|Home | About | Journals | Submit | Contact Us | Français|
Koi herpesvirus (KHV) has recently been classified as a member of the family of Alloherpesviridae within the order of Herpesvirales. One of the unique features of Herpesviridae is latent infection following a primary infection. However, KHV latency has not been recognized. To determine if latency occurs in clinically normal fish from facilities with a history of KHV infection or exposure, the presence of the KHV genome was investigated in healthy koi by PCR and Southern blotting. KHV DNA, but not infectious virus or mRNAs from lytic infection, was detected in white blood cells from investigated koi. Virus shedding was examined via tissue culture and reverse transcription-PCR (RT-PCR) testing of gill mucus and feces from six koi every other day for 1 month. No infectious virus or KHV DNA was detected in fecal secretion or gill swabs, suggesting that neither acute nor persistent infection was present. To determine if KHV latent infections can be reactivated, six koi were subjected to a temperature stress regime. KHV DNA and infectious virus were detected in both gill and fecal swabs by day 8 following temperature stress. KHV DNA was also detectable in brain, spleen, gills, heart, eye, intestine, kidney, liver, and pancreas in euthanized koi 1 month post-temperature stress. Our study suggests that KHV may become latent in leukocytes and other tissues, that it can be reactivated from latency by temperature stress, and that it may be more widespread in the koi population than previously suspected.
Koi herpesvirus (KHV), a newly identified virus, is highly contagious to fish and may cause high mortality (80 to 100%) in common carp and koi (Cyprinus carpio) (2, 10, 11). The first outbreak of KHV was reported in 1998 in Israel (7). Since then, KHV infections have been reported in the United States, Europe, and Asia (2, 10–12). This viral disease affects fish of various ages; however, it causes higher mortality in fry than in older fish (5, 22). The clinical signs of KHV infection include red and white mottling of the gills, gill hemorrhage, sunken eyes, and pale patches or blisters on the skin (7). The virus can be found in the kidney, gill, spleen, fin, intestine, and brain (8). In experimental studies, 82% of fish exposed to the virus at a water temperature of 22°C (which is 7 to 9°C above their normal environmental temperature of 13 to 15°C) died within 15 days (21).
The complete genome has been sequenced from three KHV strains from Japan, the United States, and Israel (1). The KHV genome is about 295 kbp, and contains a 22-kbp terminal direct repeat. KHV thus has the largest genome reported to date for this family (1). KHV is also known as cyprinid herpesvirus 3 (CyHV3) and has been proposed to be a member of Alloherpesviridae family in the order Herpesvirales (31). Alloherpesviruses are distinct and highly diverged from both the Herpesviridae and Malacoherpesviridae. KHV is closely related to CyHV1 (carp pox herpesvirus) and CyHV2 (goldfish hematopoietic necrosis virus).
One of the unique features of Herpesviridae is latency. Latency is the most remarkable property of herpesviruses, ensuring the maintenance of their genetic information in their hosts for an extended period in the absence of productive infection (20, 25). There are three subfamilies within Herpesviridae: Alphaherpesvirinae, Betaherpesvirinae, and Gammaherpesvirinae (25). Members of the three herpesvirus subfamilies infect a wide variety of target cells and are able to establish latent infection, which is associated with a restricted gene expression of the viral genome with no production of infectious virus (20). Most alphaherpesviruses become latent in the peripheral nervous system, such as sensory ganglia and dorsal root ganglia (15, 33). The betaherpesviruses become latent in bone marrow, lymphoid tissue, and kidney (3, 20). Some gammaherpesvirus become latent in splenic B cells (20, 23).
Another member of the Alloherpesviridae family, the channel catfish virus (CCV), has been suggested to become latent following a primary infection, based on detection of CCV DNA in peripheral blood leukocytes (9, 29). However, the question of whether KHV becomes latent or persists following a primary infection remains unresolved (4, 24, 27, 28). There is evidence showing that asymptomatic fish are KHV carriers (30). It is important to know whether KHV can become latent and reactivate from latency as this characteristic would not only constitute a cornerstone for developing detection and control strategies for the virus but also shed light on our understanding of herpesvirus evolution. Currently, the only method to screen KHV infection is via detection of KHV antibody by standard serum neutralization assay or by KHV antigen-specific enzyme-linked immunosorbent assay (ELISA) (27), which may be limited by test sensitivity. In this study, we examined the status of KHV latency in clinically normal koi with a history of probable exposure to KHV.
