INTRODUCTION
Parasite burdens in fish can be used as indicators of heavy-metal contamination in the environment due to the suppressive effects of heavy metals on the host immune response (MacKenzie et al. 1995; Lafferty, 1997). At the same time, heavy metals can be toxic to fish parasites, a notable example being Ichthyophthirius multifiliis, a protozoan pathogen responsible for ‘white-spot’ disease in freshwater species (Matthews, 1994; Dickerson & Dawe, 1995).
In the case of Ichthyophthirius, heavy metals kill water-borne stages of the parasite, and copper sulphate is sometimes used to control infections in home aquaria (Ling, Sim & Lam, 1993; Straus, 1993; Schlenk, Golon & Griffin, 1998; Tieman & Goodwin, 2001). However, use of heavy metals on food fish is not recommended and efforts to control outbreaks in commercial facilities have relied more on the use of formalin, which is also hazardous. As an alternative to chemical treatments, the development of immunoprophylactic methods of disease prevention would be highly desirable. In this regard, fish acquire long-term immunity against Ichthyophthirius in response to controlled infections and we have identified a class of abundant parasite membrane proteins, known as i-antigens, as having an important role in development of host resistance (Clark, Lin & Dickerson, 1995; Clark & Dickerson, 1997; Wang & Dickerson, 2002). While Ichthyophthirius cannot be readily cultured, i-antigens have been expressed as recombinant proteins in the related, free-living ciliate, Tetrahymena thermophila (Gaertig et al. 1999). Tetrahymena can be grown to high-density in inexpensive media, and high-level expression of parasite antigens has recently been achieved in this system using a cadmium-inducible metallothionein gene promoter (Shang et al. 2002). Furthermore, as few as 106 cells induced with Cd have been shown to elicit protective immunity against ‘white-spot’ when administered to juvenile channel catfish as live vaccines (Wang, Gaertig, Gorovsky, Dickerson and Clark, unpublished). Nevertheless, treatment with Cd would be expected to stimulate metallothionein synthesis resulting in an accumulation of bound metal ions in transformed Tetrahymena (Piccinni et al. 1987; Piccinni, Irato & Guidolin, 1990). Thus, the protection afforded by live cells may have been due to toxic effects of residual Cd introduced with the vaccine itself. To test this, we determined the levels of CdCl2 that are toxic to I. multifiliis in vitro and measured the actual concentrations of Cd in different host tissues following intraperitoneal injection with live recombinant cells. In addition, we tested the effects of non-transformed T. thermophila grown in the presence or absence of CdCl2 on both parasite load and fish survival following lethal challenge with Ichthyophthirius.
MATERIALS AND METHODS
Parasite growth and maintenance
Ichthyophthirius multifiliis (strains G5 and NY1) were maintained on juvenile channel catfish (Ictalurus punctatus) as described elsewhere (Noe & Dickerson, 1995; Lin, Clark & Dickerson, 1996). Tomonts were collected by gently rubbing fish. Parasites dislodged from the skin were harvested on wire-mesh filters and incubated overnight at room temperature in carbon-filtered water. After dividing to form theronts, parasites were collected by low speed centrifugation (2 min at 300× g) and resuspended in water.
Analysis of cadmium effects in vitro
CdCl2 made up at varying concentrations in carbon-filtered water was added in 50 μl aliquots to individual wells of a flat-bottomed 96-well polystyrene plate. Equal volumes of cell suspensions containing Ichthyophthirius theronts were then added. Cells were maintained at room temperature for 18–22 h and, at intervals, examined under a dissecting microscope (Olympus). Cells were photographed using Kodak TMAX-400 film.
Preparation of Tetrahymena cultures and vaccination of fish
T. thermophila cell lines were grown at 30 °C with constant shaking in Neff media containing 0·25% proteose peptone, 0·25% yeast extract, 0·55% glucose, and 0·033 mM FeCl3. Tetrahymena strains used in this study were a transgenic cell line harbouring parasite i-antigen gene IAG52B[G5] (Lin et al. 2002; Clark et al., unpublished), and CU522, the parental line used in the construction of the vaccine strain. Unless otherwise indicated, cell lines were grown to mid-late log phase (5–7×105 cells/ml), and then maintained in the same media either with or without 2 μg/ml CdCl2. After 16–24 h, cells were harvested by centrifugation at 300× g for 3 min, washed twice in 10 mM Tris–HCl (pH 7·5) and resuspended at a concentration of 5×106 cells/ml in PBS buffer containing 130 mM NaCl, 3 mM KCl, 10 mM Na2HPO4, pH 7·4. Aliquots containing 200 μl (1×106 cells) were injected into the peritoneal cavities (i.p.) of juvenile channel catfish weighing 6·9+2·7 g using a tuberculin syringe fitted with a 26 gauge needle. Fish were injected twice at two week intervals and either directly challenged with I. multifiliis or sacrificed for the collection of tissue samples (below). Animals used in challenge studies were maintained in 15 gallon aquaria with aeration for the duration of experiments.
