Environ. Sci. Technol. 2007, 41, 3382-3387
Characterization of a Radiation-Induced Stress Response Communicated in Vivo between Zebrafish C A R M E L M O T H E R S I L L , * ,†,‡ RICHARD W. SMITH,† NALINI AGNIHOTRI,‡ AND C O L I N B . S E Y M O U R †,‡ Medical Physics and Applied Radiation Sciences Department, McMaster University, 1280 Main Street West, Hamilton, Ontario, L8S 4K1, Canada, and Juravinski Cancer Centre, 699 Concession Street, Hamilton, Ontario, Canada L8V 5C2
Radiation-induced communication of stress signals between rainbow trout (Oncorhynchus mykiss W) have recently been described by this group and linked to the bystander effect. This paper addresses the question of whether another totally unrelated fish species (Danio rerio L) can demonstrate the effect and also looks at attenuation of both the bystander signal, from irradiated fish, and the bystander effect, in naive fish. The data show that zebrafish produce bystander signals, and that, as with rainbow trout these can affect naı¨ve (i.e., non-irradiated) fish placed in water with X-rayed fish or in water previously occupied by X-rayed fish. Skin explants from directly X-rayed fish still reduce HPV-G reporter cell growth 6 h after X-ray, but the bystander signal to naı¨ve fish is lost. Twelve h after X-ray the signal is lost in X-rayed fish. The bystander effect is also attenuated if induction was by placing naı¨ve fish in water which previously held the X-rayed fish. However, the effect is retained if induction was by placing X-rayed and naı¨ve fish together. This suggests the signal is not retained by water for long periods of time. Individual fish data reveal unique responses by bystander fish which could indicate varying levels of sensitivity to signal strength among individuals.
Introduction Nontargeted effects of radiation are of considerable importance in radiation and environmental protection. These predominantly low dose effects are defined as responses in unirradiated cells which are in receipt of signals from irradiated cells. Such effects include bystander effects, (1-4), delayed reproductive death and genomic instability (5-8), apoptosis (9), mutation (10), and transformation (11). These effects are generally adverse and are mechanistically linked to increased oxidative stress, therefore the idea has become established that they make low doses of ionizing radiation more harmful than previously thought (12, 13). Key questions include the in vivo relevance and in vivo persistence of these effects, which have been mainly studied in vitro. Our group has recently published data showing that * Corresponding author phone: 905 525 9140 ex 26227; fax: 905 522 5983; e-mail:
[email protected]. † McMaster University. ‡ Juravinski Cancer Centre. 3382
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007
when irradiated rainbow trout were allowed to swim with unirradiated trout for 2 h, the unirradiated fish produced bystander signals which could be picked up in a reporter assay (14). Five tissues were studied and although the responses were broadly similar there were quantitative differences between external tissues (fin, skin, and gill), which exhibited a stronger bystander effect than internal tissues (head kidney and spleen). In the current study, we decided to concentrate on the more responsive gill and skin. The data for rainbow trout provide the first demonstration of a radiation-induced bystander signal, transmissible through water to another live animal. Previous reports of allelopathic responses exist in the plant world where the phenomenon is well known and is due to a variety of compounds such as alkaloids, quinones, or auxins with multiple targets in the cell including DNA, mitochondria, and membranes (1518). It has been demonstrated, following ionizing radiation stress, disease, grazing, or drought (19-21). In the animal world, allelopathic effects are confounded by behavioral and instinctive responses so that proof of a chemical stress or warning signal, causing a particular response in other organisms, is difficult to prove. There are reports that in times of high density or low food supply, Daphnia species can suppress reproduction of individuals at a population level using allelopathic signals (22). In higher animals, Surinov (23) showed that irradiated mice housed with unirradiated mice could communicate signals which were clastogenic to the mice which had not been irradiated. It is possible that all these phenomena are part of a mechanism of hierarchical organization at the population level and that perturbing stresses such as radiation are communicated in order to coordinate a species-level survival response using pheromones, kairomones, or synomones, etc (24-26). This may be adverse for some individuals. As part of our investigation of these phenomena, it is important to establish whether and to what extent bystander effects can be induced in another nonrelated freshwater fish. Zebrafish are a tropical species whereas rainbow trout are cold water salmonids. However, like rainbow trout, zebrafish are rapidly becoming a well-studied and characterized laboratory model. Thus they are an ideal candidate to expand our understanding of the bystander effect in fish. Being considerably smaller (up to 2 cm long), also makes it possible to treat groups at one time (the rainbow trout used by Mothersill et al., (14), were individually irradiated or exposed to a bystander effect) so individual variation can be studied and the bystander induction methodology can be developed further. If the effect is found in two species then kairomonic processes can be addressed. In this paper we specifically address whether the bystander effect described for trout also occurs in zebrafish, we report on the longevity of the signal in water, and we ask whether individual variation in the bystander effect be detected.
