phylogeny. Phylogenetic tree of VHSV haplotypes
based the G-gene from maximum likelihood and Bayesian analyses. Values
above nodes = 2000 bootstrap pseudoreplicates/Bayesian posterior
probabilities. Values in parentheses and italics = estimated divergence time
(years). VHSV-IVa (AB179621) served as the outgroup...... 56
Figure 2-4 G-gene haplotype networks. Partial G-gene sequences (669 NT) from 176
isolates using POPART (https://popart.otago.ac.nz) and TCS (Clement et al.
2000) for (A-B) Great Lakes regions (Upper, Central, Lower), (C-D) time
periods (Early, Middle, Later), and (E-F) host species. A, C, and E are based
on nucleotide substitutions and B, D, and F on amino acid changes. Circles
are sized according to frequency of the haplotype in the population. Lines
denote a single substitution step between haplotypes, with dashed lines for
synonymous changes and solid lines for nonsynonymous changes. Small,
unlabeled black circles represent hypothesized haplotypes. The ‘Other’
category in Fig. E-F, contains all host species, in which three or fewer non-
unique isolates were detected: alewife (Alosa pseudoharengus), amphipod
(Diporeia spp.), brown bullhead (Ameiurus nebulosus), burbot (Lota lota),
channel catfish (Ictalurus punctatus), Chinook catfish (Oncorhynchus
tshawytscha), common carp (Cyprinus carpio), cisco (Coregonus artedi),
lake whitefish (C. clupeaformis), leech (Myzobdella lugubris), muskellunge
(Esox masquinongy), northern pike (E. lucius), rainbow trout
(Oncorhynchus mykiss), sea lamprey (Petromyzon marinus), shorthead
redhorse (Moxostoma macrolepidotum), and walleye (Sander vitreus) ...... 57
xiv Figure 2-5 Concatenated N, P, M, G, and Nv-gene haplotype networks. Gene sequences
(3355 NT) from 47 isolates in POPART and TCS for (A-B) Great Lakes
regions and (C-D) time periods. Figure parts A and C show nucleotide
substitutions and B and D those with amino acid changes. Circles are sized
according to population frequency of the haplotype. Lines denote single
substitution steps, with dashes denoting synonymous changes and solid lines
nonsynonymous changes. Small, unlabeled black circles represent
hypothesized haplotype steps ...... 58
Figure 2-6 Neighbor-joining genetic distance tree depicting relationships among
VHSV-IVb populations. Reynold’s (1983) genetic distances (RST) used G-
gene haplotypes and their frequencies in PHYLIP (Felsenstein, 2007).
Bootstrap percentage support for nodes from 10,000 replications are shown.
Sample sizes (N) are in parentheses. Colored boxes denote sampling groups.
*=samples from confirmed fish kill event ...... 59
Figure 2-7 Tests for relationship between genetic divergence (θST) among VHSV G-
gene sampling groups vs. geographic distance (A-C, nearest waterway
distance, km) or time (D, yrs). (A) Early time period (y=0.001x + 0.091,
R2=0.288, p=0.006*), B) Middle time period (y=1.86e-5x + 0.298,
R2=0.001, p=0.278), C) Later time period (y=-9.82e-5x + 0.860, R2=0.049,
p=0.709), and D) all samples (y=0.040x + 0.342, R2=0.149, p=0.002*) ...... 60
Figure 2-A VHSV-IVb structure and genome layout. Colors match the gene to the
structure diagram. Numbers sharing the same colors as the gene refer to the
xv nucleotide positions. Numbers in parentheses correspond to the region
sequenced within each gene. Modified with permission from Pore (2012) ....75
Figure 3-1 Map of VHSV (I-IV) full genome isolates included in this study, colored
according to genogroup and shapes denoting subgenogroups ...... 121
Figure 3-2 Map of VHSV-IVb occurrences in the Great Lakes. Shapes colored
according to time period (Early: 2003-06, Middle: 2007-10, Late: 2011-17).
Diamonds are isolates having sequenced genomes. Large circles depict
VHSV-IVb that lack whole genome data ...... 121
Figure 3-3 Haplotype network showing genetic relationships among 47 VHSV-IVb full
genomes from POPART. Circles sized according to haplotype frequency
among isolates. Numbers inside parentheses designate NT differences
between each node and the original haplotype, C03MU*. Small, unlabeled
black circles = hypothesized haplotype steps ...... 122
Figure 3-4 Novirhabdovirus phylogenetic trees, based on full genome sequences (see
Tables 3.1 and 3.2), with maximum likelihood and Bayesian analyses.
Colored squares designate support values, top left half = Bayesian posterior
probabilities, bottom right = 500 bootstrap pseudoreplicates. Tree was
rooted to the snakehead rhabdovirus (SHRV, GenBank: AF147498) ...... 123
Figure 3-5 VHSV-IVb phylogenetic tree of IVb whole genome haplotypes, with
maximum likelihood and Bayesian analyses. Colored squares = support
values, top left half = Bayesian posterior probabilities, bottom right half
=1450 bootstrap pseudoreplicates. Hashes represent cropped region for
xvi visualization. *=original IVb isolate. The tree was rooted to IVa (GenBank:
JF792424) ...... 124
Figure 3-6 Phylogenies of individual novirhabdovirus gene, based on full coding
sequences of two representatives sequences of all species (see Tables 3.1
and 3.2), with maximum likelihood and Bayesian analyses for (A) NT and
(B) AA. Colored squares designate support values, top left half = Bayesian
posterior probabilities, bottom right=2000 bootstrap pseudoreplicates. Tree
is rooted to the snakehead rhabdovirus (SHRV L-gene, GenBank:
AF147498) ...... 125
Figure 3-7 Host cell viability as measured by Sulforhodamine B (SRB) assay in EPC
cells infected with MOIs (MOI=1x10-8-1.0) of four VHSV-IVb isolates,
96hpi. Average values from a single experiment (conducted in triplicate) are
shown, and are representative of at least three independent experimental
replicates. Standard error bars, *p<0.05; **p<0.01; ***p<0.001 ...... 126
Figure 3-8 Viral yield assay comparison of infectious viral particles produced (pfu/ml)
for wild type (CellC03) and three more recent VHSV-IVb isolates, in BF2
cells 96hpi, following exposure to media collected from each of the above
times post infection. Average values from a single experiment (conducted in
triplicate) are shown, and are representative of at least three independent
experimental replicates. Standard error bars *p< 0.05; **p<0.01;
***p<0.001 ...... 126
Figure 3-9 Antiviral assay comparison of host IFN suppression between reference
(CellC03) and three more recent VHSV-IVb isolates, in EPC cells 96hpi
xvii following exposure to UV irradiated media collected at the above time
points. Values are quantified as the number of antiviral units (uIFN) per mL.
Average values from a single experiment (conducted in triplicate) are
shown, and are representative of at least three independent experimental
replicates. Standard error bars, *p< 0.05; **p<0.01; ***p< 0.001 ...... 127
Figure 3-10 Gene expression of host immune response and viral RNA produced. qPCR
comparisons between reference (CellC03) and three more recent VHSV-IVb
isolates for (A) EPC IFN and (B) virus detected for samples taken at each of
the above time points. Data were normalized to β-actin mRNA levels.
Average values from a single experiment (conducted in triplicate) are
shown, and are representative of at least three independent experimental
replicates. Significance was calculated from Ct values and transformed
values (2-ΔΔCT) are shown. Standard error bars, *p< 0.05; **p<0.01; ***p<
0.001...... 128
Figure 3-A Haplotype network showing genetic relationships among 24 VHSV-IVa full
genomes from POPART. Circles sized according to haplotype frequency
among isolates. Numbers inside parentheses designate NT differences
between each node, unlabeled black circles = hypothesized haplotype steps.
Year and location of isolation are provided below isolate names ...... 129
Figure 3-B Haplotype network showing genetic relationships among 16 VHSV-Ia, Ib,
and III full genomes from POPART. Circles sized according to haplotype
frequency among isolates. Numbers inside parentheses designate NT
differences between each node. Small, unlabeled black circles =
xviii hypothesized haplotype steps. Year and location of isolation are provided below isolate names ...... 130
xix Preface
The chapters of this dissertation are organized into presence and population analyses and complete genome phylogenetic with cell study comparisons, and are largely identical to their published/submitted manuscript versions, with only slight re-wordings.
Chapter 2 is currently in revision and submitted as:
Molecular Ecology. Niner, M.D. and Stepien, C.A.
Chapter 3 is currently in revision and submitted as:
...... Evolution. Niner, M.D., Stepien, C.A., Gorgoglione, B., and Leaman, D.W.
xx Chapter I
Introduction
RNA viruses are common pathogens of eukaryotic organisms, characterized by their rapid evolutionary rates, short generation times, lack of mutation proofreading, and small genomes, whose study can provide insight on host-pathogen coevolution at accelerated rates (Holmes, 2009; Sanjuán et al., 2010; Volz et al., 2013). Often termed an
“arms-race”, this description of host-pathogen interactions was popularized as the Red
Queen Hypothesis, in which selection pressures from pathogens lead to genetic adaptations in their hosts and vice versa (Van Valen 1973, 1974). As RNA viruses mutate barely beneath the limit of mutational meltdown, beyond which genes lose functionality, they can rapidly adapt to new challenges (Pereia and Amorium 2013). Some RNA viruses diversify following a “quasispecies” pattern, or a “cloud-like” array of closely related variants, with the oldest sequences located near the center (Belshaw et al., 2008; Pereira
& Amorim, 2013; Andino & Domingo, 2015). On the periphery of these arrays are more divergent sequences, which augment a genetic reservoir of mutants that possess various attributes, some of which are beneficial and facilitate adaptation to new hosts and novel environments (Quer et al., 1996; Lauring & Andino, 2010; Andino & Domingo, 2015).
Viral Hemorrhagic Septicemia Virus (VHSV), also known as Piscine rhabdovirus
(Walker et al. 2018), has a negative-sense RNA genome (-RNA) and belongs to the
Rhabdoviridae family in the order Mononegavirales. Rhabdoviriuses generally have bullet-shaped envelopes and similar genomes comprising five genes arranged 5’ to 3’ as: nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), and large protein (L), with some accessory genes varying among genera (Kuzmin et al. 2009;
1 Dietzgen 2012). Host ranges are broad within the family, often encompassing invertebrates, fishes, mammals, and plants. They share a monophyletic origin based on
G-gene analysis (Dietzgen & Kuzmin 2012) and our full-genome phylogeny here (Fig. 1-
1). Some of the more notorious members of this family include Vesicular Stomatitis
Virus (VSV) and Rabies Virus (RABV), both which cause fatal disease in mammals and have significant monetary costs (Blokhuis et al., 2008; Hampson et al., 2015). Multiple rhabdoviruses impact aquaculture, including Infectious Hematopoietic Necrosis Virus
(IHNV), Viral Hemorrhagic Septicemia Virus (VHSV), and Spring Viremia of Carp
Virus (SVCV), with the latter having the most economic effects (Purcell et al., 2012).
VHSV and IHNV, along with Hirame Rhabdovirus (HIRRV) and Snakehead
Rhabodovirus (SHRV), all infect fishes and comprise the genus Novirhabdovirus, which is distinguished from other rhabdoviruses by the presence of a sixth gene, non-virion (Nv) that is located between the G and L-genes (Wolf, 1988; Kurath, 2012). Nv is non- essential for replication and its experimental knockout variants did not impact SHRV
(Johnson et al. 2000; Alonso et al. 2004), whereas IHNV (Thoulouze et al. 2004) and
VHSV (Ammayappan et al 2011) had reduced virulence. Nv sequences were not conserved between the novirhabdoviruses (Kurath 2012), and reduced pathogenicity was observed in an experiment that swapped Nv between VHSV and IHNV (Einer-Jensen et al. 2014), indicating species specificity. Current evidence suggests that Nv suppresses host interferon (IFN) immune pathway elements (Kim & Kim 2012, 2013; Einer Jensen et al. 2014) and/or delays the onset of apoptotic cell death (Ammayappan and Vakharia
2011). Single nucleotide substitutions in Nv reduced viral suppression of the host cell
2 immune response (Chincilla and Gomez-Casado 2017; Ballion et al. 2018). The origins of the Nv-gene, whether from a gene duplication event or elsewise, are unknown.
VHSV infects more than 80 fish species across the Northern Hemisphere, making it one of the world’s most devastating viral fish diseases (Escobar et al. 2018). Vaccines for both VHSV and IHNV have been developed (Leong and Fryer, 1993; Xu et al., 2017;
Kole et al., 2019), but an effective strategy to vaccinate hundreds of fish at once is lacking (Purcell et al., 2012; Kim and Kim 2019). Common symptoms of VHSV infection include erratic swimming, bulging eyes, along with internal and external bleeding (Elsayed et al. 2006; Lumsden et al. 2007). Transmission among hosts occurs in the water column via bodily fluids from infected individuals (Meyers and Winton 1995;
Skall et al. 2005). Despite the first known case occurring in Europe, the hypothesized origins of VHSV lie in marine waters of the North Atlantic (Pierce & Stepien 2012). To date, four genogroups of the virus have been characterized (I–IV), with genogroups I–III found in the Northeastern Atlantic (Europe), as a distinct phylogenetic clade from IV
(Meyers & Winton, 1995; Hedrick et al., 2003; Pierce & Stepien, 2012). The oldest- known genogroup is I, which has the broadest characterized geographic range, several designated subgroups, and largest number of known host species (Kurath 2012; Pierce and Stepien 2012), and causes extensive economic losses to European aquaculture
(Abbadi et al., 2016; Ghorani et al., 2016). Genogroup II is found in the estuarine waters of the Baltic Sea, and is most closely related to I (Kurath, 2012; Pierce & Stepien, 2012).
Genogroup III first was described from an outbreak of marine rainbow trout farm in 2007 near Storfjorden, Norway and in several subsequent outbreaks from nearby areas (Dale et al., 2009).
3 Genogroup IV, endemic to North America and Asia, is a sister group to the
European clade (I-III), and contains three geographically distinct subgenogroups across the North American continent (a–c; Pierce & Stepien, 2012). IVa first was documented during outbreaks in the 1980s along the Pacific Coast of North America, where it infected many marine fishes and impacted local species of salmon (Brunson et al., 1989; Hopper,
1989; Meyers et al., 1992). IVa spread to the coastal waters of Japan and Korea in 1996
(Takano et al., 2000), primarily infecting farmed olive flounder (Paralichthys olivaceus)
(Kim et al. 2013). In the freshwater Great Lakes, large outbreaks of IVb occurred during the spring months in 2005 and 2006, resulting in widespread and massive fish kills
(Lumsden et al., 2007; Groocock et al., 2007; Thompson et al., 2011). IVb traces its apparent origin to an infected muskellunge (Esox masquinongy) collected in 2003 from
Lake St. Clair (Ammayppan & Vakharia, 2009). Less studied than IVa and b, subgenogroup IVc appears most closely related to IVb (Pierce & Stepien, 2012; Stepien et al., 2015). IVc was found in the marine/estuarine waters of the North Atlantic in 2000, pre-dating the first appearance of IVb (Gagné et al., 2007). In 2017, a possible fourth subgenogroup was recovered from lumpfish (Cyclopterus lumpus) in coastal marine waters of Iceland (Guðmundsdóttir et al., 2018).
VHSV-IVb has been in the Great Lakes for near two decades. Between 2005 and
2008, IVb caused widespread outbreaks in >30 species of native and non-native fishes
(Lumsden et al., 2007; Groocock et al., 2007; Whelan, 2007). Following those outbreaks,
IVb was detected less often, and has been less studied over recent years (Cornwell et al.,
2015; Stepien et al., 2015). A small outbreak occurred in Lake St. Clair in 2009 (Faisal et al., 2012), and scattered detections were seen in 2010 in Lakes Michigan, Huron, and
4 Ontario, Budd Lake, and the St. Lawrence River (Faisal et al., 2011; Cornwell et al.,
2015). Two additional fishes tested positive, yet were symptom free, in 2012 from Lake
Erie (Stepien et al., 2015). In 2013 and 2014, single gizzard shad (Dorosoma cepedianum) individuals tested positive from Lake Ontario and the St. Lawrence River, respectively (Getchell et al., 2017). Three small outbreaks occurred in 2017 in Lake St.
Clair (M. Faisal and G. Whelan, personal commination, 2017), Cayuga Lake (New York
Finger Lakes), and Lake Ontario (R. Getchell, personal communication, 2017).
VHSV-IVb is most prevalent during spring months, in water temperatures of 9–
12°C (Smail, 1999; Escobar et al., 2016), often coinciding with the spawning of many
Great Lakes fishes (Scott & Crossman, 1973; Trautman, 1981), further enhancing transmission (Stepien et al., 2015). Early outbreaks included large die-offs of freshwater drum (Aplodinotus grunniens), which accounted for the majority of mortalities in the
2005 Lake Ontario outbreak (Lumsden et al., 2007). In 2006 Lake Erie outbreaks, yellow perch (Perca flavescens) and largemouth bass (Micropterus salmoides) also were affected in significant numbers (Kane-Sutton et al., 2010; Stepien, personal observation). In 2007 there were many reports of VHSV in round goby (Neogobius melanostomus), an invasive species introduced to the Great Lakes in the 1990s (Jude et al. 1992) (see Lumsden et al.,
2007; Groocock et al., 2007). Continued detections of VHSV in round goby suggest that it may be a reservoir and vector (Cornwell et al., 2012a,b, 2014, 2015). The 2017 Cayuga
Lake outbreak largely infected round goby (R. Getchell, personal communication, 2017).
Invertebrates also have been shown to harbor VHSV-IVb, including an individual leech (Myzobdella lugubris) (Faisal & Schulz, 2009), Pontoporeiidae amphipods (e.g.,
Diporeia spp.; Faisal & Winters, 2011), and Hyalellidae amphipods (Throckmorton et al.,
5 2017). This suggests aquatic invertebrates may serve as potential VHSV-IVb reservoirs, which may pose added risk of accidental transfer to naïve waterways.
Most VHSV-IVb outbreaks displayed the characteristic symptoms of infection
(Elsayed et al. 2006; Lumsden et al. 2007; Whelan personal communication 2017), however, fishes from recent isolated detections have appeared symptom free (Cornwell et al. 2014; Stepien et al. 2015). It may be those are survivors of recent infections, as experimentally challenged muskellunge (Esox masquinongy) continued to actively shed
VHSV-IVb three months post infection (Kim and Faisal 2012). Alternatively, VHSV-IVb may be shifting toward less virulence. Imanse et al. (2014) examined a 2010 Lake
Ontario isolate in cell culture and found reduced virulence and mortality than in the original isolate (MI03GL). Getchell et al. (2017) found lower viral titers in round gobies infected with tested three other isolates against MI03GL, but all had similar levels of mortalities. Those previous studies did not examine the impact of more recent isolates on the host cells’ immune responses, which can serve as an underlying factor for reduced production of virus and virulence.
Genetic mutations are the root of phenotypic differences, with studies showing that single nucleotide changes can have large impacts on VHSV’s virulence (Baillon et al. 2017). Common mutations seen across multiple isolates may provide evidence of selective pressures on specific sites within genes (Murell et al. 2012; Murell et al. 2013), which in turn can indicate the importance of the gene involved, leading to better understanding of evolutionary trends as well as specific regions to target for vaccine development (Zhang & Gui 2015). Full genome analysis may reveal broader evolutionary trends across multiple genes to aid understanding of host-pathogen interactions (Bayliss
6 et al. 2017). Previous VHSV-IVb evolutionary studies were more limited in scope, either by the number of full genome sequences [4] ( Getchell et al. 2017) or by depth of examination, as none evaluated patterns across time, space, and among hosts. Thus, detailed examination of this virus aids understanding of its history, impact on the Great
Lakes, and possible future trends.
Objectives
The two major objectives of this dissertation were to: (1) determine the current extent and presence of VHSV-IVb in the Great Lakes, and genetic differences and (2) examine different evolutionary patterns of VHSV-IVb based on genetic information based on spatial and temporal relationships. These included three parts, resulting in two separate publications:
(1) Determine the current extent of VHSV-IVb in the Great Lakes by obtaining samples from areas and fishes previously affected by the virus. (Chapter 1 & 2; Niner and Stepien, in review)
(2) Examine evolutionary trends with population genetic analyses to understand how IVb is changing with respect to geographic location, isolate year, and host species (Chapter 2;
Niner and Stepien, in review)
(3) Evaluate differences in vitro of new VHSV-IVb isolates compared to the original isolate (MI03GL) through cell culture assays. (Chapter 2; Niner, Stepien, Gorgolione, &
Leaman, in review)
7 (4) Examine genomic and global trends by sequencing full genome of multiple IVb isolates and comparing and contrasting with existing global VHSV isolates. (Chapter 2;
Niner, Stepien, Gorgolione, & Leaman, in review)
8 100 Kamese virus ArB 9074 (KX497133) 88 Flanders hapavirus CT460 (MH664048) 100 Hapaviruses Wongabel virus (NC011639) 100 Ngaingan virus (NC013955) 100 Coastal plains virus DPP53 (NC025397) Tibroviruses 100 Bivens arm virus UF10 (KP688373) 100 Berrimah virus DPP63 (NC025358) 100 Malakal virus Sudr 1169 64 (NC025400) Ephemeroviruses 100 100 Koolpinyah virus DPP819 (NC028239) 87 Ekpoma virus 2 (NC038283) Tibroviruses Curionopolis virus BE-AR 440009 (KJ701190) Curioviruses 100 Drosophila obscura sigma 10A (NC022580) Sigmaviruses Drosophila melanogaster sigma virus HAP23 (NC038281) 100 Piry virus strain BeAn2423 (NC038286) 100 Perinet virus (NC025394) 98 81 Malpais Spring virus 85-488NM (NC025364) Vesiculoviruses 100 American bat vesiculovirus TFFN (JX569193) 100 Vesicular stomatitis Indiana virus (NC001560) 100 Pike fry rhabdovirus F4 (NC025356) Spriviviruses* 81 Spring viremia of carp (U18101) 100 Fikirini bat rhabdovirus KEN352 (NC025341) 100 Kolente virus DakAr K7292 (NC025342) Ledanteviruses 78 81 100 Kumasi rhabdovirus (NC028236) Eel virus European X-153311 (NC022581) Perhabdoviruses* Tupaia virus (NC007020) Tupaviruses 100 Australian bat lyssavirus 2430 (KT868956) 100 Rabies lyssavirus CNIM1703 (MH267792) 100 Khujand lyssavirus (NC025385) Lyssaviruses 100 100 European bat lyssavirus 1 (NC009527) 84 Ikoma lyssavirus (NC018629) 100 Arboretum virus Lo121 (NC025393) Puerto Almendras virus Lo39 (NC025395) Sripuviruses 100 Cabbage cytorhabdovirus 1 FERA-050726 (KY810772) Cytorhabdoviruses 100 Strawberry crinkle cytorhabdovirus A (MH129615) 100 Alfalfa associated nucleorhabdovirus (MG948563) Nucleorhabdoviruses Iranian maize mosaic nucleorhabdovirus (DQ186554) 100 Hirame rhabdovirus (NC005093) 100 Infectious hematopoietic necrosis virus 1995 WRAC Idaho (L40883) Novirhabdoviruses* 100 Viral hemmorhagic septicemia virus IVb MI03GL (GQ385941) Viral hemmorhagic septicemia virus Ia French Strain 07-71 (AJ233396)
0.7 Figure 1-1. Rhabdovirus phylogenetic tree based on full genome sequences as determined by Bayesian analyses. Bayesian posterior probabilities are on each branch. Tree is rooted to the viral hemorrhagic septicemia isolate MI03GL (VHSV, GenBank: GQ385941).
9 Chapter 2
Evolutionary trajectory of the fish Viral Hemorrhagic Septicemia virus across its history in
the Laurentian Great Lakes: Temporal and spatial patterns
Molecular Ecology. Niner, M.D. and Stepien, C.A.
2.1. Abstract
Viral Hemorrhagic Septicemia virus (VHSV) first appeared in the Laurentian Great Lakes in the early 2000s, as a new and novel subgenogroup (IVb), killing >30 fish species in large outbreaks from 2005-2010. A long interlude period followed, which preceded two small, allopatric, and constrained 2017 outbreaks; none occurred in 2018 or 2019. Our investigation employs a population genetics approach to evaluate VHS-IVb’s spatial and temporal evolutionary trajectory, analyzing G-gene sequence variation, along with a smaller five-gene concatenated data set.
Results show that the G-gene (N=185 individual fish isolates) has diversified into 36 haplotypes, stemming from two originals (“a” and “b”). Haplotypes “a” and “b” differ by just a single synonymous nucleotide substitution, remained the most abundant until 2011, before disappearing.
Group “a” descendants have been most prevalent in the Upper and Central Great Lakes, comprising 14 haplotypes, with eight (51%) having non-synonymous substitutions. Group “b” descendants primarily have been found in the Lower Great Lakes, including 22 haplotypes, of which 15 (68%) contained non-synonymous changes. The virus’ populations have significantly diverged among the Upper, Central, and Lower Great Lakes, as well as showed significant diversification over time. Spatial divergence has structured the overall patterns of nucleotide substitutions, and amino acid changes exhibited pronounced temporal increase. VHSV-IVb thus has significantly diversified across its 10 2.2. Introduction Co-evolutionary responses of viruses and their host populations offers intriguing insights into the myriad of genetic pathways that may ensue over time and space. RNA viruses possess rapid evolutionary rates due to their small genomes, lack of mutation proofreading, and short generation times (Holmes, 2009; Sanjuán et al., 2010; Volz et al., 2013). RNA viruses often diversify into a multi-directional “cloud-like” burst of closely related variants over time, from one or more central ancestral types, termed a “quasispecies” pattern (Belshaw et al., 2008; Pereira & Amorim, 2013; Andino & Domingo, 2015). The resultant pool of similar variants may serve as a genetic reservoir to facilitate adaptation to new hosts and environments (Quer et al., 1996; Lauring & Andino, 2010; Andino & Domingo, 2015). Such patterns of RNA virus diversification have been described for Avian Leukosis Virus (Meng et al., 2016), Deformed Wing Virus (Mordecai et al., 2015), and Viral Hemorrhagic Septicemia Virus (VHSV; Pierce & Stepien, 2012; Stepien et al., 2015); the latter is analyzed here using a combined phylogenetic and population genetics approach. Phylogenetic approaches have elucidated overall evolutionary patterns of emerging and resurging viruses, such as Zika (Faye et al., 2014), West Nile (May et al., 2010) and Measles (Kimura et al., 2015), and in our laboratory’s earlier work with VHSV fish virus (Pierce & Stepien, 2012; Stepien et al., 2015). Such broad scale examinations generally lacked the fine scale resolution to address recent temporal and spatial trends, which can be provided with a population genetics approach. Population genetic investigations analyze changes in gene frequencies over spatial and temporal scales to 11 elucidate the effects of natural selection, drift, and gene flow on mutational variation, as well as their respective influences on fitness and adaptations (summarized by Hedrick, 2011; Lowe et al., 2017). Most traditional virus studies have defined a viral population as an isolate from a single host individual (e.g., Beerenwinkel & Zagordi, 2011; Yang et al. 2012). Here we define a viral population as the gene pool obtained from isolates at a given geographic location and a common time point, comprising a single outbreak. We use partial sequences of the glycoprotein (G-gene) and concatenated gene regions (N, P, M, G, and Nv-genes) to evaluate the evolutionary patterns of a rhabdovirus across its geographic spatial and temporal scales. The few population genetic studies of viruses to date largely have been restricted to agriculturally important plant viruses (Tsompana et al., 2005; Alabi et al., 2010; Lin et al., 2004) or human pathogens (Bahl et al., 2009; Kearney et al., 2009; Pesko & Ebel, 2012). Most examined viral populations either within a single host species (Tsompana et al., 2005; Alabi et al., 2010; Pesko & Ebel, 2012) or a single host individual (Kearney et al. 2009), rather than focusing on broad scale evolutionary trends across a virus’ temporal and spatial history. This new research analyzes genetic variation of VHSV-IVb across the spatial and temporal scales of its nearly two-decade long evolutionary history in the North American Laurentian Great Lakes using field-collected samples of its hosts, which had naturally occurring infections. VHSV is a negative-sense RNA virus of 11,158 nucleotides (NT) that encode six genes: nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), nonvirion (Nv), and large protein (L) 5’N-P-M-G-Nv-L’3 (Wolf, 1988; Kurath, 2012). A unique Nv-gene characterizes VHSV and the three other members of its Novirhabdovirus 12 genus, including Infectious Hematopoietic Necrosis Virus (IHNV), Hirame Rhabdovirus (HIRRV), and Snakehead Rhabdovirus (SHRV); all four infect fishes (Kurath, 2012). The Novirhabdovirus genus belongs to the Rhabdoviridae family, along with Rabies Virus (RABV) and Vesicular Stomatitis Virus (VSV), which both infect mammals, with the latter primarily occurring in livestock (Dietzgen & Kuzmin, 2012). Studies of Rhabdoviridae evolution, including VHSV (Getchell et al., 2017) have shown that they do not recombine (Walker et al., 2011). VHSV and IHNV are the best-studied novirhabdoviruses, with both having broad geographic ranges and exerting economic impacts on aquaculture (Kurath et al., 2003; Kurath, 2012). 2.2.1 VHS evolution, outbreaks, and hosts. VHSV infects over 140 species of fishes in marine, estuarine, and freshwater environments across the Northern Hemisphere, rendering it one of the world’s most serious fish diseases (Escobar et al., 2018). VHSV is hypothesized to have originated from a North Atlantic marine ancestor (Pierce & Stepien 2012), diversifying in four genogroups (I–IV), with I–III forming a clade in the Northeastern Atlantic region (Europe) (Meyers & Winton, 1995; Hedrick et al., 2003; Pierce & Stepien, 2012). Genogroup II diverged in Baltic Sea estuarine waters and is the sister group of I–II, with I mostly found in freshwater and II in North Sea marine and estuarine waters. Genogroup I has a wide and diverse geographic range across Western Europe, several sub-genogroups, and infects the most fish host species (Kurath, 2012; Pierce & Stepien, 2012), having significant aquaculture impacts (Abbadi et al., 2016; Ghorani et al., 2016). Genogroup IV comprises the sister group to the I–III clade, and contains three monophyletic allopatric subgenogroups in North America (a–c; Pierce & Stepien, 2012). 13 In the 1980s, subgenogroup IVa emerged in the coastal Northeastern Pacific, infecting salmonids and many marine fishes (Brunson et al., 1989; Hopper, 1989; Meyers et al., 1992), and in 1996 appeared in the Asian Northwestern Pacific (Takano et al., 2000). The Great Lakes’ endemic subgenogroup –IVb– was back-traced to a 2003 muskellunge (Esox masquinongy) from Lake St. Clair (Ammayppan & Vakharia, 2009). Its first outbreaks occurred during the 2005 and 2006 spring months, resulting in massive and widespread fish kills across the Great Lakes (Lumsden et al., 2007; Groocock et al., 2007; Thompson et al., 2011). In 2000, IVc was discovered in marine/estuarine North Atlantic waters (Gagné et al., 2007); IVc is the sister group of IVb, which together form a clade that is the sister group of IVa (Pierce & Stepien, 2012; Stepien et al., 2015). A possible fourth undesignated subgenogroup was detected in lumpfish (Cyclopterus lumpus) from marine waters of Iceland (Guðmundsdóttir et al., 2018). From 2005 to 2008, widespread VHSV-IVb outbreaks killed >32 fish species with external and internal hemorrhaging (Lumsden et al., 2007; Groocock et al., 2007; Whelan, 2007), nicknamed “fish Ebola” (Hamblin, 2015). The virus then went “underground”, becoming less prevalent and less-studied in the past decade (Cornwell et al., 2015; Stepien et al., 2015). A smaller 2009 outbreak occurred in Lake St. Clair (Faisal et al., 2012), followed by scattered 2010 detections in Lakes Michigan, Huron, and Ontario, Budd Lake, and the St. Lawrence River (Faisal et al., 2011; Cornwell et al., 2015), and in two Lake Erie fishes in 2012 (Stepien et al., 2015). Single infected gizzard shad (Dorosoma cepedianum) individuals were reported from Lake Ontario in 2013 and the St. Lawrence River in 2014 (Getchell et al., 2017). In spring 2017, relatively minor and geographically restricted allopatric outbreaks occurred in Lake St. Clair (M. Faisal 14 and G. Whelan, personal commination, 2017), Cayuga Lake (New York Finger Lakes), and Lake Ontario (R. Getchell, personal communication, 2017). VHSV-IVb most often appears during spring, in water temperatures of 9–12°C (Smail, 1999; Escobar et al., 2016), coinciding with spawning of many Great Lakes fishes (Scott & Crossman, 1973; Trautman, 1981), which may aid transmission (Stepien et al., 2015). Susceptibility varies among fish host species. Freshwater drum (Aplodinotus grunniens) was severely affected in Lake Ontario’s 2005 outbreak (Lumsden et al., 2007) and along with yellow perch (Perca flavescens) and largemouth bass (Micropterus salmoides) experienced large die-offs in Lake Erie during the May–June 2006 outbreak (Kane-Sutton et al., 2010; Stepien, personal observation). Round goby (Neogobius melanostomus), which was introduced to the Great Lakes in the 1990s (Jude et al., 1992), also has undergone high VHS-IVb mortalities (Lumsden et al., 2007; Groocock et al., 2007). In 2010, large proportions of round gobies tested positive, suggesting persistent susceptibility and that it comprises a possible vector (Cornwell et al., 2012a,b, 2014, 2015). The 2017 Cayuga Lake NY outbreak also occurred in round goby (R. Getchell, personal communication, 2017). It has been postulated that some invertebrates may harbor VHSV-IVb, serving as possible reservoirs, including Hyalellidae amphipods (Throckmorton et al., 2017), Pontoporeiidae amphipods (e.g., Diporeia spp.; Faisal & Winters, 2011), and an individual leech (Myzobdella lugubris) (Faisal & Schulz, 2009). It has been suggested that zebra and quagga mussels (Dreissena polymorpha and D. rostriformis) might harbor the virus (M. Faisal, personal communication, 2015 and 2017). That possibility is investigated here, alongside broad scale fish population surveys. An invertebrate 15 reservoir might pose increased threat to naïve areas from transfer or transport of these smaller organisms. 