Interferon Kinetics and its effect on DNA vaccine induced immunity in
Channel Catfish (Ictalurus punctatus)
Except where reference is made to the work of others, the work described in this thesis
is my own or was done in collaboration with my advisory committee. This thesis does
not include proprietary or classified information.
Kendal Thomas Harder
Certificate of Approval:
Richard C. Bird, PhD
Professor
Pathobiology
Kenneth E. Nusbaum, Chair
Associate Professor
Pathobiology
Vicky L. van Santen
Professor
Pathobiology
Frederik W. van Ginkel
Associate Professor
Pathobiology
George T. Flowers
Interim Dean
Graduate School
Interferon Kinetics and its effect on DNA vaccine induced immunity in
Channel Catfish (Ictalurus punctatus)
Kendal Thomas Harder
A Thesis
Submitted to
the Graduate Faculty of
Auburn University
in Partial Fulfillment of the
Requirements for the
Degree of
Master of Science
Auburn, Alabama
August 9, 2008
Interferon Kinetics and its effect on DNA vaccine induced immunity in
Channel Catfish (Ictalurus punctatus)
Kendal Thomas Harder
Permission is granted to Auburn University to make copies of this thesis at its
discretion, upon the request of individuals or institutions and at their
expense. The author reserves all publication rights.
Signature of Author
Date of Graduation
iii
Vita
Kendal Dalton (Thomas) Harder, daughter of James Philip and Dalton (Robinson)
Thomas, was born April 5, 1980, in Jackson, Mississippi. She graduated magna cum
laude from The Woodlands High School in 1999. While attending Baylor University in
Waco, Texas, she worked for two summers as a summer research assistant for Dr. Dorota
Skowyra in the Edward A. Doisy Department of Biochemistry and Molecular Biology at
St. Louis University of Medicine, and worked one semester as a research assistant for Dr.
Myeongwoo Lee in the Department of Biology at Baylor University. She graduated with
a Bachelor of Science degree in Biology with minors in Chemistry and Computer Science.
She married Adam Wells Harder, son of Mike and Ann (Wells) Harder, on August 30,
2003. She worked for 4 months as a lab technician for Dr. Kenneth E. Nusbaum in
the department of Pathobiology at Auburn University College of Veterinary Medicine
before entering Graduate School at Auburn University. She currently lives in Melbourne,
Florida with her husband and is a Software Engineer at Harris Corporation.
iv
Thesis Abstract
Interferon Kinetics and its effect on DNA vaccine induced immunity in
Channel Catfish (Ictalurus punctatus)
Kendal Thomas Harder
Master of Science, August 9, 2008
(B.S., Baylor University, 2003)
101 Typed Pages
Directed by Kenneth E. Nusbaum
Increasing the effectiveness of vaccination against channel catfish diseases, including
channel catfish virus (CCV) is important economically and clinically for channel catfish
farming. Catfish interferon (IFN), an innate, nonspecific immune protein, provides im-
mediate nonspecific viral protection and can be induced with either CCV, UV-inactivated
CCV or polyinosinic-polycytidylic acid, (poly I:C). Conditions and kinetics of the IFN
response in channel catfish ovary (CCO) cells and in channel catfish are described. Vac-
cination of channel catfish with ORF 59, a cloned gene, increased total anti-CCV activity
during 4 hours to 4 weeks post vaccination. When IFN is elicited concurrently with DNA
vaccination, the effectiveness of the vaccine does not increase.
v
Acknowledgments
The author thanks Dr. Vicky van Santen and Natalia Petrenko for time and assis-
tance with real time PCR. She thanks Patricia DeInnocentes for guidance, mentorship,
counseling and friendship during the course of this study.
vi
Style manual or journal used Fish & Shellfish Immunology (together with the style
known as ?aums?). Bibliograpy follows Fish & Shellfish Immunology.
Computer software used The document preparation package TEX (specifically
LATEX) together with the departmental style-file aums.sty. Text composed in Google
Docs.
vii
Table of Contents
List of Figures x
1 Introduction 1
1.0.1 Nucleic acid vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.0.2 Mass spectrometry to identify genes encoding structural proteins
of CCV for vaccination . . . . . . . . . . . . . . . . . . . . . . . . 4
1.0.3 Using nucleic acid vaccines in channel catfish . . . . . . . . . . . . 5
1.0.4 DNA vaccines induce IFN expression . . . . . . . . . . . . . . . . . 6
2 Literature Review 11
2.1 Introduction to fish immunity . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Introduction to IFN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.1 Type I IFN antiviral pathway in higher vertebrates . . . . . . . . . 13
2.2.2 Type II IFN function in higher vertebrates . . . . . . . . . . . . . 16
2.3 Antiviral pathways in fish . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.1 Characterization of fish IFNs . . . . . . . . . . . . . . . . . . . . . 20
2.3.2 Functional studies of type I IFN in fish . . . . . . . . . . . . . . . 23
2.3.3 Alignment analysis of fish IFNs . . . . . . . . . . . . . . . . . . . . 26
2.3.4 Function of type II IFN in fish . . . . . . . . . . . . . . . . . . . . 30
2.3.5 JAK-STAT pathway and members in fish . . . . . . . . . . . . . . 31
2.3.6 Antiviral pathways induced by IFN in fish . . . . . . . . . . . . . . 32
2.3.7 IFN activity demonstrated in fish - two viruses used to demonstrate
protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3.8 IFN activity demonstrated in fish - IFN detection using Mx protein
expression induced by poly I:C . . . . . . . . . . . . . . . . . . . . 41
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3 Materials and Methods 44
3.1 Maintaining cell culture and virus propagation . . . . . . . . . . . . . . . 44
3.2 Induction and detection of IFN by poly I:C in CCO cells . . . . . . . . . . 44
3.3 Induction of IFN expression by CCV in CCO cells . . . . . . . . . . . . . 46
3.4 Induction of IFN expression by ultraviolet (UV) inactivated CCV in CCO
cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.5 Primers and conditions for rt-PCR . . . . . . . . . . . . . . . . . . . . . . 46
3.6 Induction of IFN in channel catfish . . . . . . . . . . . . . . . . . . . . . . 47
3.7 Effect of IFN induction at the time of vaccination . . . . . . . . . . . . . . 47
3.8 Antiviral assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.9 Virus neutralization assays . . . . . . . . . . . . . . . . . . . . . . . . . . 49
viii
4 Results 50
4.1 Inducing IFN expression in CCO cells . . . . . . . . . . . . . . . . . . . . 50
4.1.1 Optimization and kinetics of poly I:C . . . . . . . . . . . . . . . . 50
4.1.2 Comparison of expression of IFN genes . . . . . . . . . . . . . . . . 51
4.1.3 Inducing IFN expression using CCV . . . . . . . . . . . . . . . . . 54
4.1.4 Inducing IFN expression using inactivated CCV . . . . . . . . . . 55
4.2 Challenge Assays using Culture Medium from induced CCO cells . . . . . 57
4.3 Inducing IFN expression in channel catfish with poly I:C and ORF 59
vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.4 Antiviral activity at 4, 24, and 48 hours from poly I:C/ORF 59 vaccinated
fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.5 Antiviral assays from poly I:C/ORF 59 vaccinated fish at 2 and 4 weeks . 59
4.6 Viral neutralization assays from poly I:C/ORF 59 vaccinated fish at 4
weeks post vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.7 Inducing IFN expression in channel catfish with poly I:C and CCV . . . . 61
4.8 IFN protection in serum from CCV and poly I:C induced channel catfish . 62
5 Discussion 65
6 Overall Conclusions 80
Bibliography 81
ix
List of Figures
4.1 Antiviral activity of CCO culture medium . . . . . . . . . . . . . . . . . . 51
4.2 IFN mRNA expression in CCO cells . . . . . . . . . . . . . . . . . . . . . 52
4.3 IFN-2 cloned sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4 IFN-2 mRNA expression in CCO cells . . . . . . . . . . . . . . . . . . . . 54
4.5 CCV / UV-CCV induced IFN mRNA expression . . . . . . . . . . . . . . 55
4.6 Poly I:C/ORF59 induced IFN mRNA expression in channel catfish . . . . 63
4.7 Poly I:C / CCV induced IFN mRNA expression in channel catfish . . . . 64
x
Chapter 1
Introduction
Farmed fish are unavoidably subjected to a variety of stresses, which will lower their
innate defenses and will subsequently lead to outbreaks of viral or bacterial infections.
These stresses are caused by handling, transportation, living in over-crowded conditions,
poor water quality and exposure to pollutants. They are associated with an increased
susceptibility of fish to diseases (reviewed in [1]). The innate immune system in fish is
constantly challenged by pathogens and constitute the first line of defense (reviewed in
[2]). Some investigators have suggested that the innate immune system is more important
in fish than adaptive immune responses, because innate defenses respond quickly and
are temperature independent. Therefore, they are able to respond to and slow down
or halt viral and bacterial invasions before adaptive responses are generated (reviewed
in [2]). However, the innate system in fish alone is not enough to protect fish against
viral pathogens. In fact, viral diseases cause significant losses in aquaculture each year
[3]. In aquaculture there is an increasing demand for effective and economical vaccines.
Viral diseases in fish are of clinical and economic importance in farmed fish. Thus, many
studies have focused on ways to efficiently vaccinate fish against viral diseases.
An example of a fish disease of economic consequence is channel catfish virus (CCV)
disease. CCV, a herpesvirus (ictalurid herpesvirus 1, IHV-1), may cause fatal disease
in 40-90% of channel catfish fry each year [3]. CCV is an ?-like herpesvirus and has a
double stranded 134 kbp genome that contains 79 open reading frames (ORF), with 14
of the genes present twice [4] [5]. Even though experimental data suggest that older fish
1
are susceptible to natural outbreaks of acute CCV, the disease occurs almost exclusively
in fish that are less than 1 year of age, and generally less than 4 months of age [6]. Water
temperature is a critical environmental factor for outbreaks. When water temperature
rises above 27?C the mortality rate is high. The rate decreases as water temperature
decreases, and there is no mortality below 18?C [6]. Diseased fish exhibit ascites, ex-
ophthalmia and hemorrhage in fins and musculature. Histologically, the most damage
occurs in the kidney with extensive necrosis of renal tubules and interstitial tissue [6].
Fish that survive CCV infection may become latent carriers. Surviving fish display
protective levels of CCV-specific antibodies (reviewed in [6]). During the latent state,
the virus is undetectable by traditional culture or antigen-detecting methods, even in
immunosuppressed spawning adults (reviewed in [6]).
The main reservoir of CCV consists of carrier fish. Infectious CCV can be detected
in tanks of experimentally infected catfish, although the route of viral shedding has yet to
be determined. CCV transmission occurs horizontally and vertically. Horizontal trans-
mission may be direct or through water, the main abiotic vector. Vertical transmission
occurs commonly, though the mechanism is not known (reviewed in [6]).
Currently, control of CCV is based on maintaining relatively low stocking densities
of fish and avoiding stressful handling of young fish during the summer months. Facil-
ities used for incubating eggs and rearing catfish fry and juveniles are separated from
those used for carrier populations to prevent the occurrence of CCV in a CCV-free fish
production site. Defining CCV-free populations has to be done largely from historical
data or from identifying populations that are sero-negative to the virus, since virus is not
detected in latent carrier states but only during active outbreaks. However, recent use of
PCR and hybridization probes to detect latent CCV in fish shows that CCV is present
2
in many fish populations, including those that have no record of the disease (reviewed
in [6]). CCV is cited as the cause of deaths in fewer than 3% of cases submitted to
diagnostic laboratories (reviewed in [6]). However, due to the ubiquity of CCV in cul-
turing channel catfish, the virus is an important economic consideration for the catfish
industry.
CCV infected cells are most likely targeted by channel catfish NK-like cells, as
reported in mammalian herpesviruses. NK cells efficiently lyse herpesvirus-infected cells
in a MHC-unrestricted manner. The recognition of herpesvirus-infected cells by cytotoxic
cells requires only early virus gene expression [7]; thus, it could be very beneficial to
catfish to induce a rapid innate response against CCV. Vaccination against CCV is
currently not being used due to a lack of efficient delivery of the available vaccines. Of
the different vaccines, nucleic acid-based vaccines are the most promising. They induce
rapid innate immune responses to viruses, as well as antigen-specific protection in fish
[8] [9] [10] [11] [12] [13] [14] [15].
1.0.1 Nucleic acid vaccines
A typical DNA vaccine is composed of a bacterial plasmid that contains a com-
plementary DNA (cDNA) encoding a protein antigen from a virus. Antigen presenting
cells (APC), such as dendritic cells, are transfected by the plasmid and the cDNA is
transcribed and translated into immunogenic peptides that are presented on cell sur-
faces and elicit specific immune, B and T cell responses. DNA vaccines provide strong
protection because they produce viral proteins that are correctly folded and modified by
the host cell so that viral peptides are presented by both class I and II MHC molecules,
3
thus providing long-lived humoral and cell-mediated immune protection of the host (re-
viewed in [8]). This protection is caused by specific immunity consisting of antibodies
and CTL responses. DNA vaccines are able to induce a strong CTL response because the
DNA-encoded proteins are synthesized in the cytosol of transfected cells and presented
on MHC class I proteins on the cell membrane. Additionally, DNA vaccines provide
strong protection because plasmid DNA is rich in unmethylated CpG motifs that are
recognized by Toll-like receptor 9 (TLR9) on macrophages, T and B cells. The CpG
motifs activate APCs and increase the expression of surface molecules that induce and
enhance the adaptive immune response. These CpG motifs also induce release of inter-
leukins IL-1, IL-12, IFN-? and tumor necrosis factor alpha (TNF-?) (reviewed in [8]).
Thus, plasmid DNA vaccines are effective and their CpG motifs may act as an intrinsic
adjuvant (reviewed in [16]). Kim et al. [8], showed that CpG motif-containing DNA acts
similarly in fish as has been reported in mammals.
1.0.2 Mass spectrometry to identify genes encoding structural proteins of
CCV for vaccination
Davison and Davison [5] used mass spectrometry to compare herpes simplex virus
type 1 (HSV-1) to herpesvirus CCV because of their common virion morphology. This
commonality enabled Davison and Davison [5] to identify 12 genes, 11 viral and 1 cellular,
which encoded 16 major structural proteins that could be used as possible antigenic
determinants for CCV. Some of these proteins are late proteins which generally are large,
stable and antigenic. These structural proteins made up the mature capsid, immature
capsid, the tegument, and the envelope associated proteins.
4
Sequence analysis of the proteins found in CCV showed that CCV lacked envelope
glycoproteins that are homologous to envelope glycoproteins found in mammalian or
avian herpesviruses. Gene 59 encodes the principal envelope protein of CCV. The protein
encoded by gene 59 contains four stretches of hydrophobic residues capable of spanning
the membrane, and if it is folded in an orientation in which the termini are inside the
envelope, this would position a loop containing three potential N-linked glycosylation
sites on the external surface of the virion [5]. Although multiple hydrophobic proteins are
a feature to all herpesviruses, CCV is unusual in that a protein with so many hydrophobic
domains forms the main envelope protein. Some membrane proteins that are not major
envelope proteins are encoded by seven other CCV genes, include genes 6, 7, 8, 10, 19,
46, and 51. Gene 59 proteins are prime candidates for the major antigenic determinants
of CCV [5] and [3]. In fact, Nusbaum et al. [3] selected CCV genes 59, 6, 7, 8a, 10, 51
and 53 to be used in DNA vaccinations against CCV, because of the size, putative order
of expression, and association with membrane structures as envelope glycoproteins, or
function as viral capsid proteins of the proteins encoded by these genes.
1.0.3 Using nucleic acid vaccines in channel catfish
Nusbaum et al. [3] were able to induce protective immunity in channel catfish using
a DNA vaccine. Seven full-length transcripts from the CCV genome were selected for the
DNA vaccination. ORF 59 encodes a glycoprotein associated with the envelope; ORFs
6, 7, 10 and 51 encode membrane proteins; ORF 8a encodes a membrane-associated
protein; and ORF 53 encodes the viral capsid protein. Channel catfish were vaccinated
with either a single or a pair of ORF(s) inserted into an expression vector. Solutions
of DNA containing one or two of the seven CCV ORFs, vector alone, or PBS were
5
injected intramuscularly into 6 to 10 month old catfish [3]. Four to six weeks post
vaccination, the catfish were challenged by immersion with one LD50 of CCV/ml [3].
When compared to other single ORF vaccinations, fish vaccinated with ORFs 59 and 6
alone showed the strongest resistance to CCV challenge, with a 74% survival rate and
a 39% survival rate, respectively. Non-injected, PBS injected, or vector injected fish
showed only a 34% to 56% survival rate [3]. The combination of the vaccine pair ORF
59 and ORF 6 proved to be even more effective against challenge with a 78% survival
rate. The enhanced protection suggests that CCV possesses multiple targets encoding
neutralization antigens [3].
Two weeks after the challenge, surviving fish were bled to assay for neutralizing
antibodies. All fish injected with vaccines, including the control group, showed neu-
tralization activity, although barely detectable in control fish. All of the CCV ORF
vaccinated groups developed an increase in neutralizing antibodies following sublethal
viral challenge, however in fish vaccinated with ORFs 59 and 6, protection was seen and
neutralizing antibodies increased by two-fold.
1.0.4 DNA vaccines induce IFN expression
DNA vaccines encoding viral glycoproteins not only provide long-term specific im-
munity, but have also been found to induce interferon (IFN) expression in rainbow trout
[8]. Kim et al. [8] found that initial protection provided by DNA vaccines is due to IFN
type I expression.
Previous experiments [17] showed that injection of a plasmid expressing a viral
glycoprotein to infectious hematopoietic necrosis virus (IHNV) provided protection, al-
though it was unclear whether this protection was due to a specific immune response.
6
The first appearance of the anti-glycoprotein antibody was around 8 weeks post vaccina-
tion, yet protection against IHNV challenge was observed as early as 3 to 4 weeks post
vaccination. Also, virus-neutralizing antibodies at low titers only appeared in some fish
around 6 weeks post vaccination. To further understand if glycoprotein-encoding DNA
vaccines induce specific protection, Kim et al. [8] vaccinated fish against heterologous
viruses and challenged the fish with another virus. The DNA vaccines encoded viral gly-
coproteins chosen from three serologically distantly related fish rhabdoviruses, including
IHNV, spring viremia of carp virus (SVCV), and snakehead rhabdovirus (SHRV). The
three different groups of vaccinated fish, plus two control groups injected with either
PBS or an empty DNA plasmid, were all challenged with lethal doses of IHNV 30 days
post vaccination (dpv) and later at 70 dpv. They found that significant heterologous
protection existed against IHNV 30 dpv in fish vaccinated against SVCV and SHRV.
However, this protection did not persist against IHNV challenge at 70 dpv for the same
fish vaccinated against SVCV and SHRV. Kim et al. [8] believe their results suggest that
DNA vaccination with glycoprotein genes induce a potent IFN response in fish producing
heterologous protection [8].
At 30 dpv, the relative percent survival (RPS) of fish challenged with IHNV is 93%
for fish vaccinated against IHNV, 98% for fish vaccinated against SHRV, and 95% for
fish vaccinated against SVCV [8]. When the control groups were challenged 30 dpv, they
had a cumulative percent mortality (CPM) of 55% for PBS injected fish, and 57% for
plasmid-injected fish. At 70 dpv, the RPS value of fish challenged with IHNV was only
26% for fish vaccinated with SHRV and 17% for fish vaccinated with SVCV, whereas
IHNV vaccinated fish had a RPS value of 87%. Again, the control groups demonstrated
7
little protection; PBS injected fish had a CPM value of 96% and the plasmid injected
fish had a CPM value of 91%.
To confirm the induction of an IFN response, Kim et al. [8] measured Mx protein
production in the fish from day 0 (30 dpv) to 7 days post challenge (dpc) with IHNV. In
the fish that received vaccines, increased Mx expression was observed before challenge
with IHNV. After challenge, Mx expression in fish vaccinated against IHNV ceased by
7 dpc, and Mx expression in fish vaccinated with SHRV also decreased over 7 dpc (al-
though did not cease), while Mx expression in fish vaccinated against SVCV showed
no change over the 7 dpc. In control fish samples, Mx expression was not observed at
day 0. However, post challenge, Mx expression increased so that by day 7 there was
strong expression of Mx proteins in the control fish [8]. Thus, Kim et al. [8] concluded
that decreased Mx production upon viral challenge correlates with the specificity of the
glycoprotein in the vaccine. They postulated that this could be due to the removal of
the glycoprotein-expressing cells by specific immunity upon viral challenge [8] or it could
be due to removal of the virus.
Kim et al. [8] concluded that the fish vaccinated against IHNV, SVCV, and SHRV
were protected against IHNV challenge at 30 dpv due to induction of IFN-?/?. This
conclusion was based on the fact that the protection SVCV and SHRV vaccinated fish
had at 30 dpv was no longer present at 70 dpv. Kim et al. [8] believed that at 30
dpv IFN-? provided early and effective protection against virus infection, yet over time,
the antiviral state provided by IFN-? diminished, as a result of decreasing glycoprotein
production or as a result of some other down regulation event [8]. As the general antiviral
state lessened, the long-term specific protection provided by the DNA vaccine became
8
the important player in immunity, so that by 70 dpv only the IHNV vaccinated fish
maintained strong protection against IHNV challenge.
Other experiments (referenced unpublished data by Kim CH in [8]) helped to confirm
their conclusion that IFN-? is the source of protection in glyprotein DNA vaccinated fish.
In this study, (referenced unpublished data by Kim CH in [8]) when fish were injected
with the DNA vaccine for IHNV, IFN-inducible Mx protein expression was found in the
liver and kidney tissues of the fish. However, when fish were injected with formalin-
killed vaccines, subunit IHNV vaccines, or vaccine controls, no Mx protein expression
was found. Kim et al. [8] considered the possibility that the production of the Mx
protein itself provided nonspecific protection for the heterologous vaccines at 30 dpv.
However, cell culture experiments showed Kim et al. [8] that the over-expression of Mx
proteins did not inhibit replication of IHNV, leaving Kim et al. [8] to conclude that IFN,
not Mx proteins provided nonspecific protection.
Similarly, in their work with DNA vaccines in rainbow trout against viral hem-
orrhagic septicemia virus (VHSV) and infectious hemorrhagic necrosis virus (IHNV),
McLauchlan et al. [12] showed that these vaccines not only stimulated antibody produc-
tion in the fish, but the vaccines also induced the expression of Mx proteins in the muscle
tissues of the fish at the site of injection of the vaccine. The control DNA plasmid, which
lacked the viral glycoprotein gene insert, did not induce any of these responses in the
fish. Protection against VHSV was found as early as 1 week after the DNA vaccina-
tion was administered even though antibodies to VHSV were not detected until 4 weeks
post-vaccination. Protection was consistent with Mx protein expression. DNA vaccines
seemed to induce a rapid protection against viruses mediated by the innate nonspecific
9
IFN responses, while long-term protection is mediated by the specific immune response
[12].
10
Chapter 2
Literature Review
2.1 Introduction to fish immunity
Fish, like homeothermic vertebrates, have both innate and specific immunity. The
innate defense mechanisms are nonspecific and can be pre-existing and/or inducible.
This nonspecific innate defense system provides protection that prevents the attach-
ment, invasion or replication of a pathogen on or in the tissue of fish. Some even believe
fish depend more on their innate defense mechanisms because innate responses are es-
sentially temperature independent [2] [20]. These characteristics are very important for
ectothermic vertebrates because the specific immune defenses take more time to respond
and are temperature dependent (reviewed in [2]). For channel catfish, a species liv-
ing in a temperate climate, optimal antibody production takes 2 to 4 weeks, yet many
pathogens can kill a fish within a few days of infection. For this reason it has been
suggested that protection offered by specific immune responses is only more important
for survival than innate responses in previously immunized fish [2]. Because fish are
always in close contact with their environment, which can contain high concentrations
of bacteria or viruses, it is possible that innate defenses are more important in fish than
in endothermic vertebrates [2]. For this reason there have been several studies looking
at non-specific protection in fish. Having a better understanding of the innate immune
system in fish not only helps researchers find different ways to control fish diseases, but
also provides insight into how the adaptive immune system evolved.
11
2.2 Introduction to IFN
Infection of cells with viruses induces the expression of a family of proteins that are
known as interferons (IFNs), because they interfere with viral replication [21]. IFN was
discovered and named by Isaacs and Lindenmann [21] as an antiviral factor secreted by
chick cells after treatment with heat-inactivated influenza virus. Interferons induce an
antiviral state in vertebrate cells and therefore play a major role in the defense against
viral infections [22]. The first IFN genes cloned were human IFN-? and IFN-? in 1980
[23] [24]. IFN-like activity was detected in fish as early as 1965 [25], and has since been
detected in number of fish species after viral infection or treatment with double-stranded
RNA or poly I:C [26] [8] [27] [28], but fish IFN genes were not cloned until recently [29]
[30] [31] [32].
The three families of IFNs (type I and type II and IFN-III) can be distinguished on
the basis of gene sequences, protein structure, and functional properties [22]. Although
all three IFN types belong to the class II ?-helical cytokine (HC) family, they have
different 3-dimentional structures and bind to different receptors. Type I IFNs, including
the classical IFN-?/?s, are antiviral effector molecules induced by viruses in most cells,
encoded by intron-lacking genes, and bind to the single type I IFN receptor composed
of two chains, IFNAR1 and IFNAR2. IFN-?, which belongs to the type II IFN family,
although not directly induced by viral infection, is produced by natural killer (NK)
cells and T lymphocytes in response to interleukin-12 (IL-12), IL-18, and mitogenic and
antigenic stimuli, and binds to a receptor composed of two chains, IFNGR1 and IFNGR2
(reviewed in [33] [22]). IFN-?s are quite distinct from type I IFNs and have a more
important role in the adaptive immune response to intracellular pathogens (reviewed in
12
[22]). The human type III IFNs or IFN-?s share the same biological properties as type
I IFNs, although they possess lower antiviral activity than IFN-?/?. Also, unlike type
I IFNs, IFN-?s are encoded by intron-containing genes and bind to a receptor that is
distinct from the type I IFN receptor [34]. This receptor consists of subunits IFN-?R1
and IFN-10R2 [34].
2.2.1 Type I IFN antiviral pathway in higher vertebrates
Mammalian type I IFNs make up a multigene family with at least 8 subclasses, IFN-
?, IFN-?, IFN-?, IFN-?, IFN-?, IFN-epsilon1, IFN-?, and IFN-? (limitin) (reviewed in [35] [36]).
In humans, there are 13 IFN-? genes, 1 IFN-? gene, and 1 IFN-? gene. Mononuclear
phagocytes are the major source of IFN-? production, and IFN-? is produced by many
cells, such as fibroblasts, epithelial cells, endothelial cells, lymphoid cells and astrocytes
(reviewed in [35]). IFN-?, IFN-? and IFN-? are expressed in several different species
whereas the other type I IFNs have more exclusive expression. IFN-? is only expressed
in ruminant species during a specific stage in pregnancy [37]. IFN-? is only expressed by
human keratinocytes [38]. IFN-?, also known as limitin, is expressed only in murine cells
and shares some homology with IFN-?, IFN-?, and IFN-? [39]. IFN-? is only produced
by the trophoblasts of pigs [40].
Mammalian type I IFN vary in size (143 to 172 amino acids) and number of disulfide
bonds (0 to 2), but crystal structures show that mammalian type I IFN share a common
three-dimensional structure composed of 5 ?-helices [41] [42] [43] [44]. Type I IFN genes
encode secretory signal peptide sequences that are proteolytically cleaved before secretion
from the cell (reviewed by [35]).
13
The most potent stimulator of type I IFN production is viral infection; specifically, it
is the double-stranded RNA produced by viruses during their replication in infected cells
(reviewed in [22]). Other microbial products such as glycoproteins can stimulate IFN
production (reviewed in [1]). All type I IFNs bind to the same receptor containing the
IFNAR-1 and IFNAR-2 subunits even though all the different type I IFNs are not struc-
turally the same [35]. When IFNs bind to their receptors, a rapid and direct signaling
pathway, known as the JAK-STAT signaling pathway, effects activation and deactiva-
tion of gene expression in the nucleus. Inactive enzymes called Janus kinases (JAKs)
are loosely attached to the cytoplasmic domains of IFN receptors. When IFN binds its
receptor, the two IFN receptor molecules are brought together and this dimerization of
the receptor molecules activates the IFN receptor, bringing the associated cytoplasmic
kinases, JAK1 and Tyk2, within close proximity of each other so that they are able to
trans-phosphorylate each other (reviewed in [45]).
The activated receptor-associated JAKs phosphorylate tyrosine residues in the cyto-
plasmic portion of the clustered receptors, and the phophorylated tyrosines from the IFN
receptors recognize and bind to the Src homology 2 (SH2) domains located in monomeric
cytosolic transcription factor proteins called signal transducers and activators of tran-
scription (STATs). When the STAT proteins attach themselves to the IFN receptors,
they become phosphorylated by the receptor-associated JAK kinases. The SH2 domain
of one STAT protein is able to bind to the phosphotyrosine residues of another STAT
protein to form a dimer. Mostly STAT1 and STAT2 proteins, although to a lesser de-
gree, STAT3 and STAT4 are phosphorylated. STAT1 can homodimerize, although it is
mostly STAT1-STAT2 heterodimers that form. When two STAT proteins form a dimer,
they dissociate from the interferon receptor and migrate to the nucleus where they bind
14
to IFN?s promoter/enhancer regions and activate gene transcription (reviewed in [45]).
Oddly enough, STAT2 cannot bind directly to DNA, so STAT1-STAT2 heterodimer com-
plexes are recruited to DNA sequences through interaction with a DNA binding protein
called interferon regulatory factor (IRF)-9. The three proteins form a multimeric tran-
scription factor called ISGF3, which binds to interferon-stimulated response elements
(ISRE) in the promoter of interferon-stimulated genes and activates their transcription
(reviewed by [46][45]).
IFN type I induction in mammals includes the expression of close to one hundred
genes and some of these gene products block transcription of viral RNA or DNA and
viral replication by either activating or inhibiting protein synthesis and by degrading
mRNA in virus-infected cells. The genes from three families of IFN-inducible genes that
are best studied participate in inhibiting viral replication. They include: protein kinase
(PKR), the 2prime,5prime-oligoadenylate synthetases (OAS), and the Mx proteins (reviewed by
[45]).
PKR, when activated by viral double-stranded RNA (dsRNA) which it binds during
viral infection, phosphorylates and inactivates the eukaryotic protein synthesis initiation
factor eIF-2?, which leads to a general inhibition of protein synthesis. Thus, activation
of PKR in virus-infected cells leads to the inhibition of protein expression and viral
replication (reviewed in [45]).
OAS, also activated by viral dsRNA, catalyzes the synthesis of 2prime,5prime-oligoadenylates,
which in turn bind to inactive monomeric RNase L, which induces dimerization and
activation of RNase L. The activated RNase L degrades both viral mRNA and rRNA in
the cytoplasm of the cell, which leads to the inhibition of protein expression (reviewed
in [45]).
15
Mx proteins, only induced by type I IFN, (reviewed in [22]) are homologous to
dynamins and can be located in the cytoplasm or nucleus. Not all Mx proteins have
antiviral activity, although some Mx proteins when expressed in the cell offer some
resistance to infection by several RNA viruses. The antiviral mechanism of Mx proteins
is still not completely understood, although it is believed that Mx proteins interfere with
trafficking and/or the transcriptional activity of the viral ribonucleoprotein complexes
(reviewd in [47]).
2.2.2 Type II IFN function in higher vertebrates
Type II IFN, or IFN-?, is also called immune IFN because unlike type I IFNs,
IFN-? is produced exclusively by cells of the immune system. IFN-? plays a major role
in adaptive cell-mediated immune responses because it is produced by CD4+ T helper
(TH1) cells, and CD8+ cytotoxic T lymphocytes (CTLs) in response to MHC-presented
antigens. IFN-? also plays a role in innate responses when mononuclear phagocytes
and/or antigen presenting cells are infected with intracellular pathogens and release IL-
12 and/or IL-18 stimulating NK cells to produce IFN-?. (reviewed in [22]).
IFN-? binds to a different receptor from type I IFNs and mediates signaling through
a different, but overlapping JAK-STAT pathway. IFN-? has some antiviral activity by
inducing in most cells PKR and 2prime,5prime-oligoadenylate synthetases, yet IFN-? is not a
potent antiviral interferon, and it functions mainly as an effector cytokine in immune
responses (reviewed in [22]). Another difference between type II IFN and type I IFN is
that mammalian and avian IFN-? are encoded by a single gene containing four exons
and three introns and show no sequence similarity to type I IFNs.
16
There are several unique functions of IFN-? that are important in cell-mediated
immunity against intracellular pathogens. One function of IFN-? is that it is the
macrophage-activating cytokine. This means that IFN-? provides the means by which
T lymphocytes and NK cells activate macrophages to kill phagocytosed microbes. IFN-
? enhances the microbicidal function of macrophages by stimulating the synthesis of
reactive oxygen intermediates and nitric oxide (reviewed in [22]).
Another action of IFN-? is that it stimulates expression of class I and class II MHC
molecules and costimulators on APCs. IFN-? also stimulates the synthesis of many
proteins involved in antigen processing. Thus, IFN-? boosts MHC-associated antigen
presentation and strengthens the recognition phase of the immune response by enhancing
the expression of the ligands that T cells recognize (reviewed in [22]).
An additional function of IFN-? is that it promotes the differentiation of na??ve
CD4+ T cells to the TH1 subset and inhibits the proliferation of TH2 cells. IFN-? also
maintains TH1 effector cells by activating mononuclear phagocytes to produce IL-12,
which in turn will produce more IFN-?, the major TH1-inducing cytokine (reviewed in
[22]).
2.3 Antiviral pathways in fish
IFN genes have been recently cloned from different fish species including channel
catfish [29], zebrafish [30], Atlantic salmon [32], the Japanese pufferfish Takifugu rubripes
(Fugu), and the spotted green pufferfish Tetraodon nigrovirides [31]. Only one functional
IFN gene has been cloned from zebrafish and pufferfish genomes. Two functional genes
were found in Atlantic salmon and catfish, and it has been suggested that additional
17
IFN genes and pseudogenes exist in the Atlantic salmon genome [32], and at least 2
pseudogenes have already been found in the catfish genome.
The IFN gene for zebrafish (zfIFN) was cloned and characterized by Altmann et al.
[30] by searching the Zebrafish Information Network EST database using BLAST with
a gene for chicken IFN. Only one EST containing all but the 5prime end of the zfIFN gene
was found. Upstream RACE-PCR was used to expand the zfIFN gene.
Robertsen et al. [32] cloned two type I IFN cDNAs from Atlantic salmon TO
cells which originate from Atlantic salmon head kidneys. They induced IFN mRNA by
exposing Atlantic TO cells to poly I:C and used suppressive subtractive hybridization
(SSH) techniques to generate cDNA from the cells. Using the subtracted cDNA as a
template, and a degenerate forward primer from the conserved vertebrate type I IFN
sequence motif YSACAW at the C-terminus, PCR amplification of the IFN-like cDNA
was accomplished. These sequences allowed design of a specific gene primer that could
be used to amplify the 5prime-end of the Atlantic salmon IFN by RACE cloning. Two 5prime-end
IFN products were obtained using cDNA from the head kidney of poly I:C stimulated
fish. Based on these new sequences, new primers were designed to amplify the whole
ORF of both genes by 3prime RACE cloning. Two putative IFN cDNAs were amplified, one
829 bp and the other 1290 bp. The two clones were called SasaIFN-?1 (829 bp) and
SasaIFN-?2 (1290 bp). However, there is a possible third salmon IFN gene yet to be
cloned [32].
Lutfalla et al. [31] used a strategy to find pufferfish IFN gene based on conserved
amino acid sequence and gene structure to identify class II helical cytokines (HCII)
and their receptors (HCRII) in the genome of pufferfish, Tetraodon nigroviridis. The
18
helical cytokine family includes interferons, most interleukins, LIF, CNTF, GCSF, GM-
CSF, and thrombopoietin. These helical cytokines have no similarities at the level of
primary amino acid sequences, but they are all structured around a similar four alpha
helix bundle, and therefore share a common 3-D structure (as reviewed in [31]). The T.
nigroviridis genome was searched for exons capable of coding for molecules structurally
related to the IFNs and the IL10 related cytokines [31]. IFNs and IL-10 are not encoded
by genes that share similar intron/exon structures; type I IFN genes do not contain
introns, whereas type II IFN genes contain three introns, and IFN-? and IL-10 related
genes contain four common introns. Thus, the intron/exon structure to identify IFN
and IL-10 cytokine homologs in distant species could not be used [31]. Their strategy
identified three IL-10 related cytokines genes and a single IFN gene [31]. The Fugu
genomic data was also examined and a Fugu IFN gene was identified [31].
To clone a channel catfish IFN cDNA, Long et al. [29] created an EST library and
a single clone with sequence homology to rat and mouse IFN ? was identified. However,
this IFN-1 gene did not contain a classical signal sequence within the first 70 amino acids
of the catfish IFN-1 amino terminus [29]. Since this gene provided protection against
viral challenge, it was assumed IFN-1 was a valid expressed gene, not a pseudogene, and
that the missing signal sequence was not important for function in catfish [29]. However,
it has since been determined that without the expressed signal sequence, the IFN-1
channel catfish gene is in fact an expressed pseudogene and cDNAs representing three
additional IFN genes have been cloned from channel catfish, where two of these genes,
IFN-2 and IFN-4, are functional IFN genes including signal sequences [48].
19
2.3.1 Characterization of fish IFNs
All of the fish IFN genes translate into putative precursor proteins that contain
between 175 to 194 amino acids and have signal peptides of 22 to 24 amino acids. The
zfIFN gene codes for a protein that is 185 amino acids in length, with the first 22 amino
acids representing a putative signal sequence [30]. Both translated Atlantic salmon IFN-
? clones translate into 175-amino acid peptides that shows 95% amino acid sequence
identity and 98% nucleotide sequence identity to each other. Both Atlantic salmon
IFN-? genes encode a hydrophobic 23 amino acid signal sequence at the N-terminus
so that the mature Atlantic salmon IFNs contain 152 amino acids. When the Atlantic
salmon IFN-?1 precursor protein was aligned with 24 different type I IFNs from other
vertebrates, Robertsen et al [32] found that the signal sequence from the salmon IFN
gene showed hardly any homology with the signal sequences from higher vertebrates (0-
9%), with the exception of pig IFN-? showing 19% identity to Atlantic salmon IFN-?1
[32]. However, it was demonstrated that the signal sequences of fish IFNs display some
homology amongst themselves (around 32% identity) [32].
The properties of putative mature fish IFNs were compared and it was found that
fish IFNs are similar in size (between 152 to 170 amino acids) and similar to mammalian
IFNs [49]. Catfish IFN-2 is the largest fish IFN with 170 amino acids [48] while Atlantic
salmon IFN-?1 is the smallest with 152 amino acids [32].
The isoelectric points of catfish, zebrafish, and Atlantic salmon IFNs are basic [48],
[30], [32]. However, Fugu and pufferfish IFNs both have acidic isoelectric points [31]. This
difference in isoelectric points is not uncommon, as type I IFNs of higher vertebrates
also show large differences in isoelectric points (reviewed in [49]).
20
Potential N-linked glycosylation sites were found in all fish IFNs except those of
zebrafish [30]. The number of glycosylation sites varies between fish species with salmon
and catfish IFN-1 genes encoding 1 [32] [29], catfish IFN-4 containing 2 [48], catfish
IFN-2 and pufferfish IFN containing 3 [48] [31], and Fugu IFN containing 4 potential
glycosylation sites [31]. Again, it is not uncommon for higher vertebrates to have a
varied number of N-linked glycosylation sites between type I IFNs, from having none
in most IFN-? proteins, to 3 in mouse IFN-? (reviewed in [49]). Salmon IFN contains
its N-glycosylation motif at amino acid 124, whereas channel catfish IFN-1 contains its
glycosylation site at amino acid positions 27-30, which is characteristic of mammalian
IFN ? [29]. Thus, it was argued that glycosylation does not appear to be conserved
through evolution [49].
