INTRODUCTION
Aquatic organisms are exposed to a wide range of natural and anthropogenic stressors that have the potential to disrupt physiological homeostasis. One approach to diagnose and evaluate challenges to homeostasis, both in the field and the laboratory, is to quantify metabolic by-products indicative of physiological or cellular stress. Among such by-products are chemically-reactive molecules associated with oxidative stress. Internal or external stressors such as UV radiation, temperature extremes, contamination, tissue damage, and pathogens enhance the formation of reactive oxygen species (ROS) within specific tissue (Kelly et al. Reference Kelly, Havrilla, Brady, Abramo and Levin1998; Toyokuni, Reference Toyokuni2002). Elevated ROS cause damage to the integrity of cell lipids, DNA, and proteins, ultimately leading to pathologies associated with tissue lesions, inflammation, cancer, and other dysfunctions (Kelly et al. Reference Kelly, Havrilla, Brady, Abramo and Levin1998; Spector, Reference Spector2000; Velkova-Jordanoska and Kostoski, Reference Velkova-Jordanoska and Kostoski2005). By measuring oxidative stress biomarkers such as lipid peroxidation (LPO, a consequence of ROS-induced cell damage) or anti-oxidants (molecules that combat ROS), the mechanisms mediating the damaging effects of stressors can be determined (Kelly et al. Reference Kelly, Havrilla, Brady, Abramo and Levin1998).
Studies with fish confirm that oxidative stress is a frequent consequence of exposure to various stressors, including anthropogenic toxicants (see reviews in Di Giulio et al. Reference Di Giulio, Washburn, Wenning, Winston and Jewell1989; Kelly et al. Reference Kelly, Havrilla, Brady, Abramo and Levin1998; Oost et al. Reference Oost, Beyer and Vermeulen2003). Miller et al. (Reference Miller, Wang, Palace and Hontela2007) reported increased oxidative stress via reduced glutathione levels in juvenile rainbow trout subjected to acute selenium exposure and ROS-induced damage has been suggested as a contributing cause of selenium toxicity and pathologies such as larval deformities (Spallholz et al. Reference Spallholz, Palace and Reid2004; Muscatello et al. Reference Muscatello, Bennett, Himbeault, Belknap and Janz2006). Recent studies have shown that parasites can also cause oxidative stress in fish tissues. Kurtz et al. (Reference Kurtz, Wegner, Kalbe, Reusch, Schaschl, Hasseiquist and Milinski2006) measured a marker of oxidative protein damage (protein-bound acrolein) in liver of three-spined sticklebacks (Gasterosteus aculeatus) experimentally exposed to larval stages of nematodes and trematodes, and demonstrated a negative correlation between oxidative stress and fish body-condition, and a positive correlation with immune activation. However, most of the evidence for parasite-induced oxidative stress is indirect, arising from field studies with fish collected from sites that differ in exposure to a range of parasites. Farmed carp (Cyprinus carpio) infected with a cestode had elevated antioxidant levels compared with uninfected carp (Dautremepuits et al. Reference Dautremepuits, Betoulle and Vernet2003), and catfish (Rhamdia quelen) infected with a trematode had increased muscle LPO (Belló et al. Reference Belló, Belló-Klein, Belló, Llesuy, Robaldo and Bianchini2000). Both exposure to mercury contamination and nematode infection were associated with increased oxidative stress in yellow perch (Marcogliese et al. Reference Marcogliese, Brambilla, Gagne and Gendron2005). Taken together, these results suggest that a variety of parasites can cause oxidative stress in a range of hosts. However, experimental evidence linking oxidative stress to parasitic infections remains limited.