2.2.2 Aim and objectives. Our investigation aims to evaluate whether VHSV-IVb has diversified during its two decades in the Great Lakes, and to elucidate possible spatial and temporal evolutionary patterns using a population genetics approach. We analyze gene sequences from new samples, compiled studies, and the literature to evaluate relationships across its geographic range and time course. We then relate the results with patterns in other viruses, to augment understanding of virus and host co-evolution. 2.3 Materials and Methods 2.3.1 Sampling. We sampled 55 fish species and 2,561 individuals from 32 areas across the Great Lakes region (May–July in 2015 and 2016), focusing on locations of historic outbreaks and past positive records (Figure 2-1, Table 2-1). Locations were: Lakes Superior (N=100), Michigan (N=353), Huron (N=131), St. Clair (N=207), Erie (N=1038), and Ontario (N=273), as well as Budd Lake (Harrison, MI; N=200), Lake Oneida (Bridgeport, NY; N=74), and the St. Lawrence River (Montreal, QB; N=100). Water temperatures ranged from 2–16.5°C and multiple collection methods were employed (e.g., trawls, seines, gill nets, electrofishing) by federal, provincial, and state agency collaborators. In 2016, we used beach seines, kick nets, and electrofishing to further target the nearshore community. Following 2017 outbreaks, we sampled an additional 88 fishes in May–July from Lakes Erie, Michigan, and Ontario, and Cayuga Lake, in water temperatures of 12–22°C. No outbreaks were reported or occurred during any of our sampling. Collection data included: date, method, depth, water temperature, 16 GPS coordinates, species, total fish length (TL), sex (if determinable), and hemorrhages and/or other visual disease symptoms. Fishes were euthanized by overdose of 25mg/mL tricaine methanesulfonate (MS- 222; Argent Chemical Laboratories, Redmond, WA) and sacrificed under University of Toledo Institutional Animal Care and Use Committee (IACUC) protocol #106419. Individual surgical sites (anus–operculum) were sterilized with an alcohol wipe and a new razor blade; the spleen was removed using sterile forceps and placed in a sterile labeled tube containing 1.5 mL RNAlater (Qiagen, Valencia, CA). Each liver was wrapped individually in sterile aluminum foil, and archived at -80°C. Spleen and liver tissues later were pooled in single tubes for up to five individuals (<50 mm TL) of the same species, collection, and location. Between specimens, forceps and aluminum dissection trays were sterilized with disposable alcohol wipes, razor blades discarded in sharps containers, and gloves changed. Specimen disposal followed the respective agency protocol and/or UT IACUC#106419. In 2015, to examine potential IVb harboring, we collected 50–100 dreissenid mussels from all sites except Budd Lake, where sufficient numbers occurred, and tested them for VHS. Each mussel was smashed on a sterile surface with the back end of sterile forceps. Half of each was placed in RNAlater (as above) and the other half wrapped in foil and stored at -80°C. 2.3.2 RNA extraction and reverse transcription. Tissues from up to 11 individuals per location were pooled in sterile 1.5mL tubes for initial processing and ≤ 0.5g ground with a sterilized mortar and pestle under liquid nitrogen. RNA was extracted with the Trizol® (Molecular Research Center, Inc., Cincinnati, OH) protocol, re- 17 suspended in 30µL RNase-free water, and quantified with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). Reverse-transcription to complementary (c)DNA was performed by first incubating 1μg RNA with 100ng of random hexamer primer in 7μL at 65°C for 10 min. Reactions were cooled to 4°C before adding 13μL M-MLV-RT mixture (10X First Stand buffer (Ambion Life Technologies, Grand Island, NY), 10 mM dNTPs, 0.05mM random hexamers, 25 U/µL RNasin, and 200 U/µL M-MLV), and incubated at 42°C for 1h following Stepien et al. (2015). cDNA was labeled and stored at -80°C. 2.3.3 qPCR tests for VHSV-IVb and quantification. Presence/absence of VHSV-IVb was determined using our laboratory’s SYBR® Green quantitative PCR assay protocol (Pierce et al., 2013a,b) for the N-gene. Cell culture standards spiked with known titers (0, 10, 1x10, 1x103, and 1x104 pfu RNA/ml) provided reference VHSV-IVb readings. To circumvent false negatives, simultaneous control ß-actin housekeeping gene assays were run alongside VHSV reactions. Runs contained a positive ß-actin/negative virus control (0pfu/mL), low concentration of the virus (10pfu), high-virus control (1x104pfu), and nuclease-free water as a negative control. Pooled samples were tested in triplicate for ß-actin and VHSV. Individual samples from pools reading VHSV positive (Ct, <35) were extracted and assayed, as described above, to pinpoint individual isolates. Following identification of VHSV-positive samples, their cDNA was quantified for number of VHSV versus ß-actin molecules using our two-color fluorometric real-time PCR (2CF-qPCR) assay (Pierce et al., 2013b). This employs two internal standards (IS), respectively for the VHSV-IVb N-gene and the ß-actin gene (Pierce et al., 2013b), whose 18 output diagnoses among three alternatives: (1) reaction failure (no amplification), (2) VHSV-negative (amplification of VHSV IS alone), or (3) VHSV positive (amplification of both the sample and IS). Variable ratios of VHSV and ß-actin IS are used to quantify relative amounts of the products (Pierce et al., 2013a,b), with reactions run in triplicate, reporting means and standard errors. Viral levels were compared to prior haplotype “a” challenge experiment results (Pierce et al., 2013b). 2.3.4 Preparation of historic isolates in cell culture. Samples from historic VHSV outbreaks (2006–2011) were received from G. Kurath (USGS, Seattle, WA) as frozen media from BF2 cell culture or as RNA, to which 30µL and 150µL of serum free MEM (minimum essential media, ThermoFisher Scientific Waltham, MA) was added per well of a 12-well plate confluent with BF2 cells. Cells were incubated with media at 20°C for 1h, after which media was replaced with 10% serum MEM, and incubated at 20°C for 2.3.5 Sequencing VHSV isolates. VHSV-IVb isolates from 21 individual positive samples collected in 2015-17 were amplified using polymerase chain reactions (PCR) and sequenced for targeted regions of the G-, N-, P-, M-, and Nv-genes following Pierce & Stepien (2012) and Stepien et al. (2015). Additionally, we sequenced 23 historic isolates that lacked reported data (Supplementary Table 2-A, Gene group “G/All”). PCRs used primers (Table 2-2) designed from the original IVb isolate, MI03GL (Ammayappan 19 & Vakharia, 2009; GenBank #GQ385941.1), whose sequence is designated as haplotype “a” (Pierce & Stepien, 2012; Stepien et al., 2015). Gene regions were concatenated for analyses. Positive controls of “a” were re-sequenced to confirm accuracy, and nuclease free ddH2O served as a negative control. Results were checked on 1% agarose gels stained with ethidium bromide (EtBr) under UV light. Target PCR products were excised from gels and purified using QIAquick Gel Extraction kits (Qiagen, Valencia, CA). Sequencing was outsourced to Cornell DNA Services (Ithaca, NY) and aligned and analyzed by us. Any and all nucleotide differences from “a” in the historical and new samples were confirmed with corresponding trace files. 2.3.6 Genetic data analyses. We compiled our 44 sequence isolates with an additional 140 from NIH GenBank (https://www.ncbi.nlm.nih.gov/GenBank) that originated from the MEAP-VHSV database (http://gis.nacse.org/vhsv/) (N=114), Cornwell et al. (2014) (N=10), Stepien et al. (2015) (N=11), Getchell et al. (2017) (N=4), and M. Faisal and G. Whelan (personal communication, 2017) (N=1), for a total of 184 VHS-IVb sequences for the 669nt central G-gene segment. The G-gene was the primary focus, due to sequence availability and more robust sample sizes. We compared those results to analyses based on the multi-gene concatenated data set, totaling 47 isolates, which included 41 from our own sequencing, two others from GenBank, and four from Getchell et al. (2017). Genomic regions included coding sequences for the N (121nt), P (540nt), M (563nt), G (669nt), and Nv (365nt) genes, totaling 2,258nt. Population genetic relationships using the G-gene sequence data are tested among: (1) three time periods (Early: 2003–06, Middle: 2007–10, and Later: 2011–18), 20 (2) three primary geographic regions (Upper Great Lakes: Lake Superior, Lake Michigan, Lake Huron, Budd Lake, Lake Simcoe, and inland Wisconsin lakes; Central Great Lakes: Lakes Erie and St. Clair, Baseline Lake; and Lower Great Lakes: Lake Ontario, St. Lawrence River, and New York Finger Lakes), and (3) the top six fish species affected (freshwater drum, gizzard shad, largemouth bass, smallmouth bass (Micropterus dolomieu), round goby, and yellow perch). Time periods were formulated to contain comparable numbers of samples in each of the three groups. Three isolates from Clearfork Reservoir, Ohio were excluded from population analyses due to their location outside of the Great Lakes drainage, and small sample size. We further compare populations from individual water bodies (Lake Michigan, Lake Huron, Budd Lake, Lake St. Clair, Lake Erie, New York Finger Lakes, and St. Lawrence River), excluding Lake Superior since it comprised a single isolate at a single time point. Haplotypes are defined here as “unique gene sequences that differ by one or more nucleotide substitutions” from haplotype “a”. New haplotypes from the concatenated genes are designated with the G-gene sequence haplotype letter, followed by a number (e.g., “a1”). We calculated haplotypic and nucleotide diversity, and number and relative proportion of private haplotypes in populations with ARLEQUINv3.5 (Escoffier & Lischer, 2010). Evolutionary relationships among haplotypes are depicted with POPART (https://popart.otago.ac.nz) and TCS networks (Clement et al., 2000). Pairwise divergences between populations are analyzed using θST (FST analogue; Weir & Cockerham, 1984) in ARLEQUIN and with exact tests of differentiation (x2) in GENEPOP v4.6 (Raymond & Rousset, 1995; Rousset, 2008). The latter employs a MCMC procedure with 1,000 batches and 10,000 iterations to randomly sample allelic 21 frequencies. Probabilities are adjusted with sequential Bonferroni correction (Rice, 1989), and reported both prior and after adjustment, to identify borderline cases. Tajima’s (1989) D tests in ARLEQUIN evaluate possible influence of selection. Analysis of Molecular Variance (AMOVA) in ARLEQUIN assesses hierarchical partitioning of genetic variation among: 1) three geographic regions (Upper, Middle, and Lower Great Lakes) and their sampling events, and 2) three time periods (Early, Middle, and Later) and their sampling events. We test these scenarios for all nucleotide (NT) substitutions and separately for amino acid (AA) substitutions, using the G-gene. A neighbor-joining genetic distance tree analyzes population relationships using Reynold’s RST genetic distances (Reynolds et al., 1983) in PHYMLv3.697 (Felsenstein, 2007) with 10,000 bootstrap pseudo-replications (Felsenstein, 1985). Possible relationships between genetic distance (θST) and geographic distance are evaluated with separate Mantel (1967) tests for populations from the Early, Middle, and Later time periods, using shortest waterway distances (km) between outbreak locations or the most direct road route for land-locked locations. We also test the relationship between genetic distance (using all samples) and time (sampling years). Phylogenetic trees are derived using PHYML and MRBAYES v3.2.1 (Ronquist and Huelsenbeck, 2003) and taxon divergence times with BEAST v1.10.4 (Suchard et al., 2018), following Stepien et al. (2015). Nucleotide (NT) and amino acid (AA) substitutions are evaluated for all isolates and referenced to haplotype “a”. 22 2.4 Results 2.4.1 VHSV-IVb Detections. We collected and analyzed 2,649 individuals from 45 species in 2015–17. Our qPCR assay results identified just 21 VHSV positives (0.82%). Only two were found in 2015 – from a white perch (Morone americana) and a round goby in Lake Erie’s central basin, with each having unique G-gene haplotypes, designated as “u” and “w” (Table 2-1). In 2016, 19 tested positive: 14 from Lake Erie, including one from the central basin (largemouth bass) and 13 from a single sampling event in the western basin (one emerald shiner (Notropis atherinoides), nine gizzard shad, one pumpkinseed (Lepomis gibbosus), and two largemouth bass), and five from Lake Michigan (four round goby and one alewife (Alosa pseudoharengus)). These had two new haplotypes: 13 with “w” and six with “x” (Table 2-1). No VHSV positives were detected from 1,003 dreissenid mussels collected in May–September 2015; thus no further testing was conducted. Additionally, none of the 2017 fishes tested VHSV-positive. Our 2CF qPCR analyses (Figure 2-2) discerned VHSV-IVb concentrations (log values) for the 2015 white perch (haplotype “u”) of 1.8 x103 VHSV/106 ß-actin molecules, with the round goby 2015 individual (“v”) being much higher at 5.2 x106. The latter concentration was the highest recorded, much above the challenge experiment disease symptom threshold based on haplotype “a”, yet the fish exhibited no visible VHSV signs (Pierce et al., 2013). A wide range of viral concentrations (5.1 x10-1 to 1.9 x106) occurred in our 2016 positives (“w” and “x”), as well as in the challenged fish (haplotype “a”) (the latter from Pierce 2012). Three possessing haplotype “w” also were above the clinical disease symptom threshold (Figure 2-2). 23 2.4.2 Evolutionary patterns from the G-gene. We analyzed a total of 185 G- gene sequences, which included 36 separate haplotypes, and 47 concatenated gene sequences that had 33 haplotypes. Figures 2–3 show the consensus phylogenetic trees from maximum likelihood (PHYML) and Bayesian (MR. BAYES) analyses with their estimated divergence times (BEAST), which is congruent with the tree from Stepien et al. (2015), but expanded to include the additional haplotypes. Two major clades are shown, with the new haplotypes from 2016 and 2017 (“w”, “x”, and “bd”) contained inside clade 1 along with haplotype “a” and those from 2015 (“u” and “v”) in clade 2. Evolutionary rates for the G-gene were similar in both datasets (G-gene: 1.00x10-4 substitutions/site/year; concatenated: 1.01x10-4). The overall rate for the concatenated genes data set was slower, at 8.26x10-5. Figures 2–4 depict the haplotype networks, including separate networks (B, D, F) based on non-synonymous substitutions alone (those leading to AA changes). Two predominant haplotypes, “a” and “b”, are centrally located as the largest circles, containing 74 (40%) and 45 (24%) respective isolates. A single synonymous transition from cytosine to guanine at position 3996 separates haplotypes “a” and “b” (Sup. Table 2-B). Thirteen unique haplotypes descend from “a”, with mean divergence of 1.46±0.22 NT. Twenty-one unique haplotype descendants surround “b”, diverging by a mean of 1.67±0.19 NT. Regional patterns are apparent (Figure 2-4A, Table 2-3), with 81% of haplotype “a” occurrences in the Upper (30%) and Central Great Lakes (51%), and 93% of “b” in the Lower Great Lakes. Similar geographic separations characterize the descendants: isolates in the “a” group were more prevalent in the Upper (N=9, 25%) and Central Great Lakes (N=24, 67%) and “b” in the Lower Great Lakes (N=23, 74%). 24 Temporal patterns also are apparent in the network (Figure 2-4C, Table 2-3). Haplotypes “a” and “b” were most abundant during the Early time period, comprising 51% and 46%. During the Middle time period, “a” remained more common (56%) than “b” (19%), and both declined during the Later period (16% and 10%), disappearing after 2011. Most haplotypes from the Later time period are genetically distant from the central “a” and “b” haplotypes (mean=2.45±0.17 steps), with all five from 2015–2017 being unique. None of the “b” descendent haplotypes originated during the Early time period, with six (29%) appearing during the Middle, and the remaining 15 (71%) during the Later period. A diversity of host species (Figure 2-3E–F) had “a” and “b” haplotypes; none predominated “a” occurrences, but haplotype “b” and its descendants frequently were in round goby (38%). Overall, “a” was found in 13 host species and “b” in 11, with nine having either “a” or “b”. No genetic patterns are apparent among host species having “a” group haplotypes, however, there were multiple occurrences in gizzard shad (N=12, including haplotype “w”), round goby (N=4, all “x”), and bluegill (Lepomis macrochirus, N=3) (Figure 2-4E). All emerald shiner and white bass (Morone chrysops) contained haplotype “a” and the single white perch possessed "b". Both positive invertebrate samples (leech and amphipod) had “a”, according to Faisal & Schultz (2009) and Faisal & Winters (2011). Mean number of substitutions and relative percentage of AA changes significantly increased over time (Table 2-4A). The Later time period contained the most, averaging 1.62 substitutions and 1.05 (65%) AA changes. The Middle period averaged 1.39 and 0.85 (61%), and the Early period had just 1.00 and 0.33 (33%). Among the three 25 geographic regions, the Central Great Lakes possessed the most, with 1.69 mean NT and 1.23 (72.8%) AA changes, with fewer in the Upper (1.50, 1.10; 73.3%) and Lower Lakes (1.25, 0.50; 40%); i.e., more substitutions were synonymous in the latter. 2.4.3 Patterns from concatenated gene analyses. Lower sample sizes (N=47) and lack of available samples from some areas and temporal periods precluded in-depth population analyses using concatenated sequences of the N-, P-, M-, G-, and Nv-genes, versus the larger sample sizes available for the G-gene. Changes within the gene regions per haplotype are reported in Sup. Table 2-B, with compositions of substitutions per gene in Sup. Table 2-C. In the concatenated gene analyses (Figure 2-5), haplotype “a” comprised 13 isolates, all from the Central Great Lakes (100%) during the Early time period (77%). Haplotype “b” was 10nt steps away (5 AA), having a sole representative in the Upper Lakes during the Middle period (note that we lacked more isolates to analyze from the Lower Lakes). The “w” isolates and descendants (from 2016) were the most divergent, differing from “a” by 14–18nt substitutions (7 AA). Haplotypes “k” and “l” from 2012, “u” and “v” from 2015, and the “w” clade from 2016 also were respectively distant, showing unique trajectories (Figure 2-5). Overall relationships among geographic locations and across time were congruent between the concatenated and G-gene networks. The Later time period haplotypes diverged most from “a”, by 11.14 mean substitutions and 3.86 AA changes (35%), vs. 3.50 and 0.50 (14%) for the Middle and 1.75 and 0.26 (15%) for the Early period, using the concatenated sequence data (Table 2- 4A). Among geographic regions, the Central Great Lakes contained the most NT (7.05) 26 and AA changes (2.79; 40%). The Upper Great Lakes had fewer (5.89 and 2.44; 41.4%) and the Lower Great Lakes region had even less (4.50 and 1.96; 44%); however, fewer samples were in the latter. Overall trends agreed with the G-gene. 2.4.4 Population genetic trends. Haplotype network results were statistically supported by pairwise genetic divergence (θST) analyses. Pronounced population genetic divergence occurred over time, as shown in the G- and concatenated gene results, with the Later period differing from both the Early and Middle time periods (Table 2-5A). Populations from all three regions (Upper, Central, and Lower Great Lakes) significantly diverged (Table 2-5B), with the greatest difference between the Central and Lower Lakes. Among individual water bodies (Table 2-5C), the St. Lawrence River differed the most. Virus populations from Lakes Erie and St. Clair were more similar to each other, but very divergent from Lakes Michigan and Ontario. Hierarchical relationships among the major G-gene population groups using AMOVA (Table 2-6A) revealed significant partitioning among the three Great Lakes’ regions (18%, p<0.001) and among their component sampling events (51%, p<0.001), totaling 69%. AMOVA (Table 2-6B) found less but significant variation among the three time periods (0.52%, p<0.001), and more among their sampling events (67%, p<0.001). AMOVA analyses for AA substitutions (Table 2-6B) indicated greater genetic structuring among time periods (7%, p<0.001) and their sampling events (63%, p<0.001) than among geographic regions (Table 2-6B1). The neighbor-joining genetic distance tree (Figure 2-6) examines the relationships among virus populations, which each were collected in a single area at a single time. The resultant tree depicts two primary population clusters, with a lower cluster dominated by 27 haplotype “b” and its descendants, and an upper cluster in which haplotype “a” and its descendants predominate. The lower cluster mostly contains samples from the Lower Great Lakes region. Their earlier samples (2005–8) are the most distant, and those from more recent years, 2009 and 2011, located closer to the root. Two of the earlier samples from Lake Ontario (2006 and 2007) are found near the tree’s center. Lake St. Clair samples from 2003–2009 cluster together, along with others from the Upper and Central Great Lakes. The most recent sample from Lake St. Clair (2017) appears very distant from those. Lake Michigan and Budd Lake samples also are spread apart across the “a” top half of the tree. No trends were apparent among outbreak events (* on Figure 2-6). Mantel tests (Figure 2-7) support a positive relationship between genetic distance and geographic distance for the Early time period (Figure 7A; p=0.006), but not for the Middle (Figure 2-7B) or Later (Figure 2-7C) periods. The Early period had just four unique haplotypes (N=58), whose distribution showed increasing genetic divergence (θST) with increasing geographic distance from the original haplotype’s location in Lake St. Clair. The relationship between genetic divergence (θST) and time (years) was significant across the entire data set (R2=0.15, p=0.002; Figure 2-7D). Frequency of identical haplotypes fell off sharply after the 2005–2007 outbreaks, with most isolates from 2009 and beyond having new and divergent haplotypes, while “a” and “b” became rare (disappearing after 2011). Although the Middle and Later periods do not display correlation betweenθST and geographic distance (Figure 2-6, B–C), this likely reflects overall increasing diversity of haplotypes. Tajima’s D tests indicate that the G- and concatenated gene regions were under significant purifying selection, having negative values (Table 2-7). Only the population 28 from the Middle time period was significant in itself, with the Early and Later time periods also being negative (but not significant). The Upper and Lower Great Lakes populations likewise were significant. For tests on individual water bodies, the St. Lawrence River population alone was significant. The top eight species also were analyzed for selection pressure, with round goby being significant (Table 2-7). These variations likely reflect sample size effects, since far more round goby individuals (N=50) were collected than any other individual species (N=5-23). 2.5 Discussion 2.5.1 VHS-IV occurrences and evolutionary trajectory. VHSV-IVb has undergone extensive evolutionary changes across its Novella & Presloid 2012; Stepien et al. 2015). This quasispecies diversity may allow the virus to persist in a population under consistent and/or variable environmental conditions, and to enter new hosts and their habitats. VHSV-IVb has continued to diversify following a quasispecies pattern, with continued radiation of new variants from the central “a” and “b” haplotypes. Variants displaying high virulence frequently are outcompeted over time by less virulent ones, which can better persist in the host population (Ojosnegros & Beerenwinkel, 2010). Our 29 2015–16 sampling uncovered just a few positives, which comprised new haplotypes and a wide variety of viral titers. Their fish hosts all lacked hemorrhages and other symptoms. In contrast, the first VHSV-IVb outbreaks were characterized by high virulence and mass die-offs, which infected multiple fish species (Kim & Faisal, 2011). Similarly, the classic example of Australian Myxoma virus began with high mortality in feral rabbits, which lessened over time due to adaptations between the virus and host populations (Elsworth et al., 2014; Alves et al., 2019). This pattern also appears to characterize VHS-IVb, whose initial host fish populations may have been more susceptible than later ones, due to acclimation and resistance, while the virus has continued to diversify over time in response. 2.5.2 Evolutionary patterns across space and time. Evolutionary diversification of VHS-IVb, based on our G-gene results, radiated from the original haplotypes “a” and “b”, with the “a” descendent group primarily found in the Upper and Central Great Lakes, and the “b” group in the Lower Lakes and St. Lawrence River. They may have originated from separate introductions into the Great Lakes, with “a” in the Upper and Central Lakes, and “b” in the Lower Lakes. An alternate hypothesis is that “b” descended from “a”, with the latter first appearing in Lake St. Clair ca. 2003. Haplotypes “a” and “b” appeared nearly monotypic during the Early time period (2003–06), accounting for 90% of the known isolates, but both haplotypes disappeared after 2011. Our results reveal significant genetic divergences among VHS-IVb population groups from the Upper, Central, and Lower Great Lakes, with greatest difference between the Central and Lower Lakes. The St. Lawrence River population was the most divergent. Populations in Lakes Erie and Ontario also significantly differed from most 30 others. Most sequence diversification and new haplotypes occurred during the Middle and Later time periods, i.e., after the major 2005 and 2006 outbreaks. The closely related IHNV novirhabdovirus also exhibited geographic patterning among three distinct geographic Northeastern Pacific regions, with its” group having more substitutions than its “L” and “U” groups (Kurath et al., 2003). Similarly, IHNV first was reported in China in 1985, likely introduced from Japan, and now constitutes a distinct clade (Xu et al., 2018). Since regional differences in VHSV-IVb are based on a history of In our study, genetic divergence (θST) is strongly correlated with geographic distance. The relationship between genetic divergence and time (years) also shows significant positive correlation, indicating continuing changes over time. Increased genetic divergence has been accompanied by increasing overall diversity, consistent with the quasispecies theory. In contrast to IVb, evolutionary relatedness of IHNV G-gene sequences along the Northeastern Pacific did not correspond to time (Kurath et al., 2003). Another Northeastern Pacific IHNV study identified greater diversity in viral isolates in later years, but attributed those to increased sampling effort and surveillance (Black et al., 2016), which was not the case in our study. Like the genetic structurings of VHSV-IVb and IHNV, Balkan RABV isolates (N=210) formed five genetic groupings, whose relationships better corresponded to geographic region than to time (McElhinney et al. 2011). Unlike the more rapid evolution detected here for VHSV-IVb, McElhinney et al. (2011) found that RABV genetic 31 composition did not differ from samples collected 20 years earlier. Like VHSV-IVb, genetic similarity among isolates of another rhabdovirus, Vesicular Stomatitis virus (VSV) for cattle outbreaks from Mexico, supported greater relationship to geographic location than to time periods (Velazquez-Salinas et al., 2014). Over the course of VHSV- IVb’s evolutionary history, we discern that spatial structure exhibited more pronounced patterning than temporal structure, although both are significant. Viruses that are more evolutionarily distant from VHSV also provide spatial and temporal analogies. The VP gene of Infectious Bursal Disease Virus (IBDV) displayed more differences among regions within the Iberian Peninsula (24%), than across a 20- year time span (3%; Cortey et al., 2012). Ebola virus (EBOV) outbreaks from 2013–15 showed more genome changes earlier than later (Hoenen et al., 2015). However, Gire et al.’s (2014) examination of the longer-term relationship of EBOV isolates from the 2014 Western African outbreak uncovered significant divergence from the older 1970s outbreaks. Whether diversification occurs over time and population genetic structure develops varies among different viruses, likely reflects co-evolution with their hosts. 2.5.3. Substitution rates, types, and patterns. IVb appears to evolve at a fairly consistent rate over time. We previously determined its rate to be 2.8x10-4 substitutions/nucleotide site/year (Stepien et al., 2015), which was slightly faster than our 1.0x104 estimate here, based on 77 additional isolates. Our concatenated gene rate was slower at 8.3x10-5 vs. 6.6x10-5 (Stepien et al., 2015), however we included more of the Nv-gene and the entire N-gene, which were not assessed previously. Non-synonymous substitutions constituted 41% of the 34 substitutions in the G- gene region examined here (668 NT). Such diversification may enhance the virus’ ability 32 to infect new hosts. Most G-gene haplotypes differed by just single substitutions (62%). In comparison, using a 360nt G-gene region, Benmansour et al. (1997) identified 92% AA conservation between the Northeastern Pacific IVa genogroup (6 isolates) and the European genogroups I–III (8 isolates), and 98% similarity within IVa. Stone et al. (1997) found similar results across 741nt, with IVa (10 isolates) and European genogroups (19 isolates) being 82% similar, and 96% within IVa. Within IVb, we discerned 98.7% sequence similarity and 99.8% AA similarity based on our G-gene haplotypes, and 99.0% sequence similarity and 99.2% AA similarity with the concatenated sequences. The overall number of AA changes and evolutionary rate of IVb to date likely reflects its generalist host range and/or lack of co-evolutionary time (and host defense development) in the Great Lakes, since its first outbreaks in 2005. In comparison, IHNV displayed 91% sequence conservation for a 303nt G-gene segment of 323 Northeastern Pacific coastal isolates (Kurath et al., 2003). Complete IHNV G-gene sequences showed 94.4% sequence conservation among 38 international samples (Nishizawa et al., 2006). G-gene sequences from 61 rabies (RABV) cases in Brazilian livestock had 98% NT similarity and 97% AA conservation, with 18/27 (67%) mutations resulting in AA changes (Cargnelutti et al., 2017). A similar study in India examined 25 full-length G- gene RABV sequences from six host species, revealing 96% sequence and 96% AA conservation (Cherian et al., 2015). Those amino acid changes occurred away from antigenic sites, suggesting conservation of critical gene functions and avoidance of the host immune system (Cherian et al., 2015). Antigenic sites in VHSV are known at AA residues 140 and 430 (Béarzotti et al., 1995), however our study did not cover those 33 regions. Stone et al. (1997) examined VHSV-I and IVb isolates against an IHNV antigenic site between AA residues 230-231, but found none in VHSV. Our IVb isolates also were conserved at those positions. Analyses of concatenated sequences from the VHS-IVb N-, P-, M-, G-, and Nv- genes documented changes across the genome, finding that 38% of the 113 mutations were non-synonymous. In comparison, full genome sequencing by Getchell et al. (2017) of four isolates revealed 87 changes (0.79%) across ~11,000 NT, with 30% resulting in AA changes, similar to our percentage. A Zika virus study found just 16% nonsynonymous changes among 1030 mutations across 110 whole genomes (Metsky et al., 2017), similar to the proportion we identified here in glycoprotein alone. Expressed genetic variation in VHSV-IVb thus appears high. 2.5.4. Gene-specific variation. The G-gene encodes the glycoprotein of the viral capsid, which enables the viral particle to attach and enter the host cell via endocytosis (Kurath, 2012). In response, fish hosts mount immune defenses against glycoprotein (Dietzgen, 2012); glycoprotein’s relatively high mutation rate and diversification may help the viral population to avoid detection (Stepien et al., 2015). Adaptive radiation in the VHSV-Ia G-gene occurred in European freshwater rainbow trout (Oncorhynchus mykiss) culture following flooding of fish farms with marine water carrying marine host- based VHSV-Ia (Schönherz et al., 2018). Radiation of IVb G-gene variants may facilitate adaptation to match cell receptors of particular hosts, such as round goby and gizzard shad, in similar manner to that of Ia in rainbow trout. Of the individual genes we examined, the novirhabdovirus genus-specific Nv- gene of IVb had more substitutions (4.1%), as previously observed across all global 34 VHSV genogroups (Pierce & Stepien, 2012), within the European genogroups (Bascuro et al. 1995), and for IHNV (He et al. 2013, 2014). Conversely, Getchell et al. (2017) found a single substitution in one of their four IVb genomes, whereas Bascuro et al. (1995) found one synonymous change in one of nine IVa isolates, likely due to their smaller sample sizes. The Nv-gene encodes a small, non-structural protein (Biacchesi, 2011) and has the fastest mutation rate of the VHSV genes (Stepien et al. 2015), suggesting potential role in evading host detection and/or defense (Pierce & Stepien, 2012). Ammayappan & Vakharia (2011) found that Nv- knockout mutants induced cell apoptosis earlier than non-altered (wild-type) VHSV-IVb and IHNV, suggesting that Nv prolongs the length of infection, resulting in increased transmission potential. M- and P- encode proteins directly involved in viral replication. The M-protein inhibits host cell immune responses and promotes viral replication (Biacchesi et al., 2002; Pore, 2012; Ke et al., 2017). The M-gene sequences of VHS-IVb’s haplotype “a” differed by four AAs from IVa, with cell culture experiments showing that “a” better inhibited host cell immune response, while Ia was more effective at blocking host cell transcription (Ke et al., 2017). Similarly, the P-gene’s protein inhibits interferon activation and promotes viral replication (Biacchesi et al., 2002; Pore, 2012). P serves as a chaperone for N-protein synthesis, preventing it from binding to cellular RNA products, and also forms the RNA polymerase complex with the L-protein (Dietzgen, 2012). The M- and P-genes possessed fewer substitutions (2.8% and 2.6%, respectively). Low diversity in the P- and M-genes seen here indicates that they are relatively conserved, likely due to their importance in viral replication. M had the most non- synonymous changes (50%), despite possessing the fewest mutations, suggesting 35 selection advantage. Haplotype “u” in white perch from central Lake Erie in 2015 accounted for 25% of the M-gene substitutions. No other haplotypes had more than a single non-synonymous M-gene change, congruent with prior findings (He et al., 2014; Stepien et al., 2015). Similarly for P-, one isolate (“b”) had three AA changes, accounting for 75% of the non-synonymous changes. Such accumulations of AA changes in lone isolates may affect viral replication, and in the case of “u”, lead to less virus concentration in the host tissue (1.781x103 VHSV per 106 ß-actin molecules) and reduce the host’s symptoms. Moderate variation was discerned in the complete coding region of the N-gene (2.6%; 1215nt), having 32% non-synonymous substitutions. In comparison, partial N- gene sequences (422nt) from 16 isolates of VHSV genogroups I–IV (including only IVa) varied by 2–16% (Einer-Jensen et al., 2005). Four IVb genomes sequenced by Getchell et al. (2017) contained seven N-gene synonymous substitutions. Our sample size was ~12X greater than Getchell et al.’s (2017). The N-gene encodes the nucleocapsid protein involved with balancing viral transcription and replication. N-protein binds to viral RNA, forming a complex that serves as the template for transcription and replication, and additionally interacts directly with P- and L-proteins at multiple binding sites (Dietzgen, 2012); thus, conservation of this gene is expected. The remaining gene not examined here, L, has been relatively little studied in VHSV. Schönherz et al. (2016) sequenced the full genomes of four VHSV-Ia isolates, finding the most mutations in the G- and L-genes (5,954 NT), with L- being the largest of VHSV’s six genes. The L-protein functions in viral transcription and replication (Kurath, 2012), with its variants increasing the virus’ temperature range to >20°C, as 36 demonstrated with gene-swapping (Kim et al., 2015). Further investigations of IVb patterns should examine L-gene variation, in relation to geographic locations and ecological conditions. 2.5.5. Host species generality, specificity, and infection. Less than 1% of fishes from our 2015–2017 sampling tested VHSV positive. In comparison, 2010 samples from the Great Lakes described 13% VHSV-IVb incidence among >5000 fish individuals, with 25% in round goby, which was targeted due to its high infection incidence (Cornwell et al., 2015). That was a decrease from 16% positives in 2009 (Cornwell et al., 2012a). Our results indicate that VHSV infection has continued to decrease. More similar to our results, 2009–11 sampling in coastal Norway identified <1% of 943 fishes as VHSV-I positive (Sandlund et al., 2014). Moreover, sampling for VHSV-Id around Finland in 2005–08 detected no positives in wild fishes (1636 individuals, 17 species) despite collection near aquaculture net pens that had experienced recent outbreaks (Vennerström et al., 2018). Thus, global VHS occurrence and predictability appear sporadic. All dreissenid mussels tested VHSV-negative. Throckmorton et al. (2017) likewise discovered no VHSV in cylindrical papershell mussels (Anodontoides ferussacianus) from Budd Lake, despite positive largemouth bass and Hyalellidae amphipods from that location. Researchers in Denmark tested for a VHSV-II invertebrate reservoir, in isopods, krill, and squid, but none were positive (Skall et al., 2005). To better understand whether invertebrates may serve as a VHS-IVb reservoir or transmission vector, comprehensive sampling and testing during VHSV outbreaks should be conducted, along with challenge studies on specific taxa, focusing on species in areas experiencing outbreaks. 37 Among VHSV genogroups, IVb infects the broadest fish host range, suggesting higher ability to spread to naïve species (Escobar et al., 2018). Across genogroups I–IV, an average of three additional fish species each year since 1962 have been identified as new hosts (Escobar et al., 2018). Wide breadth of host diversity often characterizes a new, unspecialized virus. Our research discerned little support for trends in infected host species among VHSV-IVb haplotype groupings. Similarly, VHS-IVa isolates from eight fish species showed greater G-gene similarity to isolates from the same geographic region than to host species (Hedrick et al., 2003). Moreover, no host species associations were found for 63 full G-gene IVa isolates collected over a 20-year span, with just four haplotypes recovered from the same host species (Garver et al. 2013). That study’s lower numbers of common haplotypes due to its inclusion of the entire G-gene might preclude comparison with our investigation’s partial sequences. Much diversification of the IVb haplotype “b” group occurred in round goby from the Lower Great Lakes region during the Middle and Later time periods, suggesting that it might serve as a reservoir or vector (also see Cornwell et al., 2014). Many individuals recovered during and outside of large VHS-IVb outbreaks displayed classic hemorrhaging and other symptoms (Groocock et al., 2007; Cornwell et al., 2012b), yet others appeared asymptomatic (this study and Cornwell et al., 2012a). Round goby is one the most common benthic fishes in the lower Great Lakes and is a major prey item for game fishes, including yellow perch (Perca flavescens) and walleye (Sander vitreus; Johnson et al., 2005), possibly aiding VHSV transmission among species. Two round goby individuals from the peak of a small outbreak each possessed two different IVb haplotypes (Cornwell et al., 2014); such diversity within individuals could increase 38 transmission likelihood, since the host may fail to recognize some different sequences. Here we found no evidence of multiple haplotypes within an individual. Co-infections by multiple isolates are not easy to detect, as the sequence having more RNA will have greater PCR amplification (Hoofar et al. 2004). The round goby’s large population sizes and relatively high genetic diversity levels (Brown & Stepien, 2008, 2009; Snyder & Stepien, 2017) may facilitate VHSV-IVb co-evolutionary success. The goby’s high infection incidence may suggest that the virus is in the process of evolving some host specificity (Cornwell et al., 2014), notably in the “b” haplotype group indicated here. Since past research targeting round goby mostly focused on Lake Ontario (Cornwell et al., 2014) and the St. Lawrence River (Groocock et al., 2007), future efforts also should examine other populations. Gizzard shad was the second most commonly infected species (22 VHSV-IVb isolates), dominating the original 2006 Lake Erie outbreak (haplotype “a”; Thompson et al., 2011; Stepien, pers. obs.). The 2017 Lake St. Clair outbreak also primarily infected gizzard shad (M. Faisal and G. Whelan, personal communications, 2017), comprising a unique single haplotype (“be”). Many gizzard shad positives occurred during the Later time period in the Central Great Lakes, suggesting that the species might serve as a VHSV-IVb reservoir. We report the first IVb detection in alewife (Lake Michigan, 2016), which, like gizzard shad, belongs to the Clupeid family. Unlike VHSV, IHNV has a much narrower host range, exclusively infecting certain salmonid species in the Northeastern Pacific (Lapatra, 1998). Within its “U” group, subgroup “UP” is more common in sockeye salmon (Oncorhynchus nerka), whereas “UC” infects Chinook salmon (O. tshawytscha) and steelhead trout (O. mykiss), 39 revealing some geographic regional differences (Black et al., 2016). Similarly, group “b”- derived haplotypes of VHS-IVb might develop specificity in round goby over time, meriting future investigation. In comparison, high RABV genetic diversity is believed to have facilitated jumps among a wide taxonomic range of key host species, including bats, raccoons, and many canids (Rodríguez-Nevado et al., 2018). A genetically diverse host pool, whether in terms of population or species, can increase diversity of viral sequences (Ojosnegros & Beerenwinkel, 2010); this appears to be the case for VHSV-IVb. The effect of VHSV-IVb on different host species, as well as within a host species, varied considerably in our study, as measured by concentrations of the virus in infected field-collected fish tissues determined with our qPCR assay (which uniquely used IS, increasing accuracy). Similar wide variation in virus titer characterized individual fish in viral challenge experiments infected with haplotype “a” (Pierce et al., 2013a,b). Notably, four of our field-collected samples were above the threshold for clinical onset of symptoms in haplotype “a” challenged fish (Pierce et al., 2013b). The 2015 round goby positive possessed the highest recorded level of virus, yet had no physical symptoms. Concentrations in three gizzard shad individuals also were above the clinical threshold. The higher sensitivity of the Pierce et al. (2013a,b) qPCR assay over standard cell culture methodology and other qPCR methods (without IS) facilitated these results, as 67% of our 2015–2016 isolates were below the cell culture detection threshold and would have gone undetected. The virus’ titers and likely its effects on the immune system of the fish hosts, both among and within species, appears to vary considerably 40 among haplotypes and individuals. This merits further investigation with challenge experiments. 2.5.6 Selection and co-evolution. Tajima’s D analyses indicated that the G-gene and concatenated gene regions evolved under significant purifying selection. Following a quasispecies pattern, VHSV radiated into many closely related variants, with the less fit ones removed from the gene pool by purifying selection. The original haplotypes “a” and “b” disappeared over time, possibly reflecting selective response to increasing host recognition and resistance. This increasing diversification likely has enabled some VHSV variants to elude host immune recognition and delay responses. Diversification in the descendent virus populations has allowed the virus to persist over time, albeit at lower levels and with less virulence. Like VHS-IVb, populations of Spring Viremia of Carp Virus (SVCV) and IHNV underwent purifying selection (Padhi & Verghese, 2012; He et al., 2013). It is possible that purifying selection pressures on IHNV are greater than those for IV-b, due to the former’s more limited reported host species, and since it has a longer history. Analysis of 27 Rhabdoviridae taxa revealed that purifying selection characterized all members and was greatest in the Lyssavirus genus (including RABV; Kuzmin et al., 2006). In RABV, purifying selection occurred in response to reduced vaccination of potential hosts in Taiwan (Lin et al., 2016). Global RABV patterns displayed strong purifying selection across all branches of its phylogeny, appearing most pronounced in bat and dog host clades (Troupin et al., 2016). It is likely that VHSV evolution will continue to be shaped by purifying selection. 41 Illustrating local host population resistance, Millard et al. (2014) discerned that VHSV-IVb antibodies lasted up to one year in muskellunge. Throckmorton et al. (2015) detected VHSV from Budd Lake in 2011 after its apparent absence since 2007, suggesting an interlude of dormancy due to local immunity. Similar to our findings, a study of Danish VHSV-Ia isolates distinguished an almost four-year gap between fish kills, despite some detections in local fishes (Kahns et al., 2012). This may explain the relative rarity of VHSV-IVb from 2010–16 in Great Lakes populations. However, our results show that during these interludes, the virus has continued to evolve and maintains “a foothold” in the local populations. VHSV-IVb has undergone considerable diversification over time and space, whose trajectory likely will continue. 2.5.7 Summary and conclusions. VHSV-IVb remains resident in the Great Lakes, at lower incidence and virulence, and has been mutating between outbreaks. Significant divergences occurred across its spatial and temporal courses, with geographic divergence accounting for increased nucleotide genetic variance and significant temporal diversification in amino acids. This pattern appears to help circumvent immune system recognition and resistance of fish hosts, and appears correlated with reduced virulence and persistence at low levels in resident populations, meriting future investigation. It is likely that VHSV-IVb will continue to infect naïve populations and spread to other water bodies, facilitated by its quasispecies, cloud-like pattern of evolutionary diversification. 2.6 Acknowledgements Research funding was provided by grants to CAS (with D. Leaman), including NSF-DBI- 1354806 for “Gene diversity of the VHS fish virus: Evolution of cellular immune 42 response and pathogenesis” and USDA-ARS CRIS #3655-31320-002-00D, under specific cooperative agreement #58-3655-9-748, “VHS fish virus in yellow perch aquaculture”. The Genetics and Genomics Group headed by CAS relocated to NOAA Pacific Marine Environmental Laboratory (PMEL) in October 2016, and part of the NSF award was subawarded to the NOAA Joint Institute for Science of the Atmosphere and Ocean (JISAO) at the University of Washington (where CAS is an Affiliate Professor of Biological Oceanography). MDN remained at University of Toledo to complete her PhD degree under the advisement of CAS, working in the laboratory of M. Krishnamurthy. We thank S. Edwards for two seasons of field and laboratory assistance, and B. Bodamer, D. Eddins, M. Faisal, A. Haponski, G. Kurath, P. Kocovsky, O. Miller, K. Ofori, H. Scheppler, and G. Whelan for collection help. Agencies and other universities who provided samples included: Michigan Department of Natural Resources, New York State Department of Environmental Conservation, Ohio Department of Natural Resources, Wisconsin Department of Natural Resources, U.S. Fish and Wildlife Service, United States Geological Survey, Ohio Environmental Protection Agency, Ohio State University’s Stone Laboratory, Ontario Ministry of Natural Resources, Ministère des Forêts de la Faune et des Parcs, Cornell University, 1854 Treaty Authority, and Grand Valley State University. We thank T. Ackerman, K. Andrews, F. Calzonetti, J. Chaffin, E. Crawford, K. Czajkowski, A. Elz, T. Fisher, B. Gorgoglione, A. Izzi, D. Leaman, M. Krishnamurthy, E. Kramer, N. Marshall, D. Moorhead, S. McBride, L. Pierce, C. Sabine, and J. Willey for logistic help. This is contribution #4921 from the NOAA Pacific Marine Environmental Laboratory (PMEL). 43 Table 2-1. VHSV positives from our collections in 2012, 2015, and 2016. N Species N Individuals Concentration G-gene Concatenated Sampling surveyed surveyed Common Name TL Sampling Location VHSV/106 Haplotype genes Haplotype Date (VHSV (VHSV (Scientific name) (mm) Actin positives) positives) Lake Erie, Sandusky Bay, Sandusky, OH Largemouth bass 4/12/2012 1 1 349 9.572E+02 k k (41.453, -82.726) (Micropterus salmoides) Freshwater drum 5/10/2012 1 1 625 1.117E+03 l l (Aplodinotus grunniens) Lake Erie, Fairport Harbor, OH White perch 5/21/15 10 (2) 75 (2) 207 1.781E+03 u u (41.765259, -81.281594) (Morone americana) Round goby 5/26/15 179 5.244E+06 v v (Neogobius melanostomus) Lake Erie, Sandusky Bay, Sandusky, OH Emerald shiner (x6) 4/13/16 12 (4) 54 (13) >75 8.448E+01 x N/A (41.471594, -82.733653) (Notropis atherinoides) Gizzard shad 132 1.334E+05 w w (Dorosoma cepedianum) " " 109 2.598E+02 " " N/A " " 124 4.415E+04 " " N/A " " 130 7.754E+05 " " w " " 119 2.980E+02 " " N/A " " 123 1.853E+03 " " w1 " " 114 1.101E+01 " " N/A " " 109 1.875E+06 " " w2 " " 436 7.764E+01 " " w3 Pumpkinseed 134 3.388E+00 " " N/A (Lepomis gibbosus) Largemouth bass 156 5.050E-01 " " N/A " " 149 1.399E+01 " " N/A Lake Erie, Ashtabula, OH 5/18/16 16 (1) 72 (1) " " 340 1.199E+01 " " w (42.995682, -87.881905) Lake Michigan, South Shore Park 5/25-26/16 6 (2) 100 (5) Round goby (x3) <80 7.018E+02 x x1 Milwaukee, WI (42.995682, -87.881905) " " (x4) <80 5.706E+03 " " x Lake Michigan, Grant Park Milwaukee, 5/26/16 " " 73 5.465E+01 " " N/A WI (42.920696, -87.846161) " " 66 2.722E+01 " " N/A 44 Lake Michigan, South Shore Park Alewife " " 98 5.104E-01 " " N/A Milwaukee, WI (Alosa pseudoharengus) (42.995682, -87.881905) 45 Table 2-2. Primers used for PCR and sequencing per gene region, with name, reference, sequence, annealing temperature, and extension time. Annealing Extension Gene Primer name Reference Sequence (5’-3’) temp. (oC) time (min) N N.se This study ATGGAAGGAGGAATCCGTGCAG 56 2:30 N_mid.se " " CTCAATGGGACAGGAATGACCATGA " " " " N_mid.as " " GGTGAACAGTCCAATCATGGTCATT " " " " N.as " " TTAATCAGAGTCCTCGGGGTAGTC " " " " P VHSVPFor2 Stepien et al. (2015) CGCTGAGAGCTCACAATGAC 55 2:00 VHSVRev2 " " GCCTTGATTGCCTTTGAGAC " " " " M VHSVM.se " " GTCTCAAGGCAATCAAGGC " " " " VHSVM.as " " GCCTTGATTGCCTTTGAGAC " " " " G Gint1F Thompson et al. (2011) TCCCGTCAAGAGGCCAC 53 2:30 Gint4R " " TTCCAGGTGTTGTTTACCG " " " " Nv Nv_new3.se This study CGTCACTATCAGGAATGGGGG 59 " " Nv_new3.as " " ATGCTCAGTCTTTGGGGTCA " " " " 46 Table 2-3. Haplotype numbers, diversity, and private haplotypes per (A) time period and (B) and Great Lakes region. A N N N Mean N Haplotypes N Haplotypes N Haplotypes Haplotypic N Private Proportion Haplotypes Haplotypes Haplotypes Nucleotide Upper Great Central Lower Diversity Haplotypes Private (Total “a” Group “b” Group Diversity Lakes Great Lakes Great Lakes +/- SE (%) Haplotypes Isolates) (%) (%) +/- SE (Total) (Total) (Total) Early 4 3 1 0.540 0.00086 2 1 3 2 0.034 Period (58) (75%) (25%) +/-0.026 +/-0.00078 (50%) (7) (22) (29) Middle 14 7 7 0.672 0.00169 12 9 4 7 0.276 Period (58) (50%) (50%) +/-0.061 +/-0.00124 (86%) (18) (15) (25) 0.915 Later 22 6 16 0.00450 20 3 8 13 +/-0.017 0.735 Period (68) (27%) (72%) +/-0.00263 (91%) (17) (24) (27) 36 14 22 0.778 0.002671 34 11 14 17 TOTAL 0.370 (184) (39%) (61%) +/-0.024 +/-0.00172 (94%) (42) (61) (81) B N N N Mean Haplotypic N Private Proportion N Haplotypes N Haplotypes N Haplotypes Haplotypes Haplotypes Haplotypes Nucleotide Diversity Haplotypes Private Early Period Middle Period Later Period (Total “a” Group “b” Group Diversity +/- SE (%) Haplotypes (Total) (Total) (Total) Isolates) (%) (%) +/- SE Upper 11 6 5 0.642 0.00186 8 1 9 3 0.214 Great Lakes (42) (55%) (45%) +/-0.084 +/-0.00133 (73%) (7) (18) (17) Central 14 8 6 0.661 0.00260 10 3 4 8 0.426 Great Lakes (61) (57%) (43%) +/-0.057 +/-0.00170 (71%) (22) (15) (24) Lower 17 2 16 0.704 0.00153 14 2 7 13 0.284 Great Lakes (81) (11%) (89%) +/-0.045 +/-0.00115 (82%) (29) (25) (27) 36 14 22 0.778 0.002671 32 4 14 22 TOTAL 0.321 (184) (39%) (61%) +/-0.024 +/-0.00172 (89%) (58) (58) (68) 47 Table 2-4. Mean numbers of nucleotide (NT) and amino acid (AA) changes (+/- standard errors) among VHS-IVb haplotypes for (A) time periods and (B) geographic regions from the G-gene and concatenated Concat) genes. A. Time Period Early Middle Later Total G-NT 1.000+/-0.000 1.385+/-0.241 1.619+/-0.161 4.004+/-0.402 G-AA 0.333+/-0.333 0.846+/-0.274 1.048+/-0.223 2.227+/-0.830 Concat-NT 1.750+/-0.526 3.500+/-0.872 11.143+/-1.199 16.393+/-2.597 Concat-AA 0.625+/-0.263 1.900+/-0.504 3.857+/-0.582 6.382+/-1.349 B. Region Upper Central Lower Total G-NT 1.500+/-0.307 1.692+/-0.237 1.250+/-0.112 4.442+/-0.656 G-AA 1.100+/-0.314 1.231+/-0.303 0.500+/-0.183 2.831+/-0.800 Concat-NT 5.889+/-1.020 7.053+/-1.482 4.500+/-1.960 17.442+/-4.462 Concat-AA 2.444+/-0.475 2.789+/-0.575 0.750+/-0.374 5.983+/-1.424 48 Table 2-5. Pairwise genetic divergences of VHSV populations between (A) sampling time periods, Early (2003–6), Middle (2007–10), and Later (2011–18), (B) Great Lakes regions (Upper, Central, Lower), and (C) individual water bodies, based on variation for the (1) G-gene and (2) concatenated gene data sets, using exact tests (GENEPOP; above diagonal) andθST divergences (ARLEQUIN; below diagonal). *=P0.05, **=remained significant (P<α) following sequential Bonferroni correction, NS=P>0.05. A1 A2 Early Middle Later Early Middle Later N=58 N=57 N=65 N=19 N=13 N=15 Early - NS ** Early - ** ** Middle 0.019 - ** Middle 0.019** - ** Later 0.153** 0.123** - Later 0.235** 0.174** - B1 B2 Upper Central Lower Upper Central N=39 N=61 N=80 N=17 N=30 Upper - ** ** Upper - NS Central 0.111** - ** Central 0.041 - Lower 0.264** 0.348** - C1 Lake Budd Lake Lake Lake Lake St. Lawrence Finger Michigan Lake Huron St Clair Erie Ontario River Lakes N=19 N=12 N=7 N=19 N=43 N=21 N=52 N=6 Lake Michigan - * NS ** ** * ** NS Budd Lake 0.072* - * NS * * ** NS Lake Huron 0.037 0.198* - * NS NS ** NS Lake St. Clair 0.136** 0.135 0.180* - * ** ** NS Lake Erie 0.165** 0.119* 0.115 0.114* - ** ** NS Lake Ontario 0.095* 0.220* 0.001 0.230** 0.189** - ** NS St. Lawrence River 0.393** 0.561** 0.257** 0.563** 0.447** 0.207** - ** Finger Lakes 0.001 0.153 0.001 0.180 0.113 0.001 0.318** - C2 L. Erie & L. St Clair L. Michigan L. Ontario & Budd L. N=29 N=8 N=9 L. Erie & L. Ontario - NS NS L. St Clair 0.017 - NS L. Michigan & Budd L. 0.046 0.010 - 49 Table 2-6. Relative distribution of genetic variation among VHSV-IVb isolates using Analysis of Molecular Variance (AMOVA, Excoffier et al., 1992), calculated from 185 partial G-gene sequences for A) nucleotide sequences, and B) amino acid changes, using ARLEQUINv3.5.1.3 (Excoffier & Lischer, 2010). *=significant Distribution of Genetic Variation % Variation Φ p A. G-gene Nucleotide Substitutions 1. a. Among three Great Lakes Regions (Upper, Central, Lower) 17.99 0.624* <0.001 * 1. b. Sampling events within the three regions 51.14 0.691* <0.001* 1. c. Within samples 30.87 0.180* 0.016* 2. a. Among time periods (Early, Middle, Later) 0.52 0.671* <0.001* 2. b. Sampling events within the three time periods 66.79 0.673* <0.001* 2. c. Within samples 32.69 0.005 0.395 B. G-gene Amino Acid Substitutions 1. a. Among three Great Lakes Regions (Upper, Central, Lower) 0.100 0.698* <0.001* b. Sampling events within the three regions 69.77 0.699* <0.001* 1. c. Within samples 30.13 0.001 0.304 2. a. Among time periods (Early, Middle, Later) 7.17 0.682* <0.001* 2. b. Sampling events within the three time periods 63.31 0.705* <0.001* 2. c. Within samples 29.51 0.072 0.131 50 Table 2-7. Tajima’s D test values (ARLEQUIN) for selection pressures on VHSV-IVb evolution based on A) all samples, B) geographic region (Upper, Central, Lower Great Lakes), C) time period (Early (2003–6), Middle (2007–10), Later (2011–18), D) individual water bodies, and E) host species. *=p0.05 Data Set Test Group Sample size Tajima's D p A. Haplotypes 1. G-gene All samples 186 -2.008* 0.002* 2. Concat. genes All samples 052 -2.412* 0.000 * B. Regions 1. G-gene Upper Great Lakes 039 -1.982* 0.007* Central Great Lakes 064 -1.059 0.145 Lower Great Lakes 082 -1.957* 0.005* Central & Lower 2. Concat. genes 038 -2.031* 0.006* Great Lakes Upper Great Lakes 009 -1.672* 0.027* C. Time 1. G-gene Early Period 059 -0.225 0.439 Middle Period 057 -2.042* 0.004* Later Period 068 -1.063 0.139 2. Concat. genes Early Period 019 -2.369* 0.001* Middle Period 013 -2.127* 0.003* Later Period 015 -1.009 0.154 D. Water bodies 1. G-gene Lake Michigan 019 -1.223 0.108 Budd Lake 012 -1.451 0.066 Lake Huron 007 -0.598 0.332 Lake St. Clair 020 0.939 0.835 Lake Erie 044 -0.989 0.169 Lake Ontario 022 -1.102 0.153 St. Lawrence River 053 -2.099* 0.002* Finger Lakes 006 -0.050 0.443 E. Species 1. G-gene Black crappie 005 1.225 0.943 Freshwater drum 013 -0.909 0.223 Gizzard shad 023 0.124 0.594 Pumpkinseed 005 -0.175 0.437 Bluegill 006 -1.011 0.207 Round goby 050 -1.711* 0.022* Largemouth bass 014 -0.908 0.201 Smallmouth bass 014 -0.387 0.359 Yellow perch 012 -0.382 0.313 Other 049 -1.946* 0.007* 51 Figure 2-1. Maps showing locations (circles, colored by year) of VHSV-IVb isolates analyzed here, per time period (A) Early (2003–6), (B) Middle (2007–10), and (C) Later (2011–18). Figure 2-2. Concentrations of VHSV-IVb (+/- standard error) in wild-caught fish tissues, compared to results from experimental laboratory haplotype “a” challenged muskellunge, determined with our qPCR assay using internal standards (Pierce et al., 2013b). Laboratory samples (squares) are named by the number of days (6-42D) after VHSV-IVb inoculation, H=high virus dosage (1x105 pfu/mL), and L=low dosage (100 pfu/mL) (data from Pierce, 2013). Haplotype of each sample is listed above its standard error bars. * above the haplotype indicates that the sample was sequenced in both the G- gene and concatenated gene analyses. Solid line denotes the experimental symptom threshold and dashed line the cell culture detection threshold (Pierce et al., 2013b). Wild caught samples (circles) are designated by abbreviated common name, followed by collection year and sample number (Table 2-1). Fish species names: MUS=muskellunge, LMB= largemouth bass, FRD=freshwater drum, WPE=white perch, ROG=round goby, GIZ=gizzard shad, PUM=pumpkinseed, EMS=emerald shiner, ALE=alewife. Figure 2-3. VHSV-IVb G-gene phylogeny. Phylogenetic tree of VHSV haplotypes based on the G-gene from maximum likelihood and Bayesian analyses. Values above nodes = 2000 bootstrap pseudoreplicates/Bayesian posterior probabilities. Values in parentheses and italics = estimated divergence time (years). VHSV-IVa (AB179621) served as the outgroup. Figure 2-4. G-gene haplotype networks. Partial G-gene sequences (669 NT) from 176 isolates using POPART (https://popart.otago.ac.nz) and TCS (Clement et al. 2000) for (A-B) Great Lakes regions (Upper, Central, Lower), (C-D) time periods (Early, Middle, Later), and (E-F) host species. A, C, and E are based on nucleotide substitutions and B, D, and F on amino acid changes. Circles are sized according to frequency of the haplotype in the population. Lines denote a single substitution step between haplotypes, with dashed lines for synonymous changes and solid lines for nonsynonymous changes. Small, unlabeled black circles represent hypothesized haplotypes. The ‘Other’ category in E-F, contains all host species, in which three or fewer non-unique isolates were detected: alewife (Alosa pseudoharengus), amphipod (Diporeia spp.), brown bullhead (Ameiurus nebulosus), burbot (Lota lota), channel catfish (Ictalurus punctatus), Chinook catfish (Oncorhynchus tshawytscha), common carp (Cyprinus carpio), cisco (Coregonus artedi), lake whitefish (C. clupeaformis), leech (Myzobdella lugubris), muskellunge (Esox masquinongy), northern pike (E. lucius), rainbow trout (Oncorhynchus mykiss), sea lamprey (Petromyzon marinus), shorthead redhorse (Moxostoma macrolepidotum), and walleye (Sander vitreus). 52 Figure 2-5. Concatenated N, P, M, G, and Nv-gene haplotype networks. Gene sequences (3355 NT) from 47 isolates in POPART and TCS for (A-B) Great Lakes regions and (C–D) time periods. A and C show nucleotide substitutions and B and D those with amino acid changes. Circles are sized according to population frequency of the haplotype. Lines denote single substitution steps, with dashes denoting synonymous changes and solid lines nonsynonymous changes. Small, unlabeled black circles represent hypothesized haplotype steps. Figure 2-6. Neighbor-joining genetic distance tree depicting relationships among VHSV-IVb populations. Reynold’s (1983) genetic distances (RST) used on G-gene haplotypes and their frequencies in PHYLIP (Felsenstein, 2007). Bootstrap percentage support for nodes from 10,000 replications are shown. Sample sizes (N) are in parentheses. Colored boxes denote sampling groups. *=samples from confirmed fish kill events. Figure 2-7. Tests for relationship between genetic divergence (θST) among VHSV G- gene sampling groups vs. geographic distance (A–C, nearest waterway distance, km) or time (D, yrs). (A) Early time period (y=0.001x + 0.091, R2=0.288, p=0.006*), B) Middle time period (y=1.86e-5x + 0.298, R2=0.001, p=0.278), C) Later time period (y=- 9.82e-5x + 0.860, R2=0.049, p=0.709), and D) all samples (y=0.040x + 0.342, R2=0.149, p=0.002*). 53 A B Middle C Later Figure 2-1. 54 Figure 2-2. 55 Figure 2-3. 56 Figure 2-4. 57 Figure 2-5. 58 Figure 2-6. 59 Figure 2-7. 60 Supplementary Table 2-A. VHSV-IVb samples used for our analyses. Isolate names, years, location information, host species, geographic coordinates, haplotypes, GenBank Accession numbers, sources and analysis group for the (1) Early, (2) Middle, and (3) Later time periods. Supplementary Table 2-B. Substitutions in the concatenated gene sequence haplotypes (N, P, M, G, and Nv-genes). Numbered from the start of the full haplotype “a” genome (GQ385941), with corresponding haplotypes. *=non-synonymous. Supplementary Table 2-C. Compositions of substitutions in concatenated gene sequence regions (N, P, M, and Nv-genes). Proportions of substitutions for each region sequence are in parentheses. dN/dS = proportion of non-synonymous versus synonymous changes. Supplementary Figure 2-A. VHSV-IVb structure and genome layout. Colors match the gene to the structure diagram. Numbers sharing the same colors as the gene refer to the nucleotide positions. Numbers in parentheses correspond to the region sequenced within each gene. Modified with permission from Pore (2012). 