Six koi (designated K1 to K6) between 2 and 15 years old were from premises that had previous KHV infections or exposure. Three (K2, K4, and K5) were recently imported koi that had been housed in the same quarantine facility. Prior to donation, K4 and K6 tested positive for KHV by serum antibody ELISAs carried out at the Immunology and Virology Laboratory, Veterinary Medicine Teaching Hospital, University of California, Davis (29). The remaining three koi (K1, K2, and K5) were survivors from a pond associated with a suspected KHV outbreak in 1998 and a confirmed KHV outbreak in 2003, based on a positive PCR test at the University of Georgia Infectious Disease Laboratory. To investigate whether KHV becomes latent in the peripheral leukocytes, 0.5- to 2.0-ml blood samples from fish K1 to K6 were collected and stored in EDTA tubes at 2 weeks, 1 month, and 2 months following arrival of the fish at the Oregon State University, Salmon Disease Research Lab (OSU-SDL). Three sets of blood samples from these six koi were collected to ensure the consistency of KHV genome detection as persistence of the genome over time is a characteristic of latency. The OSU-SDL is specifically designed for conducting in vivo experiments with infectious diseases. The incoming water is from a deep well and is pretreated with UV irradiation and is thus not a source for KHV.
Five 2-year-old koi were obtained from facilities with no known history of KHV problems, and these were designated KI to KVI. An additional four 2-year old koi were obtained from a local pet store, and these were designated KVI to KIX. These four koi were home bred by a local pet owner, and there are no records of KHV infection in the breeding facility. Both groups of koi were kept segregated and maintained at 12°C in 4-ft-diameter tanks at OSU-SDL in accordance with the Animal Care and Use Committee regulations. All blood samples were collected via caudal vein puncture after the koi were anesthetized with MS-222 (100 ppm). To investigate whether KHV becomes latent in certain tissues, all koi in this study were euthanized via MS-222 (500 ppm) overdose. Tissues, including the brain, spleen, gills, heart, eye, intestine, kidney, liver, and pancreas, were collected at necropsy.
Both a common carp brain (CCB) cell line and koi fin cell line (KF-1) (gift of Ronald Hedrick, University of California, Davis) were cultured in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Gemini Bio-Products, West Sacramento, CA), penicillin (100 U/ml), and streptomycin (100 μg/ml) (Sigma-Aldrich, Inc., St. Louis, MO) and incubated at 22°C. The strains of KHV from the United States and Israel (KHV-U and KHV-I, respectively) were gifts of Ronald Hedrick.
The presence of KHV DNA or infectious virus in body secretions is indicative of persistent infection. To determine whether persistent infection was present, we swabbed fish K1 to K6 every other day for 1 month to see if we could detect the presence of KHV or KHV DNA on gill surfaces and within the anal vent. These Dacron swabs were placed in 0.5 ml of sterile DMEM (Invitrogen, Carlsbad, CA) containing penicillin (200 U/ml) and streptomycin (200 μg/ml) (Sigma-Aldrich, Inc., St. Louis, MO). Five-microliter aliquots of swab solution were tested by real-time PCR for KHV DNA, and 0.2 ml of swab solution was inoculated to CCB cells seeded in 12-well plates for KHV virus detection.
Blood collected from the caudal vein into a syringe previously coated with heparin (Sigma-Aldrich, Inc., St. Louis, MO) at 1,000 U/ml in phosphate-buffered saline (PBS) was transferred to an EDTA tube. Blood was centrifuged at 650 × g at 4°C for 10 min; the buffy coat was collected and exposed to 3 to 4 volumes of red blood cell lysis buffer (Tris-NH4Cl). The remaining white blood cells (WBC) were washed twice in sterile DMEM (Invitrogen) by centrifugation at 650 × g at 4°C for 10 min (Becman XJ). Then, WBC from each blood sample were subjected to a total DNA extraction (yielding WBC total DNA) using a High Pure PCR Template Preparation Kit (Roche Diagnostics, Indianapolis, IN). Total DNA was extracted similarly from the plasma pellet that was collected after ultracentrifugation of 250 to 500 μl of the plasma layer at 25,000 rpm for 1 h min at 4°C (Beckman model XL-70) in an SW28 rotor. Approximately, 0.1 to 0.5 μg/μl total DNA can be isolated from WBC obtained from each fish. The DNA concentration was adjusted to 0.1 μg/μl before use in PCR.
