j ourna l homepage: www.e lsev ie r .com/ locate /aqua-on l ine
Ivermectin blocks the nuclear location signal of parvoviruses in crayfish,Cherax quadricarinatus
Kim Y. Nguyen a, Kitikarn Sakuna a,b, Robert Kinobe c, Leigh Owens a,⁎a Microbiology and Immunology, School of Veterinary and Biomedical Sciences, James Cook University, Townsville 4811, Queensland, Australiab Faculty of Veterinary Science, Rajamangala University of Technology, Srivijaya 80240, Thailandc Physiology and Pharmacology, School of Veterinary and Biomedical Sciences, James Cook University, Townsville 4811, Queensland, Australia
Parvoviruses have been responsible for major problems in the shrimp aquaculture for decades with few optionsfor control apart from avoidance. As intranuclear viruses for some of their replication, parvoviruses need to usethe cell's nuclear transport signals for entry into the nucleus. This study was conducted to see if ivermectinwhich has recently been shown to block importins in vitro would do so against two presumptive parvovirusesin a freshwater crayfish, Cherax quadricarinatus, model. Crayfish were shown to tolerate ivermectin at 7 μg/kginjected intramuscularly and survival appeared to be enhanced with increasing dose (P ≤ 0.1). Ivermectin dra-matically decreased hypertrophied nuclei caused by presumptive gill parvovirus by ~68% (P ≤ 0.001) after 2doses of 7 μg/kg reducing from 1591 to 505 affected cells in the gills. The reduction did not increase furtherwith increasing doses. Also, ivermectin appeared to increase the survival of crayfish when challenged withC. quadricarinatusparvo-like virus (CqPV) to levels statistically equivalent to non-infected crayfish but did not ap-pear to affect the number of viral infected cells. There was a negative correlation between the size of crayfish andtheir longevity (P ≤ 0.05, R2 = 0.15) with smaller crayfish dying faster when challenged with CqPV. This is thefirst in vivo testing of ivermectin against viruses and showed that ivermectins do dramatically block some parvo-viruses, possibly by interactions with cellular importins. There may be a therapeutic role for ivermectins in viralreduction in broodstock in crustacean aquaculture.
The penaeid parvoviruses Penaeus monodonDensovirus (PmonDNV,colloquially known as HPV) and Penaeus stylirostris Brevidensovirus(PstBNV, colloquially known as IHHNV) cause many disease issues inpenaeids (see reviews Safeena et al., 2010 and Rai et al., 2012). Indeedparvoviruses cause major diseases in many animals including humans,dogs, cats, mink, pigs, cattle, crustaceans and insects. Parvoviruses areintranuclear in their replication and they need rapidly dividing cells inthe S-phase to access the cellular DNA replication enzymes. Thus theparvovirus needs to transport their proteins into the nucleus using thecell's nuclear importing molecules, karyopherin also called importin,IMPα/β linked to their nuclear location sequences or signals (NLS).Recently, Owens (2013) identified many possible signals in thesepenaeid parvoviruses and indeed this current study was spawnedfrom that analysis.
Recently, ivermectin andmifepristonewere reported to have potentantiviral activity in vitro (Wagstaff et al., 2011, 2012) by preventing ac-tive nuclear transport of the integrase molecule of human immunodefi-ciency virus (HIV)-1. Mifepristone is a specific inhibitor of the nuclear
ghts reserved.
import of the protein integrase, but ivermectin appears to act onIMPα/β-mediated nuclear import generally. This raises the intriguingpossibility that ivermectin could be an anti-parvoviral agent if parvovi-ruses use IMPα/β to transit into the nucleus.
Ivermectin is an effective antiparasiticide used widely on animalfarms including aquaculture against parasites such as sea liceLepeophtheirus salmonis and Caligus elongatus (Davies and Rodger,2000) and metacercariae of Clinostomum marginatum (Lorio, 1989).
Crustaceans are very sensitive to ivermectin. Loss of action potentialin the neuron, loss of motor function and eventual paralysis fromavermectin in the brine shrimp Artemia salina, which contains neuro-transmitter gamma-aminobutyric acid (GABA) receptors (Calcott andFatig, 1984), have been documented. The mysid shrimp, Mysidopsisbahia, was sensitive at 96 h LC50 0.022 μg/l (Wislocki et al., 1989),whilstthe no-observed effect concentration (NOEC) was 4 ng/l, but the 96 hLC50 for pink shrimp Penaeus duorarum was 1.6 μg/l. The mysid,Neomysis integer showed a 96 h LC50 of 70 (44–96, 95% CI) ng/l, whenimmersed (Davies et al., 1997). Through the digestive tract of shrimpCrangon septemspinosa, ivermectin was toxic but not via the gills(Burridge and Haya, 1993). Shrimp could tolerate ivermectin in waterat the maximum concentration 21.5 μg/l, but ivermectin was lethal at96 h LC50 = 8.5 μg ivermectin/g of food. The shrimp's average weightwas 2.76 g, and the feeding rate was 1% body weight per day. The
289K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
NOEC was approximately 2.6 μg/g of food (Burridge and Haya, 1993).Given that ivermectin was toxic in crustacea but reported concentra-tions variedwidely, it was necessary to examine the effect of concentra-tions of ivermectin in redclaw crayfish, Cherax quadricarinatus, theproposed crustacean model.