61 Table 2-A. VHSV-IVb samples used in our analyses. Isolate name, year, location information, host species, geographic coordinates, haplotype, GenBank Accession number, source and analysis group are provided for the (1) Early, (2) Middle, and (3) Later time periods. 1. Early Accession Gene Isolate Name Year Body of Water Region Nearest City Lat, Long Host Species Haplotype Source No. Data Set Esox TAVgr05-01 2003 L. St. Clair Central Detroit, MI 42.391, -82.911 a GQ385941 G/All Ammayapan et al., 2009 masquinongy Aplodinotus U 13653-1 2005 L. Ontario Lower Brighton, ON 43.968, -77.629 b HQ453209 G MEAP-VHSV database grunniens U 13653-2 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Jeannette's Creek, Lepomis 5464 Bluegill 2006 L. St. Clair Upper 42.358, -82.459 a GQ385941 “ “ “ “ ON macrochirus 5464 Drum “ “ “ “ “ “ “ “ “ “ A. grunniens “ “ “ “ “ “ “ “ 5464 Smallmouth Micropterus “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ bass dolomieu Cape Vincent, Neogobius Goby 1-5 “ “ L. Ontario Lower 44.126, -76.334 t AB672615 G/All “ “ NY melanostomus TAVgr06-01 “ “ “ “ “ “ Rochester, NY 43.216, -77.633 “ “ b HQ453209 G “ “ TAVgr06-02 “ “ St. Lawrence R. “ “ Clayton, NY 44.250, -76.016 E. masquinongy “ “ “ “ “ “ “ “ Cape Vincent, N. TAVgr06-03 “ “ “ “ “ “ 44.127, -76.333 “ “ “ “ “ “ “ “ NY melanostomus TAVgr06-04 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr06-05 “ “ L. Erie Central Sandusky, OH 41.474, -82.703 A. grunniens a GQ385941 “ “ “ “ TAVgr06-06 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Fairport Harbor, USGS Western Fisheries TAVgr06-07 “ “ “ “ “ “ 41.756, -81.287 “ “ a1 MK777870 G/All OH Research Center TAVgr06-08 “ “ “ “ “ “ “ “ “ “ Sander vitreus a2 MK777871 “ “ “ “ Morone TAVgr06-09 “ “ “ “ “ “ “ “ “ “ a GQ385941 “ “ “ “ chrysops Perca TAVgr06-10 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ flavescens TAVgr06-11 “ “ “ “ “ “ “ “ “ “ M. dolomieu a3 MK777872 “ “ “ “ TAVgr06-12 “ “ “ “ “ “ Sandusky, OH 41.492, -82.667 P. flavescens a GQ385941 “ “ “ “ TAVgr06-13 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ 62 TAVgr06-14 “ “ “ “ “ “ “ “ “ “ M. chrysops a4 MK777873 “ “ “ “ Fairport Harbor, Dorosoma TAVgr06-15 “ “ “ “ “ “ 41.755, -81.286 a GQ385941 G MEAP-VHSV database OH cepedianum USGS Western Fisheries TAVgr06-16 “ “ “ “ “ “ “ “ “ “ M. chrysops d MK777861 G/All Research Center TAVgr06-17 “ “ “ “ “ “ Sandusky, OH 41.492, -82.667 S. vitreus a GQ385941 G MEAP-VHSV database Notropis TAVgr06-18 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ atherinoids Wellesley Island, Ameiurus TAVgr06-19 “ “ St. Lawrence R. Lower 44.327, -75.937 b HQ453209 “ “ “ “ NY nebulosus Morone TAVgr06-20 “ “ “ “ “ “ “ “ 44.323, -75.935 “ “ “ “ “ “ “ “ americana TAVgr06-21 “ “ “ “ “ “ Clayton, NY 44.248, -76.098 Esox lucius “ “ “ “ “ “ “ “ TAVgr06-22 “ “ “ “ “ “ Grindstone, NY 44.254, -76.150 A. nebulosus “ “ “ “ “ “ “ “ Cape Vincent, TAVgr06-23 “ “ “ “ “ “ 44.187, -76.225 L. macrochirus “ “ “ “ “ “ “ “ NY TAVgr06-24 “ “ “ “ “ “ “ “ 44.172, -76.247 P. flavescens “ “ “ “ “ “ “ “ TAVgr06-25 “ “ “ “ “ “ Alexandria, NY 44.323, -75.935 “ “ “ “ “ “ “ “ “ “ Cape Vincent, TAVgr06-26 “ “ “ “ “ “ 44.187, -76.225 M. dolomieu “ “ “ “ “ “ “ “ NY TAVgr06-27 “ “ “ “ “ “ Clayton, NY 44.250, -76.016 Lota lota “ “ “ “ “ “ “ “ N. USGS Western Fisheries TAVgr06-28 “ “ L. Ontario “ “ Irondequoit, NY 43.200, -77.527 a5 MK777874 G/All melanostomus Research Center TAVgr06-29 “ “ “ “ “ “ Sodus Bay, NY 43.249, -76.962 M. dolomieu a GQ385941 G MEAP-VHSV database Cape Vincent, TAVgr06-30 “ “ “ “ “ “ 44.116, -76.333 “ “ b HQ453209 “ “ “ “ NY Ambloplites TAVgr06-31 “ “ St. Lawrence R. “ “ Clayton, NY 44.251, -76.134 “ “ “ “ “ “ “ “ rupestris Wellesley Island, Pomoxis TAVgr06-32 “ “ “ “ “ “ 44.323, -76.014 “ “ “ “ “ “ “ “ NY nigromaculatus TAVgr06-33 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr06-34 “ “ “ “ “ “ Clayton, NY 44.248, -76.014 M. dolomieu “ “ “ “ “ “ “ “ Ictalurus TAVgr06-35 “ “ “ “ “ “ “ “ 44.254, -76.014 “ “ “ “ “ “ “ “ punctatus TAVgr06-36 “ “ “ “ “ “ “ “ 44.187, -76.014 “ “ “ “ “ “ “ “ “ “ Greater Napanee, Micropterus TAVgr06-37 “ “ “ “ “ “ 44.172, -76.964 “ “ “ “ “ “ “ “ ON salmoides Wellesley Island, TAVgr06-38 “ “ “ “ “ “ 44.323, -76.014 M. dolomieu “ “ “ “ “ “ “ “ NY 63 TAVgr06-39 “ “ “ “ “ “ Clayton, NY 44.187, -76.014 “ “ “ “ “ “ “ “ “ “ TAVgr06-40 “ “ “ “ “ “ Orleans, NY 44.268, -76.014 “ “ “ “ “ “ “ “ “ “ TAVgr06-43 “ “ L. Huron Upper Alpena, MI 45.050, -83.200 S. vitreus a GQ385941 “ “ “ “ Oncorhynchus TAVgr06-44 “ “ “ “ “ “ Rogers City, MI 45.502, -83.783 “ “ “ “ “ “ “ “ tshawytscha Coregonus TAVgr06-45 “ “ “ “ “ “ Alpena, MI 45.050, -83.200 “ “ “ “ “ “ “ “ clupeaformis TAVgr06-46 “ “ “ “ “ “ Bois Blanc, MI 45.718, -84.374 “ “ “ “ “ “ “ “ “ “ Harrison USGS Western Fisheries TAVgr06-47 “ “ L. St. Clair Central 42.635, -82.778 E. lucius “ “ GQ385941 G/All Township, MI Research Center TAVgr06-48 “ “ “ “ “ “ Grosse Point, MI 42.343, -82.902 D. cepedianum a6 MK777875 “ “ “ “ TAVgr06-49 “ “ “ “ “ “ “ “ “ “ A. rupestris a GQ385941 “ “ “ “ Moxostoma TAVgr06-50 “ “ “ “ “ “ “ “ “ “ macrolepidotu “ “ GQ385941 “ “ “ “ m TAVgr06-51 “ “ “ “ “ “ “ “ “ “ P. flavescens “ “ GQ385941 “ “ “ “ TAVgr06-52 “ “ “ “ “ “ “ “ “ “ A. grunniens “ “ GQ385941 “ “ “ “ Percopsis TAVgr06-53 “ “ “ “ “ “ “ “ “ “ m N/A G MEAP-VHSV database omiscomaycus 2. Middle Hamilton Beach, OMNR #5577 2007 L. Ontario Lower 43.295, -79.772 A. grunniens a HQ623440 G MEAP-VHSV database ON OMNR #5583 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Jeannette's Creek, OMNR #5579 “ “ L. St. Clair Central 42.328, -82.472 M. salmoides “ “ “ “ “ “ “ “ ON USGS Western Fisheries TAVgr07-01 “ “ Budd L. Upper Harrison, MI 44.016, -84.788 L. macrochirus e MK777862 G/All Research Center P. TAVgr07-02 “ “ “ “ “ “ “ “ “ “ a HQ623440 G MEAP-VHSV database nigromaculatus TAVgr07-03 “ “ “ “ “ “ “ “ “ “ M. salmoides “ “ “ “ “ “ “ “ Lepomis USGS Western Fisheries TAVgr07-04 “ “ “ “ “ “ “ “ “ “ f MK777863 G/All gibbosus Research Center TAVgr07-05 “ “ L. Michigan Upper Oshkosh, WI 44.028, -88.421 A. grunniens a HQ623440 G MEAP-VHSV database TAVgr07-06 “ “ L. Ontario Lower Irondequoit, NY 43.233, -77.650 D. cepedianum “ “ “ “ “ “ “ “ TAVgr07-07 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr07-08 “ “ L. Erie Central Dunkirk, NY 42.490, -79.338 “ “ b HQ453209 “ “ “ “ 64 TAVgr07-09 “ “ “ “ “ “ “ “ “ “ “ “ a HQ623440 “ “ “ “ P. TAVgr07-10 “ “ L. Ontario Lower Ransomville, NY 43.240, -79.920 “ “ “ “ G/All “ “ nigromaculatus TAVgr07-11 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Cyprinus USGS Western Fisheries TAVgr07-12 “ “ L. Erie Central Dunkirk, NY 42.491, -79.338 c MK777860 “ “ carpio Research Center Salvelinus TAVgr07-13 “ “ L. Ontario Lower Skaneateles, NY 42.939, -76.425 “ “ “ “ G MEAP-VHSV database namaycush TAVgr07-14 “ “ “ “ “ “ Waterloo, NY 42.910, -76.910 L. gibbosus a HQ623440 “ “ “ “ TAVgr07-15 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr07-16 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Oncorhynchus TAVgr07-17 “ “ “ “ “ “ Mexico, NY 43.459, -76.228 c HQ623435 “ “ “ “ mykiss TAVgr07-18 “ “ “ “ “ “ Skaneateles, NY 42.939, -76.425 M. dolomieu b HQ453209 “ “ “ “ TAVgr07-19 “ “ “ “ “ “ “ “ “ “ A. rupestris “ “ “ “ “ “ “ “ Catawba Island, USGS Western Fisheries TAVgr07-20 “ “ L. Erie Central 41.542, -82.789 L. macrochirus g MK777864 G/All OH Research Center Fairport Harbor, TAVgr07-21 “ “ “ “ “ “ 41.801, -81.356 P. flavescens a7 MK777876 “ “ “ “ OH TAVgr07-22 “ “ “ “ “ “ “ “ 41.755, -81.277 “ “ a HQ623440 G MEAP-VHSV database Sturgeon Bay, USGS Western Fisheries TAVgr07-24 “ “ L. Michigan Upper 44.885, -87.389 M. dolomieu h MK777865 G/All WI Research Center TAVgr08-02 2008 “ “ “ “ Racine, WI 42.799, -87.761 P. flavescens b MK777859 “ “ “ “ TAVgr08-03 “ “ “ “ Upper Kenosha, WI 42.485, -87.800 A. rupestris i MK777866 “ “ “ “ N. TAVgr08-04 “ “ St. Lawrence R. Lower Clayton, NY 44.250, -76.016 b HQ453209 G MEAP-VHSV database melanostomus TAVgr08-05 “ “ L. Ontario “ “ Oswego, NY 43.450, -76.510 “ “ “ “ “ “ “ “ “ “ TAVgr08-06 “ “ “ “ “ “ Sterling, Ny 43.350, -76.690 “ “ “ “ “ “ “ “ “ “ St. Lawrence TAVgr08-07 “ “ St. Lawrence R. “ “ 43.340, -75.910 “ “ “ “ “ “ “ “ “ “ River Petromyzon TAVgr08-08 “ “ L. Huron Upper Cheboygan, MI 45.652, -84.468 “ “ “ “ “ “ “ “ marinus Fairport Harbor, USGS Western Fisheries TAVgr08-09 “ “ L. Erie Central 41.769, -81.294 N. atherinoides a8 MK777877 G/All OH Research Center TAVgr08-10 “ “ “ “ “ “ “ “ 41.769, -81.354 A. grunniens a HQ623440 “ “ “ “ TAVgr08-11 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Benona TAVgr09-01 “ “ Lake Michigan Upper 43.600, -86.916 Amphipod “ “ “ “ G MEAP-VHSV database Township, MI 65 TAVgr09-02 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Chesterfield TAVgr09-03 “ “ Lake St. Clair Central 42.631, -82.765 Leech “ “ “ “ “ “ “ “ Township, MI TAVgr09-04 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr09-05 “ “ L. Erie “ “ Lake Erie N/A “ “ “ “ “ “ “ “ “ “ N. TAVgr09-13 “ “ L. Michigan Upper Milwaukee, WI 42.926, -87.770 “ “ “ “ “ “ “ “ melanostomus USGS Western Fisheries TAVgr09-17 “ “ “ “ “ “ “ “ 43.040, -87.802 P. flavescens j MK777867 G/All Research Center TAVgr09-09 2009 Baseline Lake “ “ Pinckney, MI 42.427, -83.899 A. nebulosus a HQ623441 G MEAP-VHSV database Harrison USGS Western Fisheries TAVgr09-10 “ “ L. St. Clair Central 42.616, -82.757 E. masquinongy “ “ “ “ G/All Township, MI Research Center St. Clair Shores, TAVgr09-11 “ “ L. St. Clair “ “ 42.475, -82.879 M. dolomieu “ “ “ “ G MEAP-VHSV database MI Apostle Islands, Coregonus TAVGR10-01 “ “ L. Superior Upper 47.085, -90.641 “ “ “ “ “ “ “ “ WI artedi N. TAVgr10-05 “ “ Lake Ontario Lower Irondequoit, NY 43.236, -77.534 “ “ “ “ “ “ “ “ melanostomus TAVgr10-06 “ “ Lake Ontario “ “ Pulaski, NY 43.577, -76.203 “ “ b HQ453209 “ “ “ “ TAVgr10-07 “ “ “ “ “ “ “ “ “ “ “ “ o N/A “ “ “ “ TAVgr10-08 “ “ Lake Ontario “ “ “ “ “ “ “ “ p N/A “ “ “ “ TAVgr10-09 “ “ St. Lawrence R. “ “ Clayton, NY 44.266, -76.012 “ “ b HQ453209 “ “ “ “ Cape Vincent, TAVgr10-10 “ “ “ “ “ “ 44.186, -76.224 “ “ p N/A “ “ “ “ NY TAVgr10-11 “ “ L. Ontario “ “ Pulaski, NY 43.577, -76.203 “ “ o N/A “ “ “ “ TAVgr10-12 “ “ L. Ontario “ “ “ “ “ “ “ “ q N/A “ “ “ “ Sturgeon Bay, TAVgr9-12 “ “ Lake Michigan Upper 44.860, -87.393 M. dolomieu a HQ623440 “ “ “ “ WI N. GL2010-098 2010 L. Huron “ “ Barrie, ON 44.407, -79.368 s N/A “ “ “ “ melanostomus TAVgr10-02 “ “ L. Huron “ “ Huron Beach, MI 45.519, -84.087 S. namaycush n N/A “ “ “ “ 3. Later TAVgr11-01 2011 L. Michigan Upper Milwaukee, WI 43.030, -87.915 D. cepedianum r N/A G MEAP-VHSV database TAVgr11-02 “ “ “ “ “ “ “ “ “ “ “ “ r “ “ “ “ “ “ USGS Western Fisheries TAVgr11-18 “ “ “ “ “ “ “ “ 43.036, -87.853 P. flavescens a HQ453209 “ “ Research Center TAVgr11-19 “ “ “ “ “ “ “ “ “ “ “ “ a9 MK777878 G/All “ “ 66 TAVgr11-03 “ “ Budd L. “ “ Harrison, MI 44.015, -84.788 M. dolomieu a HQ453209 G “ “ TAVgr11-05 “ “ “ “ “ “ “ “ “ “ M. salmoides “ “ “ “ “ “ “ “ TAVgr11-08 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ MEAP-VHSV database USGS Western Fisheries TAVgr11-09 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ Research Center TAVgr11-11 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr11-13 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr11-15 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ TAVgr11-17 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ N. vcG017 “ “ St. Lawrence R. Lower Clayton, NY 44.243, -76.079 y N/A “ “ Cornwell et al., 2014 melanostomus “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ vcG018 “ “ “ “ “ “ “ “ “ “ “ “ z “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ vcG001 “ “ “ “ “ “ “ “ “ “ “ “ a HQ453209 “ “ “ “ vcG002 “ “ “ “ “ “ “ “ “ “ “ “ b HQ453209 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ vcG032 “ “ “ “ “ “ “ “ “ “ “ “ AA N/A “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ vcG045 “ “ “ “ “ “ “ “ “ “ “ “ ab “ “ “ “ “ “ vcG046 “ “ “ “ “ “ “ “ “ “ “ “ ac “ “ “ “ “ “ vcG047 “ “ “ “ “ “ “ “ “ “ “ “ ad “ “ “ “ “ “ vcG048 “ “ “ “ “ “ “ “ “ “ “ “ ae “ “ “ “ “ “ 67 vcG049 “ “ “ “ “ “ “ “ “ “ “ “ af “ “ “ “ “ “ vcG050 “ “ “ “ “ “ “ “ “ “ “ “ ag “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ vcG051 “ “ “ “ “ “ “ “ “ “ “ “ ah “ “ “ “ “ “ FD12 2012 L. Erie Central Sandusky, OH 41.453, -82.726 A. grunniens l MK777869 G/All Stepien et al., 2015 LMB12 “ “ “ “ “ “ “ “ “ “ M. salmoides k MK777868 “ “ “ “ Irondequoit Bay, RPL2013-002 2013 L. Ontario Lower 43.231, -77.533 D. cepedianum bc KY359356 “ “ Getchell et al., 2017 NY FPL2014-001 2014 L. Erie Central Dunkirk, NY 42.495, -79.333 “ “ bd KY359357 “ “ “ “ Fairport Harbor, N. RG-H31 2015 “ “ “ “ 41.765, -81.281 v MK777880 “ “ This study OH melanostomus WP-H6 “ “ “ “ “ “ “ “ “ “ M. americana u MK777879 “ “ “ “ B01 2016 “ “ “ “ Sandusky, OH 41.471, -82.733 N. atherinoides x MK777881 G “ “ B09 “ “ “ “ “ “ “ “ “ “ D. cepedianum w MK777881 G/All “ “ B10 “ “ “ “ “ “ “ “ “ “ “ “ “ “ MK777881 “ “ “ “ B11 “ “ “ “ “ “ “ “ “ “ “ “ “ “ MK777881 G “ “ B13 “ “ “ “ “ “ “ “ “ “ “ “ w1 MK777882 G/All “ “ B14 “ “ “ “ “ “ “ “ “ “ “ “ “ “ MK777881 “ “ “ “ B16 “ “ “ “ “ “ “ “ “ “ “ “ “ “ MK777881 “ “ “ “ B17 “ “ “ “ “ “ “ “ “ “ “ “ w2 MK777883 “ “ “ “ B18 “ “ “ “ “ “ “ “ “ “ “ “ w3 MK777884 G/All “ “ B19 “ “ “ “ “ “ “ “ “ “ “ “ w MK777881 G “ “ B20 “ “ “ “ “ “ “ “ “ “ L. gibbosus “ “ MK777881 “ “ “ “ B21 “ “ “ “ “ “ “ “ “ “ M. salmoides “ “ MK777881 “ “ “ “ B22 “ “ “ “ “ “ “ “ “ “ M. salmoides “ “ MK777881 “ “ “ “ G61 “ “ “ “ “ “ Ashtabula, OH 41.898, -80.795 M. salmoides “ “ MK777881 G/All “ “ N. L56 “ “ L. Michigan Upper Milwaukee, WI 42.996, -87.882 x1 MK777886 “ “ “ “ melanostomus L59 “ “ “ “ “ “ “ “ “ “ N. x MK777885 “ “ “ “ 68 melanostomus Alosa L75 “ “ “ “ “ “ “ “ “ “ pseudoharengu “ “ MK777885 G “ “ s N. L72 “ “ “ “ “ “ “ “ 43.920, -87.846 “ “ MK777885 “ “ “ “ melanostomus L73 “ “ “ “ “ “ “ “ “ “ “ “ “ “ MK777885 “ “ “ “ St. Clair Shores, M. Faisal, personal 1704122 2017 L. St. Clair Central 42.473, -82.880 D. cepedianum be N/A “ “ MI commination, 2017 1704123 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ 1704124 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ P. 1704125 “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ “ nigromaculatus Clay Township, 1703302 “ “ “ “ “ “ 42.663, -82.617 L. macrochirus “ “ “ “ “ “ “ “ MI 69 Table 2-B. Substitutions in concatenated gene sequence regions (N-, P-, M-, G-, and Nv-genes). Numbered from the start of the full haplotype “a” genome (GQ385941), with corresponding haplotypes that have that substitution (lettered, lower case). *=non- synonymous. Reference Genome Gene Substitution Haplotype Location N 214 G-T* d N 224 A-G* w1, u N 244 G-A a5 N 285 G-A* w2, w3, w, x N 293 G-A* v N 305 A-C* a8, k, l N 313 C-T l N 385 T-C w, w2, w3, x N 451 G-A w, w2, w3, x N 517 G-A b N 519 C-A* u, v, w1 N 547 C-T a6, bc N 568 G-A a3 N 577 T-C bd N 578 C-T x1 N 679 A-C u, v, w1, bd N 755 A-G* j N 848 C-T bc N 857 C-A* w3, x1 N 862 C-T c N 892 G-A k N 970 G-A k N 1022 G-A* b N 1060 C-A bc N 1084 G-A b N 1105 G-A a1, bc N 1105 G-C x1 N 1114 G-A f N 1117 G-A b N 1153 C-T v N 1183 G-A k, l N 1251 G-A* i N 1274 G-A* a8, k, l N 1354 T-C l P 1578 C-T a4, d P 1584 T-C w1, w2, w3, x1, w, x 70 P 1707 T-C bc P 1725 G-A u, v, bd P 1757 A-G* a9, t P 1761 G-T x P 1773 A-G v P 1782 T-C l P 1875 A-C w1, w2, w3, x1, w P 1938 T-C bc P 2019 G-T* b P 2037 C-T b P 2044 C-A* b P 2074 C-T* b M 2415 C-T x1 M 2427 G-A a9, x1, x M 2436 G-T* h M 2478 T-C k M 2496 G-A v M 2648 C-T* u M 2654 A-G* b M 2697 T-C k, l M 2742 T-A bc M 2767 T-C* a9 M 2798 A-G* bd M 2803 T-C* a6, x1, u M 2811 T-C a6, u M 2819 T-C* u M 2839 C-T* x M 2868 A-G j G 3428 A-G n G 3455 T-C y, af G 3495 A-G* ac G 3587 A-G i G 3590 A-G e G 3600 A-G* h, AA G 3647 G-T* x, x1 G 3698 A-G ad G 3704 T-C af G 3707 C-T bd G 3731 A-G ae G 3740 G-A q G 3754 C-A* f, w, w1, w2, w3 G 3755 A-G* j G 3788 T-C bc 71 G 3789 A-C* u, v, w, w1, w2, w3 G 3795 A-C* ab G 3814 A-G* i G 3816 A-G* s G 3829 G-A* h, l, o, v, ad G 3832 A-G* i G 3849 A-G* p, s, u, v, w, w1, w2, w3, bd, be G 3872 C-T l, k G 3884 C-T l G 3929 G-A c G 3932 C-T m G 3969 T-C* i, r G 3971 T-C ah G 3995 C-T z, ac G 3996 A-C* d G 3998 T-A g b, i, n, o, p, q, r, s, u, v, y, z, AA, ab, ac, G 4007 C-G ad, ae, af, ag, ah, bc, bd G 4031 T-G bc G 4037 C-T ag Nv 4594 C-T* w, w2, w3 Nv 4614 T-C* a7, e, k, l Nv 4654 T-A* w2 Nv 4654 T-C* w1 Nv 4674 T-C* a2 Nv 4691 C-T k Nv 4700 A-C w, w2, w3 Nv 4709 T-C a2 Nv 4719 T-C* a2 Nv 4736 C-A a6, a9 Nv 4790 G-A e Nv 4873 C-T* w1 Nv 4883 T-C x, x1 Nv 4898 T-C v Nv 4931 C-A a6, a9, k 72 Table 2-C. Compositions of substitutions in concatenated gene sequence regions (N-, P-, M-, and Nv-genes). Proportions of substitutions for each region sequence are in parenthesis. dN/dS = proportion of non-synonymous versus synonymous changes. Gene N P M Nv NT (%) 34(0.028) 14(0.026) 16(0.028) 15(0.041) AA (%) 11(0.027) 4(0.022) 8(0.043) 7(0.057) dN/dS 0.324 0.286 0.500 0.467 73 Figure 2-A. 74 Chapter 3 Genomic and immunogenic changes across the evolutionary history of the Viral Hemorrhagic Septicemia (VHS-IVb) fish virus (Piscine novirhabdovirus) in the Laurentian Great Lakes Evolution. Niner, M.D., Stepien, C.A., Gorgolione B., and Leaman, D.W. 3.1 Abstract Viral hemorrhagic septicemia virus (VHSV) appeared in the Laurentian Great Lakes in 2005, comprising a new subgenogroup (IVb) that killed >32 fish species during large outbreaks in 2005 and 2006, with smaller ones in 2007 and 2008, and subsequent minor occurrences. This research sequenced the entire genome of 46 isolates collected from 2003-2016, in order to discern its evolutionary diversification patterns, comparing and contrasting its nucleotide and amino acid variation, in reference to its other genogroups (I-IVa). Phylogenetic relationships are compared within and among genogroups and other Novirhabdoviruses, including investigation of the unique Nv-gene’s origins. IVb isolates had 253 genomic nucleotide substitutions (2.3% of the total), of which 85 (16.6%) were non-synonymous. The highest proportion occurred in the non-coding region (NCDS; 4.3%) followed by the Nv (3.8%), and M (2.8%) genes. The greatest proportion of amino acid changes occurred in the M gene (52.9%), followed by the Nv (50.0%), G (48.6%), N (35.7%) and L (23.1%) genes. A 2016 gizzard shad isolate differed the most (E16GSa, 38nt, 15.0%, 15 AA). VHS-IVa from the northeast Pacific exhibited the fastest substitution rate (2.01x10-3), followed by IVb (6.64x10-5), and I/III from Europe (4.09x10-5). In vitro studies compared the immunogenicity of three 2016 isolates from 75 Lake Erie to the first IVb isolate (2003) from Lake St Clair. The 2016 isolates displayed reduced virulence and impact on the innate antiviral response, suggesting that sequence changes have phenotypic effects. 3.2 Introduction Viral Hemorrhagic Septicemia Virus (VHSV; Piscine novirhabdovirus) is a widespread and highly virulent pathogen that impacts aquaculture and wild fish populations across the Northern Hemisphere. VHSV infects >140 species of marine, estuarine, and freshwater fishes, making it one of the world’s most serious viral finfish diseases (Escobar et al. 2018: Walker et al. 2018). With short genomes, lack of proofreading, and quick generation times (Holmes 2009; Sanjuán et al. 2010; Volz et al. 2013), such RNA viruses undergo rapid evolution, facilitating their adaptation to new hosts and environments (Quer et al. 1996; Lauring and Andino 2010; Andino and Domingo 2015). VHSV’s single-stranded, negative sense RNA genome of 11,158 nucleotides (NT) encodes five structural genes: nucleoprotein (N), phosphoprotein (P), matrix (M), glycoprotein (G), and large protein (L) 5’N-P-M-G-Nv-L’3 (Wolf 1988; Kurath 2012). Novirhabdoviruses, including VHSV, also possess a unique non-virion (Nv) gene that encodes a distinctive non-structural protein. Four other novirhabdoviruses are recognized, including Salmonid novirhabdovirus (IHNV), Hirame novirhabdovirus (HIRRV) and Snakehead novirhabdovirus (SHRV), which selectively infect fish species (Kurath 2012; ICTV 2018). VHSV and IHNV have the broadest geographic ranges and economic impacts on aquaculture, and thus are more often used as models (Kurath et al. 2003; Kurath 2012; Ke et al. 2017). 76 The functionality of the Novirhabdoviruses specific Nv-gene has intrigued scientists for years. Although non-essential for replication, Nv is necessary for pathogenicity in IHNV (Thoulouze et al. 2004) and VHSV (Ammayappan et al. 2011), but not SHRV (Johnson et al. 2000; Alonso et al. 2004). Interchanging Nv between VHSV and IHNV reduced pathogenicity in both recombinants (Einer-Jensen et al. 2014). Nv has been suggested to function in suppressing the host innate immune response (Kim and Kim 2012, 2013; Einer-Jensen et al. 2014), or preventing apoptosis in infected cells (Ammayappan and Vakharia 2011), but the exact mechanisms are still being investigated. Single changes in Nv sequence can impact the virus’s capacity to suppress host cell responses (Chincilla and Gomez-Casado 2017; Ballion et al. 2018). Although Nv characterizes Novirhabdoviruses, its sequence is not conserved among viral species (Kurath 2012), and our literature search found no previous studies that investigated the potential evolutionary origin of this unique gene. VHSV occurs as four geographically and genetically distinct lineages, with I–III found in Europe and IV in North America and Asia (Fig. 3-1) (Meyers and Winton 1988; Hedrick et al. 2003; Pierce and Stepien 2012). VHSV-I is the oldest-known, impacts the largest known geographic range, has the most named phylogenetic subgroups, and the greatest number of described host species (Kurath 2012; Pierce and Stepien 2012), resulting in significant economic impacts on European aquaculture (Abbadi et al. 2016; Ghorani et al. 2016). Phylogenetic analyses show that I, II, and III share common ancestry, distinct from genogroup IV (Pierce and Stepien 2012; Stepien et al. 2015). VHSV-IV contains four genogroups, of which IVa was initially described from the northeastern Pacific’s coastal region in the 1980s (Brunsen et al. 1989; Hopper1989; 77 Meyers et al. 1992) and later caused outbreaks along the Asian Pacific coast, near Korea and Japan in the late 1990s (Takano et al. 2000). VHSV-IVc is located in marine waters off the northwestern Atlantic, with isolates dating to 2000 (Gagné et al. 2009). Phylogenetic analyses place IVc as the sister group to IVb, which together (IVc and b) then is the sister group to IVa (Pierce and Stepien 2012; Stepien et al. 2015). An unclassified potential fourth IV genogroup was isolated from wild lumpfish (Cyclopterus lumpus) from Breidafjördur Bay, Iceland (Guðmundsdóttir et al. 2018). The main focus of this study, genogroup IVb, appears endemic to the freshwater Laurentian Great Lakes region, and was back-traced to its first known occurrence in Lake St. Clair in 2003 (Elsayed et al. 2006; Ammayappan and Vakharia 2009). Massive outbreaks were reported across the Great Lakes region during the spring months of 2005 and 2006, killing >30 fish species (Lumsden et al. 2007; Groocock et al. 2007; Thompson et al. 2011). Thereafter VHSV-IVb became less prevalent (Cornwell et al. 2015; Stepien et al. 2015) until several smaller and more geographically restricted outbreaks were reported in the Great Lakes region in 2009, 2011, and 2017 (Faisal et al. 2011 2012; Cornwell et al. 2015; M. Faisal, personal communication 2017; R. Getchell personal communication 2017). Additional evidence for VHSV-IVb persistence comes from non-pathogenic virus-positive fishes collected between 2012–2016 in Lakes Michigan, Erie, Ontario, and the St. Lawrence River (Stepien et al. 2015, Getchell et al. 2017, Niner and Stepien in re-review). Prior studies showed that VHS-IVb has undergone rapid evolutionary trajectories, with significant geographic differentiation and temporal diversification (Stepien et al. 2015; Niner and Stepien in re-review). Its evolution appears to follow a “quasispecies” pattern of multi-directional, “cloud-like” diversification of 78 closely related variants (Lauring and Andino 2010; Andino and Domingo 2015). Knowledge of changes across VHSV-IVb’s genome and the functional roles of genes will enhance understanding of the co-evolutionary “arms race” between this pathogen and its hosts. Only five VHSV-IVb genomes were previously fully sequenced (Ammayappan and Vakharia 2009; Getchell et al. 2017), and just one publication evaluated similarities and differences (Getchell et al. 2017). Our study is the first to compare phylogenetic trends across full VHSV-IVb genomes, analyzing a wide historical representation of isolates across the Great Lakes, and evolutionary patterns among different genes. VHSV infection is characterized by petechial hemorrhaging diffuse to external and internal organs, with surviving fishes continuing to shed virus up to three months post-infection (Kim and Faisal 2012). Studies on VHSV-IVb pathology largely have used the original 2003 isolate (C03MU; Kim and Faisal 2010; Cornwell et al. 2014; Coffee et al. 2017) or experimentally mutated versions (Ammayappan et al. 2011, Biacchesi et al. 2017; Ke et al. 2017), leaving uncertainty about the traits of new strains decades later. Just two studies compared the original isolate to other naturally occurring IVb isolates. Of these, Imanse et al. (2014) found that a 2010 isolate from Lake Ontario showed reduced virulence and mortality. Getchell et al. (2017) tested three other isolates in round goby, finding lower viral titers, which showed no mortality differences. After gaining entry to the host cell, a virus must successfully inhibit the innate immune defense in order to replicate. The cell’s recognition of viral byproducts triggers several host signaling cascades, especially the interferon (IFN) pathway (Janeway and Medzhitov 2002). An array of Type I IFNs are highly induced in VHSV infection and pathogenesis is triggered (Zou et al. 2014). IFNs bind to specific surface receptors of 79 effector cells, initiating transcription of numerous IFN-stimulated genes to initiate antiviral response (Collet and Secombes 2001). Thus, comparatively measuring IFN production provides insights on how effectively a virus slows down the host response, serving as a good proxy of potential host-pathogen evolution. Host-pathogen relationships often are referred to as an “arms-race” as described by Van Valen’s (1973, 1974) “Red Queen Hypothesis”, in which selection pressures from pathogens lead to death or adaptation in host populations and vice versa. Since RNA viruses mutate just below the maximum limit for functionality, they can readily adapt to host defense modulations (Pereia and Amorium 2013). VHSV-IVb has been changing over time (Thompsonet al. 2011; Stepien et al. 2015; Niner & Stepien in re-review), however, little is known about how its more recent isolates behave during infection. The rising importance of the aquaculture industry has raised concerns about fish pathogens from economic and sustainability perspectives (Lafferty et al. 2015; Kennedy et al. 2016). With new developments in biotechnology, including high-throughput sequencing, whole viral genomes are becoming more available for examining evolutionary patterns in pathogen lineages (Bayliss et al. 2017). Our study objective here is to better understand evolutionary relationships and trajectories across the genome of VHSV-IVb, and discern spatial and temporal patterns in relation to host-pathogen coevolution. We evaluate the evolutionary patterns within IVb, and to other VHSV genogroups and Novirhabdoviruses, as well as within their Nv-genes. We additionally examine changes in pathology of three 2016 VHSV-IVb isolates in comparison with the initial 2003 isolate, in relation to IFN induction and virulence. 80 3.3 Materials and Methods 3.3.1 Sample nomenclature. VHSV-IVb isolates are named here with unique identifiers, by first letter of lake name, last two digits of isolation year, followed by the first two letters of the host species’ common name. Example: the original isolate, collected from a Lake St. Clair muskellunge in 2003 here is C03MU. If more than one isolate shared the above identifiers, an additional lower-case letter was added (Table 3-1). 3.3.2 Virus Isolation. Viral isolates from 2015–16 were obtained as previously described (Niner and Stepien in re-review). Historical VHSV-IVb isolates from 2006–11 were provided by G. Kurath (USGS, Seattle, WA) as frozen infected media from Bluegill Fry (BF-2) cell culture. RNA was extracted using a previously optimized Trizol protocol (Gorgoglione 2014), re-suspended in RNase-free water, and quantified with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Frozen media was thawed on ice, 30 µL from each sample diluted 1:5 with serum-free MEM (ThermoFisher Scientific), and added to individual wells of a 12 well plate with confluent BF-2 monolayers. Cells were incubated with the infected media at 20°C for 1 hr, then replaced with complete MEM 10% (v/v) cosmic calf serum (GE Healthcare) and 1% penicillin/streptomycin antibiotics (Invitrogen). Infected cells were incubated at 20°C for ≤one week and sampled when ≥80% cytopathic effect (CPE) was achieved. Media was collected and added to a 1.5ml tube with 250 µL of versene (Fisher Scientific) for 10 min prior to centrifugation for 4 min at 4°C and 4000 rpm. Supernatant was discarded, and the versene/cell mixture spun again with the same settings. Supernatant was discarded, and 250 µL Trizol® (ThermoFisher Scientific) was added to the remaining pellet. 81 3.3.3 Full genome sequencing. cDNA was synthesized from total RNA extracted from tissue samples using SuperScript IV (Invitrogen), following manufacturer’s instructions. Genomic cDNA was amplified in four segments using primers from Schönherz (2016), with VHSV_Frag1I_nt18_+s substituted for a more specific primer (5’GAGAGCTGCAGCACTTCACCG C3’), and 1μl cDNA in 25µL reactions with One Taq DNA polymerase (New England Biolabs). Amplicons were examined under UV light on 1% agarose gels stained with ethidium bromide. Target PCR products were excised from gels and purified using QIAquick Gel Extraction kits (Qiagen). Additional PCRs amplified the front 700 NT and end 400 NT of the genome, with extension time of 45 s. The front segment utilized VHSV_Frag1I_nt18_+s (5’GAGTTATGTTACARGGGACAGG3’) (Schönherz et al. 2016) and anti-sense 5’TGACCGAGATGGCAGATC3’, and end primers were designed based on the VHSV- IVb genome (GenBank: GQ385941) (End sense: 5’CCCAGATGCTATCACCGAGAA3’, End anti-sense 5’ACAAAGAATCCGAGGCAGGAG3’). Cleaned products were Sanger sequenced by Cornell DNA Services (Ithaca, NY) and sequences aligned and analyzed using MEGA X (Kumar et al. 2018). Genomic sequencing was outsourced to the Molecular and Cellular Imaging Center at Ohio State University (Wooster, OH). Sequences were uploaded to the Galaxy web platform, and programs at usegalaxy.org used to analyze the data (Afgan et al. 2016). Segments were aligned to the reference VHSV-IVb genome (C03MU, GenBank: GQ385941) using MAP WITH BWA-MEM (Li et al. 2013). For each of the 46 isolates, consensus sequences were generated followed by manual checking of each single 82 nucleotide polymorphism (SNP) and coverage read using Integrative Genomics Viewer (IGV: Robinson et al. 2011, Thorvaldsdóttir et al. 2013). Consensus sequences, front, and end segments were concatenated, aligned, and trimmed in MEGA X. 3.3.4 Genetic analyses. Additional complete VHSV genomes (I–IVa; Table 3-1) and Novirhabdoviruses (Table 3-2) were downloaded from GenBank and aligned with MEGA X. Basic Local Alignment Search Tool (BLAST) was used to detect similar sequences to Nv, excluding the Novirhabdovirus genus. JMODELTEST v3.7 (Posada and Crandall 1998) determined the best-fitting evolutionary model using the Akaike Information Criterion (AIC) (Posada 2008) for IVb, Novirhabdovirus genus, and protein Coding Sequences (CDSs). Phylogenetic analyses evaluated the most parsimonious evolutionary relationships among VHSV-IVb variants in relation to outgroup IVa (per Pierce and Stepien 2012, and Stepien et al. 2015). Both Maximum Likelihood (ML) (PHYML v3.0; Guindon and Gascuel 2003) and Bayesian (MRBAYES v3.1; Ronquist and Huelsenback 2003) analyses were run for all trees. For the Bayesian approach, Metropolis-coupled Markov Chain Monte Carlo (MCMCMC) analyses were run for five million generations, sampling every 100 to obtain posterior probability values. Burn-in was determined by plotting the log likelihood values to identify stationarity, discarding the first 25% of runs. Branch support for ML analyses was determined from non- parametric bootstrapping replications. Due to software limitations, number of bootstrap replications was limited to 500 for the Novirhabdovirus tree, 1450 for IVb, 2000 for the Nv NT origin tree, and 500 for the Nv AA origin tree. We define a haplotype as “a unique gene sequence differing by one or more nucleotide substitutions” from C03MU. Haplotype networks were analyzed in POPART 83 (www.popart.otago.ac.nz) using TCS (Clement et al. 2000). To discern whether samples significantly diverged over time or space, pairwise θST (FST analogue; Weir & Cockerham 1984) comparisons were conducted in ARLEQUIN (Escoffier & Lischer 2010). Values are reported prior and after sequential Bonferroni corrections (Rice 1989). Additional haplotype networks were created for IVa and the combined group of Ia, Ib, and III (I/III). 