KHV was cultured in CCB or KF-1 cell lines that were maintained in DMEM (Invitrogen, Carlsbad, CA) supplemented with 5% fetal bovine serum (Gemini Bio-Products, West Sacramento, CA), penicillin (100 U/ml), and streptomycin (100 μg/ml) (Sigma-Aldrich, Inc., St. Louis, MO) at 22°C. KHV virus isolation from tissue samples was performed by homogenizing frozen tissue samples in DMEM (1:5 ratio [wt/vol]), centrifuging the tissue homogenate at 2,000 × g for 10 min, filtering the homogenate through a 0.45-μm-pore-size filter (Waterman), and inoculating supernatant of the tissue homogenate into 25-cm2 flasks seeded with CCB cells (0.5 ml of tissue preparation per 25-cm2 flask). To monitor virus shedding in the gill and feces, each gill or fecal swab medium (0.2 ml/swab) was inoculated onto CCB cells seeded in 12-well plates. Visible characteristic KHV cytopathogenic effect (CPE) was considered positive isolation. If initial inoculation failed to produce KHV CPE the first time, total cell lysate was reinoculated onto CCB cells seeded in 12-well plates and incubated at 22°C. Swabs or tissue preparations failing to produce KHV CPE in CCB cells the second time were considered KHV negative. Stock KHV cultures were inoculated onto one well of each plate as a positive control.
Viral DNA for positive controls was extracted from either virions or purified intracellular nucleocapsids as described previously (13). Briefly, the purified virions or nucleocapsids were digested in 10 mM Tris-HCl (pH 8.0), 100 mM EDTA, 1% N-lauroyl sarcosine, and 200 μg/ml proteinase K overnight at 55°C. The viral DNA was extracted twice with an equal volume of phenol-chloroform (1:1 [vol/vol]) and then precipitated with two volumes of ethanol and 1/10 volume of sodium acetate. The precipitate was washed once in 70% ethanol and dissolved in TE buffer (10 mM Tris-HCl [pH 8.0]-1 mM EDTA).
To determine whether KHV latency can be reactivated, the koi tank water temperature was increased from 12°C to 23°C at a rate of 1°C per day. The temperature was then held constant for 4 days at 23°C before being dropped back to 12°C at a rate of 1°C per day (see Fig. 3A). Anal vent and gill swabs were collected from all six koi every other day starting at day 2 post-temperature increase until day 16 post-temperature stress. Swabs were collected and transported in 0.5 ml of sterile PBS containing penicillin (200 U/ml) and streptomycin (200 μg/ml) (Sigma-Aldrich, Inc., St. Louis, MO). Five-microliter aliquots of swab solution were tested by real-time PCR for KHV DNA, and 0.2 ml of swab solution was inoculated onto CCB cells seeded in 12-well plates for KHV virus detection.
Two koi, designated K5 and K6, died on day 20 post-temperature stress (i.e., 20°C). These fish were necropsied, and tissue samples were collected, including the brain, spleen, gills, heart, eye, intestine, and kidney. One month after completion of the temperature stress regime, the remaining four koi were bled as described above and then euthanized to permit collection of tissues, including the brain, spleen, gills, heart, eye, intestine, kidney, liver, and pancreas, via necropsy.
Tissue samples (approximately 100 to 200 mg) obtained from freshly euthanized or dead koi were stored at −80°C. Before DNA extraction, the frozen tissues were homogenized in 800 μl of 1× lysis buffer by 2.5-mm silica beads (Biospec Product) and digested overnight at 55°C in the presence of 100 μg of proteinase K. Genomic DNA was then extracted from the tissue lysates with a High Pure PCR Template Preparation Kit according to the manufacturer's instructions (Roche Diagnostics, Indianapolis, IN). Approximately 0.1 to 1 μg/μl total DNA could be extracted from each tissue. All the tissue DNA was adjusted to 0.1 μg/μl before being used in PCRs. For each sample, 5 μl of total DNA (about 0.5 μg) was used in real-time PCR or PCR.
To determine if persistent infection occurs in the WBC, viral mRNA expressed during lytic infection was examined in total white blood collected from four healthy koi obtained from a local pet store; these fish had tested positive for KHV DNA in WBC by PCR. Total RNA was extracted from the combined WBC from the four koi using TriZol (Invitrogen). As a positive control, total cellular RNA from KHV-infected KF-1 cells was harvested at 8 days postinfection and extracted by using TriZol in accordance with the manufacturer's instructions. Total RNA from uninfected KF-1 cells served as a negative control. The isolated total RNA was resuspended in RNase-free H2O and examined for KHV lytic infection by reverse transcription-PCR (RT-PCR). The extracted total RNA from WBC (WBC total RNA) and KF-1-infected or uninfected cells was adjusted to 0.5 μg/ml before use in RT-PCRs.
Selection of primers for KHV sequence amplification was based on conserved DNA sequences of KHV (AF411803). Real-time PCR primers were selected as previously described (8): KHV-86f (5′-GACGCCGGAGACCTTGTG-3′), KHV-163r (5′-CGGGTTCTTATTTTTGTCCTTGTT-3′), and TaqMan probe KHV-109p (5′-6FAM-CTTCCTCTGCTCGGCGAGCACG-TAM-3′, where FAM is 6-carboxyfluorescein and TAM is 6-carboxytetramethylrhodamine). Another real-time PCR primer set specific for a host gene encoding glucokinase was also used as an internal control to equalize the amount of input total DNA. All the real-time PCRs for KHV DNA were run with equal amounts of DNA estimated by real-time PCR of the glucokinase gene. The primer sequence and TaqMan probe were selected as reported previously (8). The primers used for screening for the presence of viral DNA in tissue samples and tissue culture fluid were complementary to the KHV DNA polymerase and open reading frame 26 (ORF26) sequences: KHVDF-242 and KHVDR-242 for the KHV DNA polymerase gene; KHVF-447 and KHVR-447 for KHV ORF26 (Table 1). To further probe the amplified DNA sequence specific for KHV, another set of nested primers was selected: KHVNF242 and KHVNR242 as the DNA polymerase gene probe (Table 1) and KHV263F and KHV263R as the ORF26 probe (Table 1).
First-strand cDNA was synthesized from 2.5 μg of total RNA by using 10 pmol of random primer and Superscript Reverse Transcriptase III (Invitrogen) according to the manufacturer's recommendations. PCR amplification with KHV-specific primers for detection of cDNA of KHV genes expressed during lytic infection was performed using a 25-μl reaction mixture consisting of 22.5 μl of amplification buffer (Platinum PCR Supermix; Invitrogen, Carlsbad, CA), a 1.25 μM concentration of each primer, and 2.5 μl of the completed RT reaction mixture. The mixture was subjected to 34 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 45 s and then incubated at 72°C for 5 min after the final cycle.
PCR amplification with KHV-specific primers for detection of viral DNA in tissues or culture fluid was performed using a 25-μl solution consisting of 19 μl of amplification buffer (Platinum PCR Supermix; Invitrogen, Carlsbad, CA), a 0.4 μM concentration of each primer, and 5 μl of total DNA (~0.5 μg) or 5 μl of swab medium. The mixture was subjected to 94°C for 2 min, and 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min, followed by a 10-min elongation reaction at 72°C after the final cycle. Real-time PCR was performed according to the manufacturer's instructions for quantitative PCR (qPCR) Supermix (Platinum qPCR Supermix-UDG with ROX; Invitrogen, Carlsbad, CA).
The PCR product amplified from KHV-U DNA with real-time PCR primers was cloned by a TOPO 2.1 PCR cloning vector (Invitrogen, Carlsbad, CA). The correct insert was screened by restriction digestion and then sequenced by the Center for Genome Research and Biocomputing (CGRB) at Oregon State University. This plasmid product was employed to set up the standard curve for measuring the viral DNA copy number in tissue samples or swab fluid.
PCR products (40% of total PCR with 0.5 μg of DNA template) generated using WBC or tissue DNA were electrophoresed through a 1.5% agarose gel, transferred to a nylon membrane (14), and then UV cross-linked to the membrane. The DNA products were then probed with a digoxigenin (DIG)-labeled DNA probe. The probe was generated with nested PCR primers that are specific for the target genes. To make digoxigenin-labeled PCR products, digoxigenin-labeled deoxynucleoside triphosphates (Roche Diagnostics, Indianapolis, IN) were added to the PCR mixtures according to the manufacturer's instructions (Roche Diagnostics, Indianapolis, IN). The membrane was prehybridized with prehybridization buffer (Roche Diagnostics, Indianapolis, IN) at 68°C and then hybridized with the DIG-labeled DNA probes specific for the gene coding for DNA polymerase or major capsid protein at 68°C. After incubation with the probe, membranes were washed with 0.1% sodium dodecyl sulfate and 10% 20× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) before incubation with an antidigoxigenin antibody conjugated with peroxidase. The membrane was then developed by incubation with a chemiluminescent peroxidase substrate (Roche Diagnostics, Indianapolis, IN). The blots were exposed to film (Kodak) at room temperature for 30 min to 2 h. The molecular masses of the resulting bands were estimated by using a 1-kb DNA ladder (Invitrogen, Carlsbad, CA).
The PCR products were cleaned with a ChargeSwitch PCR Clean-Up Kit (Invitrogen, Carlsbad, CA) before sequencing and were sequenced by the CGRB at Oregon State University. The nucleotide sequences were analyzed with the Geneiou software.
Herpesvirus latency is characterized by an absence of infectious virus and the presence of the viral genome in the latently infected tissue. To investigate whether KHV becomes latent in koi peripheral leukocytes, WBC from fish K1 to K6 were examined by both real-time PCR (8) and PCR with primers (KHVD242F and KHVD242R) specific for the KHV DNA polymerase gene in blood taken 2, 4, and 8 weeks after arrival of the fish at OSU-SDL (8). To estimate the amount of DNA in the latently infected tissue, DNA copy number was determined by a real-time PCR standard established by using 10-fold serial dilutions of known KHV DNA from 108 to 0 copies of the plasmid bearing the KHV DNA insert (Fig. 1A). Although KHV DNA was not detected in every blood sample (Fig. 1B), KHV DNA was detected in WBC of all six koi in the third blood draw, when a minimum of 1-ml of whole blood was used for WBC isolation. The amount of detectable KHV DNA was approximately 2 to 60 copies per microgram of the total WBC DNA, which is very common for herpesvirus latency (26, 32). This suggested that KHV may become latent in koi white blood cells.
To further prove that the PCR product was specific for KHV DNA, the amplicons with primers KHVD242F and KHVD242R were then hybridized using a DNA probe specific for the DNA sequence between the region amplified by the two primers (Table 1). As shown in Fig. 1C, KHV DNA was detected in all six koi and in all three blood draws for each fish. To eliminate the possibility that the virus was present in the plasma, each koi plasma sample was also examined by PCR and Southern blotting. As shown in Fig. 1D, no detectable KHV DNA was present in the plasma, further suggesting that the viral DNA resides inside the leukocytes and is not present as free virus in the plasma.
The presence of KHV DNA and infectious virus in body secretions is indicative of persistent infections. To evaluate the possibility of KHV persistent infection in these six koi, gill swabs and vent swabs were examined by both real-time PCR and virus isolation. No KHV DNA was detected from any gill swabs or vent swabs taken from all six koi every other day for 1 month although positive and negative controls performed as expected (data not shown). No KHV was isolated from swabs while positive controls yielded viral growth. In addition, by direct inoculation of WBC culture supernatant onto CCB cells seeded in 12-well plates, no KHV infectious virus was detected in koi WBC taken before this time period (from fish bled at 2 weeks and 1 month following their arrival) and after this time period (from fish bled at 2 months following their arrival). This argues against persistent infection in these koi.