Sampling of tissues and cells by ICP
Fish were euthanazed by overdose with tricanemethanesulfonate (MS-222; 0·5 g/l) and placed on ice. Kidney, skin and muscle tissues were surgically removed from fish euthanazed at different time points, namely, 1, 5 and 14 days after the last injection with T. thermophila. A total of 10 fish were sacrificed at each time point. Tissues from individual fish were combined as follows. For kidney, samples collected from 10 fish were combined for each time point. For muscle, tissues from each of 2 fish were combined yielding 5 separate groups for each time point. For skin, tissues from 3–4 fish were combined yielding 3 separate groups for each time point. Samples were weighed after collection, and individual groups (containing 0·5–3 g tissue) were air-dried. After 14 days, samples were placed in a drying oven at 60 °C for an additional week. Combined tissues were again weighed and sent for analysis using inductively-coupled argon plasma (ICP) emission spectrometry. To determine Cd concentrations in T. thermophila and the surrounding growth media, cells were grown to ~8×105 cells/ml phase in Neff medium containing 2 μg/ml CdCl2. After 16 h, cells were harvested by centrifugation as above and the supernatant fraction collected for ICP analysis. The cell pellet was then re-suspended and washed twice in 40 ml 10 mM Tris–HCl (pH 7·5) buffer. Supernatant, pellet and wash fractions were then dried as for tissue samples. No significant difference was seen in the amount of Cd present in cell pellets before and after washing in Tris buffer (data not shown). For ICP analysis, dried samples (either tissues, cells or media) were digested for 3–5 days in 500 μl 70% perchloric acid, resuspended in 10 ml 0·5% nitric acid and subjected to elemental analysis using a model 51000 ICP spectrometer (Perkin-Elmer/Sciex).
Parasite challenge studies
For whole body infections, channel catfish that had been injected with T. thermophila strain CU522 were placed in large beakers containing 100 ml water fish. Theronts of the G5 strain were then added to a final concentration of 10000 parasites/fish. After 2 h fish were returned to aquaria along with the water containing infective theronts. For tail-fin infections, fish that had been injected with T. thermophila were anaesthetized in 100 mg/l MS-222, wrapped in moist paper and placed with their tailfins in water containing NY1 theronts at a concentration of 2000 cells/ml. After 5 min fish were removed, rinsed briefly in carbon-filtered water, and returned to aquaria. Four days later fish were re-anaesthetized and trophonts in host tissue counted under a dissecting microscope. Fish were immediately returned to aquaria where the infection was allowed to cycle.
RESULTS
To determine the concentrations of Cd that are toxic to I. multifiliis, swimming theronts were placed in individual wells of a 96-well plate that contained varying concentrations of CdCl2 (ranging from 500 ppm to 5 ppb). Parasites were observed under a dissecting microscope and alterations in their behaviour noted over time. As indicated in Table 1, Cd was toxic to theronts in a concentration and time-dependent manner. At Cd levels of 500 ppm, theronts lost motility almost immediately, and fell to the bottom of the well by 10 min of exposure. In general dead and dying cells remained intact under these conditions and were still visible at the bottom of the container the next day (Fig. 1A). At concentrations between 50 ppm and 0·5 ppm, death occurred more slowly and was accompanied by obvious swelling and eventual cell lysis (Fig. 1B). Below 0·5 ppm, Cd was still toxic and at 50 ppb most theronts appeared dead within 24 h.
Fig. 1. Effects of Cd on Ichthyophthirius in vitro. Theronts were incubated in carbon-filtered H2O containing 500 ppm CdCl2 (A), 5 ppm CdCl2 (B) and 5 ppb CdCl2 (C). At the highest Cd concentration, cells immediately ceased swimming and were clearly dead within 10 min of exposure. Hours later, cells were still intact and visible at the bottom of the well (A). When exposed to 5 ppm CdCl2, cells rounded-up and formed large aggregates at the bottom of the well. After several hours, cells lost distinct borders (B) and eventually lysed. Exposure to 5 ppb CdCl2 had little apparent affect on theronts. Cells swam rapidly and could not be captured in focus on film (C). Bar=200 μm.
Table 1. Following incubation in varying concentrations of Cd (ppm) for different lengths of time, theronts were scored for overall changes in cell motility according to an arbitrary scale listed below. Cells that ceased swimming failed to regain movement and were considered dead
(+++=100% of cells ceased swimming. ++=50–90% of cells ceased swimming. +=10–20% of cells ceased swimming. +/−=reduction in swimming velocity. −=no obvious change.)
