Goldfish (Carassius auratus L.) possess natural antibodies with trypanocidal activity towards Trypanosoma carassii in vitro

Fish & Shellfish Immunology 34 (2013) 1025e1032

Contents lists available at SciVerse ScienceDirect

Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi

Goldfish (Carassius auratus L.) possess natural antibodies with trypanocidal activity towards Trypanosoma carassii in vitro Barbara A. Katzenback a, Debbie A. Plouffe a, Miodrag Belosevic a, b, * a b

Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada School of Public Health, University of Alberta, Edmonton, Alberta, Canada

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 November 2012 Received in revised form 10 December 2012 Accepted 11 December 2012 Available online 18 January 2013

Natural infection of cyprinids, such as carp, with the extracellular protozoan parasite Trypanosoma carassii can attain up to 100% prevalence and cause significant host morbidity and mortality, particularly in aquaculture settings. Host recovery from T. carassii infection has been shown to be antibody (Immunoglobulin M; IgM)-mediated, conferring long-term immunity in recovered animals upon challenge. To assess the role of IgM in parasite clearance in the goldfish, IgM was purified by PEG-6000 precipitation from goldfish serum collected at 0 (naïve), 21 (peak parasitaemia) and 42 (recovery phase; immune) days post infection (dpi) and used for in vitro assays. Purified IgM from 0, 21, and 42 dpi serum showed dose- and time-dependent trypanocidal activity in vitro. Incubation of T. carassii with 0 dpi IgM showed the greatest reduction in trypanosome numbers after 24 h, followed by 42 dpi IgM, and finally by 21 dpi IgM. The trypanocidal activity of the PEG-purified IgM was abrogated by pre-absorption with parasites in vitro and was affected by temperature. Furthermore, studies using 0 dpi IgM purified using gel permeation chromatography showed increased trypanocidal activity, with complete elimination of parasites after 12 h when incubated with 200 mg of 0 dpi IgM, or by 24 h when incubated with 80 mg or 100 mg of 0 dpi IgM. Lastly, in vivo passive transfer experiments demonstrated that while immune serum or purified IgM from 42 dpi serum conferred protection against a challenge, neither 0 dpi serum or 0 dpi purified IgM conferred protection against challenge with T. carassii. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Trypanosoma carassii Natural antibodies Immunoglobulin M Goldfish

1. Introduction Trypanosoma carassii (syn. Trypanosoma danilewskyi), an extracellular blood-borne protozoan parasite, can be transmitted amongst a number of fish species such as goldfish (Carassius auratus) and carp (Cyprinus carpio) through the blood meal of a leech vector [1,2]. Prevalence of T. carassii infection may reach 100% in natural fish populations [3]. Parasites replicate within the blood of the fish host and reach peak parasitaemia levels approximately 2e3 weeks post infection [4]. Individuals that are able to control parasite numbers generally enter the chronic stage of infection by 6e8 weeks post infection, where very low numbers or no trypanosomes can be observed in the host [4]. Hosts that recover from infection acquire non-sterile immunity and are resistant to secondary T. carassii infections for up to 190 days [5e7]. Immunity to T. carassii is thought to be antibody-mediated as demonstrated by resistance * Corresponding author. Distinguished University Professor, CW- 405 Biological Sciences Building, University of Alberta, Edmonton, AB, Canada T6G 2E9. Tel.: þ1 780 492 1266; fax: þ1 780 492 2216. E-mail address: [email protected] (M. Belosevic). 1050-4648/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fsi.2012.12.018

to primary infection upon passive transfer of serum from recovered to naïve hosts [6,8]. The generation of antibodies towards T. carassii suggested targeting of specific parasite-derived molecules. Indeed, previous studies by our laboratory demonstrated that immunization of fish with T. carassii excretoryesecretory (ES) products conferred significant protection to naïve hosts (two to five-log reduction in parasitaemia), after challenge infection [9]. Further evaluation of the trypanosome antigens contained in the ES products revealed that T. carassii alpha and beta tubulin were recognized by serum from naïve and recovered fish [9,10]. When parasites were incubated with polyclonal rabbit antibodies generated against recombinant T. carassii tubulin molecules, a dose-dependent trypanocidal activity was observed [8,10]. Immunofluorescence imaging of parasites incubated with 80 mg and 160 mg anti-tubulin antibodies in vitro showed discrete staining of intracellular targets. Furthermore, immunization of naïve goldfish with recombinant T. carassii beta-tubulin conferred partial protection to parasite challenge [8]. These studies suggested that generation of specific host antibodies to parasite tubulin subunits are, in part, responsible for parasite recognition and elimination during infection.

1026

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

During the course of our studies on the antibody response of goldfish to T. carassii infection, we were intrigued by the presence of anti-parasite tubulin antibodies in naïve fish serum. While these results were unexpected, they are not unprecedented as natural antibodies to host molecules, such as tubulin, have been documented in a number of species including mammals and fish [11e 16]. Therefore, our objective in this study was to assess the trypanocidal activity of purified IgM antibodies from naïve and immune fish towards T. carassii in vitro, and whether passive transfer of concentrated purified IgM from naïve and immune hosts confers protection against T. carassii infection in vivo. 2. Materials and methods 2.1. Fish Goldfish (Carassius auratus L.) were purchased from either Mount Ozark Fisheries Inc. (Southland, MO, USA) or Grassy Forks Fisheries (Martinsville, IN, USA). Goldfish were fed ad libitum and housed in tanks with a continuous-flow water system at 17  C in the Aquatic Facility of the Biological Sciences Building, University of Alberta. Prior to manipulation (bleeding, injection, clipping) fish were anaesthetized by immersion in a solution of tricaine methane sulfonate (TMS; 50 mg/L). For infection and passive transfer experiments, fish were approximately 10e15 cm in length. When necessary, fish were marked by fin clipping. The animals in the Aquatic Facility were maintained according to the guidelines of the Canadian Council of Animal Care (CCAC-Canada). 2.2. Preparation of goldfish serum for parasite maintenance Following sedation with TMS, blood was withdrawn from the caudal vein of goldfish, pooled and allowed to clot overnight at 4  C. Blood was centrifuged for 30 min at 1560  g and the serum collected. Serum used for the maintenance of parasite cultures was heat inactivated for 30 min at 56  C, filter sterilized (0.22 mm, Millipore) and stored at 20  C until use.