Ornithodiplostomum sp. is an undescribed species of trematode that encysts within the body cavity of fathead minnows (Pimephales promelas; Matisz and Goater, Reference Matisz and Goater2010) while its congener O. ptychocheilus (Faust), encysts within the optic lobes (Matisz et al. Reference Matisz, Goater and Bray2010). In lakes and ponds in Alberta, these two species typically co-occur in the same individual fish (Sandland et al. Reference Sandland, Goater and Danylchuk2001; Goater, unpublished observations). Both species utilize pond snails (Physa spp.) as first intermediate host, fathead minnows as second intermediate host, and piscivorous birds as definitive hosts. The development of larval Ornithodiplostomum spp. within host tissues is complex (Matisz and Goater, Reference Matisz and Goater2010; Matisz et al. Reference Matisz, Goater and Bray2010). For both species, there is an obligate period of development that occurs within the liver or brain (for Ornithodiplostomum sp. and O. ptychocheilus, respectively), followed by a consolidation and encystment phase that occurs within adjacent tissue. The former phase is associated with rapid rates of growth and differentiation, while the latter is associated with developmental ‘resting’ and encystment.
Developing Ornithodiplostomum metacercariae cause a range of detrimental effects in individual minnows, including reduced growth (James et al. Reference James, Noyes, Stumbo, Wisenden and Goater2008), reduced optomotor performance and activity (Shirakashi and Goater, Reference Shirakashi and Goater2005), and reduced survival (Sandland et al. Reference Sandland, Goater and Danylchuk2001). Trematode-induced oxidative stress and associated pathology may provide a unifying mechanism to explain these diverse negative consequences of infection. The purpose of this experiment was to test the hypothesis that larval trematodes cause oxidative stress in minnows, by measuring LPO in liver tissue of fathead minnows exposed to Ornithodiplostomum sp. or O. ptychocheilus cercariae. First, we evaluated whether temporary development of Ornithodiplostomum sp. within liver tissue causes long-term oxidative stress, and whether these effects are dose dependent. To assess if parasite-induced oxidative stress requires direct tissue contact, or is part of a generalized response to infection, we also evaluated oxidative stress in the liver tissue of minnows exposed to O. ptychocheilus.
MATERIALS AND METHODS
Experimental infections
Naïve minnows were exposed to known numbers of Ornithodiplostomum spp. cercariae following methods described by Sandland and Goater (Reference Sandland and Goater2000). To initiate experimental infections, young of the year (uninfected) minnows were collected from Goldspring Pond, Alberta (49° 5′ 41·3514″, −111° 59′ 28·6146″), a population known to be infected with metacercariae of O. ptychocheilus and Ornithodiplostomum sp. In July 2009, 8 one-day-old chickens were force-fed minnow brain or viscera containing large numbers of metacercariae. Three days later, parasite eggs were present in the feces of the chickens. Eggs were collected and processed following the method of Sandland and Goater (Reference Sandland and Goater2000). The F1 generation of laboratory reared Physa sp. snails were exposed to the miracidia, and reared under standard conditions. For exposure to cercariae, the dilution methods described by Sandland and Goater (Reference Sandland and Goater2000) were used to estimate the volume of water containing known numbers of 2-h-old cercariae. Exposure of individual minnows occurred within 60 mm Petri dishes for a 2-h period. Previous results in our laboratory indicate that approximately 80% of cercariae used in exposures are recovered subsequently as metacercariae.
The experiment was set up as an ‘infection’ × ‘time-period’ factorial involving 108 naïve minnows separated at random into 9 groups. Treatments consisted of 36 fish exposed to dechlorinated water (control), 20 Ornithodiplostomum sp. cercariae (low-dose), or 100 Ornithodiplostomum sp. cercariae (high-dose) on 12 September 2009. Each group of 36 fish was separated at random into 3 groups of 12, corresponding to dates of dissection at 5, 10, or 28 days post-infection (p.i.). At each of these intervals, 2 minnows were prepared for histological analysis to assess metacercariae development and intensity. Livers from the remaining fish were removed under a dissecting microscope and placed on ice in homogenization microcentrifuge tubes. Samples were stored at −80 °C until analysis. It was not possible to mechanically isolate host tissue from parasite tissue due to the small size of developing metacercariae. Thus, our estimates of lipid peroxidation in host tissue assume that the contribution from parasite tissue is negligible.