The salmon, pufferfish, Fugu and zebrafish all contain 2 cysteines in their mature
putative IFNs [32] [31][30], while all catfish IFN proteins contain 3 cysteines [48]. Based
on the number of cysteines, the mature fish IFN proteins seem to show similarities to
IFN-alpha proteins from higher vertebrates.For example, the putative mature salmon
IFNs have a cysteine as the first amino acid, which is similar to mammalian IFN-?. The
two cysteine residues in zebrafish IFN are conserved among all IFN sequences except
mammalian IFN-? [42] [43] [44]. However, mammalian and avian IFN-? genes generally
code for four or more cysteine residues and the encoded proteins may form 2 or more
disulfide bonds, whereas all the fish IFNs only have the potential to form 1 disulfide
bond, which is similar to mammalian IFN-? [29] [48] [49]. Robertsen et al. [32] believe
this demonstrates that the ancestral vertebrate IFN possessed only one pair of cysteines
and that the other pair appeared at a later stage, but before the divergence of the birds.
21
All the cloned IFN genes from the different fish species contain five exons and four
phase 0 introns [48] [30] [32] [31], as opposed to mammalian and avian type I IFN genes
which do not contain introns (reviewed in [32] [49]). However, it is not uncommon for
genes to lose introns in vertebrate evolution [18]. The fish type I IFNs have the same
exon/intron structure as mammalian IL-10 and IFN-? genes. This suggests that possibly
type I IFN genes and IL-10 genes evolved from a common ancestor [31]. In fact, the
three IL-10 related cytokines discovered in pufferfish by Lutfalla et al. [31] also each
had four phase 0 introns. Long et al. [48] also showed that the locations of the second
and third introns in teleost IFN genes are the same as those of mammalian IFN-?.
When Lutfalla et al. [31] aligned their T. nigroviridis IFN and zebrafish IFN genes with
different mammalian type I IFNs and human IFN lambdas, they found that both fish
IFNs were more closely related to human IFN-?s.
Atlantic salmon and zebrafish IFN genes share similar size exons, however the in-
trons, for the most part, are smaller in the salmon than in the zebrafish [32] [30]. It was
also noted that the trend for intron sizes to differ between fish, but exon sizes between
the different fish species were similar, with the only exception of the catfish IFN-1 exon
1, which lacks a signal sequence [48].
Robertsen [49] showed that 17 amino acid positions appear to be conserved among
type I IFNs from fish and higher vertebrates. These conserved amino acids might be
important for the stabilization and/or activity of the IFNs. Leu30, Arg33 and Phe36 are
three conserved residues in human IFN that are thought to be involved in binding to the
IFN-?/?/?/?/?/? group IFN receptor [50]. Leu30 and Arg33 do not appear in the fish
or avian IFNs, while Phe36, shown to be important for biological activity, is conserved
in zebrafish and salmon [30] [49].
22
Although mammalian type I IFNs vary in size and number of disulfide bonds, crystal
structure determinations show that mammalian type I IFNs share a common three-
dimensional structure composed of 5 alpha helices (reviewed in [32]). The helical cytokine
family, which includes interferons, has similarities at the primary amino acid sequence
level and is structured around a similar four alpha helix bundle [42] and, therefore,
share a common 3-D structure. Robertsen et al. [32] analyzed the secondary structure
of salmon IFN and found that salmon IFN has 5 alpha helices which overlap with the
helices of other mammalian type I IFNs. In fact, Robertsen et al. [32] found that the
most conserved amino acids in salmon IFNs appeared to occur in and adjacent to these
5 alpha helices, with the most conserved region of the salmon type I IFN gene being in
the C-terminal-coding region.
2.3.2 Functional studies of type I IFN in fish
For each IFN gene cloned from the different fish species, studies have been performed
to show that they have the same functional properties as mammalian type I IFNs.
Double-stranded RNA poly I:C, which is a well known inducer of type I IFNs in
mammals, induced transcripts of type I IFNs at various times in channel catfish [29],
zebrafish [30], and Atlantic salmon [32]. To determine the extent to which zebrafish IFN
(zfIFN) could be induced by poly I:C in vitro, Altmann et al. [30] treated zebrafish liver
(ZFL) cells with 25 ?g of poly I:C/ml and RNA was extracted 6, 12, 24, 36 and 48 hours
later. Quantitative qPCR results showed zfIFN expression at 6 hours and an increase
in expression by 12 hours. By 24 hours zfIFN expression had ceased. It was found that
there was a low level of constitutive expression at all times [30]. When channel catfish
ovary cells (CCO) were induced with poly I:C, catfish IFN-1 mRNA peaked at 2 hours
23
and declined thereafter [29]. However, when Atlantic salmon head kidney TO cells were
stimulated with poly I:C, IFN expression occurred 12 to 24 hours after poly I:C exposure,
with the highest level of expression occurring at 24 hours post stimulation. It was also
shown that IFN was expressed in salmon [32]. Investigators used northern blot analysis
to show that two IFN transcripts of the correct sizes were expressed in the head kidney
of Atlantic salmon 14 hours after injection of poly I:C at relatively low concentrations.
The highest level of IFN expression was observed 12 hours after injection of poly I:C,
and this expression diminished by 24 and 48 hours post injection. Robertsen et al. [32]
concluded that IFN transcripts were induced by poly I:C more rapidly in live fish than in
TO cells because cell populations in fish, perhaps leukocytes, are more reactive to poly
I:C than the cell cultured head kidney TO cells.
Recombinant IFNs from pufferfish, channel catfish and Atlantic salmon have been
expressed in eukaryotic cells or by in vitro translation [31] [29] [32] to show the protective
activity of the fish IFNs. Mx protein was induced by salmon and pufferfish IFNs [32]
[31]. To study whether the Atlantic salmon IFN-?1 gene encoded a biologically active
IFN, Robertsen et al. [32] subcloned it into a eukaryotic expression vector pCR3.1
in forward and reverse orientation from the cytomegolovirus (CMV) promoter. The
constructs were transfected into human embryonic kidney 293 (HEK293) cells. The
media supernatants from the HEK293 cells were collected 48 hours later, acid treated
over night, and then assayed for antiviral activity against infectious pancreatic necrosis
virus (IPNV) in Atlantic salmon TO cells or in the Chinook salmon CHSE-214 cell line.
Supernatants from HEK293 cells transfected with the SasaIFN-?1 gene in the correct
orientation showed relatively high levels of antiviral activity when tested on TO cells and
24
CHSE-214 cells. The HEK293 cells that were transfected with the Atlantic salmon IFN-
?1 gene in the reverse orientation gave no IFN activity in either cell system. Robertsen
et al. [32] also demonstrated that supernatants of HEK293 cells transfected with the
salmon IFN-?1 gene in the forward orientation induced the Mx protein in the CHSE-214
cells and TO cells.
The biological activity of pufferfish IFN (TnIFN) was tested by assessing expression
of the TnMX gene [31]. The TnIFN ORF was cloned into the same plasmid and used
to produce recombinant TnIFN [31] . The recombinant TnIFN was used to challenge
primary cultures of T. nigroviridis cephalic kidney cells. After a 6 hour exposure to
recombinant TnIFN, total RNA was isolated from cells, and the amount of TnMX mRNA
expressed was measured. When Lutfalla et al. [31] intraperitoneally injected poly I:C
into the T. nigroviridis, a large amount of TnIFN was induced in the testis and kidney
of the pufferfish. This treatment with poly I:C was used as a positive control to induce
TnIFN in the pufferfish. Compared to poly I:C treatment, the recombinant TnIFN
molecule induced a similar level of TnMX mRNA expression in pufferfish. Expression of
the two PKR genes in T. nigroviridis was used to verify the results found with TnMX
expression; both PKR genes were induced by the recombinant TnIFN just like the single
TnMX gene [31].
When zebrafish cells were transfected with the zebrafish IFN gene, they showed
increased resistance against infection by snakehead rhabdovirus and increased Mx gene
expression [30]. Zebrafish ZF4 cells, chosen for their ability to form plaques when infected
with snakehead rhabdovirus (SHRV), were transfected with zfIFN and 16 hours later
were exposed to SHRV for 1 hour. Altmann et al. [30] were able to show that zfIFN can
protect ZF4 cells against viral infection when the gene is transfected 16 hours prior to
25
viral challenge. There was a 31% reduction in plaque number in zfIFN-transfected cells
compared to untransfected cells and a 36% reduction compared to cells transfected with
the control pcDNA3 vector. An experiment was performed to determine if the zfIFN gene
was capable of inducing the Mx gene [30]. This was done by cotransfections of ZF4 cells
using the Mx promoter-pGL3 luciferase construct and the zfIFN gene. Expression from
the Mx promoter induced by IFN expression is necessary to drive luciferase expression.
ZF4 cells transfected with Mx promoter construct alone were induced with poly I:C as a
positive control. The negative control was an Mx promoter construct cotransfected with
control pcDNA3 vector to show that random DNA does not induce the Mx gene. ZF4
cells cotransfected with zfIFN and the Mx promoter construct had greatest luciferase
activity at 5.6 X 103 relative light units (RLU), while poly I:C induction yielded an
average of 4.6 X 103 RLU [30].
The catfish and salmon IFNs demonstrated antiviral activity against channel catfish
virus (CCV) and infectious pancreatic necrosis virus (IPNV), respectively [29] [32].
2.3.3 Alignment analysis of fish IFNs
Using alignment analysis, it was demonstrated that a low homology between the fish
IFN amino acid sequences and IFN sequences from higher vertebrates existed [49], [48],
[32], [31], [30], [29]. It was also noted that fish IFNs, when compared to higher vertebrate
IFN, showed better homology with IFNs from different fish species than vertebrates. For
example, the catfish IFN-1 sequence showed only 28% identity and 41% similarity to rat
IFN ? and 30% identity and 52% similarity to mouse IFN ?, whereas the catfish IFN-1
gene showed 39% identity and 55% similarity to the zebrafish IFN gene. The different
fish IFN genes were aligned and percent identity was calculated [49]. It was found that
26
IFN sequences from catfish, zebrafish, and Atlantic salmon are most similar, showing
43-46% identity, whereas the pufferfish and Fugu IFNs showed 23-35% identity with
the other fish IFNs. The highest percent identity/similarities are shown between IFN
gene sequences from the same fish species. The fugu (T. rubripes) and pufferfish (T.
nigroviridis) IFNs are more similar to each other (67.6% similar) and less similar to the
other fish IFNs, and they share the least similarity with catfish IFNs (between 15.3 and
17.6% similarity) [48]. The two Atlantic salmon IFNs have 97% sequence identity [32]
and the catfish IFNs show between 66.9 to 81.6% similarity [48]. Interestingly, zebrafish
and goldfish IFNs share 66.1% similarity [48].
When pairwise comparisons of catfish IFN proteins with the different mammalian
and avian IFNs was performed, the highest similarity was found to be with mammalian
IFN-? at 12-19% similarity [48]. The other comparisons showed even lower similarity,
with catfish having 10-14% similarity to mammalian IFN-?, 14-15% similarity to mam-
malian IFN-?, 12-14% similarity to mammalian IFN-?, and 12-16% similarity to avian
IFN-?/? [48]. Similarly, the mature salmon IFN protein showed highest sequence iden-
tity with IFN-? sequences and lowest identity with mouse and human IFN-? [32]. On
further analysis of the alignment of catfish IFNs with other fish IFNs, it was found that
there are a limited number of sections of amino acid conservation, i.e. five, were identi-
fied among the six fish species [48]. Outside these conserved areas, amino acid identities
are scarce. When the analysis was expanded to include mammalian and avian IFN genes
or proteins, the highest similarity was found to be to mammalian IFN-? genes, although
the identity/similarity was still very low [48]. Interestingly, IFN-?s from different mam-
mals show about the same degree of homology as that shown between the salmon and
zebrafish [32].
27
Since teleost IFN genes share a similar intron/exon structure with IFN-?, it has been
suggested that fish IFN and IL-10 genes evolved from a common ancestor gene that also
had four phase 0 introns, similar to IFN-? [31]. Furthermore, it has been suggested that
all class II helical cytokine (HCII) ligands and receptors share a common ancestor [31].
The first duplication (creating ancestral IL-10 related genes and ancestral IFN) would
have taken place before the osteichthyes (bony fish) radiation [31]. Lambda IFNs seems to
have evolved from ancestral IFN, while type I IFNs evolved in a retroposition event that
occurred during vertebrate evolution [31]. Thus, the key event in evolution of the IFN
gene was a retroposition event that occurred after separation of sarcopterygians (lobe
finned fish) from actinopterygians (ray finned fish) and this event created an intronless
type I IFN gene [31]. In the mammalian lineage, this gene then underwent successive
duplications to generate the alpha and the beta interferons. Later, during the mammalian
radiation, numerous alpha IFN genes were generated (reviewed in [51]).
Whether IFN-? genes truly represent ancestral IFN genes is uncertain because as
has been pointed out, IFN-? shares very little sequence similarity to the type I IFNs of
fish or higher vertebrates [48], [49]. Even with the structure similarity between teleost
IFNs and IFN-?, phylogenetic trees suggests that teleost IFNs are most closely related
to mammalian ? and ? IFNs than the other IFNs [48]. The fish IFN genes could be
derived from a primordial gene that predated the divergence of the teleost and tetrapod
lineages [48]. Perhaps the teleost IFN genes failed to expand or diversify, whereas after
the divergence of tetrapods from bony fish, their IFN genes underwent major duplication
and diversification, evolving into the three main classes of IFN genes observed today in
mammals. Since the duplication that gave rise to the IFN ?/? gene families happened
28
after the divergence of teleosts and mammals, it is possible that fish IFN genes have
both IFN ? and ?-like properties [48].
The rather low level of similarity seen between fish IFNs may be due to the increased
separation time between existing fish species and their last common ancestor [48]. The
high percent similarity between zebrafish and goldfish (66.1%), which are members of the
same taxonomic order, Cypriniformes, supports this idea that the low level of similarity
seen between the other fish IFNs from different orders is due to increased time between
these species and their last common ancestor. The low level of similarity between fish
IFNs might otherwise be due to the possibility that the identified fish IFNs from the
different species may not be members of the same subfamily (e.g., IFN-? and IFN-?)
[48]. In mammals there are differences among type I IFN classes. For example, humans
and mice have multiple IFN-? genes and only a single IFN-? gene, whereas pigs, horses,
and cows have at least five interrelated IFN-? genes (reviewed in [48]). Within the same
species, subclasses can be very different. For example, in humans, IFN-? genes only
share 50% amino acid sequence identity, whereas IFN-?s only share 22% identity with
human IFN-? and 37% identity with human IFN-? [35]. It has been suggested that 300
million years ago, when the reptile/bird/mammal split occurred, the common ancestor
genome probably contained a single class of type I IFN (reviewed in [35]). The IFN-
?/? gene families are suspected to be the result of gene duplication in the mammalian
ancestor 250 million years ago, which is supported by the presence of distinct clusters
of IFN-? and IFN-? genes within different species (reviewed in [35]). The low level of
sequence identity between fish and mammalian IFN gene products further suggests the
emergence of different type I IFN gene classes [48]. The low level of sequence identity
between fish and mammals suggests that catfish IFNs are members of a novel IFN class
29
not represented in mammals, or possibly this low level of sequence identity is due to the
long separation between mammals and teleost species and their last common ancestor
[48].
2.3.4 Function of type II IFN in fish
IFN-? genes have recently been cloned from Fugu (pufferfish) [52], zebrafish (Gen-
Bank accession no. AB158361) and Atlantic salmon (GenBank accession no. AY795563).
Fish IFN-? genes, like the IFN-? genes of higher vertebrates, all contain three in-
trons and four exons [52], (GenBank accession no. AB158361), (GenBank accession
no. AY795563), [49]. The fish IFN-? precursors contain 180 to 189 amino acids and
appear to contain signal peptides of 23 to 24 amino acids so that the mature fish IFN-?s
contain 156 to 167 amino acids. As expected, fish IFN-? shows no sequence similarity
to fish type I IFNs. Zou et al. [52] believe the Fugu IFN-? gene to be an ortholog
to mammalian and avian IFN-? for three reasons. First, comparative analysis showed
homology using BLAST analysis and 3D modeling, second, the Fugu IFN-? gene has the
same genomic structure as other known IFN-? molecules, and finally, the synteny of the
IFN-? gene cluster is conserved between humans and Fugu. Zou et al. [52] need further
data to determine the cellular sources and biological activity of Fugu IFN-? in terms of
regulating cell mediated immunity in fish.
Robertsen [49] aligned the fish IFN-? sequences from zebrafish, salmon and Fugu
with mammalian and avian IFN-?, and found that fish IFN-? sequences displayed very
low sequence identity with mammalian and avian IFN-? at 17 to 31%. In fact, the
fish IFN-? sequences showed less identity between themselves at 21 to 33% than the
30
mammalian IFN-? sequences showed between themselves at 53 to 78% [49]. The diver-
sification rate of fish IFN-?s had just as high a rate as the diversification of type I IFNs
[49].
Cloning of the T cell receptor (TCR), MHC class I and II genes from fish suggests
that fish probably possess T cell mediated adaptive immune responses similar to higher
vertebrates (reviewed in [53]). Recently, the TCR co-receptor CD8? gene was cloned
from rainbow trout and the TCR co-receptor CD4 gene was cloned from Fugu, implying
the presence of both CTLs and TH1 cells in teleost fish [54] [55]. In fact, Nakanishi et
al. [56] demonstrated through functional studies that rainbow trout possess CTLs [56],
while it was demonstrated that catfish possess NK-like effector cells that are capable of
killing virus-infected cells [7]. The homologues of the IFN-? inducing cytokines, IL-12
and IL-18 have been cloned from Fugu and rainbow trout respectively [57] [58]. Thus,
teleost fish seem to contain the major IFN-? producing cell types and cytokines found in
mammals. Although functional studies of cloned fish IFN-? have yet to be carried out,
previous work in fish has shown that rainbow trout leukocytes produce a macrophage
activating factor with IFN-?-like properties [59], and a gene of an IFN-? induced enzyme,
nitric oxide synthase (iNOS), has also been cloned from rainbow trout [60].
2.3.5 JAK-STAT pathway and members in fish
Although less is known about the IFN-signaling system in fish than in mammals,
several mammalian JAK and STAT homologues have been found in fish. JAK1 cDNA has
been cloned from carp [61] and from zebrafish [62], the JAK1 gene has been cloned from
round-spotted pufferfish Tetraodon fluviatilis [63]. Leu et al. [64] cloned JAK2, JAK3
and TYK2 genes from the round-spotted pufferfish T. fluviatilis and believe that all four
31
genes are derived from a common ancestor gene. Amino acid sequence comparison of the
four round-spotted pufferfish JAKs with carp and zebrafish show that JAK1 is the most
conserved between the fish species with 67.9% and 70.2% sequence identity, respectively
[64]. Round-spotted pufferfish JAK1 showed 57% amino acid identity to human and
murine JAK1, and carp showed high homology to mammalian JAK1 in both the kinase-
like (JH2) and kinase (JH1) domains. Similarly, the pufferfish JAK2 displays high amino
acid identity at 66.8% with mouse JAK2 [64]. However, the round-spotted pufferfish
JAK3 and TYK2 show lower amino acid identity with their mammalian counterparts
than JAK1 and JAK2 suggesting that JAK1 and JAK2 are more conserved between fish
and mammals than TYK2 and JAK3 [64]. Leu et al. [63] also found a putative STAT-
binding sequence in the JAK1 gene. STAT1 and STAT3 genes have been cloned from
zebrafish [65], STAT5 gene has been cloned from round-spotted pufferfish T. fluviatilis
[66] and STAT1 cDNA has been cloned from crucian carp [67]. Oates et al. [65] showed
that zebrafish STAT1 protein can rescue type I IFN-signaling functions in a STAT1-
deficient human cell lines, suggesting that cytokine-signaling mechanisms are likely to
be conserved between fish and mammals. Crucian carp STAT1 protein was induced by
poly I:C, grass carp hemorrhagic virus (GCHV), and inactivated GCHV [67]. Zhang
and Gui [67] also showed that carp Mx1 transcriptional activation was under control
of the JAK-STAT pathway in carp, suggesting that the fish IFN signaling transduction
pathway and antiviral mechanisms are similar to that in mammals.