We had previously shown that T. thermophila expressing parasite antigens, under the control of a Cd-inducible promoter, protect juvenile channel catfish against a lethal challenge with Ichthyophthirius following i.p. administration of the recombinant vaccine (Wang, Gaertig, Gorovsky, Dickerson and Clark, unpublished). To determine whether Cd concentrations in host tissues become elevated following vaccination, juvenile channel catfish were injected twice with 106 transformed T. thermophila and metal concentrations in different host tissues determined at varying times thereafter. As shown in Fig. 2C, Cd levels in the kidney were elevated within 1 day, and accumulated to levels >60 ppm (dry weight) within 5 days of injection. By 14 days, Cd levels were somewhat diminished but were still high in the kidney (>40 ppm dry weight). Cadmium concentrations in the skin and muscle were also elevated, although to a lesser extent. In the skin, levels were 0·8 ppm after 24 h and declined to ~0·4 ppm (dry weight) by 14 days post-vaccination (Fig. 2B). Cadmium concentrations in muscle (Fig. 2A) were roughly one-third of that in skin over the same time period, although skin was considered more relevant since it is the primary site of Ichthyophthirius infection. Based on fresh weight, Cd concentrations in the skin and muscle were determined to be 0·152 and 0·037 ppm respectively, at 1 day and 0·074 and 0·023 ppm at 14 days post-injection.
Fig. 2. [Cd] in fish tissues following vaccination with recombinant T. thermophila. Cd concentrations in skin (panel B), muscle (panel A) and kidney (panel C) were measured by ICP spectrometry. Bars represent the average concentration (in dry weight) for groups of samples taken from fish at different time points after vaccination. Bracketed lines represent the standard error between individual groups (note that kidney samples show no lines since tissue from all fish at a given time point were combined and measured as a single group). Bars at the zero time point represent the values for control fish injected with recombinant T. thermophila that had not been induced with CdCl2. Panel D shows the level of Cd in cell pellets containing 2×106 recombinant T. thermophila (A), and in the surrounding growth medium (B) 16 h after incubation in 2 μg/ml CdCl2. The basal level of Cd associated with the CU522 host strain grown in the absence of heavy metals (C) is also shown.
The fact that Cd could easily be measured in host tissues following vaccination with T. thermophila clearly suggested that the cells themselves had taken up Cd from the growth medium following induction with CdCl2. To determine the extent to which these cells accumulate Cd, we conducted ICP analysis on T. thermophila cell pellets as well as the surrounding medium. As shown in Figure 2D, ~92% of the Cd in the original growth medium was associated with the cell pellets 16 h after induction with 2 μg/ml CdCl2.
In standard vaccine trials with channel catfish, animals are challenged 14–28 days post-vaccination. Since Cd concentrations in skin at day 14 were above those that killed Ichthyophthirius in vitro, we sought to determine the potential effects of Cd on parasites in vivo. Two groups of fish were injected with non-transformed T. thermophila strain CU522 that were grown either in the presence or absence of CdCl2. Fourteen days after injection, animals were subjected to tailfin infections with the virulent NY1 strain of parasites and the number of trophonts that established in the skin by 4 days of infection was determined. The results of this experiment are shown in Fig. 3. While the variation in parasite numbers among individual fish was relatively large, no significant difference was seen in the average number of parasites in the two groups (ANOVA). Rather than terminate the experiment, we allowed the infection to cycle and determined per cent mortality in each group following secondary re-infection. No protection was seen in either group, with all fish dying by 14 days of the initial challenge (Fig. 4A). We repeated this experiment using a slightly less virulent strain of parasite (namely, G5), and fish exposed to whole-body infections. Again, all fish died in both groups (Fig. 4B).
Fig. 3. Effects of vaccine-associated Cd on parasite establishment in vivo. Groups of fish were vaccinated with T. thermophila strain CU522 that had been incubated in the presence (dashed line) or absence (solid line) of 2 μg/ml CdCl2. Two weeks later, fish were exposed to I. multifiliis and parasites on the tailfins were counted 4 days later. Horizontal lines represent the average numbers of trophonts on fish in each group. Symbols (closed circles, or hatched squares) are the actual parasite numbers in individual fish. The average number for the group containing Cd was 122·1+20 (n=8), and for the group lacking Cd was 145·8+72 (n=10). Based on one-way analysis of variance (ANOVA), the difference in the average numbers of parasites in the two groups was not significant.