We incorporated an additional group of 36 minnows exposed to 100 O. ptychocheilus cercariae into the overall design. A limited supply of fish and cercariae restricted our ability to complete a full ‘time × dose × species’ design. The 36 fish were necropsied in the same manner as described above at corresponding post-infection time-intervals (i.e. 5, 10 and 28 days).
Lipid peroxidation assay
Lipid peroxidation in liver tissue was evaluated with the BIOXYTECH® LPO-586 Assay (OXIS International, Inc., Portland, USA; catalogue no. 21012), following the methods described by Miller et al. (Reference Miller, Wang, Palace and Hontela2007). In this assay, malondialdehyde (MDA), the end product of the LPO process, is used as an indicator for the concentration of LPO within tissue (Esterbauer et al. Reference Esterbauer, Schaur and Zollner1991). LPO concentrations were quantified by the reaction of MDA with n-methyl-2-phenylindole at 45 °C and 586 nm, and LPO is expressed as μmol MDA/mg of protein.
Histopathology
Standard histological sectioning of Ornithodiplostomum sp.-infected minnows was used to provide a qualitative assessment of ontogenic changes in metacercariae growth and site selection, and to verify infection levels at 5, 10 and 28 days p.i. Histological sections of 2 randomly selected minnows were prepared from each of the 6 treatments. To allow for complete fixative penetration, minnow heads were removed immediately rostral to the operculum, and tails were removed just caudal to the visceral cavity (Matisz and Goater, Reference Matisz and Goater2010). Bodies were fixed in 10% neutral-buffered formalin for 7 days, and decalcified in 0·1M EDTA titrant for 14 days. These samples were dehydrated in ethanol prior to paraffin embedding. Each fish was serially sectioned along its sagittal plane (10 μm thickness). Sections were deparaffinized and stained with Mayer's haematoxylin and eosin Y. All sections were examined by light microscopy using a Zeiss axiocam digital camera mounted onto a Zeiss axioskop 40 microscope.
Statistical analysis
The data were log transformed prior to analyses to meet the assumptions of normality. A two-way ANCOVA was performed using Predictive Analytics SoftWare v.18 with ‘time-period’ of liver dissections (5 days, 10 days, or 28 days p.i.) and treatment (control, Ornithodiplostomum sp.-low, or Ornithodiplostomum sp.-high) evaluated as fixed effects, and LPO (μmol MDA/mg of protein) as the dependent variable. Minnow weight at necropsy was evaluated as a covariate. Least significant difference (LSD) pairwise comparisons were used to test differences between pairs of means within significant fixed effects. A second two-way ANCOVA was used to evaluate differences in LPO concentration between O. ptychocheilus-exposed and uninfected fish.
RESULTS
Results from the first two-way ANCOVA showed that the concentration of LPO in liver tissue was not affected by the interaction between time and infection with Ornithodiplostomum sp. cercariae (F 4,69 = 1·37, P = 0·255). LPO concentrations were also not affected by time-period (F 2,69 = 1·35, P = 0·246) and the covariate was not significant (F 1,69 = 0·097, P = 0·757). However, LPO concentration in liver tissue was significantly affected by infection with Ornithodiplostomum sp. (F 2,69 = 5·66, P = 0·005; Fig. 1). The LSD pairwise comparisons indicated that the 39% difference in LPO concentration between controls and lightly infected hosts was significant (P = 0·006), as was the 35% difference between controls and heavily infected fish (P = 0·013). The difference in LPO concentration between lightly and heavily infected fish was not significant (P = 0·779). This suggests that oxidative stress caused by Ornithodiplostomum sp. infection is independent of dose and time (Fig. 1).
Lipid peroxidation (malondialdehyde, MDA, mean±s.e.) at 5, 10 or 28 days post-infection in liver tissue of fathead minnows exposed to cercariae-free water (Control), Ornithodiplostomum ptychocheilus cercariae, 20 Ornithodiplostomum sp. cercariae (Low-dose), or 100 Ornithodiplostomum sp. cercariae (High-dose).