2.3.6 Antiviral pathways induced by IFN in fish
Mx proteins appear to be present in all vertebrates and Mx genes or cDNAs have
been cloned from a number of different fish species including perch [68], rainbow trout
32
[69] [70], Atlantic salmon [71], Atlantic halibut [72], Japanese flounder [73], Fugu [74],
carp [75], channel catfish [76], gilthead sea bream [77], zebrafish [78] [31], and orange-
spotted grouper [79]. As with higher vertebrates, fish Mx is expressed in the cytoplasm
[80] or in the nucleus [81], and fish Mx transcripts and proteins are induced by poly
I:C, IFN, and/or virus infection. Interferon-stimulated response elements (ISRE) were
found in the promoter of rainbow trout [82] and zebrafish [78] Mx, similar to mammalian
Mx, showing that the fish Mx protein is under control of a typical interferon-induced
promoter.
Plant and Thune [76] cloned and sequenced a full-length cDNA of channel catfish Mx
gene using degenerate primers. The initial amplified 600 bp channel catfish fragment was
found to be highly homologous to existing mammalian and fish Mx sequences, although
greatest homology was with other fish Mx genes. A comparison to zebrafish, Atlantic
salmon, Atlantic halibut, rainbow trout, Japanese flounder, Fugu, and fragments of
perch and turbot Mx genes resulted in nucleotide percent identities ranging from 78 to
84% and amino acid percent identities ranging from 72 to 80%. Using RACE PCR to
amplify the full-length catfish Mx mRNA, Plant and Thune [76] were able to analyze the
gene. They found that the amino terminus of the encoded protein contained a tripartite
GTP binding motif and a dynamin family signature. The carboxy terminus contained a
leucine zipper. BLAST analysis of the channel catfish Mx protein compared with other
Mx proteins showed highest level of identity with perch Mx protein (79%), followed by
turbot, Atlantic salmon Mx1 and Mx3, and rainbow trout Mx3, all at 74% identity.
Human MxA and MxB were 53 and 51% identical, respectively. When a parsimony tree
was created comparing all the fish Mx genes, it showed that the zebrafish and channel
catfish Mx proteins are sufficiently different from the other fish Mx proteins to each
33
warrant their own clade [76]. The perch, turbot, Japanese flounder, Atlantic halibut
and Fugu sequences clustered together, while the Atlantic salmon and rainbow trout
sequences clustered together. Interestingly, the Mx protein is more conserved than IFN
in fish. The greatest conservation between the catfish Mx sequence and other fish Mx
sequences was found in the amino terminus. Southern blot analysis [76] showed that
there is a minimum of three Mx genes in channel catfish, which is similar to rainbow
trout and Atlantic salmon, which possess three Mx genes [70] [71] and Atlantic halibut,
which possess two Mx genes [72].
Mx mRNA synthesis was induced in vivo by injecting catfish with either poly I:C
or CCV and mRNA samples were collected 0, 1, 2, 4, 6 and 8 days after injection [76].
Interestingly, whenever an rt-PCR was performed on day 0, all catfish samples were
synthesizing low levels of Mx mRNA [76]. Expression of Mx mRNA in healthy fish was
also shown in Japanese Flounder [73]. This basal level of Mx expression is possibly
associated with low levels of IFN circulating to protect against initial exposure to viral
infection [76]. Atlantic salmon Mx protein was expressed in PBS-injected control fish
as well as non-treated fish verifying that Mx expression was not a reaction to injection
[83]. When catfish were injected with poly I:C they demonstrated the highest level of
Mx mRNA expression on day 1, and this expression decreased slightly on days 2 and 4.
By day 6, Mx mRNA expression had declined, although by day 8 there was some Mx
mRNA expression detected again. Injection with CCV induced a total mortality of 15%
and was found to increase Mx mRNA expression within 24 to 48 hours. The Mx mRNA
expression level decreased after day 2 but mRNA production was maintained through
day 8 (which was the last sampling point). CCO cells were also found to constitutively
produce a low level of the Mx mRNA. Likewise, levels of Mx mRNA were enhanced in
34
CCO cells beyond basal levels in response to both 5 and 50 ?g/ml of poly I:C. Mx mRNA
transcripts showed expression by 24 hours and continued expression by 48 hours [76].
Mx mRNA expression in channel catfish in response to CCV occurred earlier and
lasted longer than noted in any other fish species. For example, Mx mRNA could be
detected from the first sampling point at day 2 to day 4 in rainbow trout infected
with infectious hematopoietic necrosis virus (IHNV) [69] and likewise, Atlantic halibut
infected with infectious pancreatic necrosis virus (IPNV) displayed a low level of Mx
mRNA at 2 and 4 days post infection, yet by day 7 expression was not seen [72]. Similarly,
Japanese flounder infected with hirame rhabdovirus (HRV) initially showed Mx mRNA
expression at 2 days and this expression peaked by day 3 [73]. The prolonged in vivo
expression of Mx mRNA in response to CCV may indicate that the catfish Mx protein
has antiviral activity, although Plant and Thune [76] were not able to show the exact
role of the Mx protein in catfish.
In fact, until recently, most Mx protein studies in fish were not able to demonstrate
the antiviral activity of Mx proteins. For example, the effect of Atlantic salmon (AS)-IFN
(the supernatants with IFN-like activity induced by poly I:C) and poly I:C on the Mx
protein expression and antiviral activity against infectious salmon anaemia virus (ISAV)
in the Atlantic salmon cell lines TO and SHK-1 was measured, in order to see if there
was a correlation between Mx protein expression and virus inhibition [84]. This analysis
did not show that Mx protein production in Atlantic salmon had antiviral properties
[84]. Similarly, rainbow trout Mx proteins do not appear to inhibit replication of the
rhabdovirus IHNV [70].
Only recently has antiviral activity of Mx proteins in fish been demonstrated. Know-
ing that a correlation between inhibition of IPNV and Mx expression in IFN-stimulated
35
salmon cells existed and had been demonstrated [28], the antiviral action of Atlantic
salmon Mx proteins against replication of IPNV was uncovered [80]. To demonstrate
antiviral activity of Mx, a Chinook salmon embryo (CHSE-214) cell line was established
that expresses Atlantic salmon Mx1 (ASMx1) constitutively [80]. The CHSE-214 cell line
was defective in producing IFN-like activity. The ASMx1-transfected CHSE-214 cells ex-
pressing ASMx1 showed a severely reduced IPNV-induced cytopathic effect, which was
supported by a 500-fold decrease in virus yield. The antiviral activity for IPNV was fur-
ther validated by the inhibition of virus protein synthesis, especially virus capsid protein
VP2 [80]. IFN was not detected in the ASMx1-transfected CHSE-214 cells indicating
that endogenous IFN was not contributing to the protective effect of ASMx1.
Antiviral activity of Japanese flounder Mx protein (JFMx) was also demonstrated
[85]. A hirame natural embryo cell line (HINAE) was transfected to stably express the
recombinant JFMx. The JFMx-expressing cells were infected with hirame rhabdovirus
(HIRRV) and viral hemorrhagic septicemia virus (VHSV). A lower expression level of
the viral nucleoprotein transcripts in the JFMx-transfected cell line was observed [85].
Kinetics of the two viruses showed that reduced levels of glycoprotein and nucleoprotein
transcripts were due to JFMx blocking replication by interfering with transcription of
the viral subgenomic mRNAs and fewer viral particles [85].
Although the duration of Mx mRNA expression is relatively short lived in the dif-
ferent fish species, detection of Mx proteins lasts much longer. It was shown that the
halibut Mx protein expression in the liver reached a maximum at 3 days and Mx protein
levels in the liver remained elevated for 14 days after poly I:C treatment [81]. Likewise,
it was shown that IPNV infection increased halibut Mx protein expression in the liver
36
from 4 to at least 35 days [81]. After vaccinating rainbow trout against different viruses,
Mx protein levels are elevated 30 days post vaccination [8].
The eIF2? gene from rainbow trout and zebrafish has been characterized and the
predicted amino acid sequences were found to be 93 and 91% identical to the human
eIF2? protein suggesting that fish eIF2?s will exhibit the same function and regulation
as mammalian eIF2? [86]. To show that the fish eIF2?s are substrates for known eIF2?-
kinases, phosphorylation of recombinant rainbow trout and zebrafish eIF2? in vitro by
human PKR was demonstrated [86]. It was also demonstrated that rainbow trout and
zebrafish eIF2? could be phosphorylated in vivo [86]. When rainbow trout gonad cells
(RTG-2) were exposed to poly I:C or infected with IPNV, an increased phosphorylation
of eIF2? was noted, which suggests the presence of a PKR gene in rainbow trout [86].
Thus, similar to mammals, fish are able to regulate protein synthesis in response to
cellular stresses through phosphorylation of eIF2? [86].
The PKR-like gene from crucian carp Carassius auratus blastulae embryonic cells
(CAB) was cloned and sequenced following exposure to UV-inactivated grass carp hem-
orrhagic virus (GCHV) [87]. The carp PKR-like gene showed a high homology to all
family members of eIF2? kinases, although it shared the highest homology with human
PKR, and it up-regulated expression in response to active GCHV, UV-inactivated GCHV
and carp IFN [87]. Investigators found that induction of the carp PKR-like gene (also
called CaPKR-like) also required the production of cellular IFN. The CaPKR-like pro-
tein is similar in amino acid length to mammalian PKRs, has a NH2-terminal regulatory
domain and a C-terminal protein kinase catalytic domain similar to mammalian PKR
proteins, and this catalytic domain shares high sequence conservation with the catalytic
domain from mammalian PKRs [87].
37
2.3.7 IFN activity demonstrated in fish - two viruses used to demonstrate
protection
In early work done in fish, researchers could only detect the antiviral protection of
IFN without clearly understanding their mechanisms. In several studies [26], [27], two
viruses were used to demonstrate induction of antiviral protection due to viral interfer-
ence. In these studies, the first virus provided protection against subsequent infection
of the second virus. The first virus, while inhibiting host cell macromolecule synthe-
sis, also inhibited replication of co-infecting heterologous or homologous viruses. Viral
interference is a result of several different mechanisms occurring within the cell. One
possible mechanism is that the infection with one virus may inhibit replication of a sec-
ond virus by blocking virus entry into the cell. The second possible mechanism is that
when the first virus infects the host cell, it may inhibit host cell functions required by
the super-infecting virus. The third possible mechanism is that the first virus blocks
virus replication directly. Finally, the first virus may block the replication of a second
virus by inducing IFN or other anti-viral factors [26].
Prior exposure to an avirulent reovirus stimulated innate host defense to the infec-
tious rhabdovirus hematopoietic necrosis virus (IHNV) in rainbow trout [27]. Previous
experiments had demonstrated that rainbow trout pre-exposed to cutthroat trout virus
(CTV) were less susceptible to subsequent challenges with IHNV [19]. Four weeks of pro-
tection against viral infection as observed following CTV exposure [19]. Additionally, the
concentration of anti-IHNV neutralizing antibodies in the serum was significantly higher
among fish previously exposed to CTV [19]. The mechanism of anti-viral protection was
38
unclear, although it resembled interferon-like activity [19]. Thus, in subsequent exper-
iments, a reovirus was used to induce host defense against a rhabdovirus [27], because
reoviruses are known to be potent stimulators of IFN (reviewed in [27]). Rainbow trout
(O. mykiss) was exposed to chum salmon reovirus (CSV) for one hour, which resulted
in strong protection for up to 8 weeks post-exposure to CSV [27]. Survival rates ranged
from 68 to 100% when fish were challenged with IHNV over an 8 week period. Similar
to previous work, higher neutralization titers in IHNV exposed fish were observed, when
previously exposed to CSV [27]. It was concluded that nonspecific immune functions,
including IFN production and the activation of macrophages and/or NK cells, protected
the fish [27].
Similar work was done wherein a double-stranded RNA virus, channel catfish re-
ovirus (CRV), was used to inhibit channel catfish herpes-virus (CCV) following either
productive infection or exposure to UV-inactivated CRV. CRV inhibited CCV replica-
tion by two different mechanisms. First, CRV inhibited CCV replication directly, as a
consequence of its own replication [26]. In cells infected with CRV alone, cellular trans-
lation was stopped due to gradual replacement of cellular gene expression by reovirus
translation. In cells co-infected with CRV and CCV, CRV infection appeared to suppress
both host and heterologous virus (CCV) gene expression. Inhibition of cellular protein
synthesis probably prevented the synthesis of any anti-viral factors [26].
Secondly, CRV was able to block CCV replication indirectly due to the synthesis (or
release) of an anti-viral factor [26]. Experiments in which CCO cells were treated with
UV-inactivated CRV led to induction of an anti-viral factor rather than to viral interfer-
ence. CCO cells were infected with UV-CRV and the culture medium was harvested at
39
24 hours and tested for its ability to block CCV replication [26]. The cells were moni-
tored for the development of CPE for 48 hours and for the production of infectious CCV.
UV-CRV infected cultures showed little CPE, and the reduction of CPE in UV-CRV in-
fected CCO cells correlated with a reduction in sustaining CCV replication. Also, CCO
cells infected with UV-CRV synthesized or secreted an anti-viral factor that peaked in
protective ability within 24 to 48 hours after infection and did not increase significantly
over the next 2 days. This protection was short-lived, and 72 hours after CCV infection,
UV-CRV-treated cultures showed a marked increase of CPE. Alternatively, mock-treated
CCO cells were readily susceptible to CCV infection and showed CPE within 48 hours
of infection. No anti-viral activity was detected in medium harvested from CCO cells
before UV-CRV infection, or from medium harvested from mock-infected cultures during
the 4 days of the experiment [26].
The kinetics of the induced anti-viral factor produced by the CCO cells was com-
pared with the kinetics of mammalian IFN and found to be similar, suggesting that IFN
is a first line of defense following virus infection in catfish [26]. However, protection of
the cells provided by the anti-viral factor was incomplete and could be overcome by in-
creasing the amount of challenging virus. Although cultures that produced the anti-viral
factor were protected from CCV challenge for up to 48 hours, during the next 24 hours
?residual? virus replicated and caused extensive CPE, which may have been caused by
the decline of the anti-viral state over time [26].
In addition to anti-viral activity being produced in CCO cells infected with UV-CRV,
anti-viral activity was also found constitutively in medium harvested from several long-
term catfish T cell-like and macrophage lines. Levels of anti-viral activity were similar
to those seen following UV-CRV infection and experiments showed that this anti-viral
40
factor was a heat liable protein, which suggested an interferon. However, B cell clones
and fibroblast cell lines did not constitutively produce the anti-viral factor. In previous
experiments, newly established T cell and macrophage lines were non-permissive for
CCV replication, which may be due to the generation of an anti-viral factor [26].
2.3.8 IFN activity demonstrated in fish - IFN detection using Mx protein
expression induced by poly I:C
Mx protein expression was used to measure IFN expression [28]. After stimulating
Atlantic salmonid cells with poly I:C, an IFN-like activity in the cells was observed at
24 hours and Mx protein expression at 48 hours [28]. It was found that double-stranded
RNA induced Mx protein expression via induction of type I IFN production, similar to
mammalian systems [28]. It was also demonstrated that Atlantic salmon macrophages
were induced by poly I:C to produce IFN-like activity and Mx protein expression [28].
Atlantic salmon macrophages were chosen to express IFN-like activity (pooled super-
natants labeled AS-IFN) because mammalian macrophages are potent expressers of type
I IFN (reviewed in [28]). It was found that IFN antiviral activity in macrophages peaked
at 24 hours and then declined after 36 hours and this peak and decline in IFN production
was independent of the poly I:C concentration used [28].
Atlantic salmon macrophages were cultured in the presence or absence of poly I:C for
24 and 48 hours. When the macrophages were cultured in the absence of poly I:C, they
did not produce Mx proteins. In the presence of poly I:C at 48 hours, the macrophages
expressed the Mx protein, while at 24 hours the macrophages showed very weak Mx
protein expression. Atlantic salmon fibroblast cells were cultured in the presence or
absence of either poly I:C or AS-IFN. In the presence of either poly I:C or AS-IFN the
41
fibroblast cells were able to express Atlantic salmon Mx protein. CHSE-214 cells were
also cultured in the presence or absence of either poly I:C or AS-IFN and analyzed for
Mx protein expression. Only the CHSE-214 cells cultured in the presence of AS-IFN
expressed the Mx protein [28]. Mammalian type I IFN was compared with the Mx
protein inducing activity of AS-IFN from the Atlantic salmon macrophage supernatants
[28]. The pH sensitivity was tested and the AS-IFN Mx inducing ability was highly
resistant to low pH. This AS-IFN activity also diminished with trypsin or heat treatment
[28]. Based on analogy to the mammalian system, these findings suggested that the acid
stability of the Mx protein inducing activity of AS-IFN was due to type I IFN activity
rather than type II IFN activity [28]. Furthermore, in humans, type II IFN-? induced
only 1% of the MxA proteins expression when compared with type I IFN [88].
Thus, several experiments indicated that induction of Mx proteins by double-stranded
RNA in salmon cells occured via induction of IFN [28]. First, AS-IFN, not poly I:C,
induced expression of Mx proteins in CHSE-214 cells, which lack the ability to make
their own IFN. Second, Atlantic salmon IFN induced maximal Mx protein expression 24
hours earlier than poly I:C in macrophage cell lines. These observations suggested that
poly I:C induced Mx protein expression indirectly through the induction of type I IFN,
which is consistent with observations in mammalian and avian systems [28]. Moreover,
studies of the chicken Mx promoter in monkey cells that lack IFN? and ? genes have
shown that the promoter is only responsive to type I IFN, but not to poly I:C [89].
This is further supported by studies in which IFN-containing supernatant induced Mx
transcripts more rapidly than poly I:C in rainbow trout gonad cells [82]. Furthermore,
the Mx promoter in rainbow trout [82] and zebrafish [78] contained an IFN-stimulated
response element (ISRE).
42
2.4 Conclusion
It is clear that teleost fish have a type I IFN system similar to mammals. This is
based on the facts that several type I IFN-inducible genes have been cloned and sequenced
from several different fish species [67]. Thus, it is believed that fish have a complete IFN
system with similar antiviral mechanisms as mammals [87]. IFN signal transduction was
also conserved between mammals and teleost [67]. Although fish IFN genes show a low
degree of similarity with mammalian homologues, some of the other molecules important
in IFN activities are more highly conserved when compared to mammalian homologues,
such as the Mx protein [76]. Knowing more about how the innate immune system works
in fish will help in future vaccination designs to optimize induction of the innate IFN
system.
43
Chapter 3
Materials and Methods
3.1 Maintaining cell culture and virus propagation
Channel catfish ovary (CCO) cells were obtained from John Plumb, PhD, Depart-
ment of Fisheries and Allied Aquacultures, Auburn University. Cells were grown at 28?C
in HEPES-buffered Eagles minimum essential medium (MEM) plus 10% fetal calf serum
(Hyclone Biologics, Logan, UT) plus antibiotic/antimycotic. The Auburn -1 strain of
CCV (ATCC VR-665) was grown by removing medium from the flasks and inoculating
CCV directly onto cells at a multiplicity of infection (MOI) of 0.5. After two hours,
MEM was added to the cells and incubated at 28?C until widespread cytopathic effects
(CPE) could be seen. Cells and medium were decanted and sonicated on ice twice at
400W peak electrical power for 15 sec, and the suspension was cleared by slow speed
centrifugation on Beckman TJ-6 tabletop centrifuge. The supernate was aliquoted into
1 ml units and frozen at -80?C. Virus was titrated by serial dilution using the Karber
method [90].
3.2 Induction and detection of IFN by poly I:C in CCO cells
CCO cells were incubated with 1 ?g/ml, 10 ?g/ml, or 50 ?g/ml polyinosinic-
polycytidilic acid (poly I:C) in MEM to induce expression of interferon (IFN). After
1 h, the solution was decanted, the cells washed with fresh MEM, and new medium was
added to the cells, and incubation continued for times ranging from 1 to 50 h. Culture
44
medium was then harvested and frozen at -20?C and RNA was extracted from the cells
and stored at -80?C.
CCO cells were removed from the flasks using a sterile cell scraper and resuspended
in 1 ml RNase-free 1% PBS and concentrated by centrifugation. The pellet was resus-
pended in 200 ?l 1% PBS and RNA extracted using High Pure RNA Isolation kit (Roche,
Penzberg, Germany). Total RNA was frozen at -80?C.
Culture medium from poly I:C treated cells, was diluted two fold seven times and
added to fresh CCO cells in 96-well plates as described by Chinchar et al. [26]. Medium
from two different control flasks was used as an experimental control (labeled T0) to
gauge the difference in constitutive IFN production and induced IFN production. The
culture in the first control flask was 2 days old and confluent (labeled T0-2). The culture
in the second control flask was 1 day old and not completely confluent (labeled T0-1).
Cells were incubated for 1 h at 28?C, the medium was removed, and 100 ?l of fresh MEM
plus 100 TCID50 CCV added to each well. Two different controls were used on the 96-
well plates. One control that served as the positive control contained CCO cells that
did not receive culture medium from poly I:C treated cells and were not infected with
CCV. The second control that served as the negative control contained CCO cells that
did not receive culture medium from poly I:C treated cells, but were infected with CCV.
Plates were sealed with paraffin wrap to prevent evaporation of medium, and the plates
incubated for 24 or 48 hours at 28?C. Medium was then removed, the plates were fixed
and stained with 1% crystal violet in 70% ethanol, and immediately rinsed. Protection
against CCV infection was assessed by determining which titer on the 96-well plate the
cells no longer showed 50% protection.
45
3.3 Induction of IFN expression by CCV in CCO cells
CCO cells were infected with CCV for 1 h at 28?C with an MOI of 1.0. Cells were
washed with MEM after an hour, refed with fresh MEM, and harvested at times from 1
to 6 hours post infection. RNA was extracted from cells and frozen at -80?C as described
above.
3.4 Induction of IFN expression by ultraviolet (UV) inactivated CCV in
CCO cells
CCV were inactivated by using 2.4x105 ?Joules of UV in a Stratalinker 1800 (Strata-
gene, LaJolla, CA). Inactivation was confirmed by serial dilution of virus on cells and
infection was measured. CCO cells were treated the same as when induced by poly I:C,
and at time intervals 1 through 6 hours, RNA was harvested from the cells and frozen
at -80?C as described in that experiment.
3.5 Primers and conditions for rt-PCR
rt-PCR was used to detect expression of the IFN genes. Primer set 1, designed
to detect IFN-1 [29] actually detects expression of 4 different catfish IFN genes [48].
The forward primer (5prime GCCAGTACAGAGCAAGAACA 3prime) and the reverse prime (5prime
CCATTCCTGATTCGCTCCA 3prime) span nucleotides 151 through 558 of the channel cat-
fish IFN-1 cDNA in Genbank (accession number AY267538) resulting in a 408 bp ampli-
con. Previously described optimized conditions for rt-PCR [29] were 1 cycle of 48?C for
45 min; 1 cycle of 94?C for 3 min; 30 cycles of 94?C for 30 sec, 55?C for 45 sec, and 72?C
for 1 min; 1 cycle of 72?C for 5 min. The primer set for IFN-2 mRNA (Genbank accession
46
number AY 847295) (forward primer 5prime CCAATGACAGGAAACACCTCTTTCCCA 3prime;
reverse primer 5prime CAGTGGATGAACGTCTTGCTCTTTCAGC 3prime) spans 265 bp of the
IFN-2 gene. Optimum conditions were 1 cycle of 48?C for 45 min; 1 cycle of 94?C for 3
min; 35 cycles of 94?C for 1 min; 56.6?C for 1 min, 72?C for 3 min, and 1 cycle of 68 C
for 10 min. Amplicons were detected using 2 ?g of RNA electrophoresed in 1.5% (IFN-
1-4) or 2% (IFN-2) agarose with ethidium bromide. Amplicons were harvested from the
gel and purified using QIAquick Gel Extraction Kit (QIAGEN, Valencia, CA), cloned
using TA cloning kit (Invitrogen), and sequenced by the Auburn University Genomics
and Sequencing Laboratory.
3.6 Induction of IFN in channel catfish
Three groups of 12 yearling channel catfish, 10-15 cm in length, were held in 55 L
of aerated dechlorinated tap water at 20 +/- 2?C. Group 1 fish were injected intraperi-
toneally (IP) with 50 ?g poly I:C. Group 2 fish were bathed in 104 TCID50 CCV/ ml
for 1 h and then moved into a tank comparable to that housing group 1 fish. Control
fish were untreated. Times ranging from 1 to 7 hours post initiation of treatment, sera
and head kidneys were collected from 3 fish from each group. Head kidneys were frozen
in liquid nitrogen, stored at -80?C, and RNA was extracted using RNA STAT-60 and
stored at -80?C.
3.7 Effect of IFN induction at the time of vaccination
Six groups of 30 yearling channel catfish, 5-8 cm in length were held in 55 L flow-
through aquaria with aerated dechlorinated water at 20 +/- 2?C. Group 1 fish served
as the control population. Group 2 fish were injected IP with 10 ?g of poly I:C. Group
47
3 fish were injected intramuscularly with 23.3 ?g of CCV ORF 59 DNA as previously
described [3]. Group 4 fish were injected with poly I:C and vaccinated with ORF 59
DNA simultaneously; Group 5 were injected with poly I:C one hour before vaccination
with ORF 59 DNA; and Group 6 were injected with poly I:C one hour after vaccination
with ORF 59 DNA.
At 4, 24, and 48 hours post treatment, serum and RNA from head kidneys were
harvested from 5 fish in each group and treated as previously described. Sera were also
harvested from 5 fish from each group 2 and 4 weeks after treatment.
3.8 Antiviral assays
Sera from fish either treated with poly I:C, CCV vaccination, or infected with CCV,
were diluted two fold seven times and added to fresh CCO cells in 96-well plates. Sera
were not pooled; each column represented a single fish. Cells were incubated with sera
for 1 h at 28?C, before 100 ?l of medium containing 100 TCID50 CCV were added to
each well. Two controls groups were included on the 96-well plate. The cells that served
as the positive control did not receive sera and were not infected with CCV. The cells
that served as the negative control did not receive sera and were infected with CCV.
Plates were sealed with paraffin wrap to prevent evaporation of medium, and the plates
incubated for 24 or 48 hours at 28?C. Medium was then discarded, the plates were fixed
and stained with 1% crystal violet in 70% ethanol, and immediately rinsed. Antiviral
activity was evaluated by looking at which titer the cells received less than 50% protection
from virus.
48
3.9 Virus neutralization assays
An insufficient amount of serum was collected from the fish injected only with poly
I:C or ORF 59 to use in experiments. Sera from individual fish in the other groups were
initially diluted 1:5 in MEM, and then serially diluted 1:2 in MEM. 100 ?l of the dilution
were mixed with 100 ?l of medium containing 1000 TCID50 CCV and incubated at 28?C
for 2 h. 100 ?l of the serum-virus mixtures were then transferred to wells of a 96-well
plate containing confluent monolayers of CCO cells. Plates were sealed, incubated, and
stained as described above.
49
Chapter 4
Results
4.1 Inducing IFN expression in CCO cells
4.1.1 Optimization and kinetics of poly I:C
As measured by CCO protection, optimum induction of IFN occurred when either
1, 10, or 50 ?g/ml poly I:C were used (Fig. 4.1 and not shown). Peak generation of
IFN protection occurred at 8 hours following incubation with poly I:C (Fig. 4.1) and
increased IFN mRNA production in cells occurred 1 through 7 hours (Fig. 4.2). The
primers used for rtPCR in these experiments were created by Long et al. [29] and were
later confirmed by Long et al. [48] to anneal to a sequence shared by all four catfish IFN
genes (Fig. 4.2). Samples collected from control groups always showed a low background
level of mRNA IFN expression. Within the first 5 hours post incubation with poly I:C,
all three concentrations of poly I:C had the same effect on IFN expression; they all
caused an increase in expression of IFN mRNA (data not shown). However, after 5
hours post incubation with 1 ?g/ml of poly I:C, the level of IFN mRNA expression
dropped when compared to IFN mRNA levels induced by 10 ?g/ml and 50 ?g/ml poly
I:C, which continued to show increased expression of IFN mRNA up to 7 hours post
incubation with poly I:C. Thus, we chose to continue experiments using 10 ?g/ml or 50
?g/ml of poly I:C. Under these conditions, CCO cells expressed IFN mRNA within the
first 12 hours post treatment with poly I:C, although, after 7 hours, the cells no longer
expressed maximum levels of IFN mRNA, but a lesser amount that was still greater than
background levels of IFN expression (Fig. 4.2). After 13 hours, repeated experiments
50
showed that induced IFN mRNA expression stopped or greatly decreased, with IFN
levels equal to or less than background expression levels of IFN mRNA expressed by
non-induced CCO cells. Overall, there was not much of a difference between using 10
?g/ml and 50 ?g/ml poly I:C to induce IFN expression in CCO cells.
Figure 4.1: Antiviral activity of culture medium collected from CCO cells at 2, 4, 8 and 12
hours post induction with 50 ?g/ml poly I:C. The CCO cells in the assay were challenged
with 100 TCID50. The highest dilutions of culture medium collected at indicated times
that show antiviral activity are as follows: at 2 hours, 1:16; 4 hours, 1:32; 8 hours, 1:64;
and at 12 hours, 1:32.
4.1.2 Comparison of expression of IFN genes
Two different sets of primers were used to examine mRNA expression of catfish IFN.
When we cloned and sequenced an IFN cDNA expressed in CCO cells sequencing results
showed that this cloned cDNA was very similar to the CF IFN-1 posted in Genbank,
51
Figure 4.2: IFN mRNA expression in CCO cells. 10 ?g/ml poly I:C induced IFN mRNA
expression in CCO cells. Poly I:C was incubated 1 hour on cells before removal. RNA
was extracted from the cells up to 54 hours post incubation. General IFN primers were
used for rt-PCR, which show expression of all four CF IFN genes. Expression is strongest
within the first 7 hours, although there was continued expression through 12 hours.
although not 100% identical (showed only 87% similarity). The other three catfish IFN
genes had not yet been found or released to Genbank. However, months later, after new
data was released to Genbank, we found that our IFN clone showed 99.3% similarity
to the CF IFN-2 gene (Genbank accession number AY847295) except at the ends of
the cloned sequence where the primer, which was designed specifically for CF IFN-1
sequence, had annealed to the CF IFN-2 gene (Fig. 4.3). From the CF IFN-2 gene, new
primers were designed to compare the differences between CF IFN-2 cDNA expression
and IFN expression using the nonspecific primers. This second set of primers amplified
a 265 bp fragment unique to the catfish IFN-2 gene.
52
Figure 4.3: Sequence of an IFN cDNA expressed in CCO cells treated with 10?g/?l poly
I:C to induce IFN expression. cDNA was amplified using primers designed by Long et
al.[29], which anneal to all four catfish IFN genes. The clone was completely identical to
the catfish IFN-2 gene, except near the ends of the sequence where the general primers
anneal.
Subsequent experiments compared expression of all four IFN genes (amplified using
the general IFN-1 primers) to IFN-2 gene expression using the two different sets of
primers in rt-PCR reactions. Maximum IFN-2 mRNA induction was reached within the
first 3 hours at both 10 ?g/ml (Fig. 4.4) and 50 ?g/ml poly I:C concentrations (data
not shown). This differed from the results generated with general IFN-1 primers created
by Long et al. [29] showing that results generated with IFN mRNA expression was
noted up to 7 hours for both concentrations of poly I:C in rt-PCR reactions (compare
Figures 4.2 and 4.4). By 4 hours, increased IFN-2 gene expression ceases. Interestingly,
however, over the 54-hour period, strong IFN-2 expression appeared sporadically at 6
and 10 hours when 10 ?g/ml poly I:C was used and 18 and 48 hours when 50 ?g/ml poly
53
Figure 4.4: IFN-2 gene expression induced by 10 ?g/ml poly I:C in CCO cells with
the strongest expression within the first 3 hours post induction. IFN-2 gene expression
was also induced by UV-inactivated CCV, with strongest expression at 6 hours post
induction.
I:C was used (data not shown). Bands double in size of the expected 265 bp amplicon
also occurred sporadically, although, again, not at the same time for the different poly
I:C concentrations. These bands showed greatest expression at 18 hours post induction
at 50 ?g/ml of poly I:C; cloning and sequencing confirmed these to be doublets.
4.1.3 Inducing IFN expression using CCV
Experiments were done to see if CCV induced IFN expression in CCO cells. Using
the general IFN-1 primers, the results indicated that CCV induced IFN expression in
CCO cells within the first 5 hours post infection (Fig. 4.5). The strongest expression
was seen at 1, 4 and 5 hours post infection with CCV. By 6 hours post infection with
54
CCV, increased IFN mRNA expression ceased. Using the IFN-2 primers to see if CCV
induced IFN-2, no evidence of any expression of this gene was found (data not shown).
Figure 4.5: IFN mRNA expression over 6 hours induced in CCO cells by infection with
CCV or UV-inactivated CCV over 6 hours. General IFN-1 primers were used.
4.1.4 Inducing IFN expression using inactivated CCV
UV-inactivated CCV was also used to induce IFN mRNA expression, and was found
to be able to induce IFN expression in CCO cells (Fig. 4.5). This was a qualitative
conclusion that will need to be tested quantitatively to confirm this conclusion. Using
the general IFN-1 primers in an rt-PCR showed that IFN genes were expressed within
the first 6 hours post infection with inactivated CCV. While live CCV did not induce the
IFN-2 gene in CCO cells, inactivated CCV induced expression of the IFN-2 gene, with a
very low level of mRNA expression beginning at 5 hours post infection and increasing by
55
6 hours post infection (Fig. 4.4). These experiments showed that CCV and inactivated
CCV induce different levels of IFN mRNA expression in CCO cells, especially in regard
to the ability of CCO cells to express the IFN-2 gene in the presence of inactivated CCV.
CCV 0 hr 1 hr 2 hrs 3 hrs 4 hrs 5 hrs 6 hrs
IFN general - + + + + + -
IFN-2 - - - - - - -
ORF 6 - + + - - - -
ORF 53 - - - + + + +
UV-CCV 0 hr 1 hr 2 hrs 3 hrs 4 hrs 5 hrs 6 hrs
IFN general - + + + + + +
IFN-2 - - - - - + +
ORF 6 - - - - + + +
ORF 53 - - - - - + -
Table 4.1: Expression (+) or no expression (-) induced by either CCV or UV-inactivated
CCV in CCO cells using the general primers that anneal to all 4 IFN mRNA genes,
primers specific for IFN-2 mRNA, primers specific to early gene CCV ORF 6 mRNA,
and primers specific to late gene ORF 53 mRNA.
CCV gene expression in cells infected with CCV or UV-inactivated CCV was also
examined by rt-PCR. In CCV infected cells, ORF 6, a membrane protein expressed
early in the infection of CCV, expression happens within the first 2 hours post infection
with expression ceasing by 3 hours post infection. In contrast, in UV-inactivated CCV
infected cells, early gene ORF 6 expression happens at 5 and 6 hours post infection with
inactivated CCV, as compared to expression being reached within the first 2 hours post
infection with CCV. In CCV infected CCO cells, late gene ORF 53, encoding a capsid
protein, begins to show expression by 3 hours and continues expression up to 6 hours
post infection with CCV. Dissimilarly, in UV-inactivated CCV infected cells, the ORF
53 gene only shows expression at 5 hours post infection.
56
4.2 Challenge Assays using Culture Medium from induced CCO cells
Culture medium was collected from the experiments in which confluent CCO cells
were induced by poly I:C to produce IFN. This culture medium was collected from
the cells and used to measure antiviral activity. The cells incubated with poly I:C
and induced to produce IFN provided protection against CCV infection (Fig. 4.1). This
protection did not completely align with the times of IFN mRNA expression from rt-PCR
assays (compare Figures 4.2 and 4.1). However, our results were similar to work by Long
et al. [29], which showed correlation between protection and IFN mRNA expression.
In their results, poly I:C induced mRNA expression only at 2 hours, whereas challenge
assays showed strongest protection at 2, 4 and 7 hours post treatment with poly I:C,
with 4 hours having the strongest protection [29]. Our assays showed that antiviral titers
were, at 2 hours, 1:16; 4 hours, 1:32; 8 hours, 1:64; and at 12 hours, 1:32 (Fig. 4.1).
The protection against CCV in the challenge assays seemed to increase with time up to
8 hours. At 8 and 12 hours, as seen from the rt-PCR reactions, IFN expression begins
to decline.
4.3 Inducing IFN expression in channel catfish with poly I:C and ORF 59
vaccination
In these experiments, one hundred eighty channel catfish 5 to 8 cm in length were
divided into six groups of 30 fish each, and each received intraperitoneal injection with
poly I:C and/or intramuscular injection of CCV ORF 59 DNA vaccine.
Using the primers designed by Long et al. [29] to look at IFN expression, results
showed that no significant increase in the amount of IFN expressed in the catfish. In
57
fact, at 4 hours, fish injected with either poly I:C or poly I:C and ORF 59 did not show
increased expression (Fig. 4.6). At 24 hours, fish injected with only poly I:C showed
slightly decreased expression and fish injected with both poly I:C and ORF 59 vaccine
showed increased IFN mRNA expression (Fig. 4.6). At 48 hours, fish from groups either
injected with only poly I:C or injected with both poly I:C and ORF 59 expressed IFN
mRNA, although this expression was slightly above the approximated background level
of IFN mRNA expression in the experimental control fish (Fig. 4.6). The fish injected
with both the ORF 59 vaccine and poly I:C expressed more IFN mRNA at 24 hours
than at 48 hours.
4.4 Antiviral activity at 4, 24, and 48 hours from poly I:C/ORF 59 vacci-
nated fish
Catfish Groups 4 hrs 24 hrs 48 hrs
experimental control 16 16 16
Poly I:C 64 32 32
ORF 59 32 64 64
ORF 59 + Poly I:C together 16 32 64
Table 4.2: Endpoint dilution of serum-derived IFN protection of CCO cells challenged
with 100 TCID50 CCV. Serum was collected at 4, 24 and 48 hours after channel catfish
were injected with either poly I:C, ORF 59 or both.
The serum collected from the different groups of catfish at 4, 24 and 48 hours was
used in interferon protection assays. At all three time points in the experimental control
assay, protection stopped at 1:16 dilution (Table 4.2). At four hours post injection, serum
collected from fish injected with poly I:C showed the greatest protection (detectable at
1:64 dilution); serum collected from fish injected with the ORF 59 vaccine showed less
58
protection (detectable at 1:32 dilution); and serum collected from fish injected with
both ORF 59 and poly I:C showed least protection (detectable at 1:16 dilution), which
was the same level of protection seen in control fish (Table 4.2). At 24 hours, serum
collected from fish injected with only poly I:C showed decreased protection (detectable
at 1:32 dilution), serum collected from fish injected with only the ORF 59 vaccine showed
increased protection (detectable at 1:64 dilution), and serum collected from fish injected
with both ORF 59 and poly I:C at the same time showed increased protection from
background levels (detectable at 1:32 dilution) (Table 4.2). At 48 hours, protection
stayed the same for fish groups in which serum was collected from fish injected with
only poly I:C and from fish injected with only the ORF 59 vaccine, whereas protection
increased to 1:64 for fish groups in which serum was collected from fish injected with
both ORF 59 and poly I:C at the same time (Table 4.2). The rt-PCR results at 24
hours show IFN mRNA expression from fish injected with both ORF 59 and poly I:C,
while fish injected with only poly I:C do not show any increased IFN expression and yet
the antiviral titers are the same for both groups. Fish injected with both poly I:C and
ORF 59 expressed more IFN mRNA at 24 hours than 48 hours, yet antiviral protection
increased between 24 and 48 hours.
4.5 Antiviral assays from poly I:C/ORF 59 vaccinated fish at 2 and 4 weeks
At 2 and 4 weeks post injection with poly I:C and/or ORF 59, serum was collected
for antiviral assays from 5 fish from each of the 6 groups mentioned in the above experi-
ments. The results were expressed using the Geometric Mean Titer (GMT) (Table 4.3).
Nonspecific antiviral protection was greatest in fish that received the ORF 59 vaccination
1 hour prior to poly I:C, although the combination of having the ORF 59 vaccine with
59
Catfish Groups 2 weeks 4 weeks
Control 6 3
Poly I:C 20 9
ORF 59 17 40
ORF 59 + Poly I:C together 113 35
Poly I:C 1h before ORF 59 80 92
ORF 59 1h before Poly I:C 160 160
Table 4.3: GMT from antiviral activity serum collected from fish groups at 2 and 4 weeks
post treatment
Catfish Groups 4 weeks
Control 3
ORF 59 + Poly I:C together 10
Poly I:C 1h before ORF 59 10
ORF 59 1h before Poly I:C 13
Table 4.4: GMT from viral neutralization assays of serum collected from groups at 4
weeks post experiment
poly I:C injection seemed to produce stronger antiviral results than either poly I:C or
the ORF 59 vaccine alone (Table 4.3).
4.6 Viral neutralization assays from poly I:C/ORF 59 vaccinated fish at 4
weeks post vaccination
At 4 weeks post injection with poly I:C and/or ORF 59, serum was collected for
viral neutralization assays from 5 fish from each of the fish groups injected with both
poly I:C and the ORF 59 DNA vaccine. The results were expressed using the Geometric
Mean Titer (GMT) (Table 4.4). No significant neutralizing activity was detected by 4
weeks. Results for all three combinations of ORF 59 and poly I:C were similarly low.
60
4.7 Inducing IFN expression in channel catfish with poly I:C and CCV
One group of twelve 5 to 8 cm yearling channel catfish each were intraperitoneally
injected with 500 ?l of 100 ?g/ml (50 ?g) of poly I:C to ensure induction of IFN mRNA
expression in vivo, and another group of twelve yearling catfish were bathed for 1 hour in
1 L of water with 30 ml of CCV at 105.5 LD50/ml. A third group, which did not receive
any treatment was set aside as a control. IFN mRNA expression was detected in channel
catfish using the general IFN-1 primers. The rt-PCR results from this CCV/poly I:C
experiment suggest that the increased concentration of poly I:C injected into the fish is
able to induce better IFN mRNA expression in channel catfish (compare Figures 4.6 and
4.7). Results showed increased IFN mRNA expression collected from channel catfish
at all four time points, 1, 3, 5 and 7 hours post injection with poly I:C. IFN mRNA
expression was also induced in catfish exposed to CCV, although for a shorter period of
expression than catfish induced with poly I:C (Fig. 4.7). Results showed increased IFN
mRNA expression at 3 and 5 hours post exposure to CCV. Interestingly, only rt-PCR
with the primers designed by Long et al. [29] was able to show increased catfish IFN
expression in vivo. The primers designed to amplify the IFN-2 mRNA did not detect
any IFN-2 mRNA expression in vivo at all in fish induced by either poly I:C or CCV
(data not shown). This suggests that one or more of the other three IFN genes (IFN-1,
IFN-3, or IFN-4) was induced in vivo.
61
4.8 IFN protection in serum from CCV and poly I:C induced channel catfish
The serum collected from the above experiment was used in interferon protection
assays. Serum from fish injected with 50 ?g of poly I:C showed the most antiviral pro-
tection 1 hour post injection (GMT of 102), protection decreased 3 hours post injection
(GMT of 81), and protection decreased further and stayed the same for both 5 and 7
hours post injection (GMT of 64) (Fig. 4.7). Serum from catfish exposed to CCV for 1
hour showed equal protection at 1, 5 and 7 hours post CCV exposure with a GMT of 51
(Fig. 4.7). The greatest antiviral activity was seen in fish 3 hours post CCV exposure
with a GMT of 64 (Fig. 4.7). This antiviral activity data for the poly I:C injected
fish correlates with rt-PCR results, with strongest IFN mRNA expression and antiviral
protection within the first 3 hours. However, the data from the antiviral activity assay
for fish exposed to CCV showed strong protection at all time points, while the rt-PCR
data showed that IFN mRNA was induced only at 3 and 5 hours post exposure to CCV.
62
Figure 4.6: IFN mRNA expression collected at 4, 24 and 48 hours post treatment from
experimental control fish, fish injected with 10 ?g poly I:C, or fish injected with both
poly I:C and ORF 59 vaccine. The general IFN-1 primers were used for the rt-PCR
reactions shown.
63
Figure 4.7: IFN expression, as determined by rt-PCR using general IFN-1 primers, and
interferon protection, as determined by antiviral activity assays. RNA collected from
fish either injected with 50 ?g poly I:C or induced with CCV at 1, 3, 5 or 7 hours post
challenge. Serum was collected from fish at each time point shown above the lanes and
the antiviral titer is shown below each lane.
64
Chapter 5
Discussion
Based on the available information, teleost fish have a type I IFN system very simi-
lar to mammals. Many type I IFN-inducible genes have been cloned and sequenced from
a variety of fish including pufferfish, Japanese flounder, rainbow trout, Atlantic salmon,
channel catfish, zebrafish, and Atlantic halibut. These include type I and II IFN genes,
[29], [30], [31], [32], [52], Mx proteins, [70], [71], [72], [73], [76], [77], [78], [79], interleukin
genes, [31], IFN receptor genes [31], interferon regulatory factor IRF genes, [98], and
JAK/STAT signaling genes [61], [63], [64], [66], [67]. Furthermore, it was shown that
poly I:C induces type I IFN, which subsequently induces Mx proteins, ultimately pro-
viding protection against viral challenge similarly as seen in mammalian systems [28].
The type I IFN innate immune response and other innate responses play a major role in
the elimination of replicating virus during the early phase of viral infection. They also
help the specific immune response become more effective. This biphasic response (early
nonspecific protection succeeded by a later specific response) has also been observed
following DNA vaccination in fish. It was demonstrated that DNA vaccines provide non-
specific protection due to IFN production and Mx expression and Mx protein protection
persisted for the following 4 weeks [8]. It was also demonstrated that non antigen-specific
protection by Mx protection in trout persisted for 4 weeks post vaccination with a DNA
vaccine [12]. Early and innate immune responses in fish immunized with DNA vaccines
possibly help trigger an efficient, long-lasting specific immune response (as discussed in
[13]). The lack of or inefficient stimulation of the innate immune system could restrict
65
induction of an effective immune response to a vaccine [93]. The use of different genetic
adjuvants would help generate a more effective immune response [93]. In fact, Leitner et
al. [95] were able to increase the efficacy of DNA vaccines in mice by providing stronger
adjuvants generating ?danger signals? to the innate immune system. Also, IFN induc-
tion in fish by poly I:C provided protection against viral challenge [29] [83]. In fact, it
was found that injecting poly I:C into fish induced IFN mRNA expression and ultimately
Mx protein expression which protected fish against viral challenge for up to 2 weeks [83].
It is anticipated that inducing IFN expression with poly I:C during vaccination would
enhance antigen-specific protection in vaccinated catfish. Wowever, we found that induc-
ing IFN by poly I:C did not increase the effectiveness of vaccination with DNA vaccines
based on virus-neutralization titers (Table 4.4), although some evidence of nonspecific
protection was observed. Our results, although limited, show similarities to work by oth-
ers with viral protection offered by Mx proteins [91] [83], although further tests would
need to confirm this notion.
Others demonstrated that the lowest concentration of poly I:C to induce IFN mRNA
expression in vitro was 1 ?g/ml in Atlantic salmon head kidney macrophage cells [28].
Our results confirmed this; within the first hours, IFN mRNA expression can be induced
with as little as 1 ?g/ml poly I:C or as much as 50 ?g/ml. However, using the higher
doses of 10 ?g/ml or 50 ?g/ml poly I:C to induce IFN mRNA expression in CCO cells
caused increased IFN expression to last longer than with 1 ?g/ml poly I:C. There was
not a visible difference observed between the ability of 10 ?g/ml and 50 ?g/ml poly
I:C to induce IFN expression in vitro. However, our in vivo results differ from the in
vitro findings. Using 10 ?g poly I:C in vivo is too low of a concentration to induce
consistent IFN mRNA expression in channel catfish. Injection of 400 ?g of poly I:C
66
into fish has been shown to be successful [83]. In fact, when we increased the amount
of poly I:C injected into fish to 50 ?g poly I:C, more consistent results were obtained.
The in vivo kinetics from using 50 ?g poly I:C were also consistent with kinetics seen in
in vitro experiments. In fact, in CCO cells, increased IFN mRNA expression was seen
at 1 hour post induction, (our first sampling point), with 10 ?g/ml poly I:C, continues
up to 12 hours post induction, and is gone by 24 and 48 hours. Similarly, when 50 ?g
poly I:C is used in catfish IFN mRNA expression was detected as early as 1 hour post
induction with poly I:C, and continued up to 7 hours post induction. The two different
in vivo experiments demonstrate that IFN mRNA expression is poly I:C dose dependent
(compare IFN expression in figures 4.6 and 4.7 in which fish received different amounts
of poly I:C).
In the DNA vaccine experiment in which only 10 ?g poly I:C was used, at 4 hours
post vaccination or injection with poly I:C there was no measurable increased in amount
of IFN mRNA detected (Fig. 4.6). In fish injected with poly I:C, there was no increase
in IFN mRNA detected likely because there was not enough poly I:C injected into fish
to induce expression. In fish vaccinated with ORF 59, no IFN mRNA was detected 4
hours after vaccination likely because insufficient vaccine glycoprotein was produced to
induce IFN expression.
Fish that received coincidental vaccination and poly I:C injection showed increased
IFN mRNA expression at 24 hours, whereas fish injected with only poly I:C did not
(Fig. 4.6). This is similar to previous findings [8], with a DNA vaccine. The results
demonstrated that increased IFN mRNA expression in fish injected with both poly I:C
and the vaccine was probably not due to poly I:C because fish that were only injected
67
with poly I:C did not show increased expression of IFN mRNA at 24 hours. In addition,
CCO cells do not express increased IFN mRNA at 24 hours after poly I:C induction.
IFN mRNA expression was increased somewhat at 48 hours in fish injected with
poly I:C and those receiving both ORF 59 vaccination and the poly I:C injection. These
results seem inconsistent with IFN mRNA expression kinetics in CCO cells, which showed
that IFN expression was the highest within the first 12 hours after induction with poly
I:C and disappeared by 48 hours.
Even in the absence of detectable increased IFN mRNA expression at 4 hours there
was limited nonspecific antiviral protection against CCV challenge (Table 4.2). At 4
hours post injection, serum from fish injected with only poly I:C showed the best pro-
tection against CCV with a protection titer of 64. This protection was stronger than
protection seen in fish that injected with the ORF 59 DNA vaccine (protection titer 32).
Since the amount of poly I:C used gave inconsistent results with IFN mRNA expression,
it is hard to know whether the antiviral nonspecific protection shown in the titers for fish
induced with poly I:C is due to the nonspecific effects of IFN induction or some other
antiviral component. Experiments performed by Harbottle et al. [91] did not show Mx
expression this early after vaccination. So, it is unclear what is causing the nonspecific
protection in ORF 59 vaccinated fish. At 4 hours, serum from fish that were both vac-
cinated and injected with poly I:C did not show nonspecific antiviral protection against
CCV above the constitutive levels of antiviral protection seen in control fish. Our results
may be due to competition for receptor sites, or spontaneous binding of vaccine and poly
I:C.
Antiviral protection increased 24 hours after fish were vaccinated with ORF 59 and
in fish groups that received the ORF 59 vaccination and poly I:C simultaneously (Table
68
4.2). This helps support the notion that IFN expression at 24 hours could be elicited
by the vaccine [8], [12]. In fact, the early IFN response in fish vaccinated with DNA
vaccines leads to Mx protein expression [12]. It was also noted that Mx protein expression
occured 1 day after vaccination with DNA in fish [91]. In fish that only received poly I:C
injections, antiviral protection decreased, which is what our in vitro time kinetics would
predict; IFN mRNA expression ceases before 24 hours.
Fish injected with both ORF 59 and the poly I:C were the only fish group with
increased antiviral protection at 48 hours after injection, while antiviral protection from
fish injected with either only poly I:C or only ORF 59, did not change (Table 4.2). Most
likely the stronger nonspecific antiviral protection seen in fish groups that received the
DNA vaccine is due to the additive stimulation of the vaccine to induce innate anti-viral
molecules in the fish.
At four weeks post vaccination and/or poly I:C injection, the virus neutralization
titers were measured. The results were similar for all three fish groups; no neutralizing
antibodies were detected. New data suggests that specific antibody protection does not
usually show up as early as 4 weeks post vaccination [92], [91], [12]. In fact, McLauch-
lan et al. [12] demonstrated that the smaller the fish and DNA vaccine dose used, the
longer it takes for neutralization antibodies to appear. Based on a previous publication
by McLauchlan et al. [12], it is expected that neutralization antibodies will show up no
earlier than 5 weeks after immunization based on fish size and vaccine dose. Even so,
DNA vaccination against CCV has not been effective at 5 or 6 weeks post vaccination in
channel catfish with very low neutralization antibody titers detected at these time points
[91]. In fact, DNA vaccines have shown the most success against rhabdoviral diseases
in fish, including infectious hematopoietic necrosis virus (IHNV) and viral hemorrhagic
69
septicemia virus (VHSV), [9], [11], [12], [13], snakehead rhabdovirus (SHRV) and spring
viremia of carp virus (SVCV) [8]. Attempts to use DNA vaccines against other fish
diseases has been limited (reviewed in [15]). Fish rhabdoviruses are presently the only
pathogens where consistent and significant levels of protection have been reported fol-
lowing DNA vaccination (reviewed in [15]). None of the DNA vaccines for infectious
salmon anaemia virus, an aquaorthomyxovirus, have provided protection to fish (cre-
ated by E. Anderson and cited in [15]). Likewise, DNA vaccination against Atlantic
halibut nodavirus also did not provide protection for fish [93].
Although our studied catfish had low neutralization titers, we did note nonspecific
protection in antiviral assays with the fish serum. We do not have RNA evidence to sup-
port our data, but similar experiments have reported non antigen-specific protection in
the form of Mx protein expression after DNA vaccination [8] [91]. Mx protein expression
is only induced by IFN expression (reviewed in [22]). In fact, even with low neutraliza-
tion results and poor protection results following CCV challenge, Mx protein expression
was noted in catfish in all treatments and all doses as early as 1 day post vaccination and
through 35 days post vaccination [91]. In our assays, the highest nonspecific protection
was seen in fish 2 and 4 weeks after they received both the ORF 59 vaccination and
poly I:C injection. We did not perform immersion challenge on the catfish, but we can
speculate that they would not have survived CCV challenge based on low neutralization
titers and previous studies [91] [93].
In our antiviral assays, fish that received the ORF 59 vaccine 1 hour before poly
I:C injection had a GMT of 160, while the fish that received the ORF 59 vaccine and
poly I:C injection at the same time had the next highest antiviral titer with GMT 113.
The fish that received the poly I:C injection 1 hour before the ORF 59 vaccine had an
70
antiviral titer of GMT 80. Even lower antiviral titers were seen in fish that were only
vaccinated with ORF 59 or fish that only received the poly I:C injection, at GMT of 17
and 20 respectively. These results are similar to a salmon study in which salmon received
nonspecific protection up to 2 weeks after induction with poly I:C [83]. In another study,
Mx proteins induced by poly I:C in catfish maintained protection up to 8 days (their
last sampling point), even though Mx mRNA expression began to decline after 2 to 4
days post injection with poly I:C [76]. Possibly poly I:C elicits Mx mRNA expression
within the first several days, but Mx protein expression lasts much longer, as seen with
Jensen et al. [83]. Kim et al. [8] showed that at 2 weeks, nonspecific protection exist
in vaccinated trout due to glycoprotein expression and Harbottle et al. [91] also showed
Mx protein expression in vaccinated catfish. In these studies, fish were induced either by
poly I:C or by DNA vaccines to produce Mx proteins and provided protection by innate
immune mechanisms.
At 4 weeks antiviral titers from fish that received only the poly I:C injection de-
creased from GMT 20 to GMT 9 (Table 4.4). Similarly, it was suggested that nonspecific
protection decreased 2 weeks after induction with poly I:C, because the half-life of Mx
proteins were only around 2 weeks [83]. Jensen et al. [83] did not show whether protec-
tion persisted after 2 weeks. Our results suggest that without a strong antiviral titer at
4 weeks, the fish would not have antiviral protection against lethal viral challenge. Also,
our virus neutralization tests showed that there would be no protection in these fish upon
challenge. Our results suggest that although catfish IFN induces some antiviral protein
to provide nonspecific protection 2 weeks post injection, this antiviral protein decreases
between 2 and 4 weeks, so that almost no protection exists by 4 weeks. At 4 weeks,
fish vaccinated with ORF 59 showed an increased antiviral titer from GMT 17 to GMT
71
40. It is unclear why nonspecific protection would increase at 4 weeks, especially when
neutralization titers were so low. Nonspecific protection and Mx protein expression were
seen in fish as long 4 weeks or 35 days [8] [91], but generally, specific protection begins
around this time.
The results for fish that received the ORF 59 DNA vaccination and injection with
poly I:C varied. In the group that received poly I:C 1 hour before the ORF 59 vaccine,
there was a slight increase in the antiviral titer from 2 weeks GMT 80, to 4 weeks GMT
92. In the group where fish first received ORF 59 vaccination 1 hour before poly I:C
injection, the antiviral titer stayed the same at GMT 160. However, in the group where
ORF 59 and poly I:C were administered at the same time, there was a significant decrease
in the antiviral titer from GMT 113 to GMT 35. It is unclear why there would be a
strong decrease in this group while the other two groups that received both the vaccine
and poly I:C do not show a decrease. These results suggest that the most effective way
to vaccinate catfish against CCV is to administer the vaccine and poly I:C at different
times.
Adding poly I:C to our vaccination did seem to increase nonspecific protection when
compared to the ORF 59 vaccination given alone, although it remains unclear whether
this nonspecific protection is due to the poly I:C acting as an adjuvant, or if it was due
to the CpG motif in the plasmid of the vaccine, or if it was due to expressed viral glyco-
proteins. Our negative controls also show some antiviral nonspecific protection, but the
protection is much lower than the positive samples. Likewise, it was shown that a low
amount of Mx protein expression was detected in buffer injected control fish as well as
non-treated fish verifying that low Mx expression was not a reaction to the poly I:C IP
injection [83]. Viral glycoproteins are known inducers of IFN in mammals such as mice
72
[94], while CpG motifs in oligodeoxynucleotides have also been shown to provide pro-
tection shortly after administration to mice [16]. Krieg [16] speculates that CpG motifs,
found in plasmid-based DNA vaccines, induce an innate response that subsequently pro-
motes a specific response, acting as an intrinsic adjuvant. However, Lorenzen et al. [13]
argue that CpG motifs alone are unlikely responsible for the early non-specific antiviral
protection seen in DNA vaccination. In fact, most studies did not show protection in
fish that received an empty DNA vaccine [8], [14], [11]. LaPatra et al. [10] would even
go as far to argue that the early protection seen in DNA vaccinated fish is not unique
to DNA vaccination, but is a feature of the natural response to endogenous expression
of the rhabdo-viral G protein in somatic fish cells. However, Harbottle et al. [91] noted
very low levels of Mx protein expression in their vector-only negative controls, whereas
McLauchlan et al. [12] did not see Mx expression in vector-only negative controls. We
did not include a vector-only negative control in our experiment. Without RNA ev-
idence, CCV challenge data, or vector-only negative controls, we do not know if the
nonspecific protection we saw in our antiviral titers was the result of the CpG motif
found in plasmid-based DNA vaccines, or if the nonspecific protection was the direct
result of the glycoprotein expressed in the vaccine.