Two parvovirus-like cellular changes reported in C. quadricarinatuswere used to investigate whether ivermectin could block their NLS.Gill parvovirus of C. quadricarinatus (Edgerton et al., 2000), herein calledgill parvovirus, and parvo-like virus of C. quadricarinatus (Bowater et al.,2002) herein called CqPV were tested. Gill parvovirus produces signet-ring, hypertrophic nuclei in the gills without an inclusion body and isamildly pathogenic, pre-existing cellular change in a population of cray-fish (Rusaini et al., 2013) (Fig. 1a). CqPV produces basophilic, Cowdrytype A intranuclear inclusions in the gills (Fig. 1b) and other systemictissues, causes heavy mortality and infective tissue is available, so thevirus can be administered on demand for experimentation. The aimwas to see if ivermectin could influence the course of either parvo-likevirus infection.
The study was designed in three parts. Part 1 was to establish con-centrations of ivermectin that the crayfish would tolerate. Part 2 wasto see if ivermectin at concentrations established from Part 1 couldchange the course of a weakly pathogenic gill parvovirus cellularchange. Part 3 was to see if ivermectin could change the course of thehighly pathogenic CqPV.
2. Materials and methods
2.1. Shared protocols: experimental animals
Grossly healthy crayfish (9–150 g) were obtained from commercialcrayfish farms in northern Queensland and transported to the AquaticInfectious Disease Facility (Fish Laboratory) of School of Veterinaryand Biomedical Sciences (SVBS), James Cook University. The availabilityof crayfish was extremely limited and whatever crayfish were availablehad to be taken; thus size range was greater than desirable. Duringtransporting, crayfish were kept within two layers of wet cloth in theStyrofoam boxes to decrease the temperature and activity of crayfish.Crayfish were maintained in 1000 l tanks with a recirculating biofiltersystem and aerator. Four days prior to commencement of the study,crayfish were divided into small groups of two or three crayfish toreduce cannibalism and kept in 50 l dark blue tanks on racks in a
Fig. 1. (a). Different stages of hypertrophied nuclei caused by a putative parvovirus in gillof crayfish Cherax quadricarinatus. Early stage (short arrow), middle stage (arrowhead),and fully developed stage (long arrow). (b). The late-stage intranuclear inclusion bodies(INIBs) (fat arrows) and early-stage INIBs (narrow arrowheads) of Cherax quadricarinatusparvo-like virus observed in gill cells of crayfish. H&E. Scale bar = 10 μm.
recirculating system at 26 ± 2 °C. All experimental crayfish were fedand monitored daily.
As there were tank/space constraints and a requirement by theethics committee to limit the number of crayfish involved in experi-ments, crayfish were assigned randomly to treatments and treatmentswere assigned randomly to tanks. This meant that a perfect balance oftheweight of crayfish in treatments and the replication of the treatmentthemselves were sacrificed to best fit the random design to maximisestatistical power.
In this proof of concept stage, it was decided to deliver the potential-ly toxic ivermectin via injection to ensure all crayfish received the exactdose relative to their weight. All injectionswere performed using sterile1 ml syringes and 26-gauge needles, and were discarded after eachinoculation to minimise cross-infections.
2.2. Part 1. Tolerance to ivermectin; experimental design
After four days of acclimation, crayfish (64.0 ± 22.0 g) were dividedinto four groups (20 crayfish each): groups I, II, and III were intramuscu-larly injected with 0.6, 1.2, and 1.4 μg/ml ivermectin (IVOMEC Antipara-sitic Injection for Cattle 200 ml, ivermectin 1% injectable, Vet-n-PetDirect) solutions to make up doses of 3, 6, and 7 μg/kg, respectively andgroup IV controls were intramuscularly injected with the homogenisingsolution of phosphate buffered saline (PBS) and Tween 20 at a concentra-tion of 5%. Each inoculumwas divided into three portions and injected in-tramuscularly into the ventral side of the first, second and thirdabdominal segments of crayfish, away from the ventral nerve cord. Thenumber of crayfish used for each treatment was 5, divided randomly ei-ther 2 or 3 crayfish/tank and dose treatments were randomised(Table 1). Each treatment group was spread randomly over 4 differentracks. Thus, the total number of crayfish used for Part 1 was 80(Table 2). The duration of observation between each injection was20 days. This experiment was carried out with three injections (at days1, 21 and 41) and terminated at 60 days.
2.3. Part 2. Effect of ivermectin on a pre-existing parvovirus
The concentration of ivermectin determined to have no toxic effectwas the high dose (7 μg/kg) (see Results section) and therefore it wasthe therapeutic dose for testing of inhibition of the viral nucleartransportation.