The two-way ANCOVA of results from the O. ptychocheilus treatment showed that the concentration of LPO in liver tissue was not significant relative to time-period (F 2,44 = 1·109, P = 0·339), treatment (F 1,44 = 0·253, P = 0·617), or covariate (F 1,44 = 0·121, P = 0·730). The interaction between time-period and treatment was significant (F 2,44 = 3·900, P = 0·028, Fig. 1) due to the peak in LPO concentration at 5 days, followed by a decline to levels similar to controls at later time-periods.
Metacercariae counts in low and high dose treatments were significantly different (P < 0·001; 15 ± 1·6 and 68·6 ± 6·4, respectively). Large numbers of small, unencysted Ornithodiplostomum sp. metacercariae were observed within the parenchyma of the liver at 5 days p.i. (Fig. 2A). At this time, a distinctive gap was evident between the metacercariae body and adjacent liver tissue. By 10 days p.i., unencysted metacercariae were still present within liver tissue. Qualitatively, these metacercariae were larger than at 5 days p.i. and the gap between the parasite and the adjacent liver tissue was still present (Fig. 2B). By 28 days pi, the metacercariae had increased further in size (Fig. 2C). Metacercariae were now absent from the parenchyma of the liver (although some were observed along its outer edge), located instead throughout the body cavity. All 28-day old metacercariae were enveloped by a distinctive cyst wall. These qualitative observations correspond with our previous assessment of Ornithodiplostomum sp. metacercariae growth and migration in minnows (Matisz and Goater, Reference Matisz and Goater2010).
Histological cross-sections of fathead minnows infected with Ornithodiplostomum sp. at 5 days (A), 10 days (B), and 28 days (C) post-infection. Liver tissue (l), Ornithodiplostomum sp. metacercariae (m), and metacercariae migration tracks (t) are indicated by arrows.
DISCUSSION
The results from this experimental study provide the first direct evidence for metacercariae-induced alteration in tissue lipid peroxidation in fish. Since LPO is a well-characterized indicator of oxidative stress, the results suggest that Ornithodiplostomum spp. infection causes elevated concentrations of ROS in fathead minnow liver tissue. These results support findings from several correlative studies involving parasite-induced oxidative stress in aquatic organisms (e.g. Belló et al. Reference Belló, Belló-Klein, Belló, Llesuy, Robaldo and Bianchini2000; Dautremepuits et al. Reference Dautremepuits, Betoulle and Vernet2003). The results also support the use of the LPO assay for investigating interactions between parasites and environmental stressors under natural conditions.
This is the first study to evaluate oxidative stress throughout the various stages of larval parasite development. For minnows exposed to Ornithodiplostomum sp. cercariae, LPO concentrations diverged from controls at 5 days post-exposure, corresponding to the period when all metacercariae were observed within the parenchyma of the liver. At this time, the metacercariae are unencysted and an extensive network of microvilli envelope the entire worm, presumably to support feeding (Matisz and Goater, Reference Matisz and Goater2010). Thus, maximum differences in LPO concentrations between infected and uninfected minnows coincided with the period of maximum metacercariae development in the liver (Matisz and Goater, Reference Matisz and Goater2010). The difference persisted at least to 28 days post-exposure, by which time all metacercariae had reached the encysted stage outside the liver. An implication of these results is that elevated liver LPO likely occurs in fathead minnows that are exposed to Ornithodiplostomum spp. cercariae in natural lakes in late summer/early fall each year (Sandland et al. Reference Sandland, Goater and Danylchuk2001).
The elevation in LPO detected in the present study may involve direct damage to liver tissue caused by developing Ornithodiplostomum sp. We did not assess tissue damage in this study, but Matisz and Goater (Reference Matisz and Goater2010) and Matisz et al. (Reference Matisz, Goater and Bray2010) showed extensive damage during the developmental phase of Ornithodiplostomum spp., followed by rapid tissue repair or regeneration. There is no direct evidence that these metacercariae feed on liver tissue, although the transient structure of the tegument during the development phase strongly suggests a feeding function (Conn et al. Reference Conn, Goater and Bray2008; Matisz and Goater, Reference Matisz and Goater2010). Further, our observations that the gap between the tissue and developing metacercariae expands during the development phase suggests that tissue adjacent to developing worms is damaged, although perhaps temporarily. When cellular damage occurs, platelets involved in tissue repair release ROS to recruit additional platelets, a process known as redox signalling (Palmer and Paulson, Reference Palmer and Paulson1997). Thus, both species of Ornithodiplostomum may induce tissue damage and oxidative stress resulting in lipid peroxidation.
