Integrative Assessment of Benzene Exposure to Caenorhabditis

May 20, 2014 - Toxicogenomic analysis was further conducted on C. elegans exposed to the same concentration of benzene that caused behavior and ...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/est

Integrative Assessment of Benzene Exposure to Caenorhabditis elegans Using Computational Behavior and Toxicogenomic Analyses Hyun-Jeong Eom,†,# Hungsoo Kim,‡,§,# Bo-Moon Kim,‡,# Tae-Soo Chon,*,‡ and Jinhee Choi*,† †

School of Environmental Engineering, Graduate School of Energy and Environmental System Engineering, University of Seoul, 163 Siripdaero, Dongdaemun-gu, Seoul 130-743, Korea ‡ Department of Biological Sciences, Pusan National University, 2 Busandaehak-ro 63 beon-gil, Geumjeoung-gu, Busan 609-735, Korea § SPENALO, National Robotics Research Center, Pusan National University, 2 Busandaehak-ro 63 beon-gil, Geumjeoung-gu, Busan 609-735, Republic of Korea W Web-Enhanced Feature * S Supporting Information *

ABSTRACT: In this study, we investigated the toxic effects of benzene to the nematode Caenorhabditis elegans in an integrative manner, using computational behavior and toxicogenomics analyses, along with survival and reproduction. Benzene exposure led to changes in locomotive behavior and reproduction decline in C. elegans. Microarray followed by pathway analysis revealed that 228 genes were differentially expressed by benzene exposure, and cyp-35a2, pmk-1, and cep-1 were selected for further reproduction and multiparametric behavior analysis. Mutant analysis showed that benzene induced reproduction decline was rescued in cyp-35a2(gk317) mutant, whereas it was significantly exacerbated in pmk-1(km25) mutant, compared with the wildtype. The multiparametric behavior analysis on the mutants of selected genes revealed that each strain exhibits different response patterns, particularly, enhanced linear movement in the cyp-35a2(gk317) mutant, whereas the changes in partial body movement were observed in the pmk-1(km25) mutant by benzene exposure. A self-organizing map revealed that the pmk-1(km25) mutant group was the most densely clustered and located on the opposite side of the map of the cyp-35a2(gk317) mutant, each crossing that of the wildtype. Overall results suggest distinct roles of cyp-35a2 and pmk-1 genes in benzene-induced alterations in behavior and reproduction in C. elegans. This study also suggests computational behavior analysis is a suitable tool for addressing the integrative impact of chemical stress alongside with toxicogenomic approach.

■

INTRODUCTION Behavioral monitoring fills the gap in information between macroscale (e.g., population dynamics, community) and microscale (e.g., molecular response) determinations used for risk assessment.1 Since behaviors reflect integrated processes relating to subcellular, cellular, physiological, and individual level responses, behavioral measurements can reveal disruptions arising from exposure to a broad spectrum of chemicals and environmental stresses.2−6 For instance, locomotive, feeding, and mating behaviors are closely associated with multiple physiological processes, including those relating to neurological and endocrine systems. Behavioral responses affect the growth, survival, and reproduction of animals,1,7,8 and thus they reflect the effects of stimuli in a comprehensive manner (i.e., overall body response). Along with the rapid development in interfacing techniques and analytic methods for complex data, the automatic detection of behavioral responses has received considerable attention as an efficient tool for biomonitoring since the 1980s.9−11 The early warning signal based on behavior changes of organisms is considered as a prompt and relevant indicator of overall environmental quality.1,12 The free-living nematode Caenorhabditis elegans is one of the most powerful experimental models in various biological fields, including developmental biology and genetics. Studies of C. elegans allow for direct mechanistic toxicity research due to © 2014 American Chemical Society

the availability of RNAi libraries and mutants, and, as such, the use of C. elegans in toxicology has also increased greatly in recent years, using various apical, molecular, or genetic end points.13−15 C. elegans is also an ideal model for neurobehavior research examining the genetic basis of behavior,16 as 302 neurons have been determined to coordinate different behaviors, such as feeding, mating, egg-laying, defecation, swimming, and many subtle forms of locomotion.17−19 The studies on how controlling neurotransmitters affect the behavior of C. elegans have been performed since the early 2000s.20,21 Intensive further explorations were conducted on genes and pathways regulating various behaviors in C. elegans.22,23 Recently, a number of wormtracking packages and video analysis tools have been designed and developed to allow for the quantification and monitoring of different aspects of locomotion in C. elegans.24 Though C. elegans behavior-related research has been widely conducted in various aspects,25,26 only limited information is available on the relationship between chemical-induced gene level alterations and behavior responses. In a toxicological context, locomotive Received: Revised: Accepted: Published: 8143

February 5, 2014 May 18, 2014 May 20, 2014 May 20, 2014 dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

Figure 1. Schemes for recording and analysis process of behavior of C. elegans “before” and “after” exposure to benzene (A). Ten min-long line movement trajectories of C. elegans “before” treatment (B) and “after” treatment (C). The line-movement tracks of the nematode were monitored for 4 h (2 h for before and 2 h for after treatment of benzene, see Web-enhanced objects: video 1 and video 2. The trajectory patterns were determined based on 10 min monitoring after 1 h recording of “before” and “after” treatment. Color indicates sequential time recording. Green (with “s”) and red (with “e”) points indicate start and end points, respectively. The circular inset shows the sequence of each line movement in 0.3 s intervals.

behavior of C. elegans has also been proven to be a useful toxic end point,27,28 and many behavior toxicity studies of C. elegans have been conducted for metals29,30 and neurotoxic organic chemicals.31−33 Recently Salhanha et al. developed a microfluidic assay and applied it to reveal the role of genes in the hypoxia response pathway (i.e., hif-1 and egl-9) and cysteine synthase, cysl-2 in cyanide toxicity and resistance in C. elegans.34 Benzene is a ubiquitous environmental pollutant and potential carcinogen. It is widely used in industry and in commercial products and also frequently found in the environment.35 Benzene and its metabolites are associated with multiple adverse health effects primarily related to impairment of the hemopoietic system, particularly in reference to acute and chