Humans are continuously exposed to ionizing radiation (IR) through many sources. Two of the main exposures are related to background environmental sources and medical procedures. In the United States the average annual radiation dose received by the general public increased from 3.6 to 6.2 mSv since the 1980s mainly due to the significant increase of exposure to medical procedures (NCRP, 2009). The most common and well-studied health effect of IR is cancer, but recently the adverse health effects of IR exposure in other diseases are beginning to be realized. The strongest links to date include cardiovascular and nervous system diseases (Kim et al., 2008; Sanchez et al., 2009; McGale et al., 2011; Boerma, 2012). Data obtained from atomic bomb survivors, radiotherapy patients, and occupationally exposed radiation workers found an excess relative risk of circulatory disease ranging from 3 to 19% per Gy among different study populations (Little et al., 2010). Baker et al. (2011) examined the risk for developing cardiovascular disease from a wide range of radiation sources and suggested an association between cardiovascular disease and IR exposure not only at high doses, but also at low-to-moderate doses. In a recent review on the pathology, radiobiology, cardiology, radiation oncology, and epidemiology regarding radiation-related heart disease, the authors concluded that there is strong evidence of increased radiation-related heart disease with a longer latency period for lower dose levels (Darby et al., 2010). Studies also report a subtle effect on cognitive functioning in adolescents exposed to low-dose radiation in utero (Bar Joseph et al., 2004; Heiervang et al., 2010), although this effect is not shown in subjects who are exposed during adulthood (Yamada et al., 2002; Brummelman et al., 2011). Despite this strong evidence, there are major gaps in the current knowledge regarding a threshold dose of IR that induces cardiovascular disease and cognitive disorders. Current consensus in the literature indicates adverse cardiovascular effects are associated with IR doses greater than 2 Gy total body irradiation (Preston et al., 2003), while cardiovascular injury is observed at doses as low as 0.5 Gy (Shimizu et al., 2010). Less is known about the threshold dose related to neurological impacts and is being continually investigated (Mizumatsu et al., 2003; Achanta et al., 2007; Atwood et al., 2007).

The molecular and genetic mechanisms underlying the non-carcinogenic adverse health effects of IR are not well understood. Several mechanisms are suggested regarding radiation-related heart disease including macrovascular injury that accelerates age-related atherosclerosis leading to cardiovascular disease and microvascular injury that reduces capillary density leading to cardiovascular disease (Little et al., 2010). The pathways to these effects involve many biological responses including inflammation, antioxidative defense, endothelium signaling, and calcium regulation. In regards to the effects of IR on cognitive function, the proposed mechanisms include alteration of neuronal responses (Sanchez et al., 2009), modified antioxidant defenses (Di Toro et al., 2007), and other changes in specific molecular pathways (Verheyde and Benotmane, 2007). Overall a change in the normal and physiological levels of an organism to environmental stressors is ultimately a result of changes at the molecular and cellular levels. Moreover, the understanding of the immediate IR induced biological responses at the genetic level that trigger processes leading to short term and long term adverse health impacts such as those to the cardiovascular and neurological systems are not known. These genetic changes can be detected using transcriptomic technologies that permit thousands of genes to be analyzed in a single step to portray a genome-wide view of gene expression alterations. Few studies to date have used these technologies to identify transcriptomic signatures and genetic biomarkers related to IR exposure (Yin et al., 2003; Amundson et al., 2004; Dressman et al., 2007; Meadows et al., 2008, 2010; Lowe et al., 2009; Morandi et al., 2009; Kabacik et al., 2011; Wyrobek et al., 2011; Jaafar et al., 2013). These studies are starting to elucidate genetic targets associated with IR induced alterations, but were conducted in either cell culture systems, mature models, or from adult human peripheral blood samples. Thus, currently there is limited information on the genetic mechanisms of immediate and latent alterations associated with IR exposure during embryogenesis. Moreover, these studies are generally broad focused with minimal information regarding the association between IR exposure and cardiovascular or neurological alterations.

In this study, we first irradiated zebrafish embryos at different doses to establish IR toxicity and assess morphological alterations to establish baseline toxicity in our laboratory conditions following IR exposure at the specific developmental time point. The zebrafish is a complementary vertebrate model to assess the impacts of developmental exposures to environmental stressors. Zebrafish gained popularity in developmental and toxicological studies because of their rapid ex utero embryonic development that simplifies dosing (Hill et al., 2005). In addition, their near-transparent chorion permits visualization of morphological structures and internal organs including the brain, eyes, liver, kidney, and heart with a light microscope throughout the entire embryonic period. Moreover, a finished genome sequence and conserved genetic function between the zebrafish and human genomes permit translation of molecular mechanisms of toxicity observed in the zebrafish model system to humans (Barbazuk et al., 2000; Howe et al., 2013). Several zebrafish orthologs known to play key roles in human diseases are identified and mutations in these genes display phenotypes similar to those present in human diseases including cardiovascular and neurological alterations (Stainier et al., 1996; Dooley and Zon, 2000; Stainier, 2001; Incardona et al., 2004; Heideman et al., 2005; Peterson et al., 2011).