heparinized capillary tubes (40 mL). Blood samples were examined for the presence of trypanosomes using a haemocytometer fitted with a glass cover slip and a bright field microscope (400). If parasites were not detectable by this method, the heparinized capillary tubes were centrifuged (5 min in a micro-haematocrit centrifuge) and examined for the presence of parasites [19]. 2.6. Generation of goldfish plasma for in vitro assays and in vivo passive transfer experiments Fish (12 per group) were infected with 6.25  106 trypanosomes as described in Section 2.4. At 21 and 42 days post infection (dpi), fish were exsanguinated and blood stored at 4  C overnight to allow for clotting. The following day, the blood was centrifuged for 30 min at 1560  g and the plasma removed and stored at 4  C until needed. For day 0 plasma collection, fish were exsanguinated at the start of the experiment prior to infection with T. carassii. In some experiments, plasma was heat-inactivated by incubation at 56  C for 30 min in a circulating water bath. 2.7. Purification of goldfish Immunoglobulin M (IgM) using polyethylene glycol IgM was purified from goldfish plasma as previously described [20]. Briefly, the plasma sample was diluted 1:10 in 1 PBS (pH 7.4), and polyethylene glycol (PEG)-6000 powder added to a final concentration of 9% w/v with stirring over 30 min at room temperature. The solution was centrifuged for 10 min at 4000  g, the supernatant removed, and the pellet washed twice with a 9% PEG6000/PBS solution (pH 7.4). The pellet fraction was dissolved in 1 PBS (pH 7.4). The pellet fraction containing the goldfish IgM was subsequently dialyzed against 4 L of 1 PBS at 4  C overnight. The PEG purified IgM/PBS fraction was filter sterilized (0.22 mm, Millipore) and stored at 4  C until use. SDS-PAGE and Western blotting were performed on the supernatants and the re-suspended pellet to determine the presence of goldfish IgM. 2.8. Preparation of Superose 6 purified goldfish IgM

2.3. Parasites Trypanosoma carassii (strain TrCa) was isolated from a crucian carp (Carassius carassius) by Dr. J. Lom in 1977. The parasites were obtained from Dr. P.T.K. Woo, University of Guelph, Ontario, Canada. Trypanosomes were cultured in TDL-15 medium supplemented with 10% heat-inactivated goldfish serum [4] and passed (10% v/v) every 6e7 days. Trypanosomes used for all assays and infections were obtained from 7-day old stock cultures as described previously [4,17,18]. 2.4. Infection of goldfish with T. carassii Prior to infection, a 50 mL blood sample from all fish was withdrawn and examined for the presence of hemoflagellates. Fish were injected intraperitoneally (i.p.) with 6.25  106 trypanosomes in 100 mL of serum-free TDL-15 medium using a 1 mL syringe fitted with a 25-gauge needle. When required, fin clipping was performed to identify individual fish. 2.5. Course of infection Parasitaemia was monitored in infected fish by withdrawing 50 mL blood samples from the caudal vein of the fish at various time points throughout the infection. Dilutions of the sample were made in trisodium citrate anticoagulant (100 mM trisodium citrate, 40 mM glucose, pH 7.3), as well as collection of the blood sample in

In certain experiments, the PEG-6000 purified IgM from naïve goldfish was further purified using gel permeation fast performance liquid chromatography. The PEG 6000 precipitated fraction containing partially purified IgM was filter sterilized (0.22 mm), applied to a Superose 6 column and eluted in 1 PBS (pH 7.4) at a flow rate of 0.4 mL/min. The fractions were collected in 15 mL centrifuge tubes and were analyzed for the presence and purity of goldfish IgM by Western blotting and silver staining of SDS-PAGE gels. The fractions containing goldfish IgM were pooled, and concentrated by PEG-18000 chips being applied to the snakeskin containing the pooled fractions. The sample was then dialyzed in 4 L of 1 PBS (pH 7.4) overnight at 4  C. 2.9. SDS-PAGE and Western blotting Proteins were separated and visualized by reducing SDS-PAGE. Samples were dissolved in an equal volume of Laemmli sample buffer (BioRad), heated at 95  C for 5 min and electrophoresed through 12% polyacrylamide gels at 100 V for 15 min followed by 185 V for 45 min. For Western blotting, proteins were transferred to 0.2 mm nitrocellulose membranes (BioRad) at 100 V for 1 h in transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol). Membranes were blocked with 0.5% BSA in Tris-buffered saline/Tween 20 (TTBS; 0.1% Tween 20 in 100 mM TriseHCl, 0.9% NaCl, pH 7.5; TBS) for 30 min at room temperature, followed by mouse-anti-carp IgM