In the poly I:C/CCV experiments in which we injected 50 ?g poly I:C into fish,
we detected a significant amount of IFN mRNA expression in the head kidneys. IFN
mRNA expression in catfish induced with 50 ?g poly I:C was seen as early as 1 hour
post induction and lasted 7 hours post induction (our last sampling point), with the
strongest IFN mRNA expression within the first 3 hours (Fig. 4.7). These results are
also consistent with in vitro kinetics, which showed IFN expression occurred rapidly
73
within the first 7 hours after induction before decreasing almost completely by 12 hours
and ceasing by 24 and 48 hours (Fig. 4.2).
The IFN mRNA expression in fish induced with 50 ?g poly I:C correlated with their
IFN titers (Fig. 4.7). The strongest antiviral protection was seen at 1 and 3 hours post
injection with poly I:C (GMT 102 and GMT 81 respectively). At 5 and 7 hours post
injection, antiviral protection decreased slightly, although it was still strong (GMT 64).
Interestingly, this was also seen in the previous experiment in which fish were injected
with only 10 ?g poly I:C. In those assays, the strongest protection was within the first
4 hours at a titer of 64, before decreasing slightly to GMT 32 by 24 and 48 hours. The
antiviral assays for the poly I:C/CCV experiments showed antiviral protection decreased
at 5 hours to 64, which was equivalent to the poly I:C/ORF 59 antiviral titer at 4 hours
(Table 4.2). In both experiments, fish induced with poly I:C showed strongest antiviral
protection within the first several hours and this antiviral protection slowly decreased
over the next 48 hours. This decrease in antiviral protection from both experiments
correlated with the decrease in expression of IFN mRNA in the catfish, as seen when
50 ?g poly I:C was used. Results from the poly I:C/ORF 59 experiment suggest that
protection would continue to decrease over the next 2 weeks, so that at 2 weeks post
injection with poly I:C, there is still some nonspecific innate proteins that would probably
provide protection as seen by Jensen et al. [83].
Even though IFN mRNA expression in catfish induced with 50 ?g poly I:C correlated
with IFN mRNA expression in CCO cells (compare Figures 4.2 and 4.7), the antiviral
titer taken from in vitro samples did not correlate with the antiviral titers taken from
the in vivo samples (compare Figures 4.1 and 4.7). The strongest antiviral protection
from the in vivo assays was within the first several hours, which was also the time
74
the strongest IFN mRNA expression occurred. However, antiviral activity from culture
medium collected from the CCO cells exposed to poly I:C showed that antiviral activity
continued to increase after IFN mRNA expression ceased. Our in vitro results were
similar to work by Long et al. [29], which showed correlation between protection and
IFN mRNA expression. In their results, poly I:C induced IFN mRNA expression only at
2 hours, whereas protection in challenge assays induced strongest protection at 2, 4 and
7 hours post treatment with poly I:C, with 4 hours having the strongest protection [29].
In our in vitro results 8 hours post induction with poly I:C, culture medium collected
from CCO cells showed the strongest antiviral activity at a titer of 1:64, and this only
slightly decreased by 12 hours post induction to 1:32, even though increased IFN mRNA
expression had ceased. Possibly, soluble IFN in the culture medium accumulated at these
time points rather than at earlier time points because maximum IFN mRNA production
by the cells had occurred and IFN was released into the culture medium. Due to the
incubation with poly I:C, cells released expressed IFN that offered protection against
viral challenge to cells. However, Long et al. [48] found that two of the channel catfish
IFN genes lack signal sequences, which they believe causes most of the IFN mRNA
product to remain in the cytoplasm of the cell. Also, the original primers used, (created
by Long et al. [29]), for the CF IFN-1 gene, were able to anneal to the other CF IFN
genes. We were able to show with our IFN-2 specific primers that in vitro the CF IFN-
2 gene, which is not a pseudogene, was being expressed and contributed to providing
protection. Our work agreed with others [29] who demonstrated that the IFN gene was
the factor responsible for antiviral activity and protection in challenge assays. However,
it still remains to be determined whether catfish IFN inhibits viral replication through
the induction of antiviral proteins, although previous work [83] [28], suggests that when
75
most fish are induced with poly I:C to produce IFN, the fish also induces antiviral
proteins.
We also infected channel catfish with CCV to compare the kinetics of IFN mRNA
expression induced with poly I:C or CCV and the protection offered by each (Fig. 4.7).
IFN mRNA expression induced by CCV was comparable to poly I:C induction. Poly I:C
induced IFN expression from 1 to 7 hours post induction, whereas CCV only induced
IFN expression from 3 to 5 hours post infection. Antiviral titers showed that CCV
infection induced less protection than that induced in fish with poly I:C. At 1 hour post
infection with CCV, serum protection was at GMT of 51 and this protection increased
slightly to GMT of 64 at 3 hours. This is the time when an increase in IFN mRNA
expression was first seen. Subsequently, the protection decreased slightly to GMT of
51 and stayed there for 5 to 7 hours. The strongest antiviral protection was seen at 3
hours (GMT 64), which correlated with increased IFN mRNA expression. Poly I:C is
possibly a better inducer of IFN mRNA expression and nonspecific antiviral protection.
Although no increased IFN mRNA expression was observed in fish at 1 and 7 hours post
CCV infection, protection was still strong, indicating some innate defense was at work.
The kinetics of IFN induction following CCV infection in vivo in channel catfish were
compared with those following CCV infection in vitro in CCO cells (compare Figures 4.7
and 4.5). When the CCO cells were infected with CCV, they expressed incrased levels
of IFN mRNA at 1, 4 and 5 hours post infection with CCV, although not at 2, 3 and 6
hours. Similarly, when channel catfish were infected with CCV, they showed increased
IFN mRNA expression at 3 and 5 hours, although not at 1 and 7 hours. We did not
collect medium from CCO cells infected with CCV to run in an antiviral assay, however,
76
previous studies by Chinchar et al. [26] indicate that virally infected cell medium can
provide protection in challenge assays due to the induction of IFN.
Also worth noting was that IFN-2 mRNA was expressed in in vitro studies with
CCO cells (Fig. 4.4). We designed IFN-2 primers unique to the catfish IFN-2 mRNA
and used them to detect differences in kinetics of IFN-2 expression and expression of
other catfish IFN genes detected with the original primers designed by Long et al. [29].
The primers designed by Long et al. were later found to anneal to all four IFN genes [48].
We were able to induce IFN-2 mRNA expression by both poly I:C and UV inactivated
CCV in CCO cells. The UV irradiation did not completely inactivate all of mRNA from
the CCV, as seen by the ORF 6 and ORF 53 rt-PCR results, although IFN-2 mRNA
synthesis was enabled along with expression of all the catfish IFNs as seen with using
the general primers that anneal to all four catfish genes.
Live CCV did not induce IFN-2 mRNA expression in CCO cells or channel catfish.
A possible explanation for this is that CCV has evolved a way to elude the IFN system,
specifically activation of the IFN-2 gene. It could be that CCV genome, as with other
large DNA viruses, may encode proteins that inhibit IFN induction [26]. We would need
to perform more experiments to further understand why live CCV did not induce IFN-2
expression in our cells or catfish. Long et al. [48], found that the catfish IFN-2 gene
was upregulated due to live virus infection in cells, however they used catfish reovirus
(CRV). They also found that catfish IFN-2 gene expression was induced in response to
UV-CRV exposure.
Unlike in vitro experiments in CCO cells, we were unable to show induction of
IFN-2 mRNA expression by poly I:C in channel catfish. Possibly, the IFN-2 gene is
not expressed in the head kidneys of channel catfish. Alternatively, we could have been
77
unable to find IFN-2 mRNA expression because the amount of poly I:C used was too
low to induce expression. Catfish IFN mRNA expression was only detected in vivo when
the general IFN-1 primers, which anneal to all four genes were used (designed by [29]).
This could be due to the expression of pseudogenes. We would need to re-run these
experiments and clone our rt-PCR products to know for certain which IFN genes were
being expressed. Also, at the time these experiments were being performed, we did not
know that the primers were annealing to all four genes and that IFN-1 was a pseudogene.
In all experiments, controls were used to measure constitutive IFN expression in
rt-PCRs and challenge assays. What is interesting in these experiments is the amount
of IFN mRNA the experimental control expressed, since this is the gauge to compare
induced to non-induced CCO cells and catfish IFN mRNA expression. Long et al. [29]
showed that fibroblast and T cell lines constitutively synthesize low levels of IFN mRNA,
while Nygaard et al. [28] found that unstimulated macrophages produced Mx proteins
because of constitutive IFN production that was up-regulated by culturing of the cells.
In contrast, Chinchar et al. [26] found that newly established T cell and macrophage cell
lines did not allow CCV replication, which they believe to be due to the generation of
an anti-viral factor [26]. As the T cell/macrophage lines are maintained in cell culture,
they gradually became more permissive for CCV infection. This increased susceptibility
was caused by the loss of synthesis of ?non-essential proteins? (or cytokines) [26]. Other
studies have confirmed this finding, showing that long-term catfish lymphoid cell lines,
which initially secrete compounds functionally homologous to IL-2 and IL-4, eventually
lose their ability to synthesize and/or secrete these ?luxury proteins? after they have
been passed for a while [26]. In other words, the cells begin to lose their differentiation.
In our in vitro experiments with CCO cells, cells in culture for more than 4 days and
78
nearly confluent monolayers were more likely than less confluent cells to express IFN,
and displayed higher background level of IFN mRNA expression, as seen in some of our
rt-PCR and challenge assay experiments. Thus, the age and confluency of the cultures
contribute to IFN mRNA expression. Possibly, older confluent cells become ?stressed?
when they run out of room to grow and divide in flasks, and therefore, induce a low level
of IFN protection. Also, in CCO cells it is possible that constitutive levels of IFN mRNA
expression showed up because we used a larger number of amplification cycles in rt-PCR
protocols. Long et al. [29] found that by increasing the number of amplification cycles
in rt-PCR protocols to greater than 15, their seemingly negative fibroblast CCO lines
constitutively expressed low levels of IFN mRNA. Our protocols always used greater
than 15 amplification cycles. IFN is expressed within the first 12 hours of induction
and after 12 hours, IFN mRNA is expressed at levels less than or equal to constitutively
expressed control levels. We also saw that live catfish express a low level of constitutive
IFN as well. All of the experimental control fish expressed some small amount of IFN
mRNA and showed some level of antiviral protection. Even when fish were not injected
or vaccinated, they expressed a low background level of IFN mRNA. An explanation
for IFN expression in control fish is that IFN is constitutively expressed or continuously
induced due to exposure to viral particles present in water [83]. Likewise, low levels of
IFN-? mRNA are known to be constitutively produced in organs and in peripheral blood
monocytes of healthy people [96] [97].
79
Chapter 6
Overall Conclusions
We found that both 10 and 50 ?g/ml poly I:C act similarly in CCO cells, inducing
IFN mRNA expression within the first 7 hours post incubation with poly I:C. The IFN-2
gene was cloned and sequenced from CCO cells, and we found that this gene is strongly
expressed within the first 3 hours after induction with poly I:C, and that CCV infection
prevents IFN-2 mRNA expression. The minimum amount of poly I:C that can be used
to provide consistent IFN mRNA expression in vivo is 50 ?g poly I:C, because less poly
I:C does not induce IFN mRNA expression consistent with in vitro results. We were
unable to induce the INF 2 gene in vivo with poly I:C, although poly I:C does induce
IFN-2 mRNA expression in vitro in CCO cells. CCV was also able to induce IFN mRNA
expression in vivo, although not as well as poly I:C.
Our vaccination project showed that inducing IFN expression with poly I:C along
with ORF 59 vaccination does not increase the effectiveness of vaccination, although
there was some evidence of nonspecific protection. Our neutralization virus results were
consistent with Harbottle et al. [91] who also found that neutralization titers are low 4
to 6 weeks post vaccination.
80
Bibliography
[1] Robertsen B. Modulation of the non-specific defense of fish by structurally con-
served microbial polymers. Fish & Shellfish Immunology, 9:269?290, 1999.
[2] Ellis AE. Innate host mechanisms of fish against viruses and bacteria. Develop-
mental & Comparative Immunology, 25:827?839, 2001.
[3] Nusbaum KE, Smith BF, DeInnocentes P, and Bird RC. Protective immunity in-
duced by DNA vaccination of channel catfish with early and late transcripts of the
channel catfish herpervirus (IHV-1). Veterinary Immunology and Immunopathol-
ogy, 6545:1?18, 2001.
[4] Davison AJ. Channel catfish virus: a new type of herpesvirus. Virology, 186:9?14,
1992.
[5] Davison AJ and Davison MD. Identification of structural proteins of channel catfish
virus by mass spectrometry. Virology, 206:1035?1043, 1995.
[6] Diagnostic Manual for Aquatic Animal Diseases. Channel Catfish Virus Disease.
http://www.oie.int/eng/normes/fmanual/A summry.htm, (accessed on June 30,
2003), 2000.
[7] Hogan RJ, Stuge TB, Clem LW, Miller NW, and Chinchar VG. Anti-viral cytotoxic
cells in the channel catfish (Ictalurus punctatus). Developmental & Comparative
Immunology, 20:115?127, 1996.
[8] Kim CH, Johnson MC, Drennan JD, Simon BE, Thomann E, and Leong JC. DNA
vaccines encoding viral glycoproteins induce nonspecific immunity and Mx protein
synthesis in fish. Journal of Virology, 74:7048?7054, 2000.
[9] LaPatra SE, Corbeil S, Jones GR, Shewmaker WD, and Kurath G. The dose-
dependent effect on protection and humoral response to a DNA vaccine against
infectious hematopoietic necrosis (IHN) virus in subyearling rainbow trout. Journal
of Aquatic Animal Health, 12:181?188, 2000.
[10] LaPatra SE, Corbeil S, Jones GR, Shewmaker WD, Lorenzen N, Anderson ED, and
Kurath G. Protection of rainbow trout against infectious hematopoietic necrosis
virus four days after specific or semi-specific DNA vaccination. Vaccine, 19:4011?
4019, 2001.
81
[11] Lorenzen E, Einer-Jensen K, Martinussen T, LaPatra SE, and Lorenzen N. DNA
vaccination of rainbow trout against viral hemorrhagic septicemia virus: a dose-
response and time-course study. Journal of Aquatic Animal Health, 12:167?180,
2000.
[12] McLauchlan PE, Collet B, Ingerslev E, Secombes CJ, Lorenzen N, and Ellis AE.
DNA vaccination against viral haemorrhagic septicaemia (VHS) in rainbow trout:
size, dose, route of injection and duration of protection?early protection correlates
with Mx expression. Fish & Shellfish Immunology, 15:39?50, 2003.
[13] Lorenzen N, Lorenzen E, Einer-Jensen K, and LaPatra SE. DNA vaccines as a tool
for analysing the protective immune response against rhabdoviruses in rainbow
trout. Fish & Shellfish Immunology, 12:439?453, 2002.
[14] Lorenzen N, Lorenzen E, Einer-Jensen K, and LaPatra SE. Immunity induced
shortly after DNA vaccination of rainbow trout against rhabdoviruses protects
against heterologous virus but not against bacterial pathogens. Developmental &
Comparative Immunology, 26:173?179, 2002.
[15] Lorenzen N and LaPatra SE. DNA vaccines for aquacultured fish. Rev. sci. tech.
Epiz., 24:201?213, 2005.
[16] Krieg AM. The role of CpG motifs in innate immunity. Current Opinion in
Immunology, 12:35?43, 2000.
[17] Anderson ED, Mourich DV, Fahrenkrug SC, LaPatra S, Shepherd J, and Leong JA.
Genetic immunization of rainbow trout (Oncorhynchus mykiss) against infectious
hematopoietic necrosis virus. Molecular Marine Biology and Biotechnology, 5:114?
122, 1996.
[18] Roy SW, Fedorov A, and Gilbert W. Large-scale comparison of intron positions
in mammalian genes shows intron loss but no gain. Proceedings of the National
Academy of Sciences of the USA, 100:7158?7162, 2003.
[19] Hedrick RP, LaPatra SE, Yun S, Lauda KA, Jones GR, Congleton JL, and
de Kinkelin P. Induction of protection from infectious hematopoietic necrosis virus
in rainbow trout Oncorhynchus mykiss by pre-exposure to the avirulent cutthroat
trout virus. Diseases of Aquatic Organisms, 20:111?118, 1994.
[20] Robertsen B, Engstad R, and Jorgensen JB. ?-Glucans as immunostimulants in
fish. In Modulators of Fish Immune Response, pages 83?99, 1994.
[21] Isaacs A and Lindenmann J. Virus interference I. The interferon. Proceedings of
the Royal Society London B: Biological Sciences, 147:258?267, 1957.
[22] Samuel CE. Antiviral actions of interferons. Clinical Microbiology Reviews, 14:778?
809, 2001.
82
[23] Taniguchi T, Mantei N, Schwarzstein M, Nagata S, Muramatsu M, and Weissmann
C. Human leukocyte and fibroblast interferons are structurally related. Nature,
285:547?549, 1980.
[24] Taniguchi T, Fujii-Kuriyama Y, and Muramatsu M. Molecular cloning of human
interferon cDNA. Proceedings of the National Academy of Sciences USA, 77:4003?
4006, 1980.
[25] Gravell M and Malsberger RG. A permanent cell line from the fathead minnow
(Pimephales promelas). Annals of the New York Academy of Sciences, 126:555?565,
1965.
[26] Chinchar VG, Logue O, Antao A, and Chinchar GD. Channel catfish reovirus
(CRV) inhibits replication of channel catfish herpesvirus (CCV) by two distinct
mechanisms: viral interference and induction of an anti-viral factor. Diseases of
Aquatic Organisms, 33:77?85, 1998.
[27] LaPatra SE, Lauda KA, and Jones GR. Aquareovirus interference mediated resis-
tance to infectious hematopoietic necrosis virus. Veterinary Research, 26:455?459,
1995.
[28] Nygaard R, Husgard S, Sommer A, Leong JC, and Robertsen B. Induction of Mx
protein by interferon and double stranded RNA in salmonid cells. Fish & Shellfish
Immunology, 10:435?450, 2000.
[29] Long S, Wilson M, Bengten E, Bryan L, Clem LW, Miller NW, and Chinchar VG.
Identification of a cDNA encoding channel catfish interferon. Developmental &
Comparative Immunology, 28:97?111, 2004.
[30] Altmann SM, Mellon MT, Distel DL, and Kim CH. Molecular and functional
analysis of interferon gene from the zebrafish, Danio rerio. Journal of Virology,
77:1992?2002, 2003.
[31] Lutfalla G, Crollius HR, Stange-Thomann N, Jaillon O, Mogensen K, and Mon-
neron D. Comparative genomic analylsis reveals independent expansion of a
lineage-specific gene family in vertebrates: the class II cytokine receptors and their
ligands in mammals and fish. BMC Genomics, 4:29, 2003.
[32] Robertsen B, Bergan V, Rokenes T, Larsen R, and Albuquerque A. Atlantic
salmon interferon genes: cloning, sequence analysis, exression, and biological ac-
tivity. Journal of Interferon & Cytokine Research, 23:601?612, 2003.
[33] Pestka S, Langer JA, Zoon KC, and Samuel CE. Interferons and their actions.
Annual Review of Biochemistry, 56:727?777, 1987.
83
[34] Kotenko SV, Gallagher G, Baurin VV, Lewis-Antes A, Shen M, Shah NK, Langer
JA, Sheikh F, Dickensheets H, and Donnelly RP. IFN-lambdas mediate antiviral
protection through a distinct class II cytokine receptor complex. Nature Immunol-
ogy, 4:69?77, 2003.
[35] Chen J, Baig E, and Fish E. Diversity and relatedness among the type I interferons.
Journal of Interferon & Cytokine Research, 24:687?698, 2004.