Materials and Methods Fish and Husbandry. Zebrafish were supplied from a local tropical fish retailer in Hamilton Ontario. They were 8-12 months old, of mixed sex, weighed 530-620 mg, and were healthy. They were acclimated for 2 weeks in glass aquaria, in 40 L of dechlorinated Hamilton city tap water (for ionic composition, pH, and hardness, refer to 14), at a stocking density of 10 fish L-1. The water was maintained at 28 °C by submersible heaters and filtered through appropriately sized multi-stage external power filters (which included mechanical, chemical, and biological filter components). The water was returned to the aquaria by a “waterfall” return which 10.1021/es062978n CCC: $37.00
2007 American Chemical Society Published on Web 03/31/2007
FIGURE 1. Bystander effect induction, as indicated by clonogenic survival of HPV-G reporter cells treated with media from gill (open bars) and skin (solid bars) cell primary cultures from zebrafish exposed to X-radiation or to the bystander effect. Letters indicate similarities and statistical differences among the treatments shown. also ensured adequate aeration. Feeding was twice daily with a commercially available tropical fish flake diet. Fish were maintained and treated according to ethical guidelines established at McMaster University and covered by AUP 06-12-65. X-radiation and Bystander Signal Exposure. The Xradiation and bystander signal exposure regimen was based on that employed by Mothersill et al. (14). Groups of 5 zebrafish, in 200 mL of water, were irradiated at 100 kVp (kilovolts peak) for 5 min using a portable X-ray (Faxitron X-ray Corporation cabinet system X-ray system, Wheeling, IL). This delivers approximately 0.5 Gy (14). Once X-rayed the zebrafish were introduced as a group into experimental containers containing 1 L of clean water. This gives a comparable experimental biomass/water volume ratio as individual 85-125 g rainbow trout in 25 L of water (14). After 2 h the X-rayed fish were transferred to one side of a meshpartitioned (but otherwise identical) container and 5 naı¨ve (non-irradiated) partner bystander fish were place on the other side of the partition. In addition, a group of 5 naı¨ve, waterborne bystander fish were placed in the water which previously held the X-rayed fish. A series of sham controls were included. Sham X-rayed fish were placed in the portable X-ray for 5 min but the machine was not switched on. These were used as above to check the effect of sham bystanders. An additional control of completely untreated zebrafish, taken directly from the holding aquaria, was included. All experimental containers were adequately aerated and maintained at 28 °C. Bystander effect induction in the zebrafish gill and the skin was studied in 2 and 3 independent trials, respectively. Bystander Signal and Bystander Effect Attenuation. To investigate bystander signal attenuation the X-rayed fish were held for 6 h in 1 L of water immediately after X-ray. After this 6 h delay they were then transferred to 1 L of clean water for 2 h and then to a partitioned container, with a group of 5 naı¨ve, partner bystander fish for an additional 2 h (i.e., as described above). Waterborne bystander fish were placed in the container which held the X-rayed fish for the 2 h interval. To investigate bystander effect attenuation, after the bystander induction exposure protocol described above, all fish were moved to separate containers (i.e., X-rayed fish were separated from naı¨ve partner bystander fish) with clean water for 12 h before skin sample collection (see below).
FIGURE 2. Relationship between HPV-G reporter cell clonogenic survival following treatment with media from skin explants from rainbow trout, and the bcl-2 expression in the explants generating the bystander media. X-ray ) b--b. Waterborne bystander ) 0- -0. Partner bystander ) 9- - - - -9. All relationships may be described by the equation; Y ) a + bX, where Y ) % bcl-2 positive cells and X ) % clonogenic survival. X-ray: a ) 7.8(11.3), b ) 1.5(0.5), r 2 ) 0.394 (p ) 0.012). Waterborne bystander: a ) 19.3(11.3), b ) 1.7(0.2), r 2 ) 0.836 (p ) 0.000). Partner bystander: a ) 13.5(6.1), b ) 1.7(0.3), r 2 ) 0.694 (p ) 0.000). There were no direct relationships between % clonogenic survival and % bcl2 expression for sham X-ray, sham waterborne bystander, and sham partner bystander treatments; r 2 ) 0.201 (p ) 0.084), r 2 ) 0.262 (p ) 0.051), and r 2 ) 0.017 (p ) 0.648), respectively. In both the attenuation studies a full range of identically treated sham controls and completely untreated fish were included. Sampling Protocol. The fish were killed by severing the head from the body and the skin and gill were removed using sterile technique in a biological safety cabinet over ice. They were placed in 1 mL of complete growth medium (14) and transported on ice to the tissue culture laboratory (5 min walk away). All tissue was obtained and treated according to guidelines at McMaster University and the procedures were covered by AUP 06-12-65. Subsequent processing of the tissue, the reporter assay, and measurement of endpoints are described in detail in Mothersill et al. (14) and are provided as Supporting Information (SI). Statistical Analysis. Data are presented as mean ( standard error of the mean. Comparison of the effects of direct X-irradiation and the bystander effect on HPV-G colony-forming ability was done by Analysis of Variance followed by Least-Square Difference analysis, using Statistix analytical software and t-tests performed using InStat software were also used where appropriate. In all statistical analysis P values