To further evaluate the possibility of KHV persistent infection in WBC, KHV mRNA expressed during lytic infection was examined in the total RNA isolated from the combined WBC from four healthy koi (KVI to KIX) which tested positive for KHV DNA in WBC. These fish were obtained from a local pet store. The transcription of the DNA polymerase gene or the major capsid gene was examined using a total RNA extract of WBC, which is indicative of productive infection. The total RNA was extracted from 5 × 107 WBC, which is about twice the number of cells used in KHV infections in vitro. Active transcription of the DNA polymerase gene or major capsid gene can be detected in 2.5 μg of total RNA of 5 × 106 KF-1 cells infected with 1,000 PFU of KHV at day 8 postinfection (Fig. 2) and in 2.5 μg of WBC total RNA extracted from acutely infected koi (data not shown). No mRNA of the DNA polymerase gene or major capsid gene product was amplified from the WBC total RNA (Fig. 2).
This result suggests that no lytic infection or detectable lytic infection is present in the WBC and is consistent with KHV latency. The detection sensitivity is based on the lytic infection in tissue culture infection at day 8 postinfection, when the earliest KHV CPE can be seen in 10 to 20% of the infected monolayer.
To investigate whether KHV latency can be reactivated, all six koi were subjected to temperature stress as described before (28). The temperature change program is shown in Fig. 3A. KHV has very low infectivity in vitro. The minimal infection titer needs to be 1 × 104 PFU/ml in KF-1 cells or CCB cells. It grows slowly and takes a minimum of 2 weeks to grow to 105 PFU/ml. Therefore, real-time PCR and virus isolation were employed to monitor virus shedding in fecal sample and gills. The advantage of using virus isolation is that isolated virus can be recovered and cultured again for infectivity verification. All swabs were subjected to real-time PCR to estimate the amount of virus reactivation. As shown in Table 2 and Fig. 3B, KHV DNA was detected in gill swabs as early as day 2 post-temperature stress. Five out of six koi were shown to be KHV positive in gill swabs by day 8 post-temperature increase (i.e., at 21°C). The peak virus reactivation was between days 8 and 12 post-temperature increase (i.e., at 21 to 23°C). The KHV DNA detected in the swabs ranged between 10 and 103 DNA copies per swab (Table 2). In agreement with real-time PCR, KHV was also isolated from gill swabs taken on day 12 and fecal swabs collected on days 10 and 12 post-temperature stress (data not shown). This result confirms that KHV latency can be reactivated under temperature stress.
Two koi, designated K5 and K6, died on day 20 post-temperature stress (i.e., 20°C). No acute gross or microscopic changes characteristic of KHV were observed at necropsy. However, a high magnitude of KHV DNA (20- to 200-fold higher than the average amount of KHV DNA copy numbers of WBC samples taken on three different times) was detected in heart, gill, spleen, and fecal samples (Fig. 4). To see if KHV infectious virions were present in these koi, sample preparations from the gills, spleen, and feces were inoculated onto CCB cells; KHV CPE was observed in cells infected with gill, spleen, and fecal tissue preparations from both dead koi (data not shown). To investigate the isolate relationship with existing KHV isolates, the DNA sequence amplified by primers KHVDF-242 and KHVDR-242 was compared with the KHV DNA sequences deposited in the GenBank. The K5 DNA polymerase gene (between nucleotides [nt] 147218 and 147459) had one nucleotide at 147298 of the KHV-U genome (DQ657948) different from the sequence of other strains and had a close phylogenetic relationship to a Japanese strain (AP008984), while KHV DNA from K6 was 100% identical (between nt 147218 and 147459) to that of KHV-U (DQ657948), indicating a close relationship with the U.S. strain (DQ657948) (Fig. 5). The phylogenetic relationship is consistent with the source of the fish. K5 was from a donor who acquired the koi from Japan; K6 was donated from a U.S. koi hobbyist. In addition, to examine their phylogenetic relationship with our domestic koi, KHV DNA amplified from WBC pooled from four koi (Kmix) obtained from a local pet store was also compared with KHV from K5 and K6. As shown is Fig. 5, Kmix has a close phylogenetic relationship with KHV isolated from common carp, which suggests that KHV may have existed in common carp long before the KHV outbreak in koi.