2.3.1. Experimental animalsCrayfish were obtained from SVBS, JCU (Rusaini et al., 2013) and
were affected by putative gill parvovirus (Edgerton et al., 2000). Fivecrayfishwere examined by histopathology and all were heavily infected(Rusaini et al., 2013), thus it was assumed most crayfish would be in-fected and randomisation would be sufficient to limit vagaries in load.The weight of crayfish ranged from 9 to 45 g.
Table 1Part 1 (n); Part 2. A randomised design for examining the tolerance of crayfish to ivermec-tin (Part 1), and for testing the effect of ivermectin on pre-existing parvovirus infection incrayfish (Part 2). In Part 1, thefirst number is the dose of ivermectin in μg/kg. The numbersin parentheses ( ) indicate the number of crayfish in different tanks. In Part 2, the numberafter the semicolon is the number of doses of 7 μg/kg given to each crayfish. nc = no cray-fish in Part 2.
Table 2The experimental design of crayfish receiving differing doses of ivermectin.
Number of crayfish Dose of ivermectin
Controls High dose 7(μg/kg)
Medium dose 6(μg/kg)
Low dose 3(μg/kg)
Total(n = 80)
5 × 4 5 × 4 5 × 4 5 × 4
290 K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
2.3.2. Experimental designCrayfish were divided into two groups, control (no ivermectin but
injected with the placebo solution described previously) and treatmentgroup (ivermectin treatment). Inoculation was carried out up to threetimes (three doses) at days 1, 21 and 41, with the same dose of drug,thus the maximum duration of Part 2 was 60 days. At the end of20 days in each injection group, 20 crayfishwere sacrificed for histolog-ical examination. The crayfish in each group were randomly dividedinto 4 crayfish/tank and into three different system racks. Each grouphad 5 replicates. Thus, the total number of crayfish used for this exper-iment was 80 (Table 3).
2.3.3. HistologyPrior to histological preparation, crayfish were anaesthetised in
chilled water for about 5 min, then fixed by injection and immersionin Davidson's fixative for 48 h. Gills of fixed crayfish were trimmed,placed in histocassettes and stored in ethanol 70% and processed rou-tinely for histological examination by using a Shandon Elliot processorto dehydrate in graded ethanol 70%, 80%, 90%, 95% and 100% and thenclearedwith xylene, prior to the paraffin embedded procedure. Sectionswere cut at 5 μm and stained with haematoxylin and eosin.
2.3.4. Analysis of hypertrophied nucleiThe cumulative mortality rate (CMR) of treatment groups with the
numbers of doses of ivermectin was recorded. Any dead or moribundcrayfish were removed, recorded, and immediately prepared for histo-logical examination. The numbers of nuclei including normal andabnormal (hypertrophied) nuclei in gills of 76 crayfishweremicroscop-ically examined at objective 40×. Hypertrophied nuclei in 50 filamentswere counted from both treatment and control groups.
Nuclear changes were classified into three types of hypertrophiednuclei based on stages of their development. They were named early,middle and fully developed stages (Fig. 1a). In the early stage, nucleistart swelling, 2 or 3 times larger than normal nuclei, chromatin rare-fied, and then moved to the internal edge of the nuclear membrane. Inthe middle stage, chromatin continues to be moved to the peripheryof nuclear membrane, andmakes a clearing within the centre of the nu-clei. It is like a halo, or ring-structure in the nuclei, but chromatin has notcompletely migrated to the inner edge of the nuclear membrane. In thefully developed stage, chromatin has completely migrated to the pe-riphery nuclear membrane, and shows a clear zone within the nuclei(c.f. Azarkh et al., 2008).
Data recorded included the day of death, the weight (dead cray-fish during the experiment and survivors when each phase of theexperiment finished) and the number of normal and categories ofhypertrophied nuclei.
Table 3Number of crayfish infected with a putative parvovirus used in Part 2.
Number of crayfish Dose of ivermectin inoculation (7 μg/kg)
Controls 1 injection 2 injections 3 injections
Total(n = 80)
5 × 4 5 × 4 5 × 4 5 × 4
2.4. Part 3. Effect of ivermectin on a challenge infection with CqPV
2.4.1. Experimental animalsTwo or three crayfish (112.1 ± 38.7 g) per tankwere randomly dis-
tributed between five experimental groups (15 crayfish in each group):No-virus, Virus-only, Ivermectin 2 days before virus, Ivermectin 2 daysafter the virus and virus and ivermectin on the Sameday divided into sixtanks/treatment. Crayfish were maintained appropriately in indepen-dent plastic aquaria filled with 40 l of water with an air-lift corner filter.Water exchange (40% volume of the water in culture tanks) was doneevery two days to maintain appropriate water quality.