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

(IgG mAb WCI 12), diluted in blocking solution 1:20, for 3 h at room temperature. Following incubation with the primary antibody, blots were washed three times in TTBS and three times in TBS for 5 min each. Finally, membranes were incubated with alkaline phosphatase-labeled goat-anti-mouse IgG diluted 1:1500 in blocking buffer for 1 h at room temperature, followed by three washes in TTBS and three washes in TBS for 5 min each. Protein bands were visualized using the chromogenic BCIP/NBT development kit according to the manufacturer’s instructions (BioRad). 2.10. Silver staining Following protein separation on SDS-PAGE gels as described in Section 2.9, polyacrylamide gels were silver stained using a silver stain kit (BioRad) according to manufacturer’s specifications. Briefly, following Coomassie staining for 30 min, the gels were destained overnight in 40% methanol, 10% acetic acid solution. The gels were rinsed with water to remove the destaining solution and the oxidizing solution added to the gels for 5 min with rocking. The gels were rinsed with water to remove excess oxidizing solution and the silver reagent added for 20 min. Following multiple washes with water, developing solution was added, and development allowed to proceed until protein bands were visualized. Development was stopped with the addition of 5% acetic acid. 2.11. Determination of total protein concentration Total protein concentrations of purified goldfish IgM preparations were determined using the Pierce Micro BCA protein assay (BioRad) according to manufacturers instructions. Briefly, a standard curve was created with bovine serum albumin (BSA). The IgM samples were diluted in varying ratios and the working reagent provided with the kit added to all samples in a 1:1 ratio. Samples were incubated at 60  C for 1 h and read in an automated plate reader at a wavelength of 570 nm. Protein concentrations were determined using the BSA standard curve. 2.12. Enumeration of parasites in in vitro assays Trypanosomes from in vitro assays were counted by first resuspending the parasites by gently pipetting, followed by withdrawing a 10 mL sample and counting on an improved haemocytometer under a bright field microscope (40). Parasites were classified as viable if any movement was observed in the flagellum or undulating membrane. In rare cases, if a trypanosome did not appear to have any movement, the parasite was observed for at least 2 min to ensure it was not viable. In most cases, however, if a parasite was observed under the microscope during enumeration, it normally exhibited flagellar or undulating membrane movement. 2.13. The effects of IgM from naïve and infected goldfish on the growth of T. carassii in vitro In vitro-cultured trypanosomes were washed twice by centrifugation for 15 min at 400  g in serum-free TDL-15 medium and re-suspended to a concentration of 1  106 trypanosomes/mL. Parasites and treatments were then placed into wells of a 96 well plate. Each well contained 100 mL of parasites in serum free TDL-15 medium and 100 mL of the desired treatment in 1 PBS. Treatments were seeded in duplicate. Plates were incubated at 20  C in the absence of added CO2. From each well, 10 mL samples were withdrawn at every time point. The numbers of parasites remaining in each treatment were determined using a haemocytometer. For the first set of experiments, the treatments consisted of 1 PBS, 1 PBS from a mock PEG-6000 purification, and various

1027

concentrations of PEG-purified IgM from 0 dpi, 21 dpi, and 42 dpi goldfish serum. In certain experiments, IgM was purified by fast performance liquid chromatography (FPLC) from healthy goldfish (0 dpi) and was used as a treatment. For wells containing the FPLC purified goldfish IgM, the treatments were 10, 20, 40, 80, 100, and 200 mg of goldfish IgM. IgM from immune serum was not further purified by FPLC due to the limiting quantities of immune serum available. Plates were incubated at 20  C for 3 or 4 days, and samples from each well were taken daily and parasites enumerated using a haemocytometer. In certain experiments, 1  105 parasites were used to preabsorb the PEG-purified and FPLC-purified IgM to determine the effect of pre-absorbed IgM on the growth of the parasites. In these experiments, after the initial four-day incubation, the plates were centrifuged for 10 min at 400  g, and 100 mL of the supernatant removed and added to another well in a 96-well plate containing 100 mL of 1  106 parasites/mL. Parasites were enumerated daily for four days. These experiments were performed using purified IgM from two separate groups of infected goldfish. To further examine the effects of trypanocidal activity of the day 0 goldfish IgM, a series of experiments were set up. The first set of experiments examined the time kinetics of trypanocidal activity of the day 0 IgM against T. carassii. Plates were set up as described above, however, samples for parasite enumeration were taken after 1, 2, 4, 8 12 and 24 h post treatment of T. carassii with either 1 PBS or 10 mg, 20 mg, 40 mg, 80 mg or 160 mg of day 0 PEG-purified IgM. In the second set of experiments, the effect of temperature on the trypanocidal activity of day 0 IgM was examined. Plates were set up in the same manner as described above, however, in addition to a plate being incubated at 20  C, a duplicate plate was incubated at 4  C. The numbers of parasites were determined daily for four days. 2.14. Passive transfer of purified goldfish IgM Small blood samples were taken from each fish prior to infection with T. carassii, to ascertain that the fish did not have a hemoflagellate infection. Fish were injected i.p. with 6.25  106 trypanosomes in a 100 mL injection and fin clipped for identification purposes. Two days post-infection, fish were either injected i.p. with sterile 1 PBS, day 0 (non-immune) plasma, day 42 (immune) plasma, 500 mg PEG-purified day 0 IgM, 500 mg PEG-purified day 0 IgM pre-absorbed with 9.25  109 parasites, 500 mg PEG-purified day 42 IgM, or 500 mg PEG-purified day 42 IgM pre-absorbed with 9.25  109 parasites, all in a 500 mL injection volume. Preabsorption of IgM with parasites was done at 20  C for 24 h, the supernatant collected, and held at 4  C until administration to fish, 24 h later. Blood samples were taken at 0, 2, 7, 14, 21, and 31 dpi from fish in all experimental groups and the numbers of parasites per mL determined using a haemocytometer. All groups contained four fish, except for the immune IgM and immune IgM preabsorbed group that contained three fish per group due to limiting quantities of immune serum. 2.15. Statistics Statistical analysis was performed using GraphPad 5.0 software. A repeated measures two-way ANOVA was employed to compare differences in parasite numbers between treatments for the 0 dpi time-course assay (Fig. 3) the short-term time-course assay (Fig. 4), temperature-dependence assay (Fig. 5) and between parasitaemia levels in infected fish injected with different sera or purified IgM preparations following challenge with T. carassii (Fig. 6). A Boniferroni multiple comparisons post-hoc test was performed. Logarithmic transformation of the data was performed to stabilize the