To determine whether KHV has specific tissue tropism during latency, KHV DNA distribution was examined in tissues collected at necropsy. One month following the temperature stress, the remaining four koi were euthanized and examined for KHV latency in tissues. Upon necropsy, three out of four koi lacked significant gross and histological pathology; one fish, K4, had prominent gill necrosis at necropsy. Total DNA was extracted from each individual tissue and examined by both real-time PCR and PCR with primers specific for KHV ORF26 (KHV447F/KHV447R). As shown in Fig. 6A, a detectable level of KHV DNA could be found in many tissues in all four fish, and especially in heart, eye, intestine, and gonads, using real-time PCR analysis. Interestingly, a large amount of KHV DNA was also detected in both intestine and intestinal contents. To further confirm our results, total DNA from all collected tissue was examined by primers KHV447F and KHV447R, which amplify a longer sequence of the KHV genome at 447 bp, as shown in Fig. 7A. As shown in Fig. 6B, KHV-specific DNA at 447 bp was detected in all four fish in almost all of the tissues examined.
To investigate the possibility that latency is common in koi, five asymptomatic koi with no known history of exposure to KHV were also tested for latent infection by testing of WBC using both real-time PCR and PCR coupled with Southern blotting. As shown in Fig. 7A, KHV DNA was detected in WBC isolated from all five fish (KI to KV) taken 2 weeks after their arrival at OSU-SDL. The KHV DNA copy number in the WBC of these five koi was approximately 20 to 150 copies per microgram of total DNA of WBC isolated from 1 to 2 ml of total blood. To confirm the detection of KHV DNA by real-time PCR, total DNA extract from WBC was subjected to PCR with Southern blotting as described in the legend of Fig. 1. As shown in Fig. 7B, KHV DNA was detected in WBC of all five koi taken at 2 weeks following their arrival. To further confirm the real-time PCR and PCR-Southern blotting results, the PCR products amplified by KHVDF-242 and KHVDR-242 from fish KI and KIII were also sequenced. As shown in Fig. 5, KHV DNA from, K3, K6, KI, and KIII share the same lineage as U.S. strain DQ657948 and KHV strain DQ177346 from Israel. All of these fish had tested negative by KHV serum antibody ELISA. Two months after their arrival at OSU-SDL, all five koi were euthanized and necropsied. No gross lesions were observed in any of these five koi. However, real-time PCR analysis of tissues collected from these five koi revealed widespread distribution of viral DNA. As shown in Fig. 8A, KHV DNA was detected in the brain, spleen, eye, gills, heart, intestine, and kidney. The amount of detectable KHV DNA from the tissue was much lower than the amount detected in the koi that died following temperature stress (Fig. 8B), suggesting that KHV DNA is maintained at low copy numbers during latency in these tissues. The amplicons of tissue DNA recovered from fish KI to KV were also amplified and hybridized by a DNA probe, as described in the legend of Fig. 1. As shown in Fig. 8C, KHV DNA was detected by PCR with Southern blotting in almost all the tissues collected at necropsy. No infectious KHV was isolated from KI to KV tissues collected at necropsy. This result suggests that KHV may become latent in various tissues.
Our studies demonstrate that KHV, like mammalian herpesviruses, can become latent in the peripheral white blood cells and various tissues. The type of white blood cell latently infected with KHV has yet to be determined. Our preliminary studies by sorting B cells and T cells demonstrate that B cells or other non-T cells are the major latency sites of KHV (unpublished data). Our results prove that KHV latency can be reactivated by temperature stress between 17°C and 23°C, which may explain the tendency for KHV outbreaks during the summer. A previous study yielded similar findings, but the relatively brief interval between acute clinical disease and reactivation raised the possibility of persistent infection rather than latency (28). In the current study the interval was several years, and tests for active infection prior to the temperature stress treatment were negative, strongly implicating reactivation.