Fig. 4. Survival rates of fish vaccinated with non-recombinant T. thermophila. In (A), groups of fish that had been vaccinated with T. thermophila grown in the presence (diamonds) or absence (triangles) of CdCl2 were subjected to tailfin infections with I. multifiliis strain NY1. Within 13 days of the initial exposure, all fish in both groups had died. In (B), groups of fish (15 animals/group) vaccinated as in (A) were subjected to whole-body infections with I. multifiliis strain G5, and % survival measured over time. By 21 days, all fish in both groups had died.
DISCUSSION
While the recombinant cells used in this study were washed extensively in Tris buffer following induction with Cd, metal concentrations in fish tissues were initially quite high. Consistent with previous reports on kidney as being the primary site of heavy metal accumulation in fish from contaminated streams (de Conto Cinier et al. 1999; Hollis, Hogstrand & Wood, 2001; Woodling, Brinkman & Horn, 2001), Cd levels in the kidneys of vaccinated fish were >60 times that in skin and muscle. Indeed, while kidney represents less than 1% of the animal's mass, its overall contribution to Cd was roughly one-third the total, or ~410 ng per fish one day post-injection. Cadmium concentrations in the skin and muscle were 152 and 37 ng/g of fresh tissue, respectively. Assuming these values were representative of the average Cd concentrations in all tissues, then the total amount of Cd in vaccinated fish (including kidney) 1 day post-vaccination was estimated to be ~1·4 μg. The maximum amount that could be delivered to fish with the vaccine if all the Cd in the growth medium was associated with Tetrahymena would be ~4·9 μg. Thus, the actual amount of Cd present in fish was a substantial fraction of the theoretical maximum (roughly 37%, assuming a half-life in the animal of 14 days), which in turn suggested that Tetrahymena took up large amounts of Cd from the growth medium. This inference was in fact borne-out by ICP measurements on T. thermophila. Based on these measurements, ~92% of the Cd in the original medium was associated with the cell pellet after 16 h. The ability of Tetrahymena to concentrate Cd from the growth medium is likely due to the up-regulation of endogenous metallothionein synthesis in response to heavy-metal treatment (Piccinni et al. 1987, 1990). Much of this Cd must then become available to fish following intraperitoneal injection with live cells.
Perhaps not surprisingly, Ichthyophthirius theronts were found to be highly sensitive to Cd in water. Heavy metals (including copper sulphate and silver nitrate) have been previously shown to kill theronts (Farley & Heckmann, 1980; Strauss, 1993; Ling et al. 1993; Schlenk et al. 1998; Tieman & Goodwin, 2001), with copper being toxic at concentrations >50 ppb. As shown here, Cd levels in the skin of vaccinated fish were 74 ppb at the time of challenge, a concentration higher than that which killed theronts in water. Despite this, there was no evidence that parasites in host tissues were sensitive to Cd residues delivered with the vaccine. Indeed, in CU522 vaccinated fish challenged by whole-body exposure to Ichthyophthirius, fish died more rapidly in the group injected with Cd-treated T. thermophila (Fig. 4B). To determine parasite sensitivities to potential Cd residues introduced with the vaccine, we had to inject fish with the non-transformed Tetrahymena strain CU522 since an immune response to recombinant cells would otherwise compromise the results. Although we have no reason to believe that Cd levels associated with the vaccine strain are different from those associated with CU522, we are in the process of comparing Cd concentrations in Tetrahymena cell lines that either do or do not contain foreign genes. The absence of an effect of elevated Cd on parasites in vivo is most likely attributable to the fact that metal ions are sequestered in the tissue and not available to interact with parasites. As in the kidney, heavy metals in skin are almost certainly complexed with metallothioneins or other metal-chelating proteins, and not free in the intersitial fluids (Hollis et al. 2001; Olsvik et al. 2001).
Finally, while the presence of Cd residues in fish vaccinated with recombinant Tetrahymena is clearly relevant to food safety, final concentrations in muscle were 23 ppb 14 days after vaccine administration. In production facilities, these concentrations would be expected to further decline due to excretion and continued growth of the fish. Furthermore, 23 ppb is below the Cd limit proposed for fish by European Commission Regulation (III/5125/95 Rev. 3), and is only slightly higher than the normal range for human body Cd according to EBI, a Canadian environmental watchdog group (http://www.e-b-i.net/ebi/contaminants/cadmimum.html). Still, efforts to reduce Cd associated with recombinant T. thermophila vaccines are currently being considered, including the possible use of zinc as an alternative for regulating foreign gene expression.
ACKNOWLEDGEMENTS
We are grateful to Dr Leon Kochian and Mr Michael Rutzke for their generous assistance with the ICP analysis, and to Dr Donna Cassidy-Hanley for her helpful discussion during the course of the work. This work was supported by USDA NRICPG grant # 98-35204-6812 to T. G. C.