3.3.5 Evaluating evolution and selection. Comparative divergence times were estimated from BEAST v1.10.4 (Suchard et al. 2018), using JMODELTEST output. BEAST model parameters followed a relaxed molecular clock with a lognormal distribution, sampled every 50,000 of 500,000,000 generations. Outputs were assessed with TRACER v1.5 (in BEAST) to ensure stationarity. Collection dates were used as calibration points and tree branches were set following PHYML output. Numbers of nucleotide substitutions per site per year (k = substitutions site–1 yr–1) were determined from pairwise (p) distances. Nucleotide and amino acid substitutions were evaluated for all IVb isolates and compared to C03MU. IVa isolates were compared to KRRV9601 (GenBank: AB179621) and I/III was compared to the European Ia isolate Hededam (GenBank: Z93412). Two codon-based methods examined the possibility of selection pressures for each gene. Fast, unconstrained Bayesian approximation (FUBAR; Murrell et al. 2013) identified positive or purifying selection; pressures that select for benefical traits. However, FUBAR’s assumption of constant selection may not accurately represent IVb, since selection pressures may differ across hosts and environmental factors. To remedy this, we used MEME (mixed effects model of evolution; Murrell et al. 2012), which can 84 detect positive selection, under strong purifying selection, or the removal of detrimental variants. FUBAR and MEME were run with HyPHY on DataMonkey (www.datamonkey.org), with significance evaluated using posterior probability >0.95 for FUBAR and p<0.05 for MEME. Both methods were run on the IVb, IVa, and I/III datasets. The amino acid sequences of each IVb isolate’s G-gene were submitted to the Phyre2 web portal (https://www.sbg.bio.ic.ac.uk/phyre2/) for protein modeling, prediction, and analysis (Kelley et al. 2015). For open reading frame (ORF) analysis, full length cDNA nucleotide sequences for each isolate were submitted to the NCBI ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder/) to look for alternate products with >300nt in length. Sequences returned coding in reverse were automatically discounted due to the single-stranded nature of VHSV. 3.3.6 Cell lines and cell culture experiments. Epithelioma papulosum cyprinid (EPC) (ATCC: CRL-2872) and BF-2 (ATCC: CCL-91) cells were grown in conditions described above. Three viable independent viral stocks of the 2016 isolates, E16GSa (Cell16a) and E16LB (Cell16b and Cell16c) were derived from pooled organs (including kidney, liver, and spleen) of sampled fish. Cell culture amplified stocks were distinguished from their original isolates by the “Cell” designation. The original sample of E16LB was not fully sequenced due to low concentration, but was partially sequenced and matched E16GSa within those regions (Niner and Stepien in re-review). The cell culture amplified reference control (CellC03) was derived from C03MU. VHSV-IVb isolates were amplified for subsequent purification by infecting confluent monolayers of BF-2 cells in 15 cm tissue culture dishes with a 1:1000 (v/v) dilution of un-purified virus 85 stock in serum-free EMEM. Viral adsorption was allowed for 1 h and thereafter the medium was replaced with complete EMEM. Plates were incubated at 20°C, until >75% cytopathic effects (CPE) occurred, ~72 h post infection (hpi). Virus containing media and attached cells were collected and subjected to one freeze-thaw cycle before removal of debris by low speed centrifugation (4,000g for 30 min, at 4°C). Supernatant from each isolate was clarified using 0.22 μm syringe-tip filter and viral particles purified through a 25% (w/v) sucrose pad upon ultra-centrifugation (25,000 rpm for 3 h at 4°C). Virus- containing pellets were re-suspended overnight at 4°C in PBS. Virus stocks were titered by 1:10 serial dilution using confluent EPC cells, then divided into 100 μL aliquots and stored at -80°C. Virus-induced CPE were quantified using Sulforhodamine B (SRB) assay, as previously described (Ke et al. 2017). Infected EPC cells were sampled respectively at 0, 18, 48, 72, or 96 hpi at MOI=1.0 with C03MU or Cell16a, -b and -c. Cells and media were collected at each time point and stored at -80°C. A viral yield assay compared the replication ability of the three test isolates of E16GSa to that of C03MU virus, by titering media from each time point in 1:10 serial dilutions on BF-2 cells. Plaques were counted 96 hpi and final viral concentration in plaque forming units per mL (pfu/mL) was calculated for each time point. Antiviral assays were performed as previously described (Ke et al. 2017). UV- irradiated media from each of the above harvested time points were added to EPC cells in 1:3 serial dilutions for 24 h. Cells then were challenged with sucrose-purified C03MU VHSV for 96 h, fixed, and stained with crystal violet (Sigma-Aldrich). CPE plaques were counted and normalized to counts obtained from untreated and uninfected wells. One unit of IFN was defined as the dilution that conferred 50% protection from viral CPE. This 86 value then was used to measure the IFN activity (mL) in the testing culture media at each time point. Total RNA was isolated from infected EPC cells, using a previously described TRIzol protocol (Gorgoglione 2014). One μg of each RNA sample was reverse transcribed into cDNA by incubation with 100 ng of random hexamer primer (Thermo- Fisher Scientific) and water, for 7 μL total volume, at 70°C for 10 min. Reactions were cooled to 4°C before adding 13 µL of Moloney Murine Leukemia Virus Reverse Transcriptase (M-MLV-RT) mixture [10X First Stand buffer, 10 mM dNTPs, 0.05 mM random hexamers, 25 U/µL RNasin Plus (Promega), and 200 U/µL M-MLV (Promega)] and incubated at 42°C for 1 hr. cDNA was diluted 10 fold in water and archived at -80°C. 1 μL of each cDNA was tested using RT-qPCR, added to 5 μL Radiant Green Lo-ROX 2× qPCR kit (Alkali Scientific), 50 ng of each oligonucleotide and water to total 10 μL volume. Primer pairs used in this study were previously published: VHSV-Nse/as, EPC IFN se/as, and β-actin se/as (Ke et al. 2017). Reactions and data collection were performed on a C1000 Real Time Thermocycler (Bio-Rad), with initial 3 min at 95°C for denaturation, 40 cycles of 15 s at 95°C and 30 s at 60°C for elongation. Readings were recorded at the end of each elongation cycle, and threshold values obtained from an automated single point threshold within the log-linear range. The detection of VHSV-N and EPC IFN was normalized to the detection of EPC β-actin, and normalized to the gene expression recorded from uninfected cell samples. The relative gene expression was calculated using the 2-ΔΔCT method (Livak and Schmittgen 2001). 87 3.4 Results 3.4.1 Genomic and genic changes. We obtained and analyzed 43 VHSV-IVb sequences of 11,083 NT from Illumina MiSeq, which had no insertions or deletions. Of these, 35 unique sequences were discerned (0.81), differing ≥1 NT from C03MU (GenBank: MK782981–MK783014). Isolates C06NP, C06RB, C06SR, C06YP, C06FD, M08AM, C08LEa, C08LEb, and C09MU were identical, designated here as the “C06NP group” (MK782990). We analyzed all available IVb whole genome sequences from GenBank (GQ385941, KY359355-57), bringing the IVb data set to 39 unique sequences (Table 3-1). No new ORFs were detected. Predicted structures of the G-gene were all identified as glycoproteins. SNPs per gene are summarized in Table 3-3A, revealing 253 SNPs, of which 85 (0.36) were nonsynonymous, resulting in different amino acids (AA). Most SNPs and AA changes were found in L (112 SNPs, 38 AA) and the least in Nv (14 SNPs, 7 AA). This was expected for the longest (L-gene: 5955 NT) and shortest (Nv- gene: 369nt) regions. The NCDS, which does not encode amino acids, contained the highest proportion of SNPs (32 SNPs, 0.043 of the region), while P and L had the lowest (0.019). Nv encoded the highest proportion of AA changes (0.057) and P the least (0.014). The highest proportion of nonsynonymous changes (dN/dS) occurred in M (0.53) and lowest in P (0.23). The overall dN/dS for the complete sequence was 0.166. In comparison, IVa (Table 3-3B) had 787 SNPs; 203 were nonsynonymous (0.26) and I/III (Table 3-3C) contained 1292 SNPs, with 225 AA changes (0.17). Most changes for both were in L (IVa: 424 SNPs, 119 AA; I/III: 618 SNPs, 70 AA). For IVa, Nv had the fewest SNPs (31 SNPs, 10aa), and M had the fewest for I/III (65 SNPs, 15 AA). In IVa, the largest proportion of SNPs occurred in the NCDS, and the Nv and P genes had 88 the most AA changes (0.081). For I/III, Nv contained the largest proportions of SNPs (0.206) and AA changes (0.220). The lowest proportions of SNPs occurred in N (0.044) for IVa, and P (0.019) for I/III. For AA changes, the least were IVa-N (0.042) and I/III-L (0.035). For both groups, the P-gene had the highest dN/dS ratio (IVa: 0.400, I/III: 0.411) and L the lowest (IVa: 0.281, I/III: 0.113). IVb possessed the least transversions (Tv; IVb=4 2, IVa=131, I/III=254), and tied with IVa in overall ratio (Tv/Ts=0.166). I/III had a larger Tv/Ts ratio (0.197). Most Tv occurred in the L-gene for IVb (N=18, Tv/Ts=0.161), IVa (N=65, Tv/Ts=0.153), and I/III (N=106, Tv/Ts=0.172). All three groups differed in which genomic regions had the fewest Tv (IVb: P and M, N=2, Tv/Ts=0.118; IVa: Nv, N=3, Tv/Ts=0.097; I/III: M, N=10, Tv/Ts=0.154). The largest proportion of Tv/Ts (transitions) occurred in IVb-Nv (Tv=4, Tv/Ts=0.286), and NCDS for both IVa (N=0.273, Tv/Ts=0.273) and I/III (N=36, Tv/Ts=0.310). The lowest Tv/Ts ratios were: IVb-NCDS (Tv=3, Tv/Ts=0.094), IVa-Nv (Tv=3, Tv/Ts=0.097), and I/III-M (N=10, Tv/Ts=0.154). Per isolate changes were examined for the more recent IVb isolates. The average number of SNPs/isolate was 16.1, with 6.1 (0.38) AA changes (Sup Table 3-A). The C06NP group had the least SNPs, with a single non-synonymous change in L. Isolates recovered between 2012 and 2016 contained the highest proportions of SNPs (averaging 27.9 SNPs, 9.5 AA), ranging from 14 (M16RGa, 4 AA) to 38 (E16GSa, 13 AA). Isolates recovered from cell culture differed by a few SNPs from their original sequences: CellC03 was 4 SNPs from C03MU and Cell16a–c were 2, 6, and 8 differences from E16GSa, respectively. 89 SNP changes are visually represented in Fig. 3-3. C03MU is centrally located nearby a large cluster of isolates, including the C06NP group. Sequences from isolates recovered from 2006–2008 Lake Erie outbreaks cluster together, differing by 1–4 SNPs. E12FD radiates from this cluster by 31 additional SNPs. Both Budd Lake, MI isolates (B07BS, PS) form an individual branch from C03MU, sharing one synonymous SNP in G. Both Lake Ontario isolates (O06SB, O13GS, Getchell et al. 2017) group together, sharing 10 SNPs before diverging by 31 SNPs. Of these shared changes, 5 are nonsynonymous: one in both M and L and three in G. Lake Michigan isolates appear scattered throughout the network, but remain closer to C03MU, with the more recent 2016 isolates differing by 15–16 NT (8 shared). E16GSa-e and derived cell culture isolates form a distant, smaller cluster. This group shares 14 NT changes before E16GSc branches off. The remaining seven isolates share 19 NT changes before radiating into individual isolates, with E16GSb as central. More SNPs occurred between isolates of IVa (Sup. Fig. 3-A) and I/III (Sup. Fig. 3-B). In the IVa network, the two Japanese isolates were very distant (KRRV9822: 84nt, JFEh001: 253nt) from the nearest Korean isolate (KJ2008). The Korean samples, while more similar, do not form distinct clusters as seen in IVb. Two identical isolates occurred in I/III, recovered a year apart in Norway (BV06048–52, FA281107). Seven Ib Swedish isolates from 1998–2000 and a 1996 Japanese fish farm isolate form a cluster, differing by 2–22 NT. The fastest overall evolutionary rate characterized IVa (2.01x10-3 substitutions/site/year), followed by IVb (6.64x10-5) and I/III (4.09x10-5) (Table 3-3). For IVa, NCDS had the highest rate (7.79x10-3), followed by P (2.32x10-3) and Nv 90 (2.03x103). For IVb, NCDS appeared faster (1.40x10-4), followed by Nv (9.76x10-5) and G (8.51x10-5). The order for I/III was: Nv (7.78x10-5), M (5.45x10-5), and N (4.59x10-5). The slowest regions were IVa: G (1.02x10-3), M (7.68x10-4) and N (5.97x10-4), IVb: P (7.18x10-5), N (5.83x10-5), and L (5.02x10-5), and I/III: P (4.07x10-5), NCDS (3.60x10-5), and L (3.58x10-5). Selection pressures for each CDS were examined (Table 3-4) for IVb, IVa, and I/III. Purifying selection characterized IVb’s N (one codon; 313), G (one codon; 342), and L (six codons’ 8, 119, 333, 460, 1284, and 1758) genes. For IVa and I/III, FUBAR implicated purifying selection for all genes (Table 3-4B, C). In IVb, just one codon (L, 1758) showed signs of purifying selection and two codons appeared under diversifying selection – one in G (103, 431) and one in Nv (25). A IVa codon reflected diversifying selection (G, 12), along with one in I/III (N, 46). MEME implicated diversifying selection for IVb-G431 and I/III-G477, and three different L sites each in IVa (147, 593, 1154) and I/III (112, 474, 1012); none matched. 3.4.2 Evolutionary relationships. Significant genetic differentiation occurred over time in IVb, as shown by pairwise genetic divergences in the full genome analysis and individual gene regions (Table 3-5A). The later time group (2012–16) significantly diverged from both the early (2003–06) and middle (2007–11) groups. Greater divergence was found between the early and later time groups, indicating increase over time. Population region groupings (Lakes Michigan/Budd, St. Clair, and Erie/Ontario) significantly diverged from each other (Table 3-5B), with greatest differences between Michigan/Budd and Erie/Ontario. This trend was not observed between individual CDS and NCDS. Results for the Nv and P genes showed no significant differences among the 91 geographic population groupings. Pairwise regional population groupings for M and NCDS indicated significant difference between Erie/Ontario and the other regions. Michigan/Budd significantly diverged from the others in L, whereas G showed differences between St. Clair and the other regions. N differed between Michigan/Budd and St. Clair. Phylogenetic trees of Novirhabdovirus genomes (Fig. 3-4) reveal distinct separation of all four species, with each defined by 1.00 posterior probability (pp) and 100% bootstrap support (bs). IHNV and HRV are sister species, forming the sister group to VHV and SHRV. The 79 genomic VHSV sequences are divided into two major clades: European strains I–III and North American/Asian IV. Within the European clade, the single VHSV-II isolate appears basally located. Two of the five VHSV-III genotypes group with two Ia isolates (Hededam, DK35928) and thus that group is not monophyletic (i.e., not supported as a phylogenetic clade). In contrast, the seven Ib isolates form a well- defined clade. Two III isolates form another well-defined clade. Genogroup IV is well- defined (Fig. 3-3), having 1.00 pp and 100% bs support. IVa (N=25) and IVb (N=39) are located on separate branches, with 1.00 pp and 100% bs support. The IVb tree (Fig. 3-5) contains several inner clades, with M08RB appearing basal (weakly supported), followed by O06SB, CellC03, and O13GS, respectively (0.60– .90 pp/<50% bs). Two Lake Erie isolates from 2014–15 (E14GS, E15RG) group together (1.00 pp/74% bs), followed by two Lake Michigan isolates (from 2007–08), located prior to the main cluster of the remaining 39 isolates, including the original, C03MU. Within the major group are two clades containing two isolates (B07BG/B07PS: 0.80–.89 pp/66% bs; M11YP/C06GS: 0.80–.89 pp/<50% bs) and two larger clades. The first larger clade 92 incorporates the 2016 isolates (0.60–.69 pp/<50% bs), further subdivided into two clades, separating Lake Michigan (1.00 pp/98% bs) from Lake Erie and cell culture derivatives (1.00 pp/100% bs). A second well-defined clade (1.00 pp/83% bs) contains 14 isolates, including most early to middle period genotypes (2006–08) from Lake Erie, and the very distinct isolate from 2012. The genotype from 2012 is the most distant and appears as the sister haplotype to E08ES from 2008 (1.00/94%), containing four (two AA) SNPs that are shared with the clade (M: one SNP, one AA; L: two SNP, one AA; NCDS: one SNP). E12FD and E08ES share an additional five SNPs: two in N (two AA), and two in L (one AA). BLAST returned no similar sequences to Nv when novirhabdoviruses were omitted as both NT and AA sequences were used. The individual gene phylogeny based on NT (Fig. 3-6A) showed distinct, separate branches for all but Nv. Two major clades divide the genes: M and G, vs. N, P and L. SHRV-Nv forms part of the M and G clade branching, sharing a common ancestor with the G clade. IHNV-Nv and HIRRV-Nv are sister clades, sharing a common ancestor with the N clade. VHSV’s Nv genes are a divergent part of the L clade. All divisions of Nv branching from other genes were fully supported by Bayesian analysis. When AAs were considered (Fig. 3-6B), the IHNV-Nv and HIRRV-Nv clade, again well supported, was sister to VHSV-Nv. This group of Nv genes formed a sister clade to the G clade. SHRV-Nv formed its own clade and was closer related to the P and L clades, but not well supported. All non-Nv genes again formed their own clades, but with less supported compared to the NT trees. 93 3.4.3 Differences in cytopathicity and immune response. To determine whether sequence variations impacted viral function, a series of cell-based studies were performed. These included SRB assays to assess cytopathicity, virus yield assays to estimate viral production, antiviral assays to assess IFN production, and qPCR to determine levels of IFN and VHSV-IVb mRNA production. Sequencing confirmed the identity of each isolate in our study and revealed changes acquired during cell culture propagation. The cell culture amplified control, CellC03, differed from C03MU by 4 NT and 2 AA, M (2312, NT:C-T, AA:T-I) and G (3397, NT:G-A, AA:K-D; 4007, NT:C-G, AA:G-G; 4394 NT:G-A, AA:V-V). Cell16a–c differed by 34–38 SNPs (13–14aa) from C03MU (Sup Table 3-A). From E16GSa, Cell16a differed by 2 NT: N (737, NT:C-A, AA:T-T) and G (4117, NT:G-A AA:D-N), Cell16b differed by 4 additional changes (6 total) in L (6962, NT:C-A, AA:A-E; 7038, NT:G-A, AA:E-E; 7047, NT:T-C, AA:C-C; 7647, NT:T-C, AA:H-H), and Cell16c had one further NT change (1 total) in L (6456, NT:C-T, AA:L-L). SRB staining examined CPE elicited at different MOIs across the 4 isolates. The 2016 isolates exhibited less CPE than CellC03 when tested at MOIs >1x10-3 (Fig. 3-7). Although not statistically different across all isolates and at all dilutions, the 2016 viral isolates tested were generally less cytotoxic than CellC03. Consistent with this observation, viral yield assays demonstrated that at 96 hpi, CellC03 produced significantly more virus than the 2016 isolates (Fig. 3-8), yielding nearly 100-fold more infectious particles as compared to Cell16a–c. 3.4.4. IFN production and expression of VHSV and IFN mRNAs. IFN production was measured in an antiviral bioassay, which revealed earlier onset of 94 antiviral immune response in cells infected with Cell16a, yielding significantly more IFN at 24 hpi than CellC03 (Fig. 3-9). All of the isolates produced comparable IFN activities after 24 hpi (Fig. 3-9). As IFN antiviral bioassays are relatively insensitive to minimal changes, we extended the previous studies by assessing IFN mRNA synthesis in infected cells with RT-qPCR. IFN mRNA was expressed significantly more in Cell16a–c, compared with the CellC03 control across all time points beyond 24 hpi (Fig. 3-10). These data suggest significant difference in ability of the 2016 isolates to induce cellular IFN expression, or a significant reduction in their suppression of IFN expression. Consequently, by 48 hpi, CellC03 produced the highest overall amount of viral RNA at 48 hpi, which continued until 96 hpi, by which point cell death in all cultures led to reduced viral expression. All 2016 isolates produced similar amounts of viral RNA. 3.5 Discussion 3.5.1 Evolutionary trends. We analyzed 46 sequences of VHSV-IVb in the largest Novirhabdoviruses genomic study to date. The greatest proportion of SNPs was in the NCDS (4.3%) and the Nv-gene (3.8%). In IVb-Nv, none of the isolates we sampled had >two SNPs, suggesting that sequence conservation is important. An experimental study on Nv function indicated that mutations at codon positions 36, 39, and 41 had greater host immune suppression (Chincilla and Gomez-Casado 2017); our sequences lacked SNPs at those locations. As Nv is not essential for viral replication (Ammayappan et al. 2011), its SNP variations may be more tolerable for the virus. Large proportions of SNPs likewise characterized NCDS in IVa and I/III, illustrating sequence flexibility for those regions. There is no evidence that NCDS regions are transcribed or play a 95 functional role in any rhabdovirus (Walker et al. 2011, Walker et al. 2018), thus its SNPs likely are not deleterious. Diversifying selection was detected in two genes in IVb: G and Nv. A study of four IVb genome sequences (C03MU, O06SB, O13GS, and E14GS; Getchell et al. 2017) uncovered no diversifying selection for either G or Nv, but discerned variation in two codons in N and one in both M and L that were unsupported here. We found evidence of purifying selection acting on single codons in N and G, and six codons in L. Fewer codons were found to be under purifying or diversifying selection in IVb than in Getchell et al.’s (2017) study. None matched between the studies, with ours being the larger, more robust dataset. Purifying selection was the major force across all VHSV genogroups, subgenogroups, and samples. Both IVa and I/III identified ≥one codon per gene under negative selection. No NT sites under selection were in common between IVb and either of the other genogroups, with one exception: codon 1758 in L for IVb and I/III. Our dN/dS ratio further indicated purifying selection, with values being >0.5 across all genes and VHSV genogroups, excepting IVb-M (0.529). Multiple studies have indicated that purifying selection regulates VHSV evolution (He et al. 2014; Abbadi et al. 2016), as well as other rhabdoviruses (Kuzmin et al. 2009), and RNA viruses in general (Hughes & Hughes 2007). He et al. (2014) compared dN/dS ratios for individual genes among VHSV genogroups, with all six genes having lower dN/dS than our data. However, most IVb (N=45) and IVa (N=18, from Hwang et al. 2018) sequences used in our study were not available to He et al. (2014). 96 Positive selection was detected for three G-gene codons: two in IVb and one in IVa. Abbadi et al. (2016) examined 108 Ia full-length G isolates, discerning positive selection on two codons (258, 259). He et al. (2014) also elucidated positive selection across VHSV at codons 258 and 476, but not 259. None of our codons matched either analysis. Positive selection on viral genes constitutes evidence for the host-pathogen “arms race”, since their codons typically suppress host immune responses (Pereira and Amorium 2013). This is seen in Human Immunodeficiency Virus (Bush et al. 1999) and the presence of regions that produce RNA interference in RNA viruses in general (Marques and Carthew 2007). Indications of positive selection in G, in multiple studies, supplies evidence of its role in defeating the host cell immune response. Overall evolutionary rates were slower in our study than were previous estimates. We calculated slower overall evolutionary rates for four genes (P: 7.186x10-5 v. 1.2x10-4, M: 7.936x10-5 v. 1.2x10-4, G: 8.516x10-5 v. 2.8x10-4, and Nv: 9.766x10-5 v. 2.0x10-3) and a slower rate across the entire genome (6.6x10-5) as compared to the concatenated regions previously examined (4.8x10-4) (Niner and Stepien, Chapter 2). Piece and Stepien (2012) estimated N’s rate at 4.26x10-4; here our rate was much slower (5.83x10-5) for the entire gene. Our slower estimated rates stem from our present use of more samples and complete genome sequences, which included conserved areas that were previously omitted. Among the three genogroups, IVa exhibited the fastest rate – 2.01x10-3. A prior study of 48 IVa sequences estimated a rate of 5.60x10-4 for G (He et al. 2014), which was slower than our IVb rate for G of 1.02x10-3. Our inclusion of recently published diverse sequences from Korea between 2012–16 (Hwang et al. 2018) may have increased this 97 rate, as those isolates differ by >20nt. Only two complete sequences of IVa were from Japan, which were even more different from the Korean isolates. This result suggests that the Korean isolates are diverging from the Japanese IVa isolates, but additional IVa genomes from Japan are needed for a robust conclusion. IVa first was detected along the North American Pacific Coast (Brunsen et al. 1989; Hopper 1989; Meyers et al. 1992), but no full-genome sequences then were available. All IVa Korean isolates were obtained from aquaculture; such transportation of infected fishes may introduce VHSV to naïve fishes and different environmental conditions, which can enhance adaptation and dramatically alter the virus, as observed in IHNV (Xu et al. 2018) and HIRRV (Zhang et al. 2017). Genogroups I/III had overall genome rates more similar to IVb, at 4.09x10-5. Some Ia, Ib, and III isolates group together in our phylogeny, indicating that their previous phylogenetic relationships, based on partial genes (Nishizawa et al., 2006; Pierce and Stepien 2012; Ghorani et al. 2016), likely were incorrect. Previous estimates for VHS-I and III using a lower number of complete G-gene sequences were faster than our estimate for IVb (4.54x10-5) of G at 5.57x10-4, based on 201 complete I sequences, and 1.63x10-3 using seven III isolates (He et al. 2014). Subgenogroup Ia was estimated at 1.74x10-3 (N=34, Einer-Jensen et al. 2014) to 7.3x10-4 from 108 Italian isolates (Abbadi et al. 2016). Our combination of the European groups may have yielded the slower rate. Despite discrepancies, all above estimates were within the range of evolutionary rates known for RNA viruses (Duffy et al. 2008). 3.5.2 Phylogenetic patterns: novirhabdoviruses. The present phylogenies (Fig. 3-3) are congruent with those from prior partial gene studies (Pierce and Stepien 2012, 98 Stepien et al. 2015). Two main sister groups characterize the novirhabdoviruses: IHNV/HIRRV and VHSV/SHRV, with VHSV/SHRV diverging first from the common ancestor. This relationship is congruent with Kurath’s (2012) analysis of complete N sequences. Two studies examined rhabdoviruses using partial L (Bourhy et al. 2005, Kitchen et al. 2016); which likewise supported our sister pairing of IHRV/HIRRV. Our phylogenetic consensus tree has 100% ML and Bayesian support for these relationships, whereas Kurath (2012) found just 76% support for the SHRV/VHSV clade using N. Thus, our analysis of the entire genome using a larger number of isolates significantly increased confidence in resolving VHSV evolutionary relationships. Resolution within VHSV supports most known genogroups and subgenogroups, except that III and Ia are not monophyletic. Prior analyses of G-gene sequences by Dale et al. (2009) and Ghorani et al. (2016) were incorrect in separating them. Genogroup Ia occurs in freshwater hosts (Einer-Jensen et al. 2004), whereas both III and Ib are marine (Dale et al. 2009); however, all three are capable of infecting rainbow trout (Oncorhynchus mykiss), a species that spends portions of its lifecycle in both environments. Our results indicate that Ib is monophyletic, whereas III and Ia comprise a single taxon, and needs to be re-defined. Additional III and Ia isolates should be completely sequenced. 3.5.3 Phylogenetic patterns: VHSV-IVb. Our IVb tree contains several clades of closely related isolates, which corresponded to sampling location and year. We provide an in-depth view of the virus in Lake Erie, where a large portion of our samples came from. The inner-most branches of the phylogeny suggest that at least two separate 99 infections led to the mid 2000s outbreaks, as one of the 2006 isolates (E06FD) appeared more closely related to C03MU and the Lake St. Clair outbreak samples. The distinct Lake Erie clade from 2006–08 and 2012 suggests a separate evolutionary history from the 2014–15 and 2016 clades. Wide diversification characterized the 2006 outbreak, which continued along that trajectory in 2007–08, with minor variants. The lone 2012 sample, although distant, appears more closely related to the E06–08 clade, than to Lake Erie samples from subsequent years. E12FD forms a sister clade with E08ES, sharing an additional four AA changes in the coding regions of N, P, and L. These AA changes possibly provided some advantage or resulted in no deleterious effects, since the substitutions remained conserved for four years. Although most of the clade was recovered from Percidae hosts, several were in freshwater drum. In addition to the 2006 and 2008 Lake Erie cases (Thompson et al. 2011), freshwater drum die-offs were reported in 2005 from Lake Ontario (Lumsden et al. 2007), which may have been more closely related to E06FD. Both of our 2007 samples from the inland Budd Lake form a clade near C03MU. Wayne County, Michigan is adjacent to Lake St. Clair, which is a highly urbanized area having the largest number of registered anglers in the state (Burkett and Winkler 2018). Anglers traveling from that area to Budd Lake may have unknowingly transported VHSV, perhaps via live bait (Stepien et al. 2015; Throckmorton et al. 2015). The 2014–15 Lake Erie genotypes differ from one another, but have a sister relationship branching from the main part of the phylogeny, closer to C03MU. ES14GS and ES15GS shared host species of gizzard shad and round goby with the 2016 isolates, but are not closely related. The 2016 isolates branch together, with low support for the 100 Lake Erie and Michigan groupings, but are linked by SNPs shown in the genotype network. Interestingly, these groups also occurred in different host species, with E16 isolates exclusively recovered from gizzard shad, and both M16 in round goby. Other studies identified potential radiation of IVb in round goby, particularly in Lake Ontario (Cornwell et al. 2014; Niner & Stepien in re-review). There have been relatively few VHS positive round goby samples obtained from Lake Michigan, despite its presence in the 2008 outbreaks (Faisal et al. 2012). Round goby has been implicated in spreading the virus (Rude et al. 2017), which often is used as bait, and is transported among water bodies by anglers (Karsiotis et al. 2012). Other VHS-positive isolates were recovered from Lakes Michigan and Erie in 2016, whose viral titers were too low for genome sequencing (Niner and Stepien, in re- review), further indicating that infection levels differ spatially and temporally. One of the first outbreaks in 2006 intensely affected the Lake St. Clair gizzard shad population (Faisal et al. 2012), as well as its 2017 outbreak (G. Whelan, personal communication 2017). Gizzard shad frequently experience large die-offs from cold weather (Feltzer et al. 2011), water temperature fluctuations, and/or spawning activities (Miller 1960). Gizzard shad is a common host for VHSV-IVb, whose optimal temperature range (12–18 °C, Kane-Sutton et al. 2010; 10–14 °C, Ekerlin et al. 2011) coincides with these high stress events. In addition to the 2016 isolates, IVb-positive gizzard shad also occurred in Lakes St. Clair, Erie, and Ontario dating back to 2006; however, only its 2016 isolates are closely related. Possible host specialization should be examined in future investigation. All but one of our identical sequences (C06FD group) were collected from Lake St. Clair over a five-year span. Its consistency suggests that particular genome may have 101 been prevalent and adapted to local conditions and host species. It differs from C03MU by a single AA in L, and was found in six different fish host species and the sole two known invertebrate host species (Faisal and Schulz 2009; Faisal and Winters 2011). Yusuff et al. (2019) discerned no effect on host specificity when swapping L-genes between C03MU and Ia’s DK-3592B. L variants appear to differ in optimal temperature range, since swapping L from IVa into IVb led to increased virus at higher temperatures, than found in IVa (Kim et al. 2015). The single AA change between the C06FD group and C03ML was not shared with IVa; instead C03ML and IVa were identical at that NT, meriting future examination. Analyses of population divergence patterns using the IVb genome discern similar spatial and temporal patterns to those based on our prior partial G-gene analyses (Niner and Stepien, in re-review). Increased divergence characterized the later time period (2012-2016) compared to the early (2003-2007) and middle (2008-2011) time periods. This pattern is evident in all individual genes as well, suggesting that the G-gene indicates overall population trends. During the later time period, VHSV-IVb occurrences were sporadic in smaller, localized outbreaks, perhaps due to increased host population immunity and resistance, as predicted by the “Red Queen Hypothesis” (Van Valen 1973, 1974). Likewise, an experiment with viral phage Φ2 showed greater genome divergence when exposed to changing hosts versus consistent ones, particularly in genes associated with host immune suppression (Paterson et al. 2010). On the host side, Alves et al. (2019) found genomic changes in immune-related genes of wild rabbit populations following 60 years of exposure to Myoxma virus (MYXV). 102 3.5.4 Evolutionary Perspectives on the Nv-gene. Our multi-gene phylogenies of both NT and AA sequences revealed no consensus on the origin of Nv. In the NT tree, Nv appears so different among the novirhabdoviruses, that its origins and relationships cannot be discerned. All genes, both as NT and AA, depict a sister relationship between IHNV and HIRRV, including Nv. VHSV-Nv had the most different pairing between NT and AA; appearing possibly related to the P- and L-genes for the NT sequence, but paired with IHNV-Nv and SHRV-Nv based on AA. SHRV-Nv differs the most from the other novirhabdoviruses, appearing more closely related to the G-gene for the NT tree and the P- and L-genes for the AA sequences. This might reflect different functionality, since studies showed that Nv is nonessential to pathogenicity in SHRV (Johnson et al. 2000; Alonso et al. 2004). In contrast, VHSV and IHNV require Nv for suppression of host immune response (Ammayappan et al. 2011). We conclude that Nv is highly saturated and homplastic. 3.5.5 Differences in cytopathogenicity. After 24 to 72 hpi, the 2016 isolates produced less VHSV-IVb RNA and induced higher IFN transcription, compared to the reference 2003 strain (CellC03) (Fig. 3-10). Although their viral yields did not significantly differ, by 96 hpi, CellC03 produced more active viral particles (Fig. 3-8) and was more cytotoxic at higher MOIs (Fig. 3-7). The genetic changes seen between CellC03 and Cell16a–c may have affected their ability to suppress host response and cause cellular damage. This may contribute to the virus appearing less virulent over time. Cell16b–c (both isolated from E16LB) differ by a single L-gene NT and from Cell16a (isolated from ESG16a) by four L NT (one AA). Single nonsynonymous changes can have functional consequences for VHSV. Chinchilla and Gomez-Casado (2017) 103 discerned the function of SNPs in Nv by inducing mutations in VHS-Ia strain FR07-71. Another study found that when four AA M-gene mutations were introduced to C03MU, the virus was less able to suppress the transcription of host immune defenses (Ke et al. 2017). These data suggest that even subtle changes can lead to altered gene function and viral phenotypic characteristics. It appears the oldest VHS-IVb isolate, C03MU, has remained the most virulent. More recent isolates from 2006–16 showed reduced virulence, potentially resulting from virus-host coevolution. Other studies observed similar trends. For example, Imanse et al. (2014) examined vcG001 (C03MU) and vcG002 (not included in our study), finding faster growth but lower titers in cells exposed to vcG002, which had similar levels of viral RNA, suggesting that vcG002 was less efficient at producing infective particles. Getchell et al. (2017) also found reduced viral load from O06SB, O13GS, and E14GS. Viral RNA production and virulence of Cell16a–c likewise did not significantly differ from CellC03, as evidenced by lower peak virus production in the viral yield assay. Reduced virulence over time has characterized MYXV in Australian rabbits (Kerr et al. 2017), human papillomaviruses (Bernard 1994), and RABV in hyenas (East et al. 2001), leading to a longer infectious period that may aid spread of the virus. Such pattern may similarly characterize VHSV-IVb, meriting further investigation. From our data, we are not yet able to determine the mechanism underlying the observed reduction in virulence for the 2016 isolates. Reduced virulence might reflect less ability to suppress the host immune response, or it could be that the stronger immune response is the consequence of reduced virulence itself, allowing infected cells more time to produce antiviral factors. Further investigations into the impact of individual SNP 104 changes seen in the 2016 isolates might answer this lingering question. Other studies have observed large impacts on virulence from single nucleotide changes in VHSV-I (Ballion et al. 2018) and vesicular stomatitis virus (VSV) (Sanjuán et al. 2004). It is possible that reduced virulence may play a role in aiding transmission, as discussed above. 3.5.6 Summary and conclusions. As VHSV-IVb nears the end of its second decade of evolution in the Great Lakes region, it has continued to change genetically and in virulence. IVb contains fewer SNPS than IVa or I/III, with IVa appearing to evolve the fastest, and rates of IVb and I/III being similar to each other. Our in vitro studies show reduced virulence in more recent isolates, consistent with other findings and field observations. Overall, VHSV-IVb may continue to comprise a threat to fish populations in the Great Lakes region and other waterways, as it continues to evolve. 3.6 Acknowledgements Research funding was provided by grants to CAS and DWL, including NSF-DBI- 1354806 for “Gene diversity of the VHS fish virus: Evolution of cellular immune response and pathogenesis” and USDA-ARS CRIS #3655-31320-002-00D, under specific cooperative agreement #58-3655-9-748, “VHS fish virus in yellow perch aquaculture”. We thank S. Edwards for field and laboratory work, N. Marshall for computational and laboratory help, and G. Kurath and J. Winton for providing historical samples. We are grateful to F. Averick, D. Butterfield, F. Calzonetti, K. Czajkowski, A. Elz, T. Fisher, M.E. Hernandez Gonzalez, A. Izzi, M. Krishnamurthy, D. Moorhead, and 105 S. McBride for logistical assistance. This is contribution # 4974 from the NOAA Pacific Marine Environmental Laboratory (PMEL). 106 Table 3-1. VHSV isolates for full genome analysis. IVb isolates analyzed by Niner and Stepien (in re-review) and their corresponding G-gene and combined gene haplotypes are included below. Isolate Name Genogroup Host Species Location Year Gen Bank Citation Coordinates G-gene Combined haplotype haplotype DK-Hededam I North sea cod North Sea, Denmark 1972 Z93412 Stone et al. 1997 56.648, 9.271 N/A N/A French strain 07-71 Ia Rainbow trout Seine-Maritime, France 1971 AJ233396 Yan et al. unpublished 49.9070, 0.7926 “ ” “ ” DK-3592B “ ” “ ” North Sea, Denmark 1986 KC778774 Yusuff and Vakharia 2013 57.7376, 10.4160 “ ” “ ” De-Fil3 “ ” “ ” Baltic Sea, Germany 1999 Y18263 Schutze et al. 1999 54.4212, 11.3928 “ ” “ ” Cod-Ulcus Ib North sea cod Denmark 1979 Z93414 Stone et al. 1997 56.648, 9.271 “ ” “ ” SE-SVA-1033-3F “ ” Rainbow trout Kattegatt, Sweden 1998 AB839748 Mori and Ito 2013 56.9793, 12.2100 “ ” “ ” SE-SVA-1033-9C “ ” “ ” “ ” “ ” AB839747 “ ” “ ” “ ” “ ” SE-SVA-14-3D “ ” “ ” “ ” “ ” AB839745 “ ” “ ” “ ” “ ” SE-SVA-14-5G “ ” “ ” “ ” “ ” AB839746 “ ” “ ” “ ” “ ” SE-SVA-1033 “ ” “ ” “ ” 2000 FJ460591 Campbell et al. 2009 “ ” “ ” “ ” KRRV9601 “ ” Olive flounder Seto Inland Sea, Japan 1996 AB672614 Ito et al. 2012 33.7252, 132.5249 “ ” “ ” DKp37 “ ” Blue whiting North Sea, Denmark 1997 FJ460590 Campbell et al. 2009 57.18766, 8.29058 “ ” “ ” DK-1p49 II Atlantic herring Baltic Sea 1996 KM244767 Lopez-Vazquez et al. 2015 55.1418, 15.3020 “ ” “ ” 23-75 III Brown trout Eure, France 1975 FN665788 Biachessi et al. 2010 49.2235, 1.2993 “ ” “ ” 14-58 “ ” Rainbow trout France 1990 AF143863 Betts and Stone 2000 “ ” “ ” “ ” GH40 “ ” Greenland Flemish Cap, 1994 KM244768 Lopez-Vazquez et al. 2015 47.0017, -44.9999 “ ” “ ” halibut Newfoundland 4p168 “ ” Atlantic herring Skagerrack, Denmark 1996 AB672616 Ito et al. 2012 57.7376, 10.4160 “ ” “ ” FA281107 “ ” Rainbow trout Storfjorden, Norway 2007 EU481506 Dueshund et al. 2010 62.4056, 6.0446 “ ” “ ” BV060408-52 “ ” “ ” “ ” 2008 FJ362510 “ ” “ ” “ ” “ ” KRRV9822 IVa Olive flounder Kagawa, Japan 1998 AB179621 Byon et al. Unpublished 34.3739, 133.9117 “ ” “ ” JF00Ehi1 “ ” Japanese Ehime, Japan 2000 AB490792 Ito et al. 2009 33.7334, 132.6460 “ ” “ ” flounder FYeosu05 “ ” “ ” South Korea 2005 KF477302 Kim et al. 2013 34.7533, 127.6585 “ ” “ ” Paralichthys “ ” “ ” China “ ” KC685626 Zhu and Zhang 2014 37.5297, 122.1231 “ ” “ ” olivaceus rhabdovirus (VHSV) KJ2008 “ ” “ ” Jeju, Korea 2008 JF792424.1 Kim,M.S. and Kim,K.H. 33.4936, 126.5242 “ ” “ ” Unpublished JF-09 “ ” “ ” “ ” 2009 KM926343 Kim et al. 2015 33.4936, 126.5242 “ ” “ ” FP-VHS2010-1 “ ” “ ” Geoje, Korea 2010 KP334106 Hwang et al. 2016 34.8965, 128.6190 “ ” “ ” ADC-VHS2012-10 “ ” “ ” Jeju, Korea 2012 KY979950 “ ” 33.4936, 126.5242 “ ” “ ” ADC-VHS2012-11 “ ” “ ” “ ” “ ” KY979951 “ ” “ ” “ ” “ ” ADC-VHS2012-5 “ ” “ ” Gyeongbuk, Korea “ ” KY979946 “ ” 36.2894, 128.9379 “ ” “ ” ADC-VHS2012-6 “ ” “ ” Jeju, Korea “ ” KY979947 “ ” 33.4936, 126.5242 “ ” “ ” ADC-VHS2012-7 “ ” “ ” “ ” “ ” KY979948 “ ” “ ” “ ” “ ” 107 ADC-VHS2012-9 “ ” “ ” “ ” “ ” KY979949 “ ” “ ” “ ” “ ” ADC-VHS2013-1 “ ” “ ” “ ” “ ” KY979952 “ ” “ ” “ ” “ ” ADC-VHS2013-2 “ ” “ ” “ ” “ ” KY979953 “ ” “ ” “ ” “ ” ADC-VHS2013-3 IVa Olive flounder Jeju, Korea 2013 KY979954 Hwang et al. 2018 33.4936, 126.5242 N/A N/A ADC-VHS2013-4 “ ” “ ” “ ” “ ” KY979955 “ ” “ ” “ ” “ ” ADC-VHS2013-9 “ ” “ ” Gyeongbuk, Korea “ ” KY979956 “ ” 36.2894, 128.9379 “ ” “ ” ADC-VHS2014-2 “ ” “ ” Jeju, Korea 2014 KY979957 “ ” 33.4936, 126.5242 “ ” “ ” ADC-VHS2014-4 “ ” “ ” “ ” “ ” KY979958 “ ” “ ” “ ” “ ” ADC-VHS2014-5 “ ” “ ” “ ” “ ” KY979959 “ ” “ ” “ ” “ ” ADC-VHS2015-2 “ ” “ ” “ ” 2015 KY979960 “ ” “ ” “ ” “ ” ADC-VHS2015-5 “ ” “ ” “ ” “ ” KY979961 “ ” “ ” “ ” “ ” ADC-VHS2016-1 “ ” “ ” “ ” 2016 KY979962 “ ” “ ” “ ” “ ” ADC-VHS2016-2 “ ” “ ” “ ” “ ” KY979963 “ ” “ ” “ ” “ ” C03MU* IVb Muskellunge L. St. Clair, USA 2003 GQ385941 Ammayappan et al. 2009 42.3908, -82.9114 a a E06FD “ ” Freshwater L. Erie, USA 2006 MK783014 This study 41.7559, -81.2868 “ ” a1 drum E06WA “ ” Walleye “ ” “ ” MK782987 “ ” “ ” “ ” a2 E06WBa “ ” White bass “ ” “ ” MK782986 “ ” “ ” “ ” N/A E06YPa “ ” Yellow perch “ ” “ ” MK782985 “ ” “ ” “ ” “ ” E06SB “ ” Smallmouth “ ” “ ” MK782984 “ ” “ ” “ ” a3 bass E06YPb “ ” Yellow perch “ ” “ ” MK782983 “ ” 41.4922, -82.6670 “ ” N/A E06YPc “ ” “ ” “ ” “ ” MK782982 “ ” “ ” “ ” “ ” E06WBb “ ” White bass “ ” “ ” MK783013 “ ” “ ” “ ” a4 O06SB “ ” Smallmouth L. Ontario, USA 2006 KY359354 Getchell et al. 2017 44.1167, -76.3333 b b bass C06NP “ ” Northern pike L. St. Clair, USA 2006 MK782990 This study 42.6348, -82.7779 a N/A C06GS “ ” Gizzard shad “ ” “ ” “ ” “ ” “ ” “ ” “ ” C06RB “ ” Rock bass “ ” “ ” “ ” “ ” “ ” “ ” “ ” C06SR “ ” Shorthead “ ” “ ” “ ” “ ” “ ” “ ” “ ” redhorse C06YP “ ” Yellow perch “ ” “ ” “ ” “ ” 42.3430, -82.9020 “ ” “ ” C06FD “ ” Freshwater “ ” “ ” “ ” “ ” 42.6348, -82.7779 “ ” “ ” drum B07BG “ ” Bluegill Budd L., MI, USA 2007 MK783006 “ ” 44.0159, -84.7881 e “ ” B07PS “ ” Pumpkinseed “ ” “ ” MK783008 “ ” “ ” f “ ” E07CC “ ” Common carp L. Erie, USA “ ” MK783005 “ ” 42.4906, -79.3381 c “ ” E07YPa “ ” Yellow perch “ ” “ ” MK782989 “ ” 41.8013, -81.3563 a a7 E07YPb “ ” “ ” “ ” “ ” MK782988 “ ” 41.7559, -81.2777 “ ” N/A M07SB “ ” Smallmouth L. Michigan, USA “ ” MK783009 “ ” 44.8846, -87.3889 h “ ” bass M08RB “ ” Rock bass “ ” 2008 MK783010 “ ” 42.4854, -87.8000 i “ ” E08ES “ ” Emerald shiner L. Erie, USA “ ” MK783012 “ ” 41.7691, -81.2940 a a8 108 E08FDa “ ” Freshwater “ ” “ ” MK782993 “ ” 41.7691, -81.3537 “ ” N/A drum E08FDb “ ” “ ” “ ” “ ” MK782992 “ ” “ ” “ ” “ ” M08AM “ ” Amphipod L. Michigan, USA 2008 MK782990 This study 43.6002, -86.9167 a N/A C08LEa “ ” Leech L. St. Clair, USA 2008 “ ” This study 42.6318, -82.7652 a N/A C08LEb “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” M08YP “ ” Yellow perch L. Michigan, USA “ ” MK783007 “ ” 43.0397, -87.8024 j “ ” C09MU “ ” Muskellunge L. St. Clair, USA 2009 MK782990 “ ” 42.6157, -82.7570 a “ ” M11YP “ ” Yellow perch L. Michigan, USA 2011 MK782991 “ ” 43.0360, -87.8530 “ ” a9 E12FD “ ” Freshwater L. Erie, USA 2012 MK783004 “ ” 41.4530, -82.7260 x N/A drum O13GS “ ” Gizzard shad L. Ontario, USA 2013 KY359355 Getchell et al. 2017 43.2362, -77.5345 bc bc E14GS “ ” “ ” L. Erie, USA 2014 KY359356 “ ” 42.4906, -79.3381 bd bd E15RG “ ” Round goby “ ” 2015 MK783003 This study 41.7652, -81.2816 v N/A E16GSa “ ” Gizzard shad “ ” 2016 MK783011 “ ” “ ” “ ” “ ” E16GSb “ ” “ ” “ ” “ ” MK782997 “ ” “ ” “ ” “ ” E16GSc “ ” “ ” “ ” “ ” MK782996 “ ” “ ” “ ” “ ” E16GSd “ ” “ ” “ ” “ ” MK782994 “ ” “ ” “ ” “ ” E16GSe “ ” “ ” “ ” “ ” MK782995 “ ” “ ” “ ” “ ” M16RGa “ ” Round goby L. Michigan, USA “ ” MK783001 “ ” 42.9969, -87.8825 x “ ” M16RGb “ ” “ ” “ ” “ ” MK783000 “ ” “ ” “ ” “ ” CellC03 “ ” Muskellunge L. St Clair, USA 2003 MK782981 “ ” Cell culture N/A N/A Cell16a “ ” Gizzard shad Lake Erie 2016 MK782998 “ ” “ ” “ ” “ ” Cell16b “ ” Largemouth “ ” “ ” MK783002 “ ” “ ” “ ” “ ” bass Cell16c “ ” “ ” “ ” “ ” MK782999 “ ” “ ” “ ” “ ” 109 Table 3-2. Additional Rhabdoviruses sequences used in phylogenetic trees. Isolate Name Host Location Year Gen Bank Sequence Citation Infectious hematopoietic necrosis virus (IHNV) X89213 Rainbow trout Oregon, USA 1969 X89213 Schutze et al. 1995 WRAC Chinook salmon Idaho, USA 1994 L40883 Morzunov et al. 1995 “ ” “ ” “ ” “ ” NC_001652 “ ” 220-90 Rainbow trout Idaho, USA 1990 GQ413939 Ammayappan et al. 2009 HLJ-09 “ ” China 2009 JX649101 Wang et al. 2016 Ch20101008 Brook trout “ ” 2010 KJ421216 Jia et al. 2014 BjLL Rainbow trout “ ” 2012 MF509592 Sun et al. 2017 Snakehead rhabdovirus (SHRV) NC_000903 Snakehead murrel Thailand 1988 NC_000903 Johnson et al. 1999 AF147498 “ ” “ ” “ ” AF147498 Johnson et al. 2000 Hirame rhabdovirus (HIRRV) CA 9703 Japanese flounder Japan 1984 NC_005093 Kim et al. 2005 “ ” “ ” “ ” “ ” AF104985 “ ” 80113 Stone flounder China 2008 FJ376982 Yingjie et al. 2011 110 Table 3-3. Single nucleotide polymorphisms (SNPs) from each gene’s coding region and the combined non-coding regions (NCDS) in IVb sequence variants for (A) VHSV-IVb, (B) IVa, and (C) Ia, Ib, & III combined. The number of nucleotides (NT) is reported in front of the number of amino acid (AA) changes (the latter are in parentheses). The proportion of nonsynonymous (dN) to synonymous (dS) changes, number of transversions (Tv), and the proportion of transversions to transitions (Ts) are given, along with the evolutionary rate. Totals are in the final row. A. Length # changes % changes Evolutionary Region dN/dS #Tv Tv/Ts NT(AA) NT(AA) NT(AA) Rate N-gene 1215(405) 28(10) 0.023(0.025) 0.357 7 0.250 5.83E-05 P-gene 669(223) 13(3) 0.019(0.014) 0.231 2 0.154 7.18E-05 M-gene 606(202) 17(9) 0.028(0.045) 0.529 2 0.118 7.93E-05 G-gene 1524(508) 37(18) 0.024(0.035) 0.486 6 0.162 8.51E-05 Nv-gene 369(123) 14(7) 0.038(0.057) 0.500 4 0.286 9.76E-05 L-gene 5955(1985) 112(38) 0.019(0.019) 0.339 18 0.161 5.02E-05 NCDS 745(N/A) 32(N/A) 0.043(N/A) N/A 3 0.094 1.40E-04 Total 11083(3446) 253(85) 0.023(0.025) 0.166 42 0.166 6.64E-05 B. Length # changes % changes Evolutionary Region dN/dS #Tv Tv/Ts NT(AA) NT(AA) NT(AA) Rate N-gene 1215(405) 54(17) 0.044(0.042) 0.315 6 0.111 3.14E-04 P-gene 669(223) 45(18) 0.067(0.081) 0.400 10 0.222 1.22E-03 M-gene 606(202) 36(14) 0.059(0.069) 0.389 5 0.139 4.03E-04 G-gene 1524(508) 87(25) 0.057(0.049) 0.287 12 0.138 5.35E-04 Nv-gene 369(123) 31(10) 0.084(0.081) 0.323 3 0.097 1.07E-03 L-gene 5955(1985) 424(119) 0.071(0.059) 0.281 65 0.153 1.02E-03 NCDS 720 (N/A) 110(N/A) 0.153(N/A) N/A 30 0.273 4.09E-03 Total 11058 (3446) 787(203) 0.071(0.059) 0.258 131 0.166 1.06E-03 C. Length # changes % changes Evolutionary Region dN/dS #Tv Tv/Ts NT(AA) NT(AA) NT(AA) Rate N-gene 1215(405) 149(43) 0.123(0.106) 0.289 34 0.228 4.59E-05 P-gene 669(223) 73(30) 0.019(0.135) 0.411 13 0.178 4.07E-05 M-gene 606(202) 65(15) 0.107(0.074) 0.231 10 0.154 5.45E-05 G-gene 1524(508) 198(40) 0.130(0.079) 0.202 41 0.207 4.54E-05 Nv-gene 369(123) 76(27) 0.206(0.220) 0.355 14 0.184 7.78E-05 L-gene 5955(1985) 618(70) 0.104(0.035) 0.113 106 0.172 3.58E-05 NCDS 788(N/A) 113(N/A) 0.143(N/A) N/A 36 0.319 3.60E-05 Total 11126(3446) 1292(225) 0.116(0.065) 0.174 254 0.197 4.09E-05 111 Table 3-4. Positive (diversifying) or negative (purifying) selection pressures on individual codons determined by FUBAR (fast, unconstrained Bayesian approximation) and MEME (mixed effects model of evolution) analyses (Murrell et al. 2012 2013) for (A) IVb, (B) IVa, and (C) I/III. None of the codons found in IVa or I/III match the codons under selection for IVb. Results of seven or more codons under selection not displayed. pp=posterior probability. A. Gene FUBAR FUBAR MEME diversifying purifying diversifying (pp > 0.95) (pp > 0.95) (p<0.05) N 0 1 (313) 0 P “ ” 0 “ ” M “ ” “ ” “ ” G 2 (103, 431) 1 (342) 1 (431) Nv 1 (25) 0 0 6 (8, 119, 333, L 0 460, 1284, “ ” 1758) B. Gene FUBAR FUBAR MEME diversifying purifying diversifying (pp > 0.95) (pp > 0.95) (p<0.05) 8 (none in N 0 0 common) P “ ” 1 (113) “ ” M “ ” 2 (145, 166) “ ” 5 (28, 75, 157, G 1 (12) “ ” 216, 301) Nv 0 1 (24) “ ” 30 (none in L “ ” 3 (147, 593, 1154) common) C. FUBAR FUBAR MEME Gene diversifying purifying diversifying (pp > 0.95) (pp > 0.95) (p<0.05) 20 (none in N 1 (46) 0 common) 8 (none in P 0 0 common) M 0 1 (160) 0 19 (none in G 0 1 (477) common) Nv 0 3 (56, 96, 109) 0 447 (matches L 0 3 (112, 474, 1012) IVb at 1758) 112 Table 3-5. Pairwise genetic divergences between VHSV populations: (A) sampling time periods, Early (2003–6), Middle (2007–11), and Later (2012–16) and (B) Great Lakes regions (Lake Michigan and Budd Lake, Lake St. Clair, Lake Erie and Lake Ontario) using exact tests (GENEPOP; above diagonal) and θST divergences (ARLEQUIN; below diagonal). *=p0.05, **=remained significant (p<α) following sequential Bonferroni correction, NS=p>0.05. A Time Groups B Region Groups Full Early Mid Late Full L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - ** ** Mid 0.001 - ** L. Erie and Ontario 0.111 - ** Late 0.21838 0.19705 - L. Michigan and 0.264 0.348 - Budd N-gene Early Mid Late N-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - NS ** Mid 0.00115 - ** L. Erie and Ontario 0.001 - NS Late 0.23211 0.20742 - L. Michigan and 0.010 0.001 - Budd P-gene Early Mid Late P-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - NS NS Mid 0.0001 - ** L. Erie and Ontario 0.043 - NS Late 0.41168 0.38117 - L. Michigan and 0.064 0.001 - Budd M-gene Early Mid Late M-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - ** NS Mid 0.0001 - ** L. Erie and Ontario 0.243 - ** Late 0.10229 0.08371 - L. Michigan and 0.073 0.248 - Budd G-gene Early Mid Late G-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - ** ** Mid 0.0001 - ** L. Erie and Ontario 0.107 - NS Late 0.2164 0.19368 - L. Michigan and 0.083 0.049 - Budd Nv-gene Early Mid Late Nv-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - NS NS Mid 0.0001 - ** L. Erie and Ontario 0.020 - NS Late 0.17498 0.16843 - L. Michigan and 0.001 0.042 - Budd L-gene Early Mid Late L-gene L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - NS ** Mid 0.0001 - ** L. Erie and Ontario 0.001 - ** Late 0.21162 0.18709 - L. Michigan and 0.212 0.187 - Budd 113 NCDS Early Mid Late NCDS L. St. Clair L. Erie and L. Michigan (N=16) (N=16) (N=11) (N=10) Ontario (N=24) and Budd (N=9) Early - NS ** L. St. Clair - ** NS Mid 0.0001 - ** L. Erie and Ontario 0.092 - ** Late 0.204 0.182 - L. Michigan and 0.000 0.098 - Budd 114 Table 3-A. Single nucleotide polymorphisms (SNPs) and nonsynonymous changes per individual isolates. *=Group, includes C06NP, C06RB, C06SR, C06YP, C06FD, M08AM, C08LEa, C08LEb, and C09MU. N-gene P-gene M-gene G-gene Nv-gene L-gene NCDS Total Average Isolate NT AA NT AA NT AA NT AA NT AA NT AA NT NT AA NT AA dN/dS E06FD 1 0 0 0 0 0 0 0 0 0 3 0 0 4 0 1.6% 0.0% 0.000 E06WA 0 “ ” “ ” “ ” 1 1 1 1 2 1 2 1 1 7 4 2.8% 4.7% 0.571 E06WBb “ ” “ ” “ ” “ ” “ ” “ ” 0 0 0 0 3 “ ” “ ” 5 2 2.0% 2.4% 0.400 E06YPa “ ” “ ” “ ” “ ” “ ” “ ” 2 “ ” “ ” “ ” 2 “ ” “ ” 6 “ ” 2.4% “ ” 0.333 E06SB 1 “ ” “ ” “ ” “ ” “ ” 1 1 “ ” “ ” “ ” “ ” “ ” 6 3 2.4% 3.5% 0.500 E06YPb 0 “ ” “ ” “ ” “ ” “ ” 0 0 “ ” “ ” “ ” “ ” 0 3 2 1.2% 2.4% 0.667 E06YPc “ ” “ ” “ ” “ ” “ ” “ ” 1 1 “ ” “ ” “ ” “ ” 1 5 3 2.0% 3.5% 0.600 E06WBb “ ” “ ” 1 “ ” “ ” “ ” 3 3 “ ” “ ” 3 “ ” “ ” 9 5 3.6% 5.9% 0.556 O06SB 1 “ ” 0 “ ” 2 2 6 “ ” 1 1 11 3 “ ” 22 9 8.7% 10.6% 0.409 C06NP* 0 “ ” “ ” “ ” 0 0 0 0 0 0 1 1 0 1 1 0.4% 1.2% 1.000 C06GS 1 “ ” “ ” “ ” 2 1 1 1 2 “ ” 3 “ ” 5 14 3 5.5% 3.5% 0.214 B07BG 0 “ ” “ ” “ ” 0 0 2 0 “ ” 1 0 0 3 7 1 2.8% 1.2% 0.143 B07PS 1 “ ” “ ” “ ” “ ” “ ” 1 “ ” 0 0 3 2 1 6 2 2.4% 2.4% 0.333 E07CC “ ” “ ” “ ” “ ” 1 1 2 1 “ ” “ ” “ ” “ ” “ ” 8 4 3.2% 4.7% 0.500 E07YPa 0 “ ” “ ” “ ” “ ” “ ” “ ” 2 1 1 2 1 2 “ ” 5 “ ” 5.9% 0.625 E07YPb “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” 0 0 3 2 “ ” “ ” “ ” “ ” “ ” “ ” M07SB 1 “ ” “ ” “ ” “ ” “ ” 3 3 “ ” “ ” “ ” 1 0 “ ” “ ” “ ” “ ” “ ” M08RB “ ” 1 “ ” “ ” 0 0 7 4 “ ” “ ” 6 2 “ ” 14 7 5.5% 8.2% 0.500 E08ES 2 2 2 2 1 1 0 0 “ ” “ ” 8 3 1 14 8 5.5% 9.4% 0.571 E08FDa 0 0 0 0 “ ” “ ” “ ” “ ” “ ” “ ” 3 1 “ ” 5 2 2.0% 2.4% 0.400 E08FDb “ ” “ ” “ ” “ ” “ ” “ ” 1 0 “ ” “ ” 4 “ ” “ ” 7 “ ” 2.8% “ ” 0.286 M08YP 1 1 “ ” “ ” 1 0 2 1 “ ” “ ” 8 4 0 12 6 4.7% 7.1% 0.500 M11YP 0 0 1 1 1 1 0 0 2 0 3 2 2 9 4 3.6% 4.7% 0.444 E12FD 5 2 2 “ ” 2 1 1 1 1 1 18 6 6 35 12 13.8% 14.1% 0.343 O13GS 4 1 “ ” 0 “ ” “ ” 8 3 0 0 9 2 4 29 7 11.5% 8.2% 0.241 E14GS 2 0 1 “ ” 1 “ ” 6 2 “ ” “ ” 15 5 2 27 8 10.7% 9.4% 0.296 E15RG 4 2 2 “ ” “ ” 0 8 4 1 0 9 3 4 29 9 11.5% 10.6% 0.310 E16GSa 3 1 3 “ ” “ ” 1 7 7 2 1 16 5 6 38 15 15.0% 17.6% 0.395 E16GSb “ ” “ ” “ ” “ ” “ ” “ ” 6 6 “ ” “ ” “ ” “ ” “ ” 37 14 14.6% 16.5% 0.378 E16GSc “ ” 2 “ ” “ ” “ ” “ ” 3 3 “ ” 2 4 1 3 19 9 7.5% 10.6% 0.474 E16GSd “ ” 1 “ ” “ ” 0 0 “ ” “ ” “ ” “ ” 16 5 5 32 11 12.6% 12.9% 0.344 E16GSe 4 2 “ ” “ ” “ ” “ ” 5 5 “ ” 1 “ ” “ ” 6 36 13 14.2% 15.3% 0.361 M16RGa 2 1 1 “ ” 4 1 1 1 1 0 4 1 1 14 4 5.5% 4.7% 0.286 115 M16RGb 3 “ ” 4 “ ” 3 “ ” “ ” “ ” “ ” “ ” 6 0 0 18 3 7.1% 3.5% 0.167 Cell16a 4 “ ” 3 “ ” 1 “ ” 6 6 2 1 16 5 6 38 14 15.0% 16.5% 0.368 Cell16b “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” 12 4 “ ” 34 13 13.4% 15.3% 0.382 Cell 16c “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” “ ” 13 “ ” “ ” 35 “ ” 13.8% “ ” “ ” CellC03 0 0 0 “ ” “ ” “ ” 3 1 0 0 0 0 0 4 2 1.6% 2.4% 0.500 AVERAGES 1.553 0.553 1.053 0.105 1.053 0.789 2.816 2.053 0.789 0.395 6.579 2.211 2.289 16.132 6.105 0.064 0.072 0.422 116 Figure 3-1. Map of VHSV (I-IV) full genome isolates included in this study, colored by Genogroup and with shapes denoting substrains. Figure 3-2. Map of VHSV-IVb occurrences in the Great Lakes. Shapes colored for time periods (Early: 2003–06, Middle: 2007–10, Late: 2011–17). Diamonds are isolates with sequenced genomes. Large circles depict VHSV-IVb that lack whole genome data. Figure 3-3. Haplotype network showing genetic relationships among 47 VHSV-IVb full genomes from POPART. Circles sized according to haplotype frequency among isolates. Numbers inside parentheses designate NT differences between each haplotype and the original haplotype, C03MU*. Small, unlabeled black circles = hypothesized haplotype steps. Figure 3-4. Novirhabdovirus phylogenetic trees, based on full genome sequences (see Tables 3.1 and 3.2), with maximum likelihood and Bayesian analyses. Colored squares designate support values, top left half = Bayesian posterior probabilities, bottom right = 500 bootstrap pseudoreplicates. Tree is rooted to the snakehead rhabdovirus (SHRV, GenBank: AF147498). Figure 3-5. VHSV-IVb phylogenetic tree of IVb whole genome haplotypes, with maximum likelihood and Bayesian analyses. Colored squares = support values, top left half = Bayesian posterior probabilities, bottom right half = 1450 bootstrap 117 pseudoreplicates. Hashes represent cropped region for visualization. *=original IVb isolate. The tree is rooted to substrain IVa (GenBank: JF792424). Figure 3-6. Phylogenies of individual novirhabdovirus gene, based on full coding sequences of two representatives sequences of all species (see Tables 3.1 and 3.2), with maximum likelihood and Bayesian analyses for (A) NT and (B) AA. Colored squares designate support values, top left half = Bayesian posterior probabilities, bottom right=2000 bootstrap pseudoreplicates. Tree is rooted to the snakehead rhabdovirus (SHRV L-gene, GenBank: AF147498). Figure 3-7. Host cell viability as measured by Sulforhodamine B (SRB) assay in EPC cells infected with MOIs (MOI=1x10-8–1.0) of four VHSV-IVb isolates, 96 hpi. Average values from a single experiment (conducted in triplicate) are shown, and are representative of at least three independent experimental replicates. Standard error bars, *p<0.05; **p<0.01; ***p<0.001. Figure 3-8. Viral yield assay comparison of infectious viral particles produced (pfu/ml) in wild type (CellC03) and three 2016 VHSV-IVb samples, in BF2 cells 96 hpi, following exposure to media from collection time post infection. Mean values from a single experiment (conducted in triplicate) are shown, and are representative of at least three independent experimental replicates. Standard error bars are shown; *p< 0.05; **p<0.01; ***p<0.001. 118 Figure 3-9. Antiviral assay comparison of host IFN suppression between reference (CellC03) and three more recent VHSV-IVb isolates, in EPC cells 96 hpi following exposure to UV irradiated media collected at the above time points. Values are quantified as the number of antiviral units (uIFN) per mL. Mean values from a single experiment (conducted in triplicate) are shown, and are representative of at least three independent experimental replicates. Standard error bars, *p< 0.05; **p<0.01; ***p< 0.001. Figure 3-10. Gene expression of host immune response and viral RNA produced. qPCR comparisons between reference (CellC03) and three 2016 VHSV-IVb isolates for (A) EPC IFN and (B) virus detected for samples at each of the above time points. Data were normalized to β-actin mRNA levels. Mean Average values from a single experiment (conducted in triplicate) are shown, and are representative of at least three independent experimental replicates. Significance was calculated from Ct values and transformed values (2-ΔΔCT) are shown. Standard error bars, *p< 0.05; **p<0.01; ***p< 0.001. Sup Fig 3-A. Haplotype network showing genetic relationships among 24 VHSV-IVa full genomes from POPART. Circles sized according to haplotype frequency among isolates. Numbers inside parentheses designate NT differences between each node, unlabeled black circles = hypothesized haplotype steps. Year and location of isolation are below isolate names. Sup Fig 3-B. Haplotype network showing genetic relationships among 16 VHSV-Ia, Ib, and III full genomes from POPART. Circles sized according to haplotype frequency 119 among isolates. Numbers inside parentheses designate NT differences between each node. Small, unlabeled black circles = hypothesized haplotype steps. Year and location of isolation are below isolate names. 120 Figure 3-1 Figure 3-2 121 [1]: C06NP, C06GS, C06RB, E16GSa C06SR, C06YP, C06FD, E16GSe (1) Cell16a Cell16c M08AM, C08LEa-b, C09MU (1) (2) (1) (4) (1) C06GS = hypothesized haplotype steps (5) E16GSb Cell16b E16GSc E16GSd M11YP (19) (6) (12) B07BG (8) (6) B07PS (14) M16RGa (5) (2) (1) (7) [1] (1) M08YP (10) (1) (8) C03MU (8) (1) M07SB (8) (3) (2) M16RGb E06FD (2) E08ES (2) (1) E08FDa E06YPa (4) Cell03C E06YPb (3) (1) (1) (6) E06WBb (1) E06YPc (6) (1) (1) E08FDb (2) E06SB (2) (1) (3) (2) E07YPb (1) (4) E07YPc E07CCE06WA (12) (10) E06WBa (12) (25) M08RB O06SB (19) E12FD (18) (18) E15RG O13GS E14GS Figure 3-3 122 1995-WRAC-Idaho-NC001652 1995-WRAC-Idaho-L40883 Oregon-1969-X89213 HLJ_09-China-2009-JX649101 20101008-China-2010-KJ421216 IHNV BjLL-China-2012-MF509592 220_90GQ413939-Idaho-1990 220_90HM461966-Idaho-1990 HRCA9703AF104985 HRNC005093 HR080113FJ376982 HirR Ia-French Strain 07-71 Ia-Fil3 Ia-DK3592B III-2375 Ia & III III-1458 Ia-Hededam Ib-KRRV9601 Ib-SESVA1033 Ib-SESVA10333F Ib-DK4p37 Ib-SE-SVA143D Ib Ib-SESVA145G Ib-SESVA10339C III-CodUlcus III-4-p168 III III-FA281107 II-DK1p49 II IVa-FP-VHS-2010-1 IVa-ADC-VHS-2012-11 IVa-ADC-VHS-2015-5 IVa-ADC-VHS-2016-2 IVa-JF09 IVa-ADC-VHS-2012-7 IVa-ADC-VHS-2012-6 IVa-ADC-VHS-2013-9 IVa-ADC-VHS-2014-4 IVa-ADC-VHS-2014-5 IVa-ADC-VHS-2016-1 IVa IVa-ADC-VHS-2012-9 IVa-ADC-VHS-2013-2 IVa-ADC-VHS-2013-3 IVa-ADC-VHS-2012-10 IVa-ADC-VHS-2015-2 VHSV IVa-ADC-VHS-2013-4 IVa-ADC-VHS-2014-2 IVaKJ2008 IVaFYeosu05 IVa-ADC-VHS-2012-5 IVa-ADC-VHS-2013-1 IVaJF00Ehi1 IVaKRRV9822 O13GS O06SB E14GS E15RG C03MU E06YPc E06SB E06WA E07YPa E07YPa E07CC E06WB E06YPa E06WBa E08FDa E08FDb E12FD E08ES E06YPb M11YP IVb C06GS M08AM (group) E16GSd BAYES PHYML % Support E16GSe E16GSb 99-100 Cell16a Cell16c 90-98 Cell16b E16GSa 80-89 E16GSc 70-79 M16RGa M16RGb 60-69 B07BG B07PS 50-59 E06FD M08YP M07SB M08RB CellC03 SHRV-NC000903 SHRV-AF147498 SHRV 0.2 Figure 3-4 123 Figure 3-5 124 Figure 3-6 125 120 ) l o 100 r CellC03 t n Cell16a o Cell16b c 80 Cell16c % ( y 60 t i l i b a i 40 *** V *** l *** ** l *** * e 20 C 0 0 -1 -2 -3 -4 -5 -6 -7 -8 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 x x x x x x x x x 1 1 1 1 1 1 1 1 1 Multiplicity of Infection (MOI) Figure 3-7 109 CellC03 8 10 Cell16a Cell16b ) 7 ** l 10 ** ** Cell16c m / 6 u 10 f p ( 5 10 d ** l e 4 i 10 * Y ** ** l 3 *** a 10 r i ** V 102 101 100 0 8 4 8 2 6 1 2 4 7 9 Hours Post Infection (hpi) Figure 3-8 126 105 CellC03 104 Cell16a Cell16b L Cell16c 103 *** m / N F I 102 u 101 100 0 8 4 8 2 6 1 2 4 7 9 Hours Post Infection (hpi) Figure 3-9 127 A 4 *** 10 *** *** *** n o *** CellC03 i *** ****** s *** Cell16a s Cell16b e 3 r 10 Cell16c p x E 2 *** N 10 *** F I e v i 1 t 10 a l e R 100 0 18 24 48 72 96 Time Post Infection (h) B 108 CellC03 d a Cell16a * ****** o 6 *** *** L *** Cell16b 10 *** V * Cell16c S ** H 4 V 10 e v i t a l 102 e R 100 0 18 24 48 72 96 Time Post Infection (h) Figure 3-10 128 Sup Fig 3-A 129 Sup Fig 3-B 130 Chapter IV Discussion 4.1 General Conclusions This collaborative PhD dissertation provides a new analysis of the evolutionary patterns and trajectory of the Great Lakes’ Viral Hemorrhagic Septicemia Virus (VHSV) subgenogroup IVb, based on recent and historical sampling data. This research evaluates genetic and genomic changes in evolutionary and biogeographic contexts, with implications for the future. Chapter 2 (in review as Niner & Stepien) investigates the results of regional sampling efforts between 2015-17 and incorporates new sequences with historical data to analyze the population genetics of IVb based on partial G-gene sequences, with additional analyses using concatenated regions of the N, P, M, G, and Nv-genes. Patterns are discerned across time periods (Early, Middle, Later), geographic regions (Upper, Central, and Lower Great Lakes), and among host species. Chapter 3 (in review as Niner, Gorgoglione, Leaman, and Stepien) analyzes phylogenetic relationships of the VHSV genogroups and broader novirhabdoviruses based on whole genomic data, along with examining the possible origin of the Nv-gene. We compare IVb variants using new genomic sequences and evaluate their evolutionary rates and selection pressures, in reference to the Pacific subgenogroup IVa and the combined European genogroups I and III. Chapter 3 additionally examines the in vitro differences between 2016 isolates and the original 2003 isolate in virus production and cell host responses. In brief, this dissertation study shows continued diversification of VHSV-IVb, with marked patterns across time and geographic regions, using gene segments as well as the entire genome. 131 We also find differences in virulence and immune response suppression in the 2016 isolates. 4.1.1 Presence and distribution of VHSV-IVb 2015 to 2017. Updated regional detection of VHSV-IVb was one of our original major goals. Following the decline in VHSV-IVb outbreaks in the Great Lakes, less sampling effort has occurred, leaving a gap we aimed to fill. Following two years of targeting species and areas with a history of VHSV-IVb detections, including additional sampling of regions with recent outbreaks in 2017, less than 1% of our samples (N=21) were virus-positive. All positives were collected in Lakes Erie (2015-16) and Michigan (2016). Evidence of a decline in IVb prevalence also appeared during 2009-10 sampling efforts by Cornwell et al. (2012a, 2015) with a reduction from 16% to 13% of VHSV-positive fishes. In Europe, sampling efforts likewise recovered few positives in more recent years(<1%, Sandlund et al. 2014) to no detections (Vennerström et al. 2018), regardless of recent outbreaks and species tested. It appears that detecting VHSV in the wild is a significant undertaking. This dissertation also investigated the potential of invertebrates harboring VHSV- IVb. Our sampling of dreissenid mussels returned no positives, but due to the low prevalence of virus from our fish samples, this may be inconclusive. Dreissenid mussel populations were sparse in the near-shore communities during the sampling months, suggesting there may not be enough overlap between VHSV-optimal conditions and mussel populations for those species to serve as a reservoir. A similar study found no VHSV in cylindrical papershell mussels (Anodontoides ferussacianus) of Budd Lake despite positive detections in largemouth bass and Hyalellidae amphipods (Throckmorton et al. 2017). 132 Our 21 positives encompassed seven fish species, with most occurring in round gobies and gizzard shad. Detection of VHSV from a single alewife recovered from Lake Michigan in 2016 is notable, representing a new host species and signifying that IVb continues to expand its host range. Only 33% of our positive samples had viral concentrations above the cell culture detection threshold. Without the use of the IVb SYBR green assay (Pierce et al. 2013b), most of our positive samples would have been missed. Although no samples showed symptoms of VHSV, 19% were above the symptom threshold, as quantified for IVb by Pierce et al. (2013a,b). 4.1.2 Recent changes and population genetic trends of IVb. In both Chapters 2 and 3 we find continued evidence of extensive evolutionary changes in VHSV-IVb after nearly two decades in the Great Lakes. From 81 partial G-gene sequences examined in Chapter 2, two major clusters of haplotypes surrounded two older, more numerous haplotypes, “a” and “b”, showing that IVb has continued to radiate in a quasispecies pattern (Belshaw et al., 2008) as described by Stepien et al. (2015). Both “a” and “b” were prevalent during the Early time period (2003-06), but declined during the Middle (2007-10) and Later (2011-2017) time periods, disappearing after 2011. Further differentiation occurred with samples from 2011 onward containing more diverse haplotypes, radiating from the central “a” and “b” haplotypes. Population genetic analyses found that Later time period isolates were more divergent from those collected in the Early and Middle time periods. The Later time period also was significantly divergent for the 41 whole IVb genomes as reported in Chapter 3, as well as for the complete coding region of each individual gene. Among novirhabdoviruses, this temporal trend appears unique: studies of IHNV found little genetic changes reflective of time 133 (Kurath et al., 2003; Black et al. 2016). Regarding other RNA viruses, Gire et al. (2014) found significant divergence of Ebola Virus isolates between collections from 2014 versus the 1970s. Thus, VHSV-IVb’s evolution has undergone significant temporal change. Geographic divergence distinguished VHSV-IVb genotypes among three regions of the Great Lakes: Upper (Lakes Michigan, Huron, and Budd), Central (Lakes St. Clair and Erie), and Lower (Lake Ontario, the New York Finger Lakes, and the St. Lawrence River). Most of haplotype “a” and its descendants were from the Upper and Central Great Lakes, whereas the “b” group was primarily found in the Lower Great Lakes. This division suggests dual origins for IVb, with “a” emerging in the Central region in 2003 and spreading to the Upper region, and “b” in the Lower Great Lakes. Population genetics analyses of the G-gene dataset in Chapter 2 found significant divergence between each region, as did the genomic sequences of Chapter 3, however, significance levels differed between the individual genes. Geographic patterns similarly were observed in IHNV, with separations among the M, L, and U groups in the Northeastern Pacific (Kurath et al, 2003). Additionally, three decades following introduction to China, IHNV has evolved into a distinct clade from the Japanese IHNV isolates (Xu et al. 2018). We may expect VHSV-IVb to continue to diverge across the different regions in the future since it displays genetic divergences corresponding to geographic distances. Despite the large host pool of VHSV-IVb, trends in two species were found in the G-gene data. The round goby appears may be a IVb vector since many isolates and descendants of haplotype “b” were recovered in Lower Great Lakes during the Middle and Later time periods. This invasive species has become one of the most prevalent 134 benthic fishes with high genetic diversity (Brown & Stepien, 2008, 2009; Snyder & Stepien, 2017), potentially influencing the co-evolutionary success of virus and host (Cornwell et al., 2014). Gizzard shad was the second most common host, dominating the early outbreaks from 2006 and housing haplotype “a” (Thompson et al., 2011; Stepien, pers. obs.). Gizzard shad continued to be a prevalent host in Lake Erie and Lake St. Clair, including a 2017 outbreak in the latter area (M. Faisal and G. Whelan, personal communications, 2017). Although it infects a narrower range of hosts, the related IHNV appears to have different major hosts for each of its genetics subgroups (Black et al., 2016). In VHSV-IVb, descendants of haplotypes “a” and “b” may develop host specificities, meriting future investigation. 4.1.3 Genomic trends of IVb, with comparisons to IVa and I/III. In Chapter 3 we reconstructed the IVb phylogeny from 41 full-length sequences, providing more comprehensive analysis of the evolutionary history of the virus in the Great Lakes. Our results supported previously determined genogroups and subgentoypes, however, Ia and III were not monophyletic as previously hypothesized from G-gene sequences (Dale et al. 2009, Ghorani et al. 2016). This is interesting as Ia is found in freshwater species (Einer- Jensen et al. 2004), while III has solely been recovered in marine habitats (Dale et al. 2009). However, both commonly infect rainbow trout (Oncorhynchus mykiss), a popular aquaculture species, which alternates between freshwater and marine environments during its lifecycle. Ia and III should be further examined . The largest numbers of nucleotide changes occurred in the NCDS regions for IVb, IVa, and the combined I/III. We expected this result as NCDS regions are not transcribed and do not play an active role in the replication success of rhabdoviruses (Walker et al. 135 2011, Walker et al. 2018), making substitutions unlikely to be deleterious. Among the genes, Nv contained the most sequence changes among all genogroups. VHSV does not require Nv for replication (Ammayappan et al. 2011), so its mutations also may not be deleterious. Purifying selection was the driving force acting on individual genes in all genogroups; however, no codons were found to be under common selection across all groups. Strong purifying selection pressures have been reported in VHSV (He et al. 2014; Abbadi et al. 2016) and RNA viruses in general, reflecting lack of proof-reading (Kuzmin et al. 2009; Hughes & Hughes 2007). Our examination of evolutionary rates found that IVa had the fastest evolving genome at 2.01x10-3. IVb and I/III were slower, but similar to one another at 6.6x10-5 and 4.09x10-5, respectively. The rates of IVb and I/III were slower than previous studies determined based on shorter sequences (IVb: Stepien et al. 2015; Pierce & Stepien 2012, I/III: He et al. 2014; Einer-Jensen et al. 2014; Abbadi et al. 2016). We expect this result is from the inclusion of longer sequence regions that contain more conserved areas. IVa, in constrast, was faster than prior estimates (He et al. 2014). We suggest that the faster rate is an artifact of sampling; the majority of the sequences used came from Korea between 2012-16 (Hwang et al. 2018) and were vastly different from samples collected in Japan. Additionally that group lacked representation from North America, where IVa was first recorded (Brunsen et al. 1989; Hopper 1989; Meyers et al. 1992). Sample bias aside, all evolutionary estimates are within the established range of RNA viruses (Duffy et al. 2008). 4.1.4 Genomic characterization of Novirhabdoviruses and the origin of the Nv-gene. Broader phylogenetic trends across the genomes found that VHSV and SHRV 136 form a sister clade to that of IHNV and HIRRV, matching previous findings based on partial gene sequences (Kurath 2012; Pierce and Stepien 2012; Stepien et al. 2015). Our inclusion of whole genomes provide more support for the relationship of SHRV/VHSV than Kurath (2012) found using partial N-gene sequences. No consensus was found on the origin of Nv from our multi-gene novirhabdovirus phylogeny from either the NT or AA approaches. Instead the Nv-genes of the four species were spread out between the other genes, with just HIRR-Nv and IHNV-Nv forming a single clade in the NT and the addition of VHSV-Nv. This may be unsurprising as the Nv- gene lacks sequence conservation among novirhabdovirus species (Kurath 2012) and has different degrees of function (Johnson et al. 2000, Alonso et al. 2004, Ammayappan et al. 2011). We found no evidence of a gene-duplication event creating the novel Nv-gene. 4.1.5 In vitro comparison of 2016 isolates versus the original 2003 isolate. Our Chapter 3 examination of three 2016 isolates in cell culture revealed marked differences between the 2016 isolates and the 2003 reference (CellC03). During the initial 48 hours of infection, those from 2016 demonstrated increased interferon (IFN) RNA production, increased virulence, and levels of VHSV RNA as compared to the 2003 reference. Following 48 hours post infection, the 2016 isolates produced fewer viral particles and RNA than CellC03, while maintaining elevated IFN production. It is likely the 13-14 AA differences between the 2016 isolates and the reference altered the interactions between the virus and host immune response in a way that causes less damage to the host while producing higher initial amounts of VHSV. Previous studies found that amino acid changes can cause notable changes in VHSV behavior; a single amino acid change in Nv reduced the efficacy of infection (Chincilla and Gomez-Casado 2017), and four mutations 137 in M led to reduced suppression of host immune defense transcription (Ke et al. 2017). This may be the mechanism behind the reduced virulence observed in the last decade. 4.2 Future Research and Recommendations This dissertation examines the present status and evolutionary trends of a recent pathogen nearing its second decade in a novel environment. Data generated here enhance our understanding of the evolutionary progression of wildlife pathogens outside of the controlled laboratory settings and how they evolve and adapt over time. Multiple areas would benefit from additional studies, as follows: 4.2.1 Continued surveillance with narrowed focus. Surveillance for VHSV-IVb should be continued across the Great Lakes and neighboring water bodies. Our research shows that the virus remains in multiple lakes and has continued to spread into additional species. Where possible, agencies and labs testing for the virus should adopt molecular- based detection methods, such as the SYBR Green test used here (Pierce et al. 2013b) instead of, or to supplement cell culture; most of the positive samples from this study would have been missed without molecular diagnosis. Low-level infections of VHSV- IVb may indicate potential for future outbreaks and risk of potential transfer. Focus should be placed on two species in particular: gizzard shad and round goby. Both species appear as common hosts for IVb. Additionally, both have been reported in recent outbreaks (R. Getchell, personal communication 2017; G. Whelan, personal communication 2017). Testing of tissues from round goby could serve as a diagnostic of IVb presence since virus-positive individuals have been widespread in the last five years. Gizzard shad have a more restricted range, but could also serve in this capacity where found in large numbers, as in Lakes St. Clair and Erie. 138 Outreach and distribution of educational materials should be renewed or continued in local communities so that the fishing and boating communities are aware of their potential role in the spread of this virus. 4.2.2 Resolving the relationship between genogroups Ia and III. In our genomic examination of VHSV genogroups, we found lack of distinction between subgenogroup Ia and genogroup III. Since we only had four sequences of Ia and five of III, increased sample size would be beneficial. Two III samples from France form a clade with Ia, whereas the samples from Denmark and Norway form a distinct clade separate from genogroup I samples. Thus, their separation appears incorrect. Further population genetic analyses of VHSV across its various genogroups and geographic realms are warranted to better understand it evolutionary trajectory in relation to geographic region, isolation year, and host species. 4.2.3 Genomics: Sequencing of Additional Samples. Chapter 3 lacked breadth of samples from different VHSV genogroups. Future efforts should continue to add genomes for II and IVc, since our study had limited sample availability. Those two genogroups are particularly understudied. Furthermore, complete sequences of IVa from the North American coast were entirely lacking from our analyses. VHSV-IVa is a continuing issue for Asian aquaculture, but its origins lie on the other side of the Pacific; the examination of IVa sequences from North America could evaluate different patterns in transmission and evolution. Moreover, IVa continues to infect and kill large numbers of herring in marine waters off the eastern Pacific. Our faster evolutionary rate for IVa raises the question of 139 how this virus is being spread, namely from aquaculture or from reoccurring transmission from wild fishes near coastal fish pens. 4.2.4 Pathogenicity of new isolates and changes in host immune suppression. We found significant differences between the 2016 isolates and the original reference from 2003. Few studies have examined differences between naturally occurring isolates, instead opting for laboratory-created mutants to study specific pathways. Although this is useful information, more effort should be made towards understanding recent VHSV- IVb isolates, as these are representative of current risks and threats to the Great Lakes and surrounding ecosystems. Fish challenge studies involving susceptible fishes including round goby and gizzard shad, should be considered. We also recommend future cell culture work targeting individual mutations, such as those seen in the 2016 genomes and other isolates as they become available. 140 References Abbadi, M., Fusaro, A., Ceolin, C., Casarotto, C., Quartesan, R., Dalla Pozza, M., ... & Panzarin, V. (2016). Molecular evolution and phylogeography of co-circulating IHNV and VHSV in Italy. Frontiers in Microbiology, 7, 1306. https://doi.org/10.3389/fmicb.2016.01306 Afgan, E., Baker, D., Batut, B., Van Den Beek, M., Bouvier, D., Čech, M., ... & Guerler, A. (2018). The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Research, 46(W1), W537- W544. https://doi.org/10.1093/nar/gky379 Alabi, O. J., Martin, R. R., & Naidu, R. A. (2010). Sequence diversity, population genetics and potential recombination events in Grapevine Rupestris Stem Pitting- associated Virus in Pacific North-West vineyards. Journal of General Virology, 91(1), 265–276. https://doi.org/10.1099/vir.0.014423-0 Alonso, M., Kim, C. H., Johnson, M. C., Pressley, M., & Leong, J. A. (2004). The NV gene of snakehead rhabdovirus (SHRV) is not required for pathogenesis, and a heterologous glycoprotein can be incorporated into the SHRV envelope. Journal of Virology, 78(11), 5875-5882. http://doi.org/10.1128/JVI.78.11.5875-5882.2004 Alves, J. M., Carneiro, M., Cheng, J. Y., de Matos, A. L., Rahman, M. M., Loog, L., ... & Strive, T. (2019). Parallel adaptation of rabbit populations to myxoma virus. Science, 363(6433), 1319–1326. https://doi.org/10.1126/science.aau7285 141 Ammayappan, A., & Vakharia, V. N. (2009). Molecular characterization of the Great Lakes Viral Hemorrhagic Septicemia virus (VHSV) isolate from USA. Virology Journal, 6(1), 171. https://doi.org/10.1186/1743-422X-6-171 Ammayappan, A., Kurath, G., Thompson, T. M., & Vakharia, V. N. (2011). A reverse genetics system for the Great Lakes strain of Viral Hemorrhagic Septicemia virus: the Nv gene is required for pathogenicity. Marine Biotechnology, 13(4), 672–683. https://doi.org/10.1007/s10126-010-9329-4 Ammayappan, A., & Vakharia, V. N. (2011). Nonvirion protein of novirhabdovirus suppresses apoptosis at the early stage of virus infection. Journal of Virology, JVI-00597. https://doi.org/10.1128/JVI.00597-11 Andino, R., & Domingo, E. (2015). Viral quasispecies. Virology, 479, 46–51. https://doi.org/10.1016/j.virol.2015.03.022 Bahl, J., Vijaykrishna, D., Holmes, E. C., Smith, G. J., & Guan, Y. (2009). Gene flow and competitive exclusion of Avian Influenza A virus in natural reservoir hosts. Virology, 390(2), 289–297. https://doi.org/10.1016/j.virol.2009.05.002 Basurco, B., Vende, P., Monnier, A. F., Winton, J. R., De Kinkelin, P., & Benmansour, A. (1995). Genetic diversity and phylogenetic classification of Viral Hemorrhagic Septicemia virus (VHSV). Veterinary Research, 26(5–6), 460–463. PMID: 8581023 Baillon, L., Mérour, E., Cabon, J., Louboutin, L., Quenault, H., Touzain, F., ... & Brémont, M. (2017). A single amino acid change in the non-structural NV protein impacts the virulence phenotype of Viral hemorrhagic septicemia virus in trout. 142 Journal of General Virology, 98(6), 1181-1184. http://doi.org/10.1099/jgv.0.000830 Bayliss, S. C., Verner-Jeffreys, D. W., Bartie, K. L., Aanensen, D. M., Sheppard, S. K., Adams, A., & Feil, E. J. (2017). The promise of whole genome pathogen sequencing for the molecular epidemiology of emerging aquaculture pathogens. Frontiers in Microbiology, 8, 121. https://doi.org/10.3389/fmicb.2017.00121 Béarzotti, M., Monnier, A. F., Vende, P., Grosclaude, J., de Kinkelin, P. & Benmansour, A. (1995). The glycoprotein of viral hemorrhagic septicemia virus (VHSV): antigenicity and role in virulence. Veterinary Research Communications, 26, 413– 422. Beerenwinkel, N., & Zagordi, O. (2011). Ultra-deep sequencing for the analysis of viral populations. Current Opinion in Virology, 1(5), 413–418. http://doi.org/10.1016/j.coviro.2011.07.008 Belshaw, R., Gardner, A., Rambaut, A., & Pybus, O. G. (2008). Pacing a small cage: mutation and RNA viruses. Trends in Ecology & Evolution, 23(4), 188–193. https://doi.org/10.1016/j.tree.2007.11.010 Benmansour, A., Basurco, B., Monnier, A. F., Vende, P., Winton, J. R., & de Kinkelin, P. (1997). Sequence variation of the glycoprotein gene identifies three distinct lineages within field isolates of Viral Haemorrhagic Septicaemia virus, a fish rhabdovirus. Journal of General Virology, 78(11), 2837–2846. https://doi.org/10.1099/0022-1317-78-11-2837 Bernard, H. U. (1994). Coevolution of papiliomaviruses with human populations. Trends in Microbiology, 2(4), 140-143. https://doi.org/10.1016/0966-842X(94)90602-5 143 Bertolotti, L., Kitron, U., & Goldberg, T. L. (2007). Diversity and evolution of West Nile virus in Illinois and the United States, 2002–2005. Virology, 360(1), 143–149. https://doi.org/10.1016/j.virol.2006.10.030 Biacchesi, S., Béarzotti, M., Bouguyon, E., & Brémont, M. (2002). Heterologous exchanges of the glycoprotein and the matrix protein in a Novirhabdovirus. Journal of Virology, 76(6), 2881–2889. https://doi.org/10.1128/JVI.76.6.2881- 2889.2002 Biacchesi, S. (2011). The reverse genetics applied to fish RNA viruses. Veterinary Research, 42(1), 12. https://doi.org/10.1186/1297-9716-42-12 Biacchesi, S., Mérour, E., Chevret, D., Lamoureux, A., Bernard, J., & Brémont, M. (2017). NV proteins of fish novirhabdovirus recruit cellular PPM1Bb protein phosphatase and antagonize RIG-I-mediated IFN induction. Scientific Reports, 7, 44025. https://doi.org/10.1038/srep44025 Black, A., Breyta, R., Bedford, T., & Kurath, G. (2016). Geography and host species shape the evolutionary dynamics of U genogroup infectious hematopoietic necrosis virus. Virus Evolution, 2(2), vew034. https://doi.org/10.1093/ve/vew034 Blokhuis, H. J., Keeling, L. J., Gavinelli, A., & Serratosa, J. (2008). Animal welfare's impact on the food chain. Trends in Food Science & Technology, 19, S79-S87. https://doi.org/10.1016/j.tifs.2008.09.007 Bourhy, H., Cowley, J. A., Larrous, F., Holmes, E. C., & Walker, P. J. (2005). Phylogenetic relationships among rhabdoviruses inferred using the L polymerase gene. Journal of General Virology, 86(10), 2849-2858. https://doi.org/10.1099/vir.0.81128-0 144 Brown, J. E., & Stepien, C. A. (2008). Ancient divisions, recent expansions: phylogeography and population genetics of the round goby Apollonia melanostoma. Molecular Ecology, 17(11), 2598–2615. https://doi.org/10.1111/j.1365-294X.2008.03777.x Brown, J. E., & Stepien, C. A. (2009). Invasion genetics of the Eurasian round goby in North America: tracing sources and spread patterns. Molecular Ecology, 18(1), 64–79. https://doi.org/10.1111/j.1365-294X.2008.04014.x Brunson, R. (1989). VHS virus isolated at Makah National Fish Hatchery. Fish Health Section of the American Fish Society Newsletter, 17, 3–4. Burkett E. M. & Winkler R. L. 2018. Recreational Fishing in Michigan: Using an Age- Period-Cohort approach to understand fishing participation. Michigan Technological University. Houghton, MI, USA. Available at https://www.mtu.edu/greatlakes/fishery/state-level/pdf/michigan-angler- demographics-2000-2014.pdf. Accessed 15 March, 2019. Bush, R. M., Fitch, W. M., Bender, C. A., & Cox, N. J. (1999). Positive selection on the H3 hemagglutinin gene of human influenza virus A. Molecular Biology and Evolution, 16(11), 1457-1465. https://doi.org/10.1093/oxfordjournals.molbev.a026057 Cano, I., Collet, B., Pereira, C., Paley, R., van Aerle, R., Stone, D., & Taylor, N. G. (2016). In vivo virulence of viral haemorrhagic septicaemia virus (VHSV) in rainbow trout Oncorhynchus mykiss correlates inversely with in vitro Mx gene expression. Veterinary Microbiology, 187, 31-40. https://doi.org/10.1016/j.vetmic.2016.02.012 145 Cargnelutti, J. F., de Quadros, J. M., Martins, M., Batista, H. B., Weiblen, R., & Flores, E. F. (2017). Glycoprotein-G-gene-based molecular and phylogenetic analysis of rabies viruses associated with a large outbreak of bovine rabies in southern Brazil. Archives of Virology, 162(12), 3697–3704. https://doi.org/10.1007/s00705-017- 3533-8 Cherian, S., Singh, R., Singh, K. P., Reddy, G. M., Kumar, G. R., Sumithra, T. G., & Singh, R. P. (2015). Phylogenetic analysis of Indian rabies virus isolates targeting the complete glycoprotein gene. Infection, Genetics and Evolution, 36, 333–338. https://doi.org/10.1016/j.meegid.2015.09.024 Chinchilla, B., & Gomez-Casado, E. (2017). Identification of the functional regions of the viral haemorrhagic septicaemia virus (VHSV) NV protein: variants that improve function. Fish & shellfish immunology, 70, 343-350. https://dx.doi.org/10.1016/j.fsi.2017.09.021 Choi, M. K., Moon, C. H., Ko, M. S., Lee, U. H., Cho, W. J., Cha, S. J., ... & Harmache, A. (2011). A nuclear localization of the infectious haematopoietic necrosis virus NV protein is necessary for optimal viral growth. PLoS One, 6(7), e22362. https://doi.org/10.1371/journal.pone.0022362 Clement, M., Posada, D. C. K. A., & Crandall, K. A. (2000). TCS: a computer program to estimate gene genealogies. Molecular Ecology, 9(10), 1657–1659. https://doi.org/10.1046/j.1365-294x.2000.01020.x Coffee, L. L., Getchell, R. G., Groocock, G. H., Cornwell, E. R., Wooster, G. A., Lumsden, J. S., & Bowser, P. R. (2017). Pathogenesis of experimental viral hemorrhagic septicemia virus IVb infection in adult sea lamprey (Petromyzon 146 marinus). Journal of Great Lakes Research, 43(3), 119-126. https://doi.org/10.1016/j.jglr.2017.03.013 Collet, B. (2014). Innate immune responses of salmonid fish to viral infections. Developmental & Comparative Immunology, 43(2), 160-173. https://doi.org/10.1016/j.dci.2013.08.017 Collet, B. and Secombes C. J. (2001). The rainbow trout (Oncorhynchus mykiss) Mx1 promoter: structural and functional characterization. European Journal Biochemistry. 268(6):1577–1584. http://doi.org/ 10.1046/j.1432- 1033.2001.02021.x Cornwell, E. R., Eckerlin, G. E., Thompson, T. M., Batts, W. N., Getchell, R. G., Groocock, G. H., ... & Bain, M. B. (2012a). Predictive factors and viral genetic diversity for Viral Hemorrhagic Septicemia virus infection in Lake Ontario and the St. Lawrence River. Journal of Great Lakes Research, 38(2), 278–288. https://doi.org/10.3354/dao076187 Cornwell, E. R., Getchell, R. G., Groocock, G. H., Walsh, M. G., & Bowser, P. R. (2012b). Low prevalence of VHSV detected in round goby collected in offshore regions of Lake Ontario. Journal of Great Lakes Research, 38(3), 575–579. https://doi.org/10.1016/j.jglr.2012.06.008 Cornwell, E. R., Primus, A., Wong, P. T., Anderson, G. B., Thompson, T. M., Kurath, G., ... & Getchell, R. G. (2014). Round gobies are an important part of VHSV genotype IVb ecology in the St. Lawrence River and eastern Lake Ontario. Journal of Great Lakes Research, 40(4), 1002–1009. https://doi.org/10.1016/j.jglr.2014.09.006 147 Cornwell, E. R., Anderson, G. B., Coleman, D., Getchell, R. G., Groocock, G. H., Warg, J. V., ... & Bowser, P. R. (2015). Applying multi–scale occupancy models to infer host and site occupancy of an emerging viral fish pathogen in the Great Lakes. Journal of Great Lakes Research, 41(2), 520–529. https://doi.org/10.1016/j.jglr.2015.01.002 Cortey, M., Bertran, K., Toskano, J., Majó, N., & Dolz, R. (2012). Phylogeographic distribution of very virulent Infectious Bursal Disease Virus isolates in the Iberian Peninsula. Avian Pathology, 41(3), 277–284. https://doi.org/10.1080/03079457.2012.682562 Dale, O. B., Ørpetveit, I., Lyngstad, T. M., Kahns, S., Skall, H. F., Olesen, N. J., & Dannevig, B. H. (2009). Outbreak of viral haemorrhagic septicaemia (VHS) in seawater-farmed rainbow trout in Norway caused by VHS virus Genotype III. Diseases of Aquatic Organisms, 85(2), 93-103. https://doi.org/10.3354/dao02065 Dietzgen R. G. (2012) Morphology, genome organization, transcription and replication of rhabdoviruses. In: Rhabdoviruses: Molecular Taxonomy, Evolution, Genomics, Ecology, Host-vector Interactions, Cytopathology and Control, (eds Dietzgen RG, Kuzmin IV), pp. 5–12. Caister Academic Press, Norfolk, UK. Dietzgen R. G., & Kuzmin I. V. (2012) Taxonomy of Rhabodviruses. In: Rhabdoviruses: Molecular Taxonomy, Evolution, Genomics, Ecology, Host-vector Interactions, Cytopathology and Control, (eds Dietzgen RG, Kuzmin IV), pp. 5–12. Caister Academic Press, Norfolk, UK. 148 Duffy, S., Shackelton, L. A., & Holmes, E. C. (2008). Rates of evolutionary change in viruses: patterns and determinants. Nature Reviews Genetics, 9(4), 267. https://doi.org/10.1038/nrg2323 East, M. L., Hofer, H., Cox, J. H., Wulle, U., Wiik, H., & Pitra, C. (2001). Regular exposure to rabies virus and lack of symptomatic disease in Serengeti spotted hyenas. Proceedings of the National Academy of Sciences, 98(26), 15026-15031. https://doi.org/10.1073/pnas.261411898 Eckerlin, G. E., Farrell, J. M., Casey, R. N., Hope, K. M., Groocock, G. H., Bowser, P. R., & Casey, J. (2011). Temporal variation in prevalence of viral hemorrhagic septicemia virus type IVb among upper St. Lawrence River smallmouth bass. Transactions of the American Fisheries Society, 140(3), 529-536. https://doi.org/10.1080/00028487.2011.581975 Einer-Jensen, K., Ahrens, P., Forsberg, R., & Lorenzen, N. (2004). Evolution of the fish rhabdovirus viral haemorrhagic septicaemia virus. Journal of General Virology, 85(5), 1167-1179. http://doi.org/10.1099/vir.0.79820-0 Einer-Jensen, K., Ahrens, P., & Lorenzen, N. (2005). Parallel phylogenetic analyses using the N, G or Nv gene from a fixed group of VHSV isolates reveal the same overall genetic typing. Diseases of Aquatic Organisms, 67(1–2), 39–45. https://doi.org/10.3354/dao067039 Einer-Jensen, K., Harmache, A., Biacchesi, S., Bremont, M., Stegmann, A., & Lorenzen, N. (2014). High virulence differences among phylogenetically distinct isolates of the fish rhabdovirus viral hemorrhagic septicaemia virus are not explained by 149 variability of the surface glycoprotein G or the non-virion protein Nv. Journal of General Virology, 95(2), 307–316. http://doi.org/10.1099/vir.0.057448-0 Elsayed, E., Faisal, M., Thomas, M., Whelan, G., Batts, W., & Winton, J. (2006). Isolation of viral haemorrhagic septicaemia virus from muskellunge, Esox masquinongy (Mitchill), in Lake St Clair, Michigan, USA reveals a new sublineage of the North American genotype. Journal of Fish Diseases, 29(10), 611–619. https://doi.org/10.1111/j.1365-2761.2006.00755.x Elsworth, P., Cooke, B. D., Kovaliski, J., Sinclair, R., Holmes, E. C., & Strive, T. (2014). Increased virulence of Rabbit Haemorrhagic Disease virus associated with genetic resistance in wild Australian rabbits (Oryctolagus cuniculus). Virology, 464, 415– 423. https://doi.org/10.1016/j.virol.2014.06.037 Escobar, L. E., Kurath, G., Escobar Dodero, J., Craft, M. E., & Phelps, N. B. (2017). Potential distribution of the Viral Haemorrhagic Septicaemia virus in the Great Lakes region. Journal of Fish Diseases, 40(1), 11–28. https://doi.org/10.1111/jfd.12490 Escobar, L. E., Escobar-Dodero, J., & Phelps, N. B. (2018). Infectious disease in fish: global risk of Viral Hemorrhagic Septicemia virus. Reviews in Fish Biology and Fisheries, 28(3), 637–655. https://doi.org/10.1007/s11160-018-9524-3 Excoffier, L., Smouse, P. E., & Quattro, J. M. (1992). Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics, 131(2), 479–491. https://www.genetics.org/content/131/2/479.article-info 150 Excoffier, L., & Lischer, H. E. (2010). ARLEQUIN suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10(3), 564–567. https://doi.org/10.1111/j.1755- 0998.2010.02847.x Faisal, M., & Schulz, C. A. (2009). Detection of Viral Hemorrhagic Septicemia virus (VHSV) from the leech Myzobdella lugubris Leidy, 1851. Parasites & Vectors, 2(1), 45. https://doi.org/10.1186/1756-3305-2-45 Faisal, M., & Winters, A. D. (2011). Detection of Viral Hemorrhagic Septicemia virus (VHSV) from Diporeia spp. (Pontoporeiidae, Amphipoda) in the Laurentian Great Lakes, USA. Parasites & Vectors, 4(1), 2. https://doi.org/10.1186/1756- 3305-4-2 Faisal, M., Shavalier, M., Kim, R. K., Millard, E. V., Gunn, M. R., Winters, A. D., ... & Whelan, G. E. (2012). Spread of the emerging Viral Hemorrhagic Septicemia virus strain, genotype IVb, in Michigan, USA. Viruses, 4(5), 734–760. Viruses, 4, 734–760. https://doi.org/10.3390/v4050734 Faye, O., Freire, C. C., Iamarino, A., Faye, O., de Oliveira, J. V. C., Diallo, M., & Zanotto, P. M. (2014). Molecular evolution of Zika virus during its emergence in the 20th century. PLoS Neglected Tropical Diseases, 8(1), e2636. https://doi.org/10.1371/journal.pntd.0002636 Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39(4), 783–791. https://doi.org/10.1111/j.1558- 5646.1985.tb00420.x 151 Felsenstein, J. (2007) PHYLIP (Phylogeny Inference Package) version 3.697. Distributed by the author. Seattle: Department of Genome Sciences, University of Washington, Seattle, WA, USA. https://evolution.genetics.washington.edu/phylip.html. Fetzer, W. W., Brooking, T. E., Jackson, J. R., & Rudstam, L. G. (2011). Overwinter mortality of gizzard shad: evaluation of starvation and cold temperature stress. Transactions of the American Fisheries Society, 140(6), 1460–1471. https://doi.org/10.1080/00028487.2011.630281 Gagné, N., MacKinnon, A. M., Boston, L., Souter, B., Cook Versloot, M., Griffiths, S., & Olivier, G. (2007). Isolation of Viral Haemorrhagic Septicaemia virus from mummichog, stickleback, striped bass and brown trout in eastern Canada. Journal of Fish Diseases, 30(4), 213–223. https://doi.org/10.1111/j.1365- 2761.2007.00802.x Garver, K. A., Traxler, G. S., Hawley, L. M., Richard, J., Ross, J. P., & Lovy, J. (2013). Molecular epidemiology of Viral Haemorrhagic Septicaemia virus (VHSV) in British Columbia, Canada, reveals transmission from wild to farmed fish. Diseases of Aquatic Organisms, 104(2), 93–104. https://doi.org/10.3354/dao02588 Gorgoglione, B. 2014. Heterogeneous infections in fish: transcriptomic studies on the trout immune response to single and co-infections (Doctoral dissertation, University of Aberdeen). Avaliable at: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.606427 152 Getchell, R. G., Cornwell, E. R., Bogdanowicz, S., Andrés, J., Batts, W. N., Kurath, G., ... & Bowser, P. R. (2017). Complete sequences of four Viral Hemorrhagic Septicemia virus IVb isolates and their virulence in northern pike fry. Diseases of Aquatic Organisms, 126(3), 211–227. https://doi.org/10.3354/dao03171 Ghorani, M., Adel, M., Dadar, M., Langeroudi, A. G., Kamyabi, R., Vakharia, V. N., & Einer-Jensen, K. (2016). Phylogenetic analysis of the glycoprotein gene of Viral Hemorrhagic Septicemia virus from Iranian trout farms points towards a common European origin. Veterinary Microbiology, 186, 97–101. https://doi.org/10.1016/j.vetmic.2016.02.019 Gire, S. K., Goba, A., Andersen, K. G., Sealfon, R. S., Park, D. J., Kanneh, L., ... & Wohl, S. (2014). Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak. Science, 345(6202), 1369-1372. https://doi.org/10.1126/science.1259657 Groocock, G. H., Getchell, R. G., Wooster, G. A., Britt, K. L., Batts, W. N., Winton, J. R., ... & Bowser, P. R. (2007). Detection of Viral Hemorrhagic Septicemia in round gobies in New York State (USA) waters of Lake Ontario and the St. Lawrence River. Diseases of Aquatic Organisms, 76(3), 187–192. https://doi.org/10.3354/dao076187 Guðmundsdóttir, S., Vendramin, N., Cuenca, A., Sigurðardóttir, H., Kristmundsson, A., Iburg, T. M., & Olesen, N. J. (2019). Outbreak of Viral Haemorrhagic Septicaemia (VHS) in lumpfish (Cyclopterus lumpus) in Iceland caused by VHS virus genotype IV. Journal of Fish Diseases, 42(1), 47–62. http://doi.org/10.1111/jfd.12910 153 Hamblin, J. (2015) 21 days. The Atlantic. Retrieved from https://www.theatlantic.com/health/archive/2014/10/21-days/381901/. Hampson, K., Coudeville, L., Lembo, T., Sambo, M., Kieffer, A., Attlan, M., ... & Costa, P. (2015). Estimating the global burden of endemic canine rabies. PLoS neglected tropical diseases, 9(4), e0003709. https://doi.org/10.1371/journal.pntd.0003709 He, M., Ding, N. Z., He, C. Q., Yan, X. C., & Teng, C. B. (2013). Dating the divergence of the Infectious Hematopoietic Necrosis Virus. Infection, Genetics and Evolution, 18, 145–150. https://doi.org/10.1016/j.meegid.2013.05.014 He, M., Yan, X. C., Liang, Y., Sun, X. W., & Teng, C. B. (2014). Evolution of the Viral Hemorrhagic Septicemia virus: divergence, selection and origin. Molecular Phylogenetics and Evolution, 77, 34–40. https://doi.org/10.1016/j.ympev.2014.04.002 Hedrick, R. P., Batts, W. N., Yun, S., Traxler, G. S., Kaufman, J., & Winton, J. R. (2003). Host and geographic range extensions of the North American strain of Viral Hemorrhagic Septicemia virus. Diseases of Aquatic Organisms, 55(3), 211– 220. https://doi.org/10.3354/dao055211 Hedrick, P. W. (2011) Genetics of Populations. Fourth Edition. Sudbury, MA: Jones & Bartlett Learning. Sudbury. Hoorfar, J., Malorny, B., Abdulmawjood, A., Cook, N., Wagner, M., & Fach, P. (2004). Practical considerations in design of internal amplification controls for diagnostic PCR assays. Journal of Clinical Microbiology, 42(5), 1863–1868. https://doi.org/10.1128/JCM.42.5.1863-1868.2004 154 Hoenen, T., Safronetz, D., Groseth, A., Wollenberg, K. R., Koita, O. A., Diarra, B., ... & Sarro, Y. S. (2015). Mutation rate and genotype variation of Ebola virus from Mali case sequences. Science, 348(6230), 117–119. https://doi.org/10.1126/science.aaa5646 Holmes, E. (2009) The evolution and emergence of RNA viruses. Oxford University Press, Oxford, NY, USA. Hopper, K. (1989). The isolation of VHSV from chinook salmon at Glenwood Springs, Orcas Island, Washington. Fish Health Section of the American Fisheries Society Newsletter. 17, 1. Hughes, A. L., & Hughes, M. A. K. (2007). More effective purifying selection on RNA viruses than in DNA viruses. Gene, 404(1-2), 117–125. https://doi.org/10.1016/j.gene.2007.09.013 Hwang, J. Y., Lee, S., Priyathilaka, T. T., Yang, H., Kwon, H., Kwon, M. G., ... & Lee, J. (2018). Phylogenetic analysis and duplex RT-PCR detection of viral hemorrhagic septicemia virus in olive flounder (Paralichthys olivaceus) from Korea. Aquaculture, 484, 242–249. https://doi.org/10.1016/j.aquaculture.2017.11.038 ICTV Virus Taxonomy, 2018 release, available online at https://talk.ictvonline.org/taxonomy. Imanse, S. M., Cornwell, E. R., Getchell, R. G., Kurath, G., & Bowser, P. R. (2014). In vivo and in vitro phenotypic differences between Great Lakes VHSV genotype IVb isolates with sequence types vcG001 and vcG002. Journal of Great Lakes Research, 40(4), 879–885. https://doi.org/10.1016/j.jglr.2014.08.004 155 Ito, T., Kurita, J., Mori, K. I., & Olesen, N. J. (2016). Virulence of viral haemorrhagic septicaemia virus (VHSV) genotype III in rainbow trout. Veterinary Research, 47(1), 4. https://doi.org/10.1186/s13567-015-0303-z Janeway Jr, C.A. and Medzhitov, R. (2002). Innate immune recognition. Annual review of immunology. 20(1):197–216. https://doi.org/10.1146/annurev.immunol.20.083001.084359 Johnson, M. C., Simon, B. E., Kim, C. H., & Leong, J. A. C. (2000). Production of recombinant snakehead rhabdovirus: the NV protein is not required for viral replication. Journal of Virology, 74(5), 2343–2350. http://doi.org/ 10.1128/JVI.74.5.2343-2350.2000 Johnson, T. B., Bunnell, D. B., & Knight, C. T. (2005). A potential new energy pathway in central Lake Erie: the round goby connection. Journal of Great Lakes Research, 31, 238–251. https://doi.org/10.1016/S0380-1330(05)70317-8 Jude, D. J., Reider, R. H., & Smith, G. R. (1992). Establishment of Gobiidae in the Great Lakes basin. Canadian Journal of Fisheries and Aquatic Sciences, 49(2), 416– 421. https://doi.org/10.1139/f92-047 Kahns, S., Skall, H. F., Kaas, R. S., Korsholm, H., Jensen, B. B., Jonstrup, S. P., ... & Olesen, N. J. (2012). European freshwater VHSV genotype Ia isolates divide into two distinct subpopulations. Diseases of Aquatic Organisms, 99(1), 23–35. https://doi.org/10.3354/dao02444 Kane-Sutton, M., Kinter, B., Dennis, P. M., & Koonce, J. F. (2010). Viral hemorrhagic septicemia virus infection in yellow perch, Perca flavescens, in Lake Erie. 156 Journal of Great Lakes Research, 36(1), 37–43. https://doi.org/10.1016/j.jglr.2009.11.004 Ke, Q., Weaver, W., Pore, A., Gorgoglione, B., Wildschutte, J. H., Xiao, P., ... & Vakharia, V. N. (2017). Role of viral hemorrhagic septicemia virus matrix (M) protein in suppressing host transcription. Journal of Virology, JVI-00279. http://doi.org/10.1128/JVI.00279-17 Kearney, M., Maldarelli, F., Shao, W., Margolick, J. B., Daar, E. S., Mellors, J. W., ... & Palmer, S. (2009). Human Immunodeficiency Virus type 1 population genetics and adaptation in newly infected individuals. Journal of Virology, 83(6), 2715– 2727. https://doi.org/10.1128/JVI.01960-08 Kennedy, D. A., Kurath, G., Brito, I. L., Purcell, M. K., Read, A. F., Winton, J. R., & Wargo, A. R. (2016). Potential drivers of virulence evolution in aquaculture. Evolutionary Applications, 9(2), 344–354. https://doi.org/10.1111/eva.12342 Kerr, P. J., Cattadori, I. M., Liu, J., Sim, D. G., Dodds, J. W., Brooks, J. W., ... & Read, A. F. (2017). Next step in the ongoing arms race between myxoma virus and wild rabbits in Australia is a novel disease phenotype. Proceedings of the National Academy of Sciences, 114(35), 9397–9402. https://doi.org/10.1073/pnas.1710336114 Kibenge, F. S. (2019). Emerging viruses in aquaculture. Current opinion in virology, 34, 97–103. https://doi.org/10.1016/j.coviro.2018.12.008 Kim, R., & Faisal, M. (2010). Experimental studies confirm the wide host range of the Great Lakes viral haemorrhagic septicaemia virus genotype IVb. Journal of Fish Diseases, 33(1), 83–88. https://doi.org/10.1111/j.1365-2761.2009.01093.x 157 Kim, R., & Faisal, M. (2011). Emergence and resurgence of the viral hemorrhagic septicemia virus (Novirhabdovirus, Rhabdoviridae, Mononegavirales). Journal of Advanced Research, 2(1), 9–23. https://doi.org/10.1016/j.jare.2010.05.007 Kim, M. S., & Kim, K. H. (2012). Effects of NV gene knock-out recombinant viral hemorrhagic septicemia virus (VHSV) on Mx gene expression in epithelioma papulosum cyprini (EPC) cells and olive flounder (Paralichthys olivaceus). Fish & shellfish immunology, 32(3), 459–463. https://doi.org/10.1016/j.fsi.2011.12.014 Kim, R. K., & Faisal, M. (2012). Shedding of viral hemorrhagic septicemia virus (Genotype IVb) by experimentally infected muskellunge (Esox masquinongy). The Journal of Microbiology, 50(2), 278–284. https://doi.org/10.1007/s12275- 012-1145-2 Kim, M. S., & Kim, K. H. (2013). The role of viral hemorrhagic septicemia virus (VHSV) NV gene in TNF-α-and VHSV infection-mediated NF-κB activation. Fish & Shellfish Immunology, 34(5), 1315–1319. Kim, S. H., Yusuff, S., Vakharia, V. N., & Evensen, Ø. (2015). Interchange of L polymerase protein between two strains of viral hemorrhagic septicemia virus (VHSV) genotype IV alters temperature sensitivities in vitro. Virus Research, 195, 203–206. https://doi.org/10.1016/j.virusres.2014.10.013 Kim, M. S., & Kim, K. H. (2019). Generation of VHSV replicon particles carrying transmembrane and C-terminal cytoplasmic region-deleted G gene (rVHSV- GΔTM) and comparison of vaccine efficacy with G gene-deleted VHSV (rVHSV- ΔG). Fish & Shellfish Immunology, 88, 231–236. https://doi.org/10.1016/j.fsi.2019.02.048 158 Kimura, H., Saitoh, M., Kobayashi, M., Ishii, H., Saraya, T., Kurai, D., ... & Kuroda, M. (2015). Molecular evolution of haemagglutinin (H) gene in Measles Virus. Scientific Reports, 5, 11648. https://doi.org/10.1038/srep11648 Kitchen, A., Shackelton, L. A., & Holmes, E. C. (2011). Family level phylogenies reveal modes of macroevolution in RNA viruses. Proceedings of the National Academy of Sciences, 108(1), 238–243. https://doi.org/10.1073/pnas.1011090108 Kole, S., Qadiri, S. S. N., Shin, S. M., Kim, W. S., Lee, J., & Jung, S. J. (2019). PLGA encapsulated inactivated-viral vaccine: Formulation and evaluation of its protective efficacy against viral haemorrhagic septicaemia virus (VHSV) infection in olive flounder (Paralichthys olivaceus) vaccinated by mucosal delivery routes. Vaccine, 37(7), 973-983. https://doi.org/10.1016/j.vaccine.2018.12.063 Kosakovsky Pond, S. L., Posada, D., Gravenor, M. B., Woelk, C. H., & Frost, S. D. (2006). Automated phylogenetic detection of recombination using a genetic algorithm. Molecular Biology and Evolution, 23(10), 1891–1901. https://doi.org/10.1093/molbev/msl051 Kumar, S., Stecher, G., Li, M., Knyaz, C., & Tamura, K. (2018). MEGA X: molecular evolutionary genetics analysis across computing platforms. Molecular Biology and Evolution, 35(6), 1547–1549. https://doi.org/10.1093/molbev/msy096 Kurath, G., Garver, K. A., Troyer, R. M., Emmenegger, E. J., Einer-Jensen, K., & Anderson, E. D. (2003). Phylogeography of Infectious Haematopoietic Necrosis Virus in North America. Journal of General Virology, 84(4), 803–814. https://doi.org/10.1099/vir.0.18771-0 159 Kurath, G. (2012) Fish novirhabdoviruses. Rhabdoviruses: Molecular Taxonomy, Evolution, Genomics, Ecology, Host-vector Interactions, Cytopathology, and Control (eds. Dietzgen, R. G., Kuzmin, I. V.), pp. 89–116) Caister Academic Press, Norfolk, United Kingdom. https://doi.org/10.1128/JVI.01960-08 Kuzmin, I. V., Hughes, G. J., & Rupprecht, C. E. (2006). Phylogenetic relationships of seven previously unclassified viruses within the family Rhabdoviridae using partial nucleoprotein gene sequences. Journal of General Virology, 87(8), 2323– 2331. https://doi.org/10.1099/vir.0.81879-0 Kuzmin, I. V., Novella, I. S., Dietzgen, R. G., Padhi, A., & Rupprecht, C. E. (2009). The rhabdoviruses: biodiversity, phylogenetics, and evolution. Infection, Genetics and Evolution, 9(4), 541–553. https://doi.org/10.1016/j.meegid.2009.02.005 Lafferty, K. D., Harvell, C. D., Conrad, J. M., Friedman, C. S., Kent, M. L., Kuris, A. M., ... & Saksida, S. M. (2015). Infectious diseases affect marine fisheries and aquaculture economics. Annual Review of Marine Science, 7, 471–496. https://doi.org/10.1146/annurev-marine-010814-015646 Lapatra, S. E. (1998). Factors affecting pathogenicity of Infectious Hematopoietic Necrosis Virus (IHNV) for salmonid fish. Journal of Aquatic Animal Health, 10(2), 121–131. https://doi.org/10.1577/1548- 8667(1998)010<0121:FAPOIH>2.0.CO;2 Lauring, A. S., & Andino, R. (2010). Quasispecies theory and the behavior of RNA viruses. PLoS Pathogens, 6(7), e1001005. https://doi.org/10.1371/journal.ppat.1001005 160 Leong, J. C., & Fryer, J. L. (1993). Viral vaccines for aquaculture. Annual Review of Fish Diseases, 3, 225-240. https://doi.org/10.1016/0959-8030(93)90036-B Li, H. (2013). Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv preprint arXiv:1303.3997. Lin, H. X., Rubio, L., Smythe, A. B., & Falk, B. W. (2004). Molecular population genetics of Cucumber Mosaic Virus in California: evidence for founder effects and reassortment. Journal of Virology, 78(12), 6666–6675. https://doi.org/10.1128/JVI.78.12.6666-6675.2004 Lin, Y. C., Chu, P. Y., Chang, M. Y., Hsiao, K. L., Lin, J. H., & Liu, H. F. (2016). Spatial temporal dynamics and molecular evolution of re-emerging rabies virus in Taiwan. International Journal of Molecular Sciences, 17(3), 392. https://doi.org/10.3390/ijms17030392 Livak, K. J., & Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2− ΔΔCT method. Methods, 25(4), 402–408. https://doi.org/10.1006/meth.2001.1262 Lowe, W. H., Kovach, R. P., & Allendorf, F. W. (2017). Population genetics and demography unite ecology and evolution. Trends in Ecology & Evolution, 32(2), 141–152. https://dx.doi.org/10.1016/j.tree.2016.12.002 Lumsden, J. S., Morrison, B., Yason, C., Russell, S., Young, K., Yazdanpanah, A., ... & Way, K. (2007). Mortality event in freshwater drum Aplodinotus grunniens from Lake Ontario, Canada, associated with Viral Haemorrhagic Septicemia virus, Type IV. Diseases of Aquatic Organisms, 76(2), 99–111. https://doi.org/10.3354/dao076099 161 Mantel, N. (1967). The detection of disease clustering and a generalized regression approach. Cancer Research, 27(2 Part 1), 209–220. Marques, J. T., & Carthew, R. W. (2007). A call to arms: coevolution of animal viruses and host innate immune responses. Trends in Genetics, 23(7), 359-364. https://doi.org/10.1099/vir.0.057448-0 May, F. J., Li, L., Davis, C. T., Galbraith, S. E., & Barrett, A. D. (2010). Multiple pathways to the attenuation of West Nile Virus in southeast Texas in 2003. Virology, 405(1), 8–14. https://doi.org/10.1016/j.virol.2010.04.022 McElhinney, L. M., Marston, D. A., Freuling, C. M., Cragg, W., Stankov, S., Lalosević, D., ... & Fooks, A. R. (2011). Molecular diversity and evolutionary history of Rabies Virus strains circulating in the Balkans. Journal of General Virology, 92(9), 2171–2180. https://doi.org/10.1099/vir.0.032748-0 Molecular Epidemiology of Aquatic Pathogens. Viral Hemorrhagic Septicemia virus (MEAP-VHSV); 2013. Available: http://gis.nacse.org/vhsv/ Meng, F., Dong, X., Hu, T., Liu, Y., Zhao, Y., Lv, Y., ... & Cui, Z. (2016). Analysis of quasispecies of Avian Leukosis Virus subgroup J using Sanger and high- throughput sequencing. Virology Journal, 13(1), 112. https://doi.org/10.1186/s12985-016-0559-6 Metsky, H. C., Matranga, C. B., Wohl, S., Schaffner, S. F., Freije, C. A., Winnicki, S. M., ... & Lin, A. E. (2017). Zika Virus evolution and spread in the Americas. Nature, 546(7658), 411. https://doi.org/10.1038/nature22402 Meyers, T. R., Sullivan, J., Emmenegger, E., Follett, J., Short, S., & Batts, W. N. (1992). Identification of Viral Hemorrhagic Septicemia virus isolated from Pacific cod 162 Gadus macrocephalus in Prince William Sound, Alaska, USA. Diseases of Aquatic Organisms, 12(3), 167–175. https://doi.org/10.3354/dao012167 Meyers, T. R., & Winton, J. R. (1995). Viral Hemorrhagic Septicemia virus in North America. Annual Review of Fish Diseases, 5, 3–24. https://doi.org/10.1016/0959- 8030(95)00002-X Millard, E. V., Brenden, T. O., LaPatra, S. E., Marcquenski, S., & Faisal, M. (2014). Detection of Viral Hemorrhagic Septicemia virus-IVb antibodies in sera of muskellunge Esox masquinongy using competitive ELISA. Diseases of Aquatic Organisms, 108(3), 187–199. https://doi.org/10.3354/dao02712 Miller, R. R. (1960). Systematics and biology of the gizzard shad (Dorosoma cepedianum) and related fishes. Fisheries, 1(2), 3. Mordecai, G. J., Wilfert, L., Martin, S. J., Jones, I. M., & Schroeder, D. C. (2015). Diversity in a honeybee pathogen: first report of a third master variant of the Deformed Wing Virus quasispecies. The ISME Journal, 10(5), 1264. https://doi.org/10.1038/ismej.2015.178 Moya, A., Holmes, E. C., & González-Candelas, F. (2004). The population genetics and evolutionary epidemiology of RNA viruses. Nature Reviews Microbiology, 2(4), 279. https://doi.org/10.1038/nrmicro863 Murrell, B., Wertheim, J. O., Moola, S., Weighill, T., Scheffler, K., & Pond, S. L. K. (2012). Detecting individual sites subject to episodic diversifying selection. PLoS Genetics, 8(7), e1002764. https://doi.org/10.1371/journal.pgen.1002764 Murrell, B., Moola, S., Mabona, A., Weighill, T., Sheward, D., Kosakovsky Pond, S. L., & Scheffler, K. (2013). FUBAR: a fast, unconstrained bayesian approximation for 163 inferring selection. Molecular Biology and Evolution, 30(5), 1196–1205. https://doi.org/10.1093/molbev/mst030 Nishizawa, T., Iida, H., Takano, R., Isshiki, T., Nakajima, K., & Muroga, K. (2002). Genetic relatedness among Japanese, American and European isolates of viral hemorrhagic septicemia virus (VHSV) based on partial G and P genes. Diseases of Aquatic Organisms, 48(2), 143-148. http://doi.org/10.3354/dao048143 Nishizawa, T., Kinoshita, S., Kim, W. S., Higashi, S., & Yoshimizu, M. (2006). Nucleotide diversity of Japanese isolates of Infectious Hematopoietic Necrosis virus (IHNv) based on the glycoprotein gene. Diseases of Aquatic Organisms, 71(3), 267–272. https://doi.org/10.3354/dao071267 Novella, I. S., & Presloid, J. B. (2012) Experimental evolution of rhabdoviruses. Rhabdoviruses: Molecular Taxonomy, Evolution, Genomics, Ecology, Host- vector Interactions, Cytopathology, and Control (eds. Dietzgen, R. G., Kuzmin, I. V.), pp. 89–116, Caister Academic Press, Norfolk, United Kingdom. https://doi.org/10.1128/JVI.01960-08 Ojosnegros, S., & Beerenwinkel, N. (2010). Models of RNA virus evolution and their roles in vaccine design. Immunome Research, 6(2), S5. https://doi.org/10.1186/1745-7580-6-S2-S5 Padhi, A., & Verghese, B. (2012). Molecular evolutionary and epidemiological dynamics of a highly pathogenic fish rhabdovirus, the Spring Viremia of Carp virus (SVCv). Veterinary Microbiology, 156(1–2), 54–63. https://doi.org/10.1016/j.vetmic.2011.10.005 164 Paterson, S., Vogwill, T., Buckling, A., Benmayor, R., Spiers, A. J., Thomson, N. R., ... & Fenton, A. (2010). Antagonistic coevolution accelerates molecular evolution. Nature, 464(7286), 275. https://doi.org/10.1038/nature08798 Pereira, F., & Amorim, A. (2013) Evolution: viruses. (eds. Maloy, S., & Hughes, K.), p 566. Brenners Encyclopedia of Genetics. Elsevier, New York, NY, USA. https://doi.org/10.1016/B978-0-12-374984-0.00499-X Pesko, K. N., & Ebel, G. D. (2012). West Nile Virus population genetics and evolution. Infection, Genetics and Evolution, 12(2), 181–190. https://doi.org/10.1016/j.meegid.2011.11.014 Pierce, L. R., & Stepien, C. A. (2012). Evolution and biogeography of an emerging quasispecies: diversity patterns of the fish Viral Hemorrhagic Septicemia virus (VHSV). Molecular Phylogenetics and Evolution, 63(2), 327–341. https://doi.org/10.1016/j.ympev.2011.12.024 Pierce, L. R. (2013). The evolution and detection of the fish viral hemorrhagic septicemia virus (VHSV) (Doctoral dissertation, The University of Toledo, Toledo, OH, 253 pp.) Available at: http://ohioseagrant.osu.edu/archive/_documents/publications/TD/TD- 1392013%20Evolution%20and%20detection%20of%20fish%20VHSV.pdf Pierce, L. R., Willey, J. C., Crawford, E. L., Palsule, V. V., Leaman, D. W., Faisal, M., ... & Stepien, C. A. (2013a). A new StaRT-PCR approach to detect and quantify fish Viral Hemorrhagic Septicemia virus (VHSV): Enhanced quality control with internal standards. Journal of Virological Methods, 189(1), 129–142. https://doi.org/10.1016/j.jviromet.2013.01.006 165 Pierce, L. R., Willey, J. C., Palsule, V. V., Yeo, J., Shepherd, B. S., Crawford, E. L., & Stepien, C. A. (2013b). Accurate detection and quantification of the fish Viral Hemorrhagic Septicemia virus (VHSV) with a two-color fluorometric real-time PCR assay. PloS one, 8(8), e71851. https://doi.org/10.1371/journal.pone.0071851 Purcell, M. K., Laing, K. J., & Winton, J. R. (2012). Immunity to fish rhabdoviruses. Viruses, 4(1), 140-166. https://doi.org/10.3390/v4010140 Pore, A. (2012). Studies on host-virus interaction for Viral Hemorrhagic Septicemia virus (VHSV) (Master’ thesis, University of Toledo, Toledo, OH, USA, 83pp.). Available at: https://etd.ohiolink.edu/pg_10?0::NO:10:P10_ACCESSION_NUM:toledo133676 6667 Pratheek, B. M., Saha, S., Maiti, P. K., Chattopadhyay, S., & Chattopadhyay, S. (2013). Immune regulation and evasion of mammalian host cell immunity during viral infection. Indian Journal of Virology, 24(1), 1–15. Quer, J., Huerta, R., Novella, I. S., Tsimring, L., Domingo, E., & Holland, J. J. (1996). Reproducible nonlinear population dynamics and critical points during replicative competitions of RNA virus quasispecies. Journal of Molecular Biology, 264(3), 465–471. https://doi.org/10.1006/jmbi.1996.0654 Raymond, M., & Rousset, F. (1995). An exact test for population differentiation. Evolution, 49(6), 1280–1283. https://doi.org/10.1111/j.1558-5646.1995.tb04456.x Reynolds, J., Weir, B. S., & Cockerham, C. C. (1983). Estimation of the coancestry coefficient: basis for a short-term genetic distance. Genetics, 105(3), 767–779. Rice, W. R. (1989). The sequential Bonferroni test. Evolution, 43, 223–225. 166 Rude, N. P., Yung, A. J., & Whitledge, G. W. (2017). Movement of small bodied fishes from Lake Michigan into Chicago Area Waterways: Insights from otolith chemistry. Journal of Applied Ichthyology, 33(6), 1166–1172. https://doi.org/10.1111/jai.13499 Robinson, J. T., Thorvaldsdóttir, H., Winckler, W., Guttman, M., Lander, E. S., Getz, G., & Mesirov, J. P. (2011). Integrative genomics viewer. Nature Biotechnology, 29(1), 24. https://doi.org/10.1038/nbt.1754 Rodríguez Nevado, C., Lam, T. T. Y., Holmes, E. C., & Pagán, I. (2018). The impact of host genetic diversity on virus evolution and emergence. Ecology Letters, 21(2), 253–263. https://doi.org/10.1111/ele.12890 Rousset, F. (2008). GENEPOP’007: a complete re implementation of the GENEPOP software for Windows and Linux. Molecular Ecology Resources, 8(1), 103–106. https://doi.org/10.1111/j.1471-8286.2007.01931.x Sandlund, N., Gjerset, B., Bergh, Ø., Modahl, I., Olesen, N. J., & Johansen, R. (2014). Screening for Viral Hemorrhagic Septicemia virus in marine fish along the Norwegian coastal line. PloS One, 9(9), e108529. https://doi.org/10.1371/journal.pone.0108529 Sanjuán, R., Nebot, M. R., Chirico, N., Mansky, L. M., & Belshaw, R. (2010). Viral mutation rates. Journal of Virology, 84(19), 9733–9748. https://doi.org/10.1128/JVI.00694-10 Schoebel, C. N., Botella, L., Lygis, V., & Rigling, D. (2017). Population genetic analysis of a parasitic mycovirus to infer the invasion history of its fungal host. Molecular Ecology, 26(9), 2482–2497. https://doi.org/10.1111/mec.14048 167 Schönherz, A. A., Lorenzen, N., Guldbrandtsen, B., Buitenhuis, B., & Einer-Jensen, K. (2016). Ultra-deep sequencing of VHSV isolates contributes to understanding the role of viral quasispecies. Veterinary Research, 47(1), 10. https://doi.org/10.1186/s13567-015-0298-5 Schönherz, A. A., Forsberg, R., Guldbrandtsen, B., Buitenhuis, A. J., & Einer-Jensen, K. (2018). Introduction of Viral Haemorrhagic Septicaemia virus (VHSV) to freshwater cultured rainbow trout is followed by bursts of adaptive evolution. Journal of Virology, JVI-00436. https://doi.org/10.1128/JVI.00436-18 Scott, W.B., & Crossman, E. J. (1973). Freshwater Fishes of Canada. Fisheries Research Board of Canada Bulletin. 184. Sieracki, J. L., Bossenbroek, J. M., & Chadderton, W. L. (2014). A spatial modeling approach to predicting the secondary spread of invasive species due to ballast water discharge. PloS One, 9(12), e114217. https://doi.org/10.1371/journal.pone.0114217 Skall, H. F., Olesen, N. J., & Mellergaard, S. (2005). Prevalence of Viral Haemorrhagic Septicaemia virus in Danish marine fishes and its occurrence in new host species. Diseases of Aquatic Organisms, 66(2), 145–151. https://doi.org/10.3354/dao066145 Smail, D. (1999) Viral Haemorrhagic Septicaemia. In: Fish Diseases and Disorders: Viral, Bacterial and Fungal Infections. (eds. Woo, P., & Bruno, D.), pp 123–147. Fisheries Research Services, Aberdeen, UK. Snow, M., Cunningham, C. O., Melvin, W. T., & Kurath, G. (1999). Analysis of the nucleoprotein gene identifies distinct lineages of viral haemorrhagic septicaemia 168 virus within the European marine environment. Virus Research, 63(1–2), 35–44. https://doi.org/10.1016/S0168-1702(99)00056-8 Snyder, M. R., & Stepien, C. A. (2017). Genetic patterns across an invasion's history: a test of change versus stasis for the Eurasian round goby in North America. Molecular Ecology, 26(4), 1075–1090. https://doi.org/10.1111/mec.13997 Stepien, C. A., Pierce, L. R., Leaman, D. W., Niner, M. D., & Shepherd, B. S. (2015). Gene diversification of an emerging pathogen: a decade of mutation in a novel fish Viral Hemorrhagic Septicemia (VHS) substrain since its first appearance in the Laurentian Great Lakes. PloS One, 10(8), e0135146. https://doi.org/10.1371/journal.pone.0135146 Stone, D. M., Way, K., & Dixon, P. F. (1997). Nucleotide sequence of the glycoprotein gene of Viral Haemorrhagic Septicaemia (VHS) viruses from different geographical areas: a link between VHS in farmed fish species and viruses isolated from North Sea cod (Gadus morhua L.). Journal of General Virology, 78(6), 1319–1326. http://doi.org/10.1099/0022-1317-78-6-1319 Suchard, M. A., Lemey, P., Baele, G., Ayres, D. L., Drummond, A. J., & Rambaut, A. (2018). Bayesian phylogenetic and phylodynamic data integration using BEAST 1.10. Virus Evolution, 4(1), vey016. https://doi.org/10.1093/ve/vey016 Takano, R., Nishizawa, T., Arimoto, M., & Muroga, K. (2000). Isolation of Viral Haemorrhagic Septicaemia virus (VHSV) from wild Japanese flounder, Paralichthys olivaceus. Bulletin-European Association of Fish Pathologists, 20(5), 186-192. https://doi.org/10.3354/dao047087 169 Tajima, F. (1989). Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics, 123(3), 585–595. https://www.genetics.org/content/genetics/123/3/585.full.pdf Thompson, T. M., Batts, W. N., Faisal, M., Bowser, P., Casey, J. W., Phillips, K., ... & Kurath, G. (2011). Emergence of Viral Hemorrhagic Septicemia virus in the North American Great Lakes region is associated with low viral genetic diversity. Diseases of Aquatic Organisms, 96(1), 29–43. https://doi.org/10.3354/dao02362 Thorvaldsdóttir, H., Robinson, J. T., & Mesirov, J. P. (2013). Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Briefings in Bioinformatics, 14(2), 178–192. https://doi.org/10.1093/bib/bbs017 Thoulouze, M. I., Bouguyon, E., Carpentier, C., & Brémont, M. (2004). Essential role of the NV protein of Novirhabdovirus for pathogenicity in rainbow trout. Journal of virology, 78(8), 4098–4107. http://doi.org/10.1128/JVI.78.8.4098-4107.2004 Throckmorton, E., Peters, A., Brenden, T., & Faisal, M. (2015). Direct and indirect evidence suggests continuous presence of Viral Hemorrhagic Septicemia virus (Genotype IVb) in Budd Lake, Michigan: management implications. North American Journal of Fisheries Management, 35(3), 503–511. https://doi.org/10.1080/02755947.2015.1020077 Throckmorton, E., Brenden, T., Peters, A. K., Newcomb, T. J., Whelan, G. E., & Faisal, M. (2017). Potential reservoirs and risk factors for VHSV-IVb in an enzootic system: Budd Lake, Michigan. Journal of Aquatic Animal Health, 29(1), 31–42. https://doi.org/10.1080/08997659.2016.1254121 170 Trautman, M. B. (1981) The Fishes of Ohio: with illustrated keys. Ohio State University Press, Columbus, OH, USA. Troupin, C., Dacheux, L., Tanguy, M., Sabeta, C., Blanc, H., Bouchier, C., ... & Bourhy, H. (2016). Large-scale phylogenomic analysis reveals the complex evolutionary history of Rabies Virus in multiple carnivore hosts. PLoS Pathogens, 12(12), e1006041. https://doi.org/10.1371/journal.ppat.1006041 Tsompana, M., Abad, J., Purugganan, M., & Moyer, J. W. (2005). The molecular population genetics of the Tomato Spotted Wilt virus (TSWv) genome. Molecular Ecology, 14(1), 53–66. https://doi.org/10.1111/j.1365-294X.2004.02392.x Van Valen, L. (1973). A new evolutionary law. Evolutionary Theory, 1, 1-30. Van Valen, L. (1974). Molecular evolution as predicted by natural selection. Journal of Molecular Evolution, 3(2), 89-101. https://doi.org/10.1007/BF01796554 Vennerström, P., Välimäki, E., Hautaniemi, M., Lyytikäinen, T., Kapiainen, S., Vidgren, G., & Virtala, A. M. (2018). Wild fish are negligible transmitters of Viral Haemorrhagic Septicaemia virus (VHSV) genotype Id in the VHS restriction zone in Finland. Diseases of Aquatic Organisms, 131(3), 187–197. https://doi.org/10.3354/dao03301 Volz, E. M., Koelle, K., & Bedford, T. (2013). Viral phylodynamics. PLoS Computational Biology, 9(3), e1002947. https://doi.org/10.1371/journal.pcbi.1002947 Walker, P. J., Blasdell, K. R., Calisher, C. H., Dietzgen, R. G., Kondo, H., Kurath, G., ... & Vasilakis, N. (2018). ICTV virus taxonomy profile: Rhabdoviridae. Journal of General Virology, 99(4), 447–448. https://doi.org/10.1099/jgv.0.001020 171 Whelan, G. E. (2007). Viral Hemorrhagic Septicemia (VHS) briefing paper. Michigan Department of Natural Resources, Ann Arbor, MI, USA. Weir, B. S., & Cockerham, C. C. (1984). Estimating F statistics for the analysis of population structure. Evolution, 38(6), 1358–1370. https://doi.org/10.1111/j.1558- 5646.1984.tb05657.x Wolf, K. (1988) Hemorrhagic septicemia virus. In: Fish Viruses and Fish Viral Diseases. (ed. Wolf, K.), pp. 217–248. Cornell University Press, Ithaca, NY, USA. Workenhe, S. T., Rise, M. L., Kibenge, M. J., & Kibenge, F. S. (2010). The fight between the teleost fish immune response and aquatic viruses. Molecular Immunology, 47(16), 2525–2536. https://doi.org/10.1016/j.molimm.2010.06.009 Volz, E. M., Koelle, K., & Bedford, T. (2013). Viral phylodynamics. PLoS Computational Biology, 9(3), e1002947. https://doi.org/10.1371/journal.pcbi.1002947 Xu, L., Zhao, J., Liu, M., Kurath, G., Ren, G., Lapatra, S. E., ... & Lu, T. (2017). A effective DNA vaccine against diverse genotype J infectious hematopoietic necrosis virus strains prevalent in China. Vaccine, 35(18), 2420-2426. https://doi.org/10.1016/j.vaccine.2017.03.047 Xu, L., Zhao, J., Liu, M., Kurath, G., Breyta, R. B., Ren, G., ... & Lu, T. (2018). Phylogeography and evolution of infectious hematopoietic necrosis virus in China. Molecular Phylogenetics and Evolution, 131, 19–28. https://doi.org/10.1016/j.ympev.2018.10.030 172 Yang, X., Charlebois, P., Gnerre, S., Coole, M. G., Lennon, N. J., Levin, J. Z., ... & Henn, M. R. (2012). De novo assembly of highly diverse viral populations. BMC Genomics, 13(1), 475. https://doi.org/10.1186/1471-2164-13-475 Yusuff, S., Kurath, G., Kim, M. S., Tesfaye, T. M., Li, J., McKenney, D. G., & Vakharia, V. N. (2019). The glycoprotein, non-virion protein, and polymerase of viral hemorrhagic septicemia virus are not determinants of host-specific virulence in rainbow trout. Virology Journal, 16(1), 31. https://doi.org/10.1186/s12985-019- 1139-3 Zhang, Q., & Gui, J. F. (2015). Virus genomes and virus-host interactions in aquaculture animals. Science China Life Sciences, 58(2), 156–169. https://doi.org/10.1007/s11427-015-4802-y Zhang, J., Tang, X., Sheng, X., Xing, J. and Zhan W. 2017. Isolation and identification of a new strain of hirame rhabdovirus (HIRRV) from Japanese flounder Paralichthys olivaceus in China. Virology Journal. 14(1):73. https://doi.org/10.1186/s12985- 017-0742-4 Zou, J., Gorgoglione, B., Taylor, N. G., Summathed, T., Lee, P. T., Panigrahi, A., ... & Mughal, S. M. (2014). Salmonids have an extraordinary complex type I IFN system: characterization of the IFN locus in rainbow trout Oncorhynchus mykiss reveals two novel IFN subgroups. The Journal of Immunology, 193(5), 2273– 2286. https://doi.org/10.4049/jimmunol.1301796 Zienius, D., Pridotkas, G., Morkūnas, M., & Ružauskas, M. (2016). Phylogenetic investigation of partial G gene sequences in Lithuanian Rabies Virus field isolates. Veterinarski Arhiv, 86(1), 23–34. ISSN: 0372-5480 173