Furthermore, zebrafish may serve as a complementary and alternative model in light of current international discussions to reduce, refine, and replace animal bioassays (e.g., those discussed in Marone et al., 2014). Animal models are currently widely used in biological experiments. A dichotomy exists among these animal models. Many higher phylogenetic animals are used including mouse, rat, rabbit, canine, and non-human primates, mainly due to the fact that they display strong similarities to humans both in terms of their genetics and systems morphology. Animal models represent a significant burden not only financially, but also ethically in terms of the large number of higher order animals manipulated and sacrificed. On the other end of the spectrum, cell culture models serve as alternatives to replace animal model studies and have provided useful information, but are limited by the fact that only specific genetic pathways are applicable to human health. The zebrafish has the potential to satisfy both the need of an in vivo vertebrate model and to reduce the burden of mammalian animal bioassays. Thus, in this study, after establishing a dose at which no gross morphological alterations were observed, global gene expression analysis was performed using the zebrafish model system to identify altered genetic targets and pathways following the developmental IR exposure. The study of gene expression is an important aspect of understanding the genetic mechanisms associated with IR injury as differential gene expression plays a major role during critical periods of rapid cell growth in early development. Lastly, the functional implication of the immediate and later in development impacts of an embryonic IR exposure on heart rate was assessed to link identified genetic targets and functional consequences.

IR is long established as a carcinogen with most studies focusing on DNA damage and mutations. Currently there are major knowledge gaps in characterizing the genetic and molecular mechanisms governing the relationship between IR exposure and diseases other than cancer. Recently, additional adverse health effects of IR were realized including alterations to the cardiovascular and nervous systems. Moreover, there is limited information on the immediate and latent effects of an embryonic exposure to these biological systems. To begin to elucidate IR toxicity in this study we first exposed embryonic zebrafish at 26 hpf to a range of IR doses to establish toxicity at doses that represent sublethal to potentially lethal whole body doses in humans. Zebrafish were observed through the developmental time course through 120 hpf and no significant increase in mortality or alterations in hatching rate was observed at any of the doses. In addition, an increase in malformations and a significant difference in morphology were only observed at 10 Gy. This initial toxicity assay provided baseline toxicity of IR in our laboratory at the specific doses at the specific developmental time point of exposure for comparison to previous completed studies with the zebrafish model system. Our findings are similar to those reported in the literature for radiation exposure at this developmental time point, while exposure at earlier developmental stages is reported as more toxic (McAleer et al., 2005; Geiger et al., 2006; Pereira et al., 2011). McAleer et al. (2005) and Geiger et al. (2006) observed that lethal effects of IR were inversely proportional to embryonic age with earlier embryonic stages (e.g., 2–6 hpf) being more sensitive to lethality than later embryonic stages (e.g., 8–24 hpf).

Previous studies using the zebrafish model system to evaluate ionizing radiation exposure have been used to assess general toxicity (e.g., mortality, hatching rates, gross morphological alterations, DNA damage, and apoptosis) (e.g., Bladen et al., 2007; Choi et al., 2010a,b, 2012a, 2013; Pereira et al., 2011; Sorrells et al., 2012; Yu et al., 2012; Toruno et al., 2014), to screen radiation protectors (McAleer et al., 2005; Geiger et al., 2006), and to assess the long-term effects on liver gene expression (Jaafar et al., 2013). To the best of our knowledge, no work is published on the effect of IR to heart rate using the zebrafish model. Heart rate is closely associated with cardiovascular function and the zebrafish near-transparent chorion permits visualization of internal organs including the heart during embryogenesis; hence, it is desirable to investigate how IR affects heart rate. During zebrafish development the initial peristaltic waves of contraction drive circulation and a heartbeat is visible by 24 hpf. Heart rate increases through the developmental time course until reaching 180–200 bpm between 48–72 hpf similar to what was observed in the fish in the control treatment (Glickman and Yelon, 2002; Incardona et al., 2004). At 120 hpf the heart has two chambers, the atrium and the ventricle. The atrium is located to the left and slightly posterior to the ventricle and the bulbus arteriosus is next the ventricle. Blood flows from the atrium to the ventricle and then through the bulbus ateriosus. Although the zebrafish heart is two-chambered compared to the four-chambered mammalian heart, the many processes of cell migration and differentiation involved in forming the zebrafish heart are parallel to those in mammalian heart development (Stainier, 2001). In this study an initial heart rate was attained at 24 hpf to attain a baseline heart rate prior to the IR exposure at 26 hpf. Heart rate analysis was conducted at multiple time points throughout the developmental time course and a significant decrease observed at 10 Gy, while a significant increase in heart rate was observed at 1, 2, and 5 Gy. While it can be concluded that the decrease in heart rate at 10 Gy is related to the gross malformations observed, the mechanism for the increase of heart rate at lower doses requires further investigation. Another potential mechanism for the increase and then decrease in heart rate could be explained by hormesis. Hormetic effects associated with exposure to alpha particles and microbeam photons were recently observed in zebrafish embryos (Choi et al., 2012a,b). Nonetheless, the increase in heart rate indicates a lasting cardiovascular impact of the developmental IR exposure and potential subtle alterations.