Previous research demonstrated high KHV DNA concentrations in the gill, kidney, and spleen at the beginning of infection and suggested that KHV is lymphoid tropic (8). It is possible that KHV becomes latent in lymphoid cells because T cells and B cells are part of the piscine immune system (6, 16–18). Since no KHV CPE was observed in CCB cells inoculated with plasma or WBC culture supernatant from KHV-positive fish prior to the temperature stress treatment, KHV DNA, but not infectious virions, was detected, as expected, in peripheral WBC. Our study also suggests that KHV DNA remains at low copy numbers in the latently infected tissues, ranging between 2 to 150 copies per microgram (Fig. 1B, B,6A,6A, and and8A.).8A.). As shown in Fig. 1, only 30% of the KHV-positive fish were detected by real-time PCR if only 0.5 ml of blood was collected for WBC isolation. The variation comes from variations in sampling as blood collection from each koi varies between each time point. However, all fish 50 to 65 cm long tested positive when a minimum of 1 ml of blood was used for WBC isolation (third blood sample from K1 to K6). For fish under 25 cm long, KHV DNA could be detected in WBC from 1 ml of blood (data not shown). However, PCR coupled with Southern blotting was shown to be a highly sensitive method for detecting KHV latency in WBC and is more consistent and reliable than the real-time PCR method (Fig. 1C and and77B).
Our studies find that KHV becomes latent in WBC and various tissues. Whether the widespread tissue distribution of KHV DNA reflects circulating latently infected leukocytes or an extremely broad tissue tropism remains uncertain. It is possible that detection of KHV DNA varies with the numbers of WBC in the tissues during collection at necropsy. It is also possible that erythrocytes or other tissue constituents can interfere with PCR-based detection of KHV in some tissues since we noticed that KHV DNA amplification was decreased in some tissues spiked with positive-control DNA (data not shown). The quality of the viral DNA from each tissue may also play an important role in the detection of the virus. If the viral DNA is extensively degraded during necropsy and sample preparation, it will reduce detection sensitivity. Thus, it is possible that all the tissues contain KHV DNA but that our PCR failed to detect the viral genome due to degradation of KHV DNA in the degraded tissue.
A large amount of KHV DNA was detected in both intestine and intestinal contents collected at necropsy from some of the clinically normal koi that had recovered from temperature stress and from the koi with no history of KHV infection (Fig. 6A and and8A).8A). This may reflect latency within gastrointestinal epithelial cells that are sloughed into the gut lumen or latency in lymphocytes undergoing transmigration into the lumen or perhaps simply contamination of the sample with blood due to the method of “milking” content from the intestinal segment during collection.
Koi exposed to KHV do mount an antibody response to the virus (27). KHV antibody detection by serum neutralization or serum ELISA is the only method used by veterinarians to detect KHV exposure to prevent KHV transmission. However, this assay may miss some fish latently infected with KHV if the antibody titer is below the detection level of the serum neutralization or the ELISA method. There are examples of other herpesvirus infections in which animals have latent infections but do not have detectable antibodies by certain ELISAs (19). Our current study found that the ELISA has very limited sensitivity in the detection of KHV-positive koi. In addition, although K4 and K6 tested positive prior to our study, the KHV antibody ELISA did not detect antibody in any of the six koi sampled at three different times, with the exception of the third blood sample from K6 (Fig. 9). However, our PCR and Southern blotting results prove that all six fish were latently infected with KHV (Fig. 1C). These latent infections were subsequently confirmed via reactivation. Furthermore, the detection of KHV DNA in multiple clinically normal fish with no known history of exposure to a KHV outbreak raises the possibility that the distribution of this virus in the koi population is far wider than previously believed (Fig. 2 and and8).8). Our study suggests that the KHV serum ELISA does not always detect KHV-exposed koi and that real-time PCR provides more accuracy when WBC are examined directly. Additionally, the PCR coupled with Southern blotting of WBC was shown to be a very sensitive method to detect koi latently infected with KHV. This may allow efficient screening for koi latently infected with KHV.
In addition, it is possible that there are many different strains of KHV present in koi populations. The KHV DNA from fish at a local pet store evolved from the U.S. strain and is closely related to common carp KHV (Fig. 5). This suggests that KHV may have originated from a common carp KHV that has yet to be well characterized.
We thank the American Koi Club Association for funding this study.
We thank Ronald Hedrick (University of California, Davis) for providing the KF-1 cell line and the KHV-I and KHV-U strains used in this study. We also thank George Rohrmann for helping to edit the last two versions of the paper.
Published ahead of print on 9 March 2011.