2.4.2. CqPV inoculum preparationInocula were prepared from infected gills of several frozen cephalo-
thoraxes of C. quadricarinatus confirmed by histopathology to contain aputative gill parvo-like virus (Bowater et al., 2002). One gramme of tis-sue was harvested, cut into 5 mm2 pieces, then homogenised at a 1:4ratio with homogenising solution (phosphate buffered saline (PBS)pH 7.0 containing 2 ml/l of 4-hexylresorcinol from Sigma Chemical,H6250). The homogenate was centrifuged in an Eppendorf centrifugeat 3300 relative centrifugal force (rcf) for 30 min at 4 °C to removecoarse cellular debris and the supernatant was further clarified by cen-trifugation at 15,200 rcf for 30 min at 4 °C before being filtered througha 0.45 μm syringe filter into a sterile container. One drop of the filteredsupernatant was plated onto blood agar and tryptic soy agar and incu-bated at 25 °C for 48 h to confirm the absence of bacterial contamina-tion. Control, non-infected ultrafiltrate was prepared from the gillsof several healthy frozen cephalothoraxes obtained from the samefarm as the experimental crayfish. The same methodology used to pre-pare the non-infected ultrafiltrate was used to prepare the infectedultrafiltrate.
2.4.3. The protective efficacy of ivermectin against CqPV in crayfishCrayfish were injected with ivermectin concurrent with CqPV inoc-
ulation (days 2 and 23), or at two (days 4 and 25) days post-CqPV inoc-ulation, or two days prior (days 0 and 21) to CqPV inoculation.
For the first inoculation (day 2), 100 μl of CqPV or no CqPV inoculumwas usedwhilst the second viral or no viral inoculation (day 23), 200 μlwas used. The inocula were injected intramuscularly into the right sideof the first abdominal segments of crayfish in the appropriate treat-ments, just to the side of the ventral nerve cord. Ivermectin (7 μg/kg)was divided into three doses and injected intramuscularly into the leftside of the first, second and third abdominal segments of crayfish inthe appropriate treatments. The experimental period began on the dayof the first injection (day 0) and concluded on day 40.
Crayfish that succumbed during the trial underwent post-mortem.Organs and tissue were recovered for histological examination. Priorto organ collection, crayfish were anaesthetized by placing in chilledwater for about 5 min. The gills of each crayfish were collected aftertreatment.
2.4.4. Data analysisThe definition of traumatic deathwas those crayfish that diedwithin
three days after each injection. Crayfish that succumbed by trauma fromthe injection were not analysed histologically. Quantification of histo-pathological changes in gills was conducted by using light microscopy(Olympus BH-2) with magnification 400× (10 × 40). The number ofeach of the three types (normal, early-stage intranuclear inclusion bod-ies (INIBs) and late-stage INIBs) in a sample of 100 gill cells was exam-ined twice. There was no statistically significant difference between thetwo assessments of histopathological examination, so the number ofnormal cells and CqPV infected cells (early-stage INIBs and late-stageINIBs) was combined for subsequent analysis.
291K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
2.5. Statistical analysis
Recorded data were the day of death, starting weight, the weight ofdead crayfish during the experiment, and the weight of all survivorswhen the experiment finished. A univariate analysis of variance(ANOVA) was used to analyse the effect of crayfish weights andhuman operators on survival. The Least Significant Difference (LSD)post-hoc test was used to separate means unless specified separately.In Part 1, itwasused to examine the effect of dose andhusbandry factors(system racks, rows and columns) on the log of the day of death. Theday of death was found to be not normally distributed, so that log 10of the day of death was used. As a precaution, nonparametric statisticswere conducted as well, but the significant results were identical, sothey are not reported. The type of death was divided into two groups:traumatic death from the injection procedure and non-traumaticdeath. The effect of treatment on the day of death was examinedusing analysis of covariance (ANCOVA) to examine effect of startingweight; the randomised variableswere examined to understand the ex-perimental noise and partition that variation where appropriate. Thelinear regression between dose of ivermectin and the day of deathwas examined (two-tailed) after removal of those crayfish categorisedas dying of traumatic death.
The relationship between ivermectin treatments and the log of theday of death or number of CqPV infected in 200 counted cells was exam-ined using analysis of covariance (ANCOVA) to examine the effect of thestarting weight and the randomised block design. Linear regression be-tween starting weight and number of CqPV infected cells in 200characterised cells was examined (two-tailed).
3. Results
3.1. Part 1. Tolerance to ivermectin
Crayfish mortality was 26 (Table 4) out of 80 animals (32.5%). Thepercentage of dead crayfish caused by trauma was 50% of the totaldeaths. There was no statistically significant difference in the means ofthe log of the day of death amongst the four treatments (F = 0.596,df = 2, 80, P N 0.05) but there was a significant difference betweenthe death by trauma group and the others (F = 5.1, df = 1, 80,P = 0.032). The means of the day of death of the traumatic deathgroup and the others were 37.23 ± 10.67 days and 54.20 ± 14.0 days,respectively. When the traumatic death was removed, a suggestion ofa protective effect of ivermectin with increasing dose from 0 to 7 μg/kgwas found (F = 3.277, df = 1, 65, P b 0.1, N0.05) accounting for ~5%of the variability (R2 = 4.96), but not statistically significant at P = 0.05.