Zebrafish larvae at 120 hpf exposed to the highest dose of IR at which no significant difference was observed in malformations and morphology (5 Gy) were used for transcriptomic analysis to determine genetic targets and molecular pathways altered. Applying criteria from the Microarray Quality Control Consortium (Guo et al., 2006; Shi et al., 2006) a multi-tiered statistical processing approach was used and identified a differentially expressed gene list of 253 genes orthologous to human genes with established functions. As observed in past studies in our laboratory when evaluating transcriptome changes at doses below those that cause gross morphological malformations, gene expression changes were not of high magnitude (Peterson et al., 2011; Weber et al., 2013). As such to eliminate inflation of the false negative rate for this data set a Bonferroni correction test was not applied (please see Norris and Kahn, 2006 for a discussion of multiple correction tests not being applied in the analysis when gene expression changes are subtle and of smaller magnitude such as those observed in this study).

LIN7B, GPR24, RLN3, and GANC were the genes with the greatest magnitude of expression differences. LIN7B encodes a protein involved in protein binding and maintaining distribution of channels and receptors in the cell membrane and is involved in neurotransmitter secretion with previous association with schizophrenia. The protein encoded by GPR24 is a member of the G protein-coupled receptor family 1 and is an integral plasma membrane protein which binds to melanin-concentrating hormone. The protein can inhibit cAMP accumulation and stimulate intracellular calcium flux. RLN3 belongs to the family of relaxins that are endocrine and autocrine/paracrine hormones belonging to the insulin gene superfamily. In humans there are three non-allelic relaxin genes, RLN1, RLN2, and RLN3. While RLN1 and RLN2 share high sequence homology, RLN3 does not. Relaxin has roles in the male and female reproductive system and also in regulating blood pressure, controlling heart rate, and releasing oxytocin and vasopressin. GANC encodes a glycosyl hydrolase enzyme that hydrolyzes the glycosidic bond between two or more carbohydrates or between a carbohydrate and non-carbohydrate moiety. This enzyme is key in glycogen metabolism and is associated with susceptibility to diabetes.

Few studies to date utilized transcriptomic technologies to investigate genetic mechanisms of IR response on non-cancer related health effects. These studies were conducted at a range of doses and in a variety of systems. All studies were conducted in either cell culture systems, in mature rodent models, or from adult human blood samples. Regardless, the differentially expressed gene list from this study was compared to each of these studies. There was minimal overlap between our differentially expressed gene list and those from the other studies. These differences are most likely due to the age of individuals, tissue analyzed, or in relation to dose response. There were similarities in up regulation of CCNG1, MDM2, and RPS27L and down regulation of MAPK10 and PAM in previous studies with human lymphocytes and whole blood analysis (Fachin et al., 2007; Kabacik et al., 2011; Wyrobek et al., 2011). In addition, there was some overlap in genetic targets associated with cognitive function and neurological disease (e.g., GRIA3 and GRIN1) in one study that analyzed 8–10 weeks male mice brain tissue 4 h after exposure to 0.1 or 2 Gy (Lowe et al., 2009). Furthermore, although there is limited overlap in the specific genetic targets, the top canonical pathways identified in our study and in the lower dose of 0.1 Gy are similar (i.e., synaptic long term depression and synaptic long term potentiation).

The dose of 5 Gy, albeit a higher dose of IR, was chosen for transcriptomic analysis as the highest dose for which no gross morphological measurements were observed and served as a starting point for this study for the assessment of gene expression alterations. To further investigate gene expression alterations, targeted expression analysis of five genes associated with cardiovascular functions and LIN7B was completed at 1 and 2 Gy. A significant decrease in expression of LIN7B was observed at both 1 and 2 Gy, while expression alterations were no longer observed for CACNA1D, GUCY1B3, NTRK3, PBX3, and PTPRE. Further work is needed to determine if overall enrichment of gene expression alteration categories are similar at the lower IR doses.

Overall this study indicates a developmental IR exposure at doses below those that induce gross morphological changes alters heart rate and expression of genes associated with cardiovascular and neurological development, function, and disease. This study furthers our understanding of the immediate IR induced biological responses during early development at the genetic level that trigger processes leading to short term cardiovascular effects. Future work is needed to define enrichment of gene expression alterations at the lower IR doses to provide additional information on transcriptome changes, to further investigate the functional impacts on the cardiovascular and nervous systems, and the lifespan consequences of the developmental IR exposure.