1028

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

variances, 1 was added to each value prior to transformation. The probability level of P  0.05 was considered significant. 3. Results 3.1. Purification of goldfish IgM from serum IgM from goldfish serum was purified using a protocol adapted from [20]. The addition of 9% PEG-6000 powder resulted in the formation of a precipitate. Analysis of the precipitate and supernatant by SDS-PAGE (Fig. 1A) and Western blotting (Fig. 1B) in comparison to whole serum showed goldfish IgM to be detected in the precipitate as demonstrated by the presence of bands at w80 kDa and w25 kDa following Coomassie staining, representative of the heavy and light chains of fish IgM (Fig. 1A). The identity of goldfish IgM in the precipitate fraction was confirmed using a monoclonal antibody that recognizes the heavy chain of carp IgM that also recognizes goldfish IgM (Fig. 1B). Fig. 1. Polyethylene glycol (PEG) precipitation of goldfish serum. Serum was pooled from either healthy goldfish or goldfish infected with T. carassii. The immunoglobulin fraction was isolated using 9% PEG. Whole serum (1), 9% PEG pellet (2), and the 9% PEG supernatant (3) were analyzed for the presence of goldfish IgM using SDS-PAGE run under reducing conditions followed by Coomassie blue staining (A) and Western blotting (B). IgM was detected in (B) using a monoclonal antibody to an epitope of the heavy chain of carp IgM (80 kDa).

3.2. Inhibition of parasite growth in vitro by PEG-purified goldfish IgM To assess the effects of PEG-purified IgM on T. carassii in vitro, varying concentrations of IgM, purified from healthy (0 dpi) or T. carassii-infected goldfish at 21 dpi and 42 dpi, were incubated with 1  105 parasites for four days and the number of parasites determined daily. The 21 dpi and 42 dpi time points were chosen

Fig. 2. In vitro trypanocidal assays using PEG-purified IgM from goldfish serum after 0, 21 and 42 days post infection with T. carassii. PEG-purified IgM from healthy goldfish serum (0 dpi; A, B), serum from T. carassii infected goldfish after 21 days post infection (21 dpi; C, D) or 42 days post infection (42 dpi; E, F) were incubated with in vitro cultivated T. carassii for four days at 20  C. Parasites were seeded at a concentration of 1  106 parasites/mL in a volume of 100 mL. An additional 100 mL of treatment (PBS or 10e200 mg of PEG-purified IgM in PBS) was added in duplicate wells and parasite numbers enumerated daily using a haemocytometer. Shown are two representative experiments (A, C, E and B, D, F), with the IgM being generated from two different batches of infected fish. Serum was pooled from 5 to 6 fish at each time point prior to IgM isolation.

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

Fig. 3. In vitro trypanocidal assay using gel permeation chromatography purified IgM from 0 dpi goldfish serum. Duplicate wells containing 1  105 parasites were treated with 10e200 mg of gel permeation chromatography purified 0 dpi IgM in PBS (A) or 10e200 mg of pre-absorbed purified 0 dpi IgM that had been previously incubated for four days with 1  105 parasites (B). An equal volume of PBS was used as a negative control. Small samples were taken from each well daily for three days and enumerated using a haemocytometer. Two independent experiments were performed. Asterisks indicate a significant difference from PBS controls, P < 0.05.

for collection of goldfish serum and subsequent IgM purification as these time points represent distinct phases during T. carassii infection; the acute stage of infection where peak parasitaemia levels are observed, and the chronic stage of infection characterized by resolution of infection, respectively. Upon performing these assays, batch-to-batch variability in the potency of trypanocidal activity was observed. This variability was likely the result of the inherent variation in the initial immune status of animals and the

Fig. 4. Time course of trypanocidal activity of purified IgM. Trypanosomes (1  105) were treated with PBS or 80 mg, 100 mg or 200 mg of purified IgM in duplicate wells. Parasites were enumerated after various time points (0e24 h) using a haemocytometer. Two independent experiments were performed. A two-way repeated measures ANOVA with a Bonferroni post hoc test was performed and P < 0.05 was considered significant. (a) denotes significance from PBS, (b) denotes significance from 80 mg, (c) denotes significant difference from 100 mg and (d) denotes significance from 200 mg.

1029

Fig. 5. Effect of temperature on trypanocidal activity of purified IgM from 0 dpi goldfish serum. Trypanosomes (1  105) were incubated with varying concentrations of purified IgM at 20  C (A) or 4  C (B) and parasite numbers enumerated daily for three days. Two independent experiments were performed. A two way repeated measures ANOVA with a Bonferroni post hoc test, with all pair-wise comparisons being made to the PBS control, was performed and P < 0.05 was considered significant. Asterisks denotes significance from the PBS control.

generation of antibodies during the course of infection. Therefore, instead of pooling the data from all in vitro assays performed, we chose to show two experiments performed using PEG-purified IgM generated from the plasma of two independent groups of healthy or previously infected fish (Fig. 2). PEG-purified IgM from 0 dpi plasma (Fig. 2A and B), 21 dpi plasma (Fig. 2C and D) and 42 dpi plasma (Fig. 2E and F) blocked the growth of T. carassii in vitro in a dose- and time-dependent manner. Surprisingly, the purified IgM from 0 dpi plasma had the greatest trypanocidal activity against T. carassii (Fig. 2A and B). Incubation of parasites with 80 mg, 100 mg or 200 mg of PEG-purified IgM from 0 dpi plasma was trypanocidal with no trypanosomes remaining after 1 (Fig. 2B) or 2 days (Fig. 2B). Incubation of parasites with 40 mg of 0 dpi IgM showed a decline in parasite numbers (Fig. 2A and B), and in some cases, parasites were completely eliminated after 4 days of incubation (Fig. 2B). Treatment of parasites with 10 mg or 20 mg of 0 dpi IgM did not have trypanocidal activity compared to the PBS controls (Fig. 2B). In contrast, trypanocidal activity was only observed for IgM from 21 dpi plasma at the highest dose, 200 mg, with no trypanosomes remaining after 2 (Fig. 2D) or 4 days (Fig. 2C). Lastly, when parasites were incubated with PEG-purified IgM from 42 dpi plasma, trypanocidal activity was observed at the 80 mg, 100 mg and 200 mg IgM treatments (Fig. 2E and F). In the first experiment shown (Fig. 2E), no trypanosomes were present after 2 days of incubation with 200 mg 42 dpi IgM and after 4 days of incubation with 80 mg and 100 mg of 42 dpi IgM (Fig. 2E). In the second experiment shown, no parasites were present in wells containing 80 mg, 100 mg or 200 mg of 42 dpi IgM after 2 days (Fig. 2F). It should be noted, however, that despite the batch-to-batch variability, the observed trend in trypanocidal

1030

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

observed compared to the PBS control, regardless of incubation time or IgM concentration (Supplementary Fig. 1A, B and C). 3.4. Gel permeation chromatography purification goldfish IgM

Fig. 6. In vivo passive transfer of PEG-purified IgM from 0 dpi and 42 dpi goldfish serum. Goldfish were infected i.p. with 6.25  106 trypanosomes. Two days post infection, goldfish were injected i.p. with 500 mL of PBS, 42 dpi immune serum (I serum), 42 dpi purified IgM (I IgM) or 42 dpi purified IgM pre-absorbed against 9.25  109 parasites (I IgM pre-abs) (A) and an equivalent study performed using 0 dpi serum (NI serum), 0 dpi PEG-purified IgM (NI IgM) or 0 dpi PEG-purified IgM preabsorbed against 9.25  109 parasites prior to injection (NI IgM pre-abs) (B). Four fish per treatment were used except for the I IgM and I IgM pre-abs groups that contained three fish per group. A two way repeated measures ANOVA with a Bonferroni post hoc test was performed and P < 0.05 was considered significant. (a) denotes significance from PBS, (b) denotes significance from serum, (c) denotes significant difference from IgM and (d) denotes significance from IgM pre-abs.

activity of PEG-purified IgM from different days post infection was consistent. To rule out the possibility of osmotic lysis due to PEG-6000 contamination, controls were set up in which parasites were incubated with 100 mL of a mock-PEG precipitation control or with 100 mL of a 9% PEG-6000 in PBS solution. No decline in parasite numbers were observed compared to the PBS controls (data not shown). Furthermore, the process of heat-inactivation of plasma by incubation at 56  C for 30 min did not affect the ability of PEGpurified goldfish IgM to inhibit growth of trypanosomes in vitro (data not shown) indicating that the trypanocidal effect of IgM was not due complement. In fact, we have previously shown T. carassii to be resistant to complement-mediated lysis in vitro [21]. These results suggest that both IgM from naïve and immune goldfish were effective in decreasing T. carassii numbers in vitro. 3.3. Pre-absorption of PEG-purified IgM against T. carassii To assess whether the molecule responsible for parasite decline in the PEG-purified IgM preparations from 0 dpi, 21 dpi, or 42 dpi plasma could be depleted, 100 mL of supernatant was taken from each well from the in vitro trypanocidal assays set up as described in Section 3.2 after 4 days of parasite incubation with IgM treatments. Plates were first centrifuged to pellet any remaining trypanosomes, 100 mL of supernatant withdrawn, and then transferred to the wells of a new plate each containing 1  105 parasites in 100 mL. During the four days of incubation with the pre-absorbed 10 mg, 20 mg, 40 mg, 80 mg, 100 mg and 200 mg PEG-purified IgM from 0 dpi, 21 dpi or 42 dpi goldfish plasma, no declines in the parasite numbers were

Since the partially purified IgM from 0 dpi goldfish plasma was effective in inhibiting parasite growth in vitro, and to a greater extent than IgM preparations from 42 dpi goldfish plasma, we decided to further purify the IgM from healthy, naive fish to determine if there was a co-precipitating molecule responsible for the trypanocidal activity observed. To this end, IgM from naïve goldfish plasma (0 dpi) was PEG-precipitated and then purified to homogeneity using gel permeation chromatography by running the samples on a Superose 6 column. The presence of proteins in individual elution fractions was monitored by measuring optical density at 280 nm. A representative trace of an IgM purification run on a Superose 6 column is shown in Supplementary Fig. 2A. Three major peaks were observed during gel permeation chromatography, consisting of fractions 13e14, 16e20 and 23e26 (Supplementary Fig. 2A). A sample of the pooled fractions was run on a SDS-PAGE gels and subsequently silver stained (Supplementary Fig. 2B) or used for Western blotting with a monoclonal to the heavy chain of carp IgM, known to cross react with the heavy chain of goldfish IgM, to detect the presence of goldfish IgM in the fractions (Supplementary Fig. 2C). Fractions 16e20 were determined to contain the purified IgM based on silver staining and Western blotting (Supplementary Fig. 2B and C). 3.5. In vitro trypanocidal activity of purified IgM from naïve fish on T. carassii There was a dose- and time-dependent decrease in trypanosome numbers in the presence of 20 mg, 40 mg, 80 mg, 100 mg and 200 mg of gel permeation chromatography purified 0 dpi IgM. There was a significant reduction in parasite numbers in the 40 mg, 80 mg, 100 mg and 200 mg purified 0 dpi IgM treated groups compared to the PBS control group after one day of incubation (p < 0.001), with no parasites remaining in the wells treated with 80 mg, 100 mg or 200 mg of IgM (Fig. 3A). At the lower concentrations of IgM, no parasites remained after two days of incubation with 40 mg of IgM or after three days with 20 mg of IgM (Fig. 3A). No differences in trypanocidal activity were observed between PBS and 10 mg IgM treated parasites over the course of incubation (Fig. 3A). Furthermore, incubation of trypanosomes with parasite pre-absorbed IgM abrogated the observed trypanocidal activity (Fig. 3B). 3.6. Time course of 0 dpi purified IgM To examine the kinetics of trypanocidal activity of purified IgM prior to 24 h of incubation, parasites were incubated with 80 mg, 100 mg or 200 mg of purified IgM and parasites enumerated after 1 h, 2 h, 4 h, 8 h, 12 h and 24 h. After 8 h, cultures treated with 200 mg of IgM showed a significant decrease in parasite numbers compared to the PBS, 80 mg and 100 mg IgM treated groups (p < 0.0001), with no parasites remaining after 12 h (Fig. 4). A similar trend was observed with the 100 mg IgM treated group, with a significant 3-log decrease in parasite numbers after 12 h of incubation (p < 0.0001), and no parasites remaining after 24 h (Fig. 4). In the 80 mg IgM group, no viable parasites were observed after 24 h of incubation (Fig. 4). 3.7. Effect of temperature on trypanocidal activity of purified IgM on T. carassii To assess whether the incubation temperature influenced the trypanocidal activity of purified IgM, the assays were performed at

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

20  C and at 4  C. At 20  C, no parasites were observed in the wells containing 100 mg and 200 mg of IgM after one day (Fig. 5A). After two days of incubation, no parasites were observed in the 80 mg IgM treatment group, and after three days, no parasites were observed in the 40 mg IgM group (Fig. 5A). In comparison, the complete elimination of parasites in the 80 mg, 100 mg and 200 mg IgM groups was not observed until two days of incubation, with no parasites after three days in the 40 mg IgM group at a temperature of 4  C (Fig. 5B). The delay in trypanocidal activity of purified IgM when the assay was performed at 4  C suggests a temperature-dependent mechanism of trypanosome killing. 3.8. Passive transfer of PEG-purified IgM from naïve and immune goldfish plasma To assess the ability of IgM from non-immune (0 dpi) and immune (42 dpi) animals to confer protection, passive transfer experiments were performed. Fish were first injected with 6.25  106 T. carassii and two days post infection, were injected with 500 mL of sterile 1 PBS (sham-injected control), 500 mL of immune serum (42 dpi), 500 mg of PEG-purified immune IgM in 500 mL of PBS, or 500 mg of PEG-purified immune IgM previously pre-absorbed with trypanosomes (Fig. 6A). Parasitaemia was monitored over the course of infection, up to 31 dpi. Passive transfer of immune serum (p < 0.0001) and purified IgM from immune serum (p < 0.05 at day 7, p < 0.001 at day 14, p < 0.01 at day 21 and p < 0.05 at day 31) conferred significant protection against parasite challenge at all time points, with approximately a 3e5 log decrease in parasitaemia compared to the sham-injected PBS controls, (Fig. 6A). Although not statistically significant (p > 0.05), there was a general trend of passively transferred immune serum conferring a greater protection compared to passively transferred purified IgM from immune serum. The average parasitaemias of PBS-injected control fish and those injected with purified IgM pre-absorbed against parasites in vitro were not significantly different (p > 0.05, Fig. 6A). However, there was a generalized decrease in parasitaemia in the IgM-preabsorbed group compared to fish in the PBS control group during the course of infection (Fig. 6A), suggesting not all IgM was preabsorbed by the parasites, thereby conferring a low level of protection to fish challenged with T. carassii (Fig. 6A). The second experiment using non-immune serum (0 dpi) and IgM from non-immune serum was set up in an identical fashion to the first study with immune serum and IgM, as described above (Fig. 6B). In contrast to the first experiment where immune serum and purified IgM from immune serum were able to confer significant levels of protection against subsequent parasite challenge, no significant differences were observed between the fish from the PBS control, non-immune serum, purified IgM from non-immune serum, or non-immune IgM pre-absorbed with parasites in vitro groups (p > 0.05, Fig. 6B). 4. Discussion Natural antibodies, also termed residual, background, innate or pre-existing antibodies, are primarily of IgM isotype [22] and are found in all animals, such as fish, as a result of vertical transfer of maternal antibodies [23,24], exposure to environmental antigens [25], or through a germline-encoded product [25,26]. Natural antibodies are considered to be one of the first lines of defense, are non-specific in their recognition-often cross reacting with a number of host epitopes such as tubulin, actin, myosin, thyroglobulin and pathogen epitopes [14e16], and have low affinity [27]. This is in contrast to antibodies developed during an acquired immune response that act as the second line of antibody defense and are antigen specific with high affinity binding [27]. In fish, there is

1031

a great variability in natural antibody levels amongst individuals, their levels often increasing with the age of the fish and the environmental conditions they were reared in [16,25]. Studies in cyprinids have shown that fish with higher natural antibody levels, compared to those with lower levels, to certain pathogens were associated with lower morbidity rates upon pathogen challenge [13,28]. We examined the trypanocidal activity of purified IgM from the serum of naïve and T. carassii e infected goldfish. IgM from both naïve and infected goldfish showed trypanocidal activity towards T. carassii in vitro, in a time- and dose-dependent manner. IgM from naïve goldfish appeared to have the greatest trypanocidal activity in vitro, similar to IgM obtained from 42 dpi goldfish. Preabsorption of the IgM with parasites abrogated the trypanocidal activity, suggesting that the responsible molecule was binding to or being taken up by the parasite, thereby depleting the responsible molecule from the preparations. The results from the in vitro assays using 21 and 42 dpi IgM are consistent with the time course of the infection and agree with previous findings [6,8]. In particular, Overath et al. [3] reported that the relative level of IgM binding to T. carassii lysates in an ELISA increased by w15% in day 20 serum and by w60% in day 40 serum compared to day 0 serum, suggesting that amount, specificity or both, were increasing during the course of T. carassii infection. In addition, we have previously reported that IgM from day 42 serum can bind to the surface of T. carassii whereas day 0 serum does not bind to T. carassii at an appreciable level [8] and that passive transfer of serum or IgM from recovered hosts confers protection to naïve hosts [6,8], again speaking to the specificity of 42 dpi IgM. Therefore, the trypanocidal activity of 42 dpi IgM in vitro was expected. In comparison, the 21 dpi IgM had considerably lower trypanocidal activity. This may be due to polyclonal lymphocyte activation during the acute phase of infection when parasites levels are highest. Polyclonal activation of lymphocytes is a characteristic of other mammalian trypanosome infections and has been implicated in the immunosuppression associated with Trypanoplasma borreli infection in carp [29,30]. Polyclonal activation by T. carassii in goldfish is further supported by the observation that goldfish exhibit high antibody titres during the peak parasitaemia [9]. Production of large quantities of irrelevant, non-specific antibodies would dilute the concentration of specific antibodies required for the elimination of the trypanosomes. Another possibility for the low trypanocidal activity of 21 dpi IgM is that the effective antiparasite antibodies may be bound to the parasite surface, resulting in decreased circulating levels of parasite specific IgM in the plasma during the acute phase of infection. Alternatively, studies performed in our laboratory have shown T. carassii to secrete a protease, gp63 [31], and demonstrated gp63 to dampen cellular antimicrobial responses. It would be interesting to pursue studies examining the effect of T. carassii excretory/secretory products, such as gp63, on antibody structure during the course of infection. It may be that during peak parasitaemia gp63 may act to affect the structure of fish IgM, thus altering the efficacy of IgM against T. carassii, and warranting future study. Regardless of the mechanism it is apparent that larger quantities of purified IgM from 21 dpi fish, compared to those from 42 dpi fish, were required to achieve similar trypanocidal activity in vitro. We were surprised to observe the highest trypanocidal activity with 0 dpi IgM, as passive transfer of non-immune serum does not protect animals against parasite challenge [8]. The 0 dpi IgM trypanocidal activity towards T. carassii in vitro was dose-, time- and temperature-dependent. These data suggest that a specific molecule may be responsible for the observed trypanocidal activity. Further purification of the 0 dpi IgM by gel permeation chromatography, as well as heat-inactivation of the samples, did not abrogate the

1032

B.A. Katzenback et al. / Fish & Shellfish Immunology 34 (2013) 1025e1032

observed trypanocidal activity, supporting the role of the 0 dpi IgM in trypanosome elimination. Furthermore, temperature dependence of the 0 dpi IgM trypanocidal activity suggests an active uptake mechanism of antibodies by the parasite [32,33]. Antibody uptake by the trypanosomes may occur at the site of the flagellar pocket, as this region is involved in active uptake of macromolecules by trypanosomes and has been proposed as the mechanism for antibodymediated trypanocidal activity in previous studies (reviewed in [34]). Taken together, these data suggest that the 0 dpi IgM must be bound by, or taken up by the parasite in order to exert its trypanocidal effect. The mechanism of trypanocidal activity by 0 dpi or 42 dpi IgM is largely unknown and requires further study. Based on our results, we believe the 0 dpi IgM fraction must contain natural antibodies of the goldfish that recognize and are trypanocidal to T. carassii. As mentioned previously, natural antibodies have a broad cross-reactivity with a number of host and pathogen derived molecules, such as Aeromonas salmonicida Aprotein, a major component of the bacterial cell membrane [13,35]. We suspect that one of the targets of natural antibodies in goldfish may be T. carassii tubulin subunits, as previously shown by Western blot and ELISA [8e10]. Furthermore, these anti-tubulin antibodies may contribute to the lytic activity of the 0 dpi IgM in vitro, as antitubulin antibodies were shown to be trypanolytic to T. carassii in vitro, and immunization of goldfish with T. carassii beta tubulin conferred partial protection against T. carassii infection in vivo [8]. While the 0 dpi IgM showed lytic effects in vitro, passive transfer of purified 0 dpi IgM did not confer protection to naïve goldfish upon T. carassii challenge. We believe this to be due to the broad crossreactivity and low affinity of natural antibodies. Thus, it is possible that after injection of 0 dpi IgM to fish, these natural antibodies may have bound a number of host epitopes, thus precluding them from binding parasite epitopes. This was not the case with passive transfer of 42 dpi IgM as the IgM antibodies present in this fraction were acquired antibodies that specifically recognize T. carassii with high affinity [3,8], and likely do not cross-react with host molecules, thus leaving them available to bind to and eliminate the parasites, in vivo. Thus, the natural antibodies from goldfish, once purified and concentrated in the absence of host-epitopes, are able to recognize and bind to T. carassii surface molecules. They are likely internalized through the flagellar pocket, and exert their trypanocidal effects once internalized. However, natural antibodies, while likely playing a role in initial host defense against pathogens in vivo, are blocked from efficiently recognizing and destroying trypanosomes, due to their low affinity and broad specificity permitting cross-reactivity with a number of host molecules. Our findings provide the first evidence for trypanocidal effects of teleost natural antibodies. Acknowledgements This work was supported by the Natural Sciences and Engineering Council of Canada (NSERC) to MB. BAK and DAP were supported by NSERC PhD scholarships. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.fsi.2012.12.018. References [1] Lunmsden WHR, Evans DA. Biology of fish trypanosomes and trypanoplasms, biology of kinetoplastida. New York: Academic Press; 1979. [2] Woo PT, Black GA. Trypanosoma danilewskyi: host specificity and host’s effect on morphometrics. J Parasitol 1984;70(5):788e93. [3] Overath P, Haag J, Mameza MG, Lischke A. Freshwater fish trypanosomes: definition of two types, host control by antibodies and lack of antigenic variation. Parasitology 1999;119(Pt 6):591e601.