[36] Pestka S, Krause CD, and Walter MR. Interferons, interferon-like cytokines, and
their receptors. Immunological Reviews, 202:8?32, 2004.
[37] Martal JL, Chene NM, Huynh LP, L?Haridon RM, Reinaud PB, Guillomot MW,
Charlier MA, and Charpigny SY. IFN-tau: a novel subtype I IFN1. Struc-
tural characteristics, non-ubiquitious expression, structure-function relationships,
a pregnancy hormonal embryonic signal and cross-species therapeutic potentiali-
ties. Biochimie, 80:755?777, 1998.
[38] LaFleur DW, Nardelli B, Tsareva T, Mather D, Feng P, Semenuk M, Taylor K,
Buergin M, Chinchilla D, Roshke V, Chen G, Ruben SM, Pitha PM, Coleman TA,
and Moore PA. Interferon-kappa, a novel type I interferon expressed in human
keratinocytes. Journal of Biological Chemistry, 276:39765?39771, 2001.
[39] Oritani K, Medina KL, Tomiyama Y, Ishikawa J, Okajima Y, Ogawa M, Yokota
T, Aoyama K, Takahashi I, Kincade PW, and Matsuzawa Y. Limitin: an
interferon-like cytokine that preferentially influences B-lymphocyte precursors. Na-
ture Medicine, 6:659?666, 2000.
[40] Lefevre F and Boulay V. A novel and atypical type one interferon gene expressed by
trophoblast during early pregnancy. Journal of Biological Chemistry, 268:19760?
19768, 1993.
[41] Karpusas M, Nolte M, Benton CB, Meier W, Lipscomb WN, and Goelz S. The
crystal structure of human interferon beta at 2.2-A resolution. Proceedings of the
National Academy of Sciences USA, 94:11813?11818, 1997.
[42] Radhakrishnan R, Walter LJ, Hruza A, Reichert P, Trotta PP, Nagabhushan TL,
and Walter MR. Zinc mediated dimer of human interferon-? 2b revealed by X-ray
crystallography. Structure, 4:1453?1463, 1996.
[43] Radhakrishnan R, Walter LJ, Subramaniam PS, Johnson HM, and Walter MR.
Crystal structure of ovine interferon-? at 2.1 A resolution. Journal of Molecular
Biology, 286:151?162, 1999.
[44] Senda T, Shimazu T, Matsuda S, Kawano G, Shimizu H, Nakamura KT, and Mitsui
Y. Three-dimentional crystal structure of recombinant murine interferon-?. EMBO
Journal, 11:3193?3201, 1992.
84
[45] Levy DE and Garcia-Sastre A. The virus battles: IFN induction of the antiviral
state and mechanisms of viral evasion. Cytokine & Growth Factor Reviews, 12:143?
156, 2001.
[46] Collet B and Secombes CJ. Type I-interferon signaling in fish. Fish & Shellfish
Immunology, 12:389?397, 2002.
[47] Haller O and Kochs G. Interferon-induced Mx proteins: dynamin-like GTPases
with antiviral activity. Traffic, 3:710?717, 2002.
[48] Long S, Milev-Milovanovic I, Wilson M, Bengten E, Clem LW, Miller NW, and
Chinchar VG. Identification and expression analysis of cDNAs encoding channel
catfish type I interferons. Fish & Shellfish Immunology, 21:42?59, 2006.
[49] Robertsen B. The interferon system of teleost fish. Fish & Shellfish Immunology,
20:172?191, 2006.
[50] Waine GJ, Tymms MJ, Brandt ER, Cheetham BF, and Linnane AW. Structure-
fuction study of the region encompassing residues 26-40 of human interferon-?4:
identification of residues important for antiviral and antiproliferative activities.
Journal of Interferon Research, 12:43?48, 1992.
[51] Roberts RM, Liu L, Guo Q, Leaman D, and Bixby J. The evolution of the type I
interferons. Journal of Interferon & Cytokine Research, 18:805?816, 1998.
[52] Zou J, Yoshiura Y, Dijkstra JM, Sakai M, Ototake M, and Secombes C. Iden-
tification of an interferon gamma homologue in Fugu, Takifugu rubripes. Fish &
Shellfish Immunology, 17:403?409, 2004.
[53] Charlemagne J, Fellah JS, De Guerra A, Kerfourn F, and Partula S. T-cell recep-
tors in ectothermic vertebrates. Immunology Reviews, 166:87?102, 1998.
[54] Hansen JD and Strassburger P. Description of an ectothermic TCR coreceptor,
CD8 alpha, in rainbow trout. Journal of Immunology, 164:3132?3139, 2000.
[55] Suetake H, Araki K, and Suzuki Y. Cloning, expression, and characterization of
fugu CD4, the first ectothermic animal CD4. Immunogenetics, 56:368?374, 2004.
[56] Nakanishi T, Fischer U, Dijkstra JM, Hasegawa S, Somamoto T, Okamoto N,
and Ototake M. Cytotoxic T cell function in fish. Developmental & Comparative
Immunology, 26:131?139, 2002.
[57] Yoshiura Y, Kiryu I, Fujiwara A, Suetake H, Suzuki Y, Nakanishi T, and Oto-
take M. Identification and characterization of fugu orthologues of mammalian
interleukin-12 subunits. Immunogenetics, 55:296?306, 2003.
85
[58] Zou J, Bird S, Truckle J, Bols N, Horne M, and Secombes C. Identification and
expression analysis of an IL-18 homologue and its alternatively spliced form in
rainbow trout Oncorhynchus mykiss. European Journal of Biochemistry, 271:1913?
1923, 2004.
[59] Graham S and Secombes CJ. The production of a macrophage-activating factor
from rainbow trout Salmo gairdneri leucocytes. Immunology, 65:293?297, 1988.
[60] Wang T, Ward M, Grabowski P, and Secombes CJ. Molecular cloning, gene orga-
nization and expression of rainbow trout (Oncorhynchus mykiss) inducible nitric
oxide synthase (iNOS) gene. Biochemical Journal, 358:747?755, 2001.
[61] Chang MS, Chang GD, Leu JH, Huang FL, Chou CK, Huang CJ, and Lo TB.
Expression, characterization, and genomic structure of carp JAK1 kinase gene.
DNA and Cell Biology, 10:827?844, 1996.
[62] Conway G, Margoliath A, Wong-Madden S, Roberts RJ, and Gilbert W. Jak1
kinase is required for cell migrations and anterior specification in zebrafish embryos.
Proceedings of the National Academy of Sciences of the USA, 94:3082?3087, 1997.
[63] Leu JH, Chang MS, Yao CW, Chou CK, Chen ST, and Huang CJ. Genomic organi-
zation and characterization of the promoter region of the round-spotted pufferfish
(Tetraodon fluviatilis) JAK1 kinase gene. Biochimica et Biophysica Acta, 1395:50?
56, 1998.
[64] Leu JH, Yan SJ, Lee TF, Chou CM, Chen ST, Hwang PP, Chou CK, and Huang
CJ. Complete genomic organization and promoter analysis of the round-spotted
pufferfish JAK1, JAK2, JAK3, and TYK2 genes. DNA and Cell Biology, 19:431?
446, 2000.
[65] Oates AC, Wollberg P, Pratt SJ, Paw BH, Johnson SL, Ho RK, Postlethwait JH,
Zon LI, and Wilks AF. Zebrafish stat3 is expressed in restricted tissues during
embryogenesis and stat1 rescues cytokine signaling in a STAT1-deficient human
cell line. Developmental Dynamics, 215:352?370, 1999.
[66] Sung SC, Fan TJ, Chou CM, Leu JH, Hsu YL, Chen ST, Hsieh YC, and Huang CJ.
Genomic structure, expression and characterization of a STAT5 homologue from
pufferfish (Tetraodon fluviatilis). European Journal of Biochemistry, 270:239?252,
2003.
[67] Zhang Y and Gui J. Molecular characterization and IFN signal pathway analysis
of Carassius auratus CaSTAT1 identified from the cultured cells in response to
virus infection. Developmental & Comparative Immunology, 28:211?227, 2004.
[68] Staeheli P, Yu YX, Grob R, and Haller O. A double-stranded RNA-inducible fish
gene homologous to the murine influenza virus resistance gene Mx. Molecular and
Cellular Biology, 9:3117?3121, 1989.
86
[69] Trobridge GD and Leong JAC. Characterization of a rainbow trout Mx gene.
Journal of Interferon & Cytokine Research, 15:691?702, 1995.
[70] Trobridge GD, Chiou PP, and Leong JC. Cloning of the rainbow trout (On-
corhynchus mykiss) Mx2 and Mx3 cDNAs and characterization of trout Mx protein
expression in salmon cells. Journal of Virology, 71:5304?5311, 1997.
[71] Robertsen B, Trobridge G, and Leong J. Molecular cloning of double-stranded
RNA inducible Mx genes from Atlantic salmon (Salmo salar L.). Developmental
& Comparative Immunology, 21:397?412, 1997.
[72] Jensen V and Robertsen B. Cloning of an Mx cDNA from Atlantic halibut (Hip-
poglussus hippoglossus) and characterization of Mx mRNA expression in response
to double-stranded RNA or infectious pancreatic necrosis virus. Journal of Inter-
feron & Cytokine Research, 20:701?710, 2000.
[73] Lee JY, Hirono I, and Aoki T. Cloning and analysis of expression of Mx cDNA in
Japanese flounder, Paralichtys olivaceus. Developmental & Comparative Immunol-
ogy, 24:407?415, 2000.
[74] Yap WH, Tay A, Brenner S, and Venkatesh B. Molecular cloning of the pufferfish
(Takifugu rubripes) and functional characterization of its promoter. Immunogenet-
ics, 54:705?713, 2003.
[75] Zhang YB, Li Q, and Gui JF. Differential expression of two Carassius auratus Mx
genes in cultured CAB cells induced by grass carp hemorrhage virus and interferon.
Immunogenetics, 56:68?75, 2004.
[76] Plant KP and Thune RL. Cloning and characterisation of a channel catfish (Ic-
talurus punctatus) Mx gene. Fish & Shellfish Immunology, 16:391?405, 2004.
[77] Tafalla C, Aranguren R, Secombes CJ, Figueras A, and Novoa B. Cloning and
analysis of expression of a gilthead sea bream (Sparus aurata) Mx cDNA. Fish &
Shellfish Immunology, 16:11?24, 2004.
[78] Altmann SM, Mellon MT, Johnson MC, Paw BH, Trede NS, Zon LI, and Kim CH.
Cloning and characterization of an Mx gene and its corresponding promoter from
the zebrafish, Danio rerio. Developmental & Comparative Immunology, 28:295?
306, 2004.
[79] Chen YM, Su YL, Lin JH, Yang HL, and Chen TY. Cloning of an orange-spotted
grouper (Epinephelus coioides) Mx cDNA and characterisation of its expression in
response to nodavirus. Fish & Shellfish Immunology, 20:58?71, 2006.
[80] Larsen R, Rokenes TP, and Robertsen B. Inhibition of infectious pancreatic necro-
sis virus replication by Atlantic salmon Mx1 protein. Journal of Virology, 78:7938?
7944, 2004.
87
[81] Bergan V and Robertsen B. Characterization of Atlantic halibut (Hippoglossus
hippoglossus) Mx protein expression. Developmental & Comparative Immunology,
28:1037?1047, 2004.
[82] Collet B and Secombes CJ. The rainbow trout (Oncorhynchus mykiss) Mx1 promo-
tor. Structural and functional characterization. European Journal of Biochemistry,
268:1577?1584, 2001.
[83] Jensen I, Albuquerque A, Sommer AI, and Robertsen B. Effect of poly I:C on
the expression of Mx proteins and resistance against infection by infectious salmon
anaemia virus in Atlantic salmon. Fish & Shellfish Immunology, 13:311?326, 2002.
[84] Jensen I and Robertsen B. Effect of double-stranded RNA and interferon on the
antiviral activity of Atlantic salmon cells against infectious salmon anemia virus
and infectious pancreatic necrosis virus. Fish & Shellfish Immunology, 13:221?241,
2002.
[85] Caipang CM, Hirono I, and Aoki T. In vitro inhibition of fish rhabodviruses by
Japanese flounder, Paralichthys olivaceus Mx. Virology, 317:373?382, 2003.
[86] Garner JN, Joshi B, and Jagus R. Characterization of rainbow trout and zebrafish
eukaryotic initiation factor 2? and its response to endoplasmic reticulum stress and
IPNV infection. Developmental & Comparative Immunology, 27:217?231, 2003.
[87] Hu CY, Zhang YB, Huang GP, Zhang QY, and Gui JF. Molecular cloning and
characterisation of a fish PKR-like gene from cultured CAB cells induced by UV-
inactivated virus. Fish & Shellfish Immunology, 17:353?366, 2004.
[88] von Wussow P, Jakschies D, Hochkeppel HK, Fibich C, Penner L, and Deicher H.
The human intracellular Mx-homologous protein is specifically induced by type I
interferons. European Journal of Immunology, 20:2015?2019, 1990.
[89] Schumacher B, Bernasconi D, Schultz U, and Staeheli P. The chicken Mx promoter
contains an ISRE motif and confers interferon inducibility to a reporter gene in
chick and monkey cells. Virology, 203:144?148, 1994.
[90] Schmidt NJ and Emmons RW. Calculation of the LD50 by the Karber method. in
Diagnostic Procedures for Viral, Rickettsial, and Chlamydial Infections. American
Public Health Association, 6:20?21, 1989.
[91] Harbottle H, Plant KP, and Thune RL. DNA vaccination against channel catfish
virus results in minimal immune response and is not efficacious against challenge.
Journal of Aquatic Animal Health, 17:251?262, 2005.
[92] Bromage ES, Kaattari IM, Zwollo P, and Kaattari SL. Plasmablast and plasma cell
production and distribution in trout immune tissues. The Journal of Immunology,
173:7317?7323, 2004.
88
[93] Sommerset I, Skern R, Biering E, Bleie H, Fiksdal U, Grove S, and Nerland AH.
Protection against Atlantic halibut nodavirus in turbot is induced by recombinant
capsid protein vaccination but not following DNA vaccination. Fish & Shellfish
Immunology, 18:13?29, 2005.
[94] Ito Y. Induction of interferon by virus glycoprotein(s) in lymphoid cells through in-
teraction with the cellular receptors via lectin-like action: an alternative interferon
induction mechanism. Archives of Virology, 138:187?198, 1994.
[95] Leitner WW, Hwang LN, deVeer MJ, Zhou A, Silverman RH, Williams BRG,
Dubensky TW, Ying H, and Restifo NP. Alphavirus-based DNA vaccine breaks
immunological tolerance by activating innate antiviral pathways. Nature Medicine,
9:33?39, 2002.
[96] Tovey MG, Streuli M, Gresser I, Gugenheim J, and Blanchard B. Interferon mes-
senger RNA is produced constitutively in the organs of normal individuals. Pro-
ceedings of the National Academy of Sciences USA, 84:5038?5042, 1987.
[97] Lallemand C, Lebon P, Rizza P, Blanchard B, and Tovey MG. Constitutive ex-
pression of specific interferon isotypes in peripheral blood leukocytes from normal
individuals and in promonocytic U937 cells. Journal of Leukocyte Biology, 60:137?
146, 1996.
[98] Collet B, Hovens GC, Mazzoni D, Hirono I, Aoki T, and Secombes CJ. Cloning and
expression analysis of rainbow trout Oncorhynchus mykiss interferon regulatory
factor 1 and 2 (IRF-1 and IRF-2). Developmental & Comparative Immunology,
27:111?126, 2003.
[99] Alonso M, Johnson M, Simon B, and Leong J. A fish specific expression vec-
tor containing the interferon regulatory factor 1A (IRF1A) promoter for genetic
immunization of fish. Vaccine, 21:1591?1600, 2003.
[100] Alonso M and Leong J. Suppressive subtraction libraries to identify interferon-
inducible genes in fish. Marine Biotechnology, 4:74?80, 2002.
[101] Aparicio S, Chapman J, Stupka E, Putnam N, Chia JM, Dehal P, Christoffels A,
Rash S, Hoon S, Smit A, Gelpke MDS, Roach J, Oh T, Ho IY, Wong M, Detter C,
Verhoef F, Predki P, Tay A, Lucas S, Richardson P, Smith SF, Clark MS, Edwards
YJK Doggett N, Zharkikh A, Tavtigian SV, Pruss D, Barnstead M, Evans C,
Baden H, Powell J, Glusman G, Rowen L, Hood L, Tan YH, Elgar G, Hawkins T,
Venkatesh B, Rokhsar D, and Brenner S. Whole-genome shotgun assembly and
analysis of the genome of (Fugu rubripes). Science, 297:1301?1310, 2002.
[102] Awad MA, Nusbaum KE, and Brady YJ. Preliminary studies of a newly developed
subunit vaccine for channel catfish virus disease. Journal of Aquatic Animal Health,
1:233?237, 1989.
89
[103] Bazzigher L, Pavlovic J, Haller O, and Staeheli P. Mx genes show weaker primary
response to virus than other interferon-regulated genes. Virology, 186:154?160,
1992.
[104] Brierley MM and Fish EN. IFN-?/? receptor interactions to biolgic outcomes:
understanding the circuitry. Journal of Interferon & Cytokine Research, 22:835?
845, 2002.
[105] Grunewald K, Desai P, Winkler DC, Heymann JB, Belnap DM, Baumeister W, and
Steven AC. Three-dimensional structure of herpes simplex virus from cyro-electron
tomography. Science, 302:1396?1398, 2003.
[106] Hedrick RP and McDowell T. Passive transfer of sera with antivirus neutralizing
activity from adult channel catfish protects juveniles from channel catfish virus
disease. Transactions of the American Fisheries Society, 116:277?281, 1987.
[107] Hedrick RP, Groff JM, and McDowell T. Response of adult channel catfish to wa-
terborne exposure of channel catfish virus. The Progressive Fish-Culturist, 49:181?
187, 1987.
[108] Jordan IK, Makarova KS, Spouge JL, Wolf YI, and Koonin EV. Lineage-specific
gene expressions in bacterial and archaeal genomes. Genome Research, 11:555?565,
2001.
[109] Kochs G and Haller O. Interferon-induced human MxA GTPase blocks nuclear
import of Thogoto virus nucleocapsids. Proceedings of the National Academy of
Sciences of the USA, 96:2082?2086, 1999.
[110] LaPatra SE, Lauda KA, Jones GR, Shewmaker WS, and Bayne CJ. Resistance to
IHN virus infection in rainbow trout is increased by glucan while subsequent pro-
duction of serum neutralizing activity is decreased. Fish & Shellfish Immunology,
8:435?446, 1998.
[111] Lunde H and Robertsen B. HP8 The complement component C3 is produced by
hepatocytes and behaves as an acute-phase protein in Atlantic salmon, Salmo salar
L., after injection of ?-glucan. Developmental & Comparative Immunology, 21:150,
1997.
[112] Noga EJ and Hartmann JX. Establishment of walking catfish (Clarius batrachus)
cell lines and development of a channel catfish (Ictalurus punctatus) virus vaccine.
Canadian Journal of Fisheries and Aquatic Sciences, 38:925?930, 1981.
[113] Shen L, Stuge TB, Zhou H, Khayat M, Barker KS, Quiniou SM, Wilson M, Bengten
E, Chinchar VG, Clem LW, and Miller NW. Channel catfish cytotoxic cells: a
mini-review. Developmental & Comparative Immunology, 26:141?149, 2002.
90
[114] Stark GR, Kerr IM, Williams BR, Silverman RH, and Schreiber RD. How cells
respond to interferons. Annual Review of Biochemistry, 67:227?264, 1998.
[115] Zhang HG, Hanson LA, and Kancharla SR. Channel catfish virus recombinants
display reduced pathogenesis yet induce specific immunity in channel catfish fin-
gerlings. Proceedings of the International Symposium on Aquatic Animal Health,
1994.
[116] Abbas AK, Lichtman AH, and Pober JS. Cellular and Molecular Immunology.
W.B. Saunders Company, 5 edition, 2003.
[117] Janeway C, Shlomchik MJ, Walport M, and Travers P. Immunobiology. Garland,
6 edition, 2005.
91