Hydrogen peroxide is not the cause of fish kills associated with Chattonella marina: Cytological and physiological evidence

Abstract

Chattonella marina, a harmful algal bloom (HAB) causative species, was used to study the mortality, physiology, and pathology of a marine stenohaline fish, goldlined seabream exposed to the toxic alga. The median lethal time (LT₅₀) was 3 h upon exposure to 8000 cells/ml of C. marina. Significant induction of filamental chloride cells (CCs) [i.e. increases in CC fractional area and in the volume density of CCs], concomitant with significant reduction of blood osmolality, were found in C. marina treated fish. To verify whether the toxicity of C. marina was mediated through oxidative stress, a hydrogen peroxide exposure experiment was carried out and the toxicity as well as cytological and physiological changes were compared with the C. marina treatment. Hydrogen peroxide at a concentration of 500 μM H₂O₂, (i.e. 25 times higher than that produced by 8000 cells/ml of C. marina (20 μM H₂O₂)) was unable to induce similar CC alterations and osmoregulatory impairment in fish as observed in the C. marina treatment. Non-specific membrane damage such as severe loss of microvilli projections on the CC apical opening and rupture of epithelial membranes in the lamellae were observed. The LT₅₀ was 6 h, two times longer than that with 8000 cells/ml of C. marina. Based on the cytological and physiological evidence and toxicity data, the mechanism by which C. marina kills fish appears to be very different from that caused by H₂O₂/ROS. Osmoregulatory distress is the major cause of fish death upon exposure to C. marina.

Introduction

Global increases in the frequency and severity of harmful algal blooms (HABs) have posed a significant threat to the world's coastal environment (Anderson, 1989, Hallegraeff, 1993, Landsberg, 2002). The raphidophycean flagellates, particularly Chattonella marina, have caused massive fish kills worldwide, including in Hong Kong, Japan, Canada and Australia (Hallegraeff et al., 1998, Tiffany et al., 2001, Landsberg, 2002).

Over the past decade, reactive oxygen species (ROS) have been considered a likely mechanism for fish kills by raphidophytes. Laboratory studies repeatedly demonstrated that C. marina generates the highest levels of ROS, such as superoxide anion (O₂⁻) and hydrogen peroxide (H₂O₂), compared to other raphidophytes (Oda et al., 1992a, Oda et al., 1994, Oda et al., 1995, Oda et al., 1997, Oda et al., 1998, Shimada et al., 1991, Shimada et al., 1993, Tanaka et al., 1994). The ROS produced by C. marina inhibited the growth of the bacterium Vibrio alginolyticus in culture (Oda et al., 1992b, Oda et al., 1997). Ishimatsu, 1996, Ishimatsu, 1996 reported a correlation between the ability of C. marina cells to produce O₂⁻ (determined by cytochrome c reduction) and their toxicity to yellowtails (Seriola quinqueradiata). However, no measurements were made on the levels of ROS produced during exposure and the authors did not provide experimental evidence to demonstrate the casual relationship between ROS production by C. marina and fish mortality. Thus, it is not known if the observed fish mortalities were actually caused by ROS.

ROS-mediated gill tissue damage has often been claimed to be the major cause of fish kills by C. marina. For instance, studies by Ishimatsu et al. (1997) and Kim et al. (2001) postulated that fish gill mucus would induce O₂⁻ generation from the glycocalyx of flagellates. Sustained O₂⁻ generation due to mucus and glycocalyx interactions may then cause severe damage to gill tissue. Marshall et al. (2003) proposed that the high levels of ROS produced by C. marina (in the presence of free fatty acid) would trigger lipid peroxidative damage of gill membranes, resulting in reduced respiratory and osmoregulatory capacity. However, none of the above studies nor any studies to date have provided direct cytological evidence to demonstrate that the levels of ROS produced by C. marina are sufficient to trigger significant gill damage, and subsequent osmoregulatory and/or respiratory impairments and eventual fish mortality.

The toxicity of ROS to biological systems has been well documented (Fridovich, 1978; Cunningham and Capone, 1992). Given that ROS-induced lipid peroxidation may lead to destruction of membrane integrity (Cunningham and Capone, 1992), if the amount of ROS produced by C. marina is sufficiently high, oxidative damage of gill epithelia would probably occur. Powell and Perry (1997) reported hypertrophy of epithelial cells and thickening of gill lamella when rainbow trout were exposed to high doses of H₂O₂ (100–500 mg/l) for 6 h.

Our recent quantitative cytological studies (Tang and Au, 2003, Tang and Au, 2004) have clearly demonstrated significant induction of chloride cells (number and size) in the gills of moribund goldlined seabream (Rhabdosargus sarba) exposed to C. marina, and the resulting cytological changes were similar to gill chloride cells undergoing active ions excretion. A concomitant 70% reduction of blood osmolality was also detected in the moribund fish (Tang and Au, 2004).

To decipher whether ROS is the cause of toxicity by C. marina, we investigated gill cytopathology and osmolality of fish exposed to a bloom concentration of C. marina (8000 cell/ml), and compared to fish exposed to added ROS at a concentration similar to that released by C. marina. Information obtained from this study will help to elucidate the role of ROS in the toxicity of C. marina, which is important for risk assessment and management of this HAB species.