[4] Wang R, Belosevic M. Cultivation of Trypanosoma danilewskyi (Laveran & Mesnil 1904) in serum-free medium and assessment of the course of infection in goldfish, Carassius auratus (L.). J Fish Dis 1994:1747e56. [5] Woo PT. Immune response of fish to parasitic protozoa. Parasitol Today 1987; 3(6):186e8. [6] Islam AK, Woo PT. Trypanosoma danilewskyi in Carassius auratus: the nature of protective immunity in recovered goldfish. J Parasitol 1991;77(2):258e62. [7] Woo PT. Acquired immunity against Trypanosoma danilewskyi in goldfish, Carassius auratus. Parasitology 1981;83(Pt 2):343e6. [8] Katzenback BA, Plouffe DA, Haddad G, Belosevic M. Administration of recombinant parasite beta-tubulin to goldfish (Carassius auratus L.) confers partial protection against challenge infection with Trypanosoma danilewskyi Laveran and Mesnil, 1904. Vet Parasitol 2008;151(1):36e45. [9] Bienek DR, Plouffe DA, Wiegertjes GF, Belosevic M. Immunization of goldfish with excretory/secretory molecules of Trypanosoma danilewskyi confers protection against infection. Dev Comp Immunol 2002;26(7):649e57. [10] Plouffe DA, Belosevic M. Antibodies that recognize alpha- and beta-tubulin inhibit in vitro growth of the fish parasite Trypanosoma danilewskyi, Laveran and Mesnil, 1904. Dev Comp Immunol 2006;30(8):685e97. [11] Karsenti E, Guilbert B, Bornens M, Avrameas S. Antibodies to tubulin in normal nonimmunized animals. Proc Natl Acad Sci U S A 1977;74(9):3997e4001. [12] Pateraki E, Portocala R, Labrousse H, Guesdon JL. Antiactin and antitubulin antibodies in canine visceral leishmaniasis. Infect Immun 1983;42(2):496e500. [13] Sinyakov MS, Dror M, Zhevelev HM, Margel S, Avtalion RR. Natural antibodies and their significance in active immunization and protection against a defined pathogen in fish. Vaccine 2002;20(31e32):3668e74. [14] Gonzalez R, Charlemagne J, Mahana W, Avrameas S. Specificity of natural serum antibodies present in phylogenetically distinct fish species. Immunology 1988;63(1):31e6. [15] Ardo L, Jeney Z, Adams A, Jeney G. Immune responses of resistant and sensitive common carp families following experimental challenge with Aeromonas hydrophila. Fish Shellfish Immunol 2010;29(1):111e6. [16] Magnadottir B, Gudmundsdottir S, Gudmundsdottir BK, Helgason S. Natural antibodies of cod (Gadus morhua L.): specificity, activity and affinity. Comp Biochem Physiol B Biochem Mol Biol 2009;154(3):309e16. [17] Bienek D, Belosevic M. Comparative assessment of growth of Trypanosoma danilewskyi Laveran and Mesnil 1904 in medium containing fish or mammalian serum. J Fish Dis 1997:209. [18] Bienek A. Biology of fish trypanosomes. Edmonton: University of Alberta; 2001. [19] Woo PT. The haematocrit centrifuge for the detection of trypanosomes in blood. Can J Zool 1969;47(5):921e3. [20] Zhong MC, Mor A, Avtalion RR. One-step procedure for the purification of goldfish (Carrasius auratus) and carp (Cyprinus carpio, L.) serum immunoglobulin by precipitation with 9% polyethylene glycol 6000. Israeli J Aquacult e Bamidgeh 1999;51(1):3e9. [21] Plouffe DA, Belosevic M. Enzyme treatment of Trypanosoma danilewskyi (Laveran and Mesnil) increases its susceptibility to lysis by the alternative complement pathway of goldfish, Carassius auratus (L.). J Fish Dis 2004;27(5): 277e85. [22] Boes M. Role of natural and immune IgM antibodies in immune responses. Mol Immunol 2000;37(18):1141e9. [23] Mor A, Avtalion RR. Transfer of antibody activity from immunized mother to embryo in tilapias. J Fish Biol 1990:37249e55. [24] Swain P, Dash S, Bal J, Routray P, Sahoo PK, Sahoo SK, et al. Passive transfer of maternal antibodies and their existence in eggs, larvae and fry of Indian major carp, Labeo rohita (Ham.). Fish Shellfish Immunol 2006;20(4):519e27. [25] Kachamakova NM, Irnazarow I, Parmentier HK, Savelkoul HF, Pilarczyk A, Wiegertjes GF. Genetic differences in natural antibody levels in common carp (Cyprinus carpio L.). Fish Shellfish Immunol 2006;21(4):404e13. [26] Carroll MC, Prodeus AP. Linkages of innate and adaptive immunity. Curr Opin Immunol 1998;10(1):36e40. [27] Baumgarth N, Tung JW, Herzenberg LA. Inherent specificities in natural antibodies: a key to immune defense against pathogen invasion. Springer Semin Immunopathol 2005;26(4):347e62. [28] Gonzalez R, Matsiota P, Torchy C, De Kinkelin P, Avrameas S. Natural anti-TNP antibodies from rainbow trout interfere with viral infection in vitro. Res Immunol 1989;140(7):675e84. [29] Minoprio P. Parasite polyclonal activators: new targets for vaccination approaches? Int J Parasitol 2001;31(5e6):588e91. [30] Saeij JP, de Vries BJ, Wiegertjes GF. The immune response of carp to Trypanoplasma borreli: kinetics of immune gene expression and polyclonal lymphocyte activation. Dev Comp Immunol 2003;27(10):859e74. [31] Oladiran A, Belosevic M. Recombinant glycoprotein 63 (Gp63) of Trypanosoma carassii suppresses antimicrobial responses of goldfish (Carassius auratus L.) monocytes and macrophages. Int J Parasitol 2012;42(7):621e33. [32] Cherian PV, Dusanic DG. Trypanosoma lewisi: ultrastructural observations of surface antigen movement induced by antibody. Exp Parasitol 1978;44(1): 14e25. [33] Barry JD. Capping of variable antigen on Trypanosoma brucei, and its immunological and biological significance. J Cell Sci 1979:37287e302. [34] Field MC, Carrington M. The trypanosome flagellar pocket. Nat Rev Microbiol 2009;7(11):775e86. [35] Sinyakov M, Dror M, Margel M, Avtalion RR. Immunogenicity of Aeromonas salmonicida A-protein in goldfish (Carassius auratus L.). Israeli J Aquacult e Bamidgeh 2001:53110e4.