The effect of randomised husbandry factors on log of the day ofdeath of crayfishwas not significant. The factors in the experimental de-sign that were analysed were system/rack (F = 0.590, df = 3, 80,P N 0.05), row (F = 0.63, df = 2, 80, P N 0.05), and column (df = 2,80, P N 0.05). Human operators also had no significant effect on logof the day of death (F = 0.420, df = 1, 80, P N 0.05). The effect ofdose of ivermectin on crayfish weights prior to each injection had nosignificant effect (P N 0.05), but the finishing weight was significant
Table 4Number of dead crayfish and survivors. The dead crayfishwere classified into two groups:traumatic and non-traumatic death. Traumatic death is defined as crayfish that died with-in three days of injection, and non-traumatic death is crayfish that died after three days ofinjection.
Number of animals Dose Total
Control Low Medium High
(0 μg/kg) (3 μg/kg) (6 μg/kg) (7 μg/kg)
Traumatic death 1 0 7 5 13Non-traumatic death 6 4 1 2 13Survivors 13 16 12 13 54
(F = 3.779, df = 1, 80, P b 0.05). The mean of the starting weightwas the lowest (63.98 ± SD 22) (Table 5), whilst the mean of theweight at the third injection was the highest (74.51 ± SD 23.86). Themean weight of crayfish increased throughout the three injections, butthe mean of the finishing weight decreased considerably, just higherthan that of the starting weight (65.13 ± SD 23.81), close to meanweight at the second injection (65.58 ± SD 23.12). This was due tocrayfish breeding and carrying their eggs on their abdomen, but releas-ing juveniles before the finalweigh-in, thus reducing their finalweights.Anecdotally, the fact that the crayfish were breeding in the experimentalso suggests that the crayfish were not too detrimentally affected bythe ivermectin at the doses used otherwise they would not have bred.
3.2. Part 2. Effect of ivermectin on pre-existing putative parvovirus
The mean of the number of hypertrophied nuclei in the controlgroup was the highest (1591 ± 392.33), compared to other groups, 1dose (1039.85 ± 383.96), 2 doses (505.89 ± 284.68) and 3 doses(671 ± 379.07) (Table 6). The early stage of hypertrophied nucleiaccounted for the largest proportion (~71%), whilst the middle and de-veloped stageswere of similar percentage (~9%). Ivermectinwas able tostatistically decrease (F = 27.58, df = 3, 76, P = 0.000) the log of thenumber of hypertrophied nuclei caused by a putative parvovirus incrayfishwith the control and three other doses being statistically signif-icantly different (F = 14.04, df = 3, 76, P b 0.001) (Fig. 2). However,the LSD comparison between 2 doses and 3 doses on the log of thehypertrophied nuclei was not significantly different (P = 0.055).There were no significant differences between the means of the log ofthe counts of hypertrophied nuclei caused by experimental design orhusbandry factors, including system racks (F = 2.34, df = 2, 76,P N 0.05), row (F = 1.72, df = 2, 76, P N 0.05), or column (F = 1.13,df = 2, 76, P N 0.05).
Within the stages of nuclear hypertrophy, there were some statisti-cally significant effects. Racks had a significant effect on log of the num-ber of early stage hypertrophied nuclei at the (F = 3.50, df = 2, 76,P = 0.037). Comparing the three racks, only racks 1 and 2 were signif-icantly different (P b 0.05). Analysing the log of the number of devel-oped hypertrophied nuclei, rack had a statistically significant effect(F = 4.49, df = 2, 76, P = 0.016), and the rack by row interactionhad significant effects (F = 2.94, df = 3, 76, P b 0.05).
3.3. Part 3. Challenge with CqPV
There was no bacterial growth over 48 h on both blood agar andtryptic soy agar.
In the CqPV-injected treatment groups, the first mortality displayingCqPV histopathology in the gills occurred 16 days post-injection. Dis-eased crayfish showed gross signs of malaise, anorexia and lethargy be-fore dying. On post-mortem examination, no gross changes to tissueswere seen.
Therewere statistically significant differences betweenmeans of thestarting weight amongst the five treatments (F = 3.142, df = 4, 66,P = 0.02). The Same-day group (x = 91.73 ± 9.56) was significantlylighter than the Virus-only group (x = 122.36 ± 11.09) and the No-virus group (x = 135.43 ± 10.92) (P = 0.027, P = 0.002 respectively)(Fig. 3).
Table 5Means of crayfish weights prior to each injection and termination of the experimentexamining the tolerance to ivermectin.
Crayfish weights (g) N Mean Std. deviation
Starting weight 80 63.98 22.00Weight of the second injection 80 65.58 23.12Weight of the third injection 74 74.51 23.86Finishing weight 67 65.13 23.81
Table 6Means and standard deviations of the number of normal nuclei and hypertrophied nuclei caused by putative gill parvovirus with different developed stages. Each dose was 7 μg/kg bodyweight inoculated intramuscularly.
Dose The number of normal cells Means of the number of hypertrophied nuclei
Total hypertrophied nuclei Early stage Middle stage Fully developed stage
292 K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
There were statistically significant differences in means of the log ofthe day of death amongst the five treatments (F = 2.98, df = 4, 66,P = 0.032). The No-virus group (x = 1.6 ± 0.01; 39.79 days) hadsignificantly higher survival compared to the Same-day group (x =1.5 ± 0.17; 33.4 days) and the Virus-only group (x = 1.51 ± 0.16;34.21 days) (P = 0.009, P = 0.023 respectively). However, there wasno significant difference between the log of the day of death amongstthe four treatments receiving CqPV inoculation (P N 0.05) (Table 7).
There was a statistically significant difference in the means of thenumber of CqPV infected cells amongst the five treatments (F = 2.98,df = 4, 66, P = 0.032) (Table 7). The average number of CqPV infectedcells of the No-virus group (x = 0.0) was significantly less than theIvermectin after (x = 56.08 ± 37.87, P = 0.00), the Ivermectin before(x = 46.09 ± 27.09, P = 0.003), the Same-day (x = 48.33 ± 33.97,P = 0.00), and the Virus-only groups (x = 36.43 ± 32.84, P = 0.018).However, there was no significant difference amongst the four treat-ments receiving CqPV injection. A significant negative linear regressionbetween starting weight and number of CqPV infected in 200categorised cells was found (F = 8.720, df = 1, 50, P = 0.005)(Fig. 4). R2 was 0.149 demonstrating that ~15% of the variability in in-fected nuclei across treatments was due to the starting weight.
4. Discussion
Ivermectin decreases the number of hypertrophied nuclei caused bya putative gill parvovirus in crayfishwhen exposed at 7 μg/kg intramus-cularly. The total number of hypertrophied nuclei of the control groupwas the highest and differs significantly from the other groups receivingivermectin (Fig. 2), whilst those groups receiving 2 doses and 3 doseswere similar. The number of hypertrophied nuclei decreased signifi-cantly in crayfish treated with 1 injection of ivermectin, although itwas not as large a decrease as compared to the decrease with 2 and 3doses.
Fig. 2. Themean of the hypertrophied nuclei at four different doses of ivermectin (7 μg/kgbody weight) injected intramuscularly into crayfish infected with a putative parvovirus.Each superscript indicates statistically homogenous groups. Error bars 95% CI.
Studies of Wagstaff et al. (2011, 2012) reported that ivermectin wasa non-specific inhibitor of viral replication (HIV-1 and dengue virus) byinterfering with the binding of integrase to its nuclear import receptorIMPα/β. Ivermectin was identified to inhibit heterodimer IMPα/β rec-ognition of nuclear co-localisation signal (NLS)-containing proteins(Wagstaff et al., 2011), but it does not affect independent IMPβ. The in-hibitor ivermectin probably binds to the NLS-binding pocket of IMPα,preventing the recognition NLS-containing cargo proteins (Wagstaffet al., 2012). However, in putative gill parvovirus in crayfish, we haveno information on how antiviral chemicals act on the virus. Wehypothesise that in the putative gill parvovirus in crayfish, the modeof action of ivermectin is similar and ivermectin probably blocks theNLS-binding pocket of IMPα limiting nuclear importation of viralcargo proteins NS1 and NS2. Crustacean hepandensoviruses have theDNA helicase Q1 NLS (Owens, 2013) which interacts with α-importin(Seki et al. 1997) supporting this hypothesis.
It is remarkably difficult to achieve 100% potency to block viral rep-lication completely (Flint et al., 2004). This incomplete block of viralreplication could be seen in crayfish as there was a further decrease inhypertrophied nuclei after two and three injections of ivermectin com-pared to one dose but the decrease appeared to plateau. In cells infectedwith mosquito Densovirus, a parvovirus can be detected at 11 h post-infection (Azarkh et al., 2008) compared to ivermectin which is gener-ally absorbed slowly. Perhaps, the hypertrophied cells that were seenat higher repeated doses had already developed in the crayfish beforethe administration of the ivermectin and the effects cannot be reversed.This may partially reflect to some extent, the timing dynamics of thevirus and the drug.
Ivermectin has a long half-life excretion/degradation period and itwas estimated to be approximately 22 days in mussels (Davies et al.,1997). In this study 20 days was assumed to be the best estimate ofthe half-life depuration period for each injection. However, we do nothave any data on the concentration still in the experimental crayfish
Fig. 3. Themeans and standard errors of weight (g) of the crayfish in each treatment at thestart of the experiment. Groupsmarked a and bwere significantly different (P b 0.05) fromeach other.
Table 7The means and standard deviations of starting weight, day of death, log of the day of death and number of CqPV infected cells in 200 categorised cells/crayfish from the gills.
Group Starting weight Day of death Log of the “day of death” Number of CqPV infected cells N
Note: groups marked a and b were significantly different (P b 0.05) from each other.
293K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
after 20 days post-injection. It is possible that the ivermectin wasstarting to build up in the bodies of the crayfish and this needs tobe studied further before widespread application of ivermectin isundertaken.
In Part 2 when examining the effect of ivermectin on a pre-existingputative gill parvovirus, the experimental design apparently affectedthe development stages of hypertrophied nuclei with some statisticallysignificant effects in racks, rows and their interaction. This is because ofthe different numbers of tanks in rows and columns with two-dosecrayfish due to randomisation of a non-balanced design (Table 1) buttwo doses had the strongest minimising effect on hypertrophied nuclei.This can happen by chance when resources are limited and a balanceddesign cannot be achieved. Racks 1 and 3 had two tanks of the two-dose crayfish, which had the largest decrease in the number ofhypertrophied nuclei, but rack 2 had only one tank of two-dose crayfish.Similarly, rows 1 and 2 across the 3 racks had 2 tanks of two-dose cray-fish, whilst row 3 had one tank of two-dose crayfish. The differences inthe numbers of tanks of two-dose crayfish between racks and rows leadto the statistically significant results. This is a reflection of the experi-mental set-up caused by restricted resources rather than the interactionof the ivermectin and the virus in the crayfish.
4.1. Part 3. Challenge with CqPV
The first mortality displaying CqPV histopathology occurred at16 days post-injection similar to Bowater et al. (2002) who reportedthefirst death at 17 days after injection. The presence of intranuclear in-clusion bodies and clinical signs was similar between the studies but nored discoloration of the carapace before dying nor oedema between theinner and outer cuticle of the carapace over the gill reported in theirstudy was noticed in our study.
Fig. 4. Linear regression between starting weight (g) and number of Cherax quadricarinatusquadricarinatus.
Both the Same-day and the Virus-only groups died faster than theNo-virus group. In the case of the Same-day group (91.73 ± 37.01 g)which was the lightest group, the effect of size could be a confoundingfactor. As weight decreased, crayfish were more susceptible to CqPV(y = 83.915 − 0.354x) (Fig. 4). By accident, the randomization of indi-viduals meant that the Same-day group had the smallest crayfish. Smallanimals are likely to receive higher stress by handling and injectionwhen compared with the larger animals.
Except the Same-day group, groups receiving ivermectin did notstatistically differ in the means of the log of the day of death from theNo-virus group. If ivermectin had no effect, then the Virus-only groupwhich had the second largest weights should have survived the longestof the viral infected groups. However, it was the Ivermectin before andafter viral exposure that were statistically most similar to the No virusgroup. Therefore given there is some confounding due to the differentweight of crayfish, logically, there is a suggestion of an ivermectin-linked protective effect to longevity in viral infected crayfish.
No significant difference amongst the four treatments receivingCqPV inoculation suggests ivermectin has no effect on the number ofCqPV-infected cells. Histopathology is less sensitive than say, quantita-tive PCR, but there is no molecular data on this virus whatsoever soqPCR could not be undertaken.
Despite confounding brought on by size discrepancies, this part ofthe study did find a therapeutic effect of ivermectin on mortality, butno strong evidence for reduction in the number of cells infected withCqPV in the gills of crayfish. These negative results might be due toCqPV proteins not imported into the nucleus via IMPα/β; the viral pro-teins might be imported by other nuclear import pathways, includingthatmediated by IMPβ1 alonewhich cannot be inhibited by ivermectin.
This is the first in vivo experiment that shows the potential of iver-mectin in blocking NLS. It opens the door for careful titrating of doses
parvo-like virus infected cells in 200 categorised cells/crayfish from the gills of Cherax
294 K.Y. Nguyen et al. / Aquaculture 420–421 (2014) 288–294
of ivermectin that can be used in crustacea, particularly broodstock andmay encourage pharmaceutical companies to look at ivermectin ana-logues that might be less toxic to crustacea. Penaeid broodstock are in-dividually handled for eyestalk ablation to induce spawning and thismight be an appropriate time for an injection of ivermectin based onthe broodstock's weight. Further studies on a more practical deliverymethod, such as in microencapsulated feeds in the hatchery or diet ingrow out, clearly need to be undertaken. Also, this proof of conceptstudy opens the intriguing possibility of studying ivermectins for usein blocking NLS to parvovirus in other domestic animals, perhaps asan adjunct in clinical therapy.
5. Conclusions
There was a suggestion (P b 0.1, N0.05) of a protective effect withivermectin as doses increased up to 7 μg/kg, so did survival. Ivermectinat 7 μg/kg dramatically (~68%) decreased the number of hypertrophiednuclei in pre-infected crayfish expressing parvo-like virus in the gills.When taking size of crayfish into account, then logically, ivermectin at7 μg/kg did appear to slightly extend the longevity of crayfish when ex-perimentally infected with the virulent CqPV.
Despite confounding brought on by shortages and timing of re-sources that added “noise” to the experiments, it is clear that there is atherapeutic advantage in delivering ivermectin to crayfish in generaland in some cases of viral infection where the virus probably uses aIMPα/β NLS for transport of proteins into the nucleus.
Acknowledgements
This study was carried out under the James Cook University AnimalEthics permit numbers A1888 and A1889. Kim Y Nguyen was therecipient of a scholarship from the Australian Agency for InternationalDevelopment (AusAID). Kitikarn Sakuna received a scholarship fromthe Rajamangala University of Technology, Srivijaya, Thailand.
References
Azarkh, E., Robinson, E., Hirunkanokpun, S., Afanasiev, B., Kittayapong, P., Carlson, J.,Corsini, J., 2008. Mosquito densonucleosis virus non-structural protein NS2 is neces-sary for a productive infection. Virology 374, 128–137.
Bowater, R.O., Wingfield, M., Fisk, A., Condon, K.M.L., Reid, A., Prior, H., Kulpa, E.C., 2002. Aparvo-like virus in cultured redclaw crayfish Cherax quadricarinatus from Queens-land, Australia. Dis. Aquat. Org. 50, 79–86.
Burridge, L.E., Haya, K., 1993. The lethality of ivermectin, a potential agent for treatment ofsalmonids against sea lice, to the shrimp Crangon septemspinosa. Aquaculture 117,9–14.
Calcott, P.H., Fatig 3rd, R.O., 1984. Inhibition of chitin metabolism by avermectin insusceptible organisms. J. Antibiot. (Tokyo) 37, 253–259.
Davies, I.M., Rodger, G.K., 2000. A review of the use of ivermectin as a treatment for sealice [Lepeophtheirus salmonis (Kroyer) and Caligus elongatus (Nordmann)] infestationin farmed Atlantic salmon (Salmo salar). Aquac. Res. 31, 869–883.
Edgerton, B.F., Webb, R., Anderson, I.G., Kulpa, E.C., 2000. Description of a presumptivehepatopancreatic reovirus, and a putative gill parvovirus, in the freshwater crayfishCherax quadricarinatus. Dis. Aquat. Org. 41 (2), 83–90.
Flint, S.J., Enquist, L.W., Racaniello, V.R., Skalka, M.A., 2004. Principles of Virology:Molecular Biology, Pathogenesis, and Control of Animal Viruses, 164. ASM press,Washington, DC 8 (Vol).
Lorio, W.J., 1989. Experimental control of metacercariae of the yellow grub Clinostomummarginatum in channel catfish. J. Aquat. Anim. Health 1 (4), 269–271.
Owens, L., 2013. Bioinformatical analysis of nuclear localisation sequences in penaeiddensoviruses. Mar. Genomics (accepted for publication 11 Oct).
Rai, P., Safeena, M.P., Krabsetsve, K., La Fauce, K., Owens, L., Karunasagar, I., 2012. Geno-mics, molecular epidemiology and diagnostics of infectious hypodermal andhaematopoietic necrosis virus. Indian J. Virol. 23, 203–214.
Rusaini, Ariel, E., Burgess, G.W., Owens, L., 2013. Investigation of an idiopathic lesion inredclaw crayfish Cherax quadricarinatus using suppression subtractive hybridization.J. Virol. Microbiol. 2013. http://dx.doi.org/10.5171/2013.569032.
Safeena, M.P., Tyagi, A., Rai, P., Karunasagar, I., 2010. Complete nucleic acid sequence ofPenaeus monodon densovirus (PmDNV) from India. Virus Res. 150, 1–11.
Seki, T., Tada, S., Katada, T., Enomoto, T., 1997. Cloning of a cDNA Encoding a NovelImportin-a Homologue, Qip1: Discrimination of Qip1 and Rch1 from hSrp1 by TheirAbilityto Interact with DNA Helicase Q1/RecQL. Biophys. Res. Commun. 234, 48–53.
Wagstaff, K.M., Rawlinson, S.M., Hearps, A.C., Jans, D.A., 2011. An AlphaScreen(R)-basedassay for high-throughput screening for specific inhibitors of nuclear import.J. Biomol. Screen. 16, 192–200.
Wagstaff, K.M., Sivakumaran, H., Heaton, S.M., Harrich, D., Jans, D.A., 2012. Ivermectin isa specific inhibitor of importin alpha/beta-mediated nuclear import able to inhibitreplication of HIV-1 and dengue virus. Biochem. J. 443, 851–856.
Wislocki, P.G., Grosso, L.S., Dybas, R.A., 1989. Environmental aspects of abamectin use incrop protection. Ivermectin and Abamectin.Springer Verlag, New York 182–200.