An alternative explanation for the observed increase in oxidative stress in infected fish is host immunity. Cercariae and metacercariae of strigeiod trematodes elicit a variety of complex immune responses in fish (Stables and Chappell, Reference Stables and Chappell1986; Whyte et al. Reference Whyte, Chappell and Secombes1990), although the nature of the fathead immune response to Ornithodiplostomum spp. is not known. Further, sticklebacks exposed to parasites showed increased immune activation via respiratory burst (Scharsack et al. Reference Scharsack, Kalbe, Harrod and Rauch2007). Respiratory bursts occur when leukocytes come in contact with foreign organisms such as bacteria, fungi, and parasites, resulting in the release of ROS (Muñoz et al. Reference Muñoz, Calduch-Giner, Sitja-Bobadilla, Alvarez-Pellitero and Perez- Sanchez1998; Wang et al. Reference Wang, Malo and Hekimi2010) reducing the viability and affecting the development of certain parasites (Wilson et al. Reference Wilson, Andersen and Britigan1994; Allen and Fetterer, Reference Allen and Fetterer2002). A positive correlation between the levels of immune activation and levels of oxidative stress has been demonstrated in sticklebacks exposed to various endoparasites (Kurtz et al. Reference Kurtz, Wegner, Kalbe, Reusch, Schaschl, Hasseiquist and Milinski2006). Further experimental studies are required to confirm the mechanistic link between parasite infection, the immune response, and oxidative stress.
Exposure to the brain-encysting trematode, O. ptychocheilus, also caused an increase in LPO in minnows. This is an important finding because it indicates that oxidative stress can be induced in non-target host tissue. Marcogliese et al. (Reference Marcogliese, Brambilla, Gagne and Gendron2005) described a similar off-target effect in the livers of perch infected with a muscle-encysting metacercariae. These authors suggested a general inflammation response to encysting worms as a possible mechanism. If this is the case for O. ptychocheilus, then development of pre-encysted worms within the optic lobes may lead to a general inflammation response that is observed within liver tissue. An important follow-up question is to evaluate the singular versus combined effects of both species of Ornithodiplostomum spp. on oxidative stress in minnows. More generally, we need to understand whether parasite-induced oxidative stress increases as the number of different types of parasites in host tissue increases.
In addition to providing support for the LPO assay as an indicator of stress due to parasitism, our experimental results also support field-based approaches that seek to uncover the importance of cumulative or combined stressors on natural fish populations. Marcogliese et al. (Reference Marcogliese, Brambilla, Gagne and Gendron2005) showed that a combination of parasites and pollutants such as mercury lead to higher levels of ROS than either stressor alone. Jacobson et al. (Reference Jacobson, Arkoosh, Kagley, Clemons, Collier and Casillas2003) showed synergistic effects leading to lowered immune function in chinook salmon when exposed to polychlorinated biphenyls (PCBs) and trematodes. Stress due to parasites is often neglected in studies of cumulative effects, but as our results indicate, specialist parasites that are a regular feature of most freshwater fish populations can reduce physiological performance (for review see Marcogliese and Pietrock, Reference Marcogliese and Pietrock2011), even at low rates of exposure. Thus, oxidative stress due to exposure to multiple environmental stressors, including parasites and pollutants, may be an underlying mechanism of decreased performance in natural fish populations.
ACKNOWLEDGEMENTS
We thank Heather Bird, Chelsea Matisz, and Lana Miller for laboratory assistance. This study was performed in accordance with regulations of the University of Lethbridge Animal Care Committee (permit # 1015).
FINANCIAL SUPPORT
Funding support from the Natural Sciences and Engineering Research Council of Canada to C.P.G. (grant number 40130) and A.H. (grant number 40220) is gratefully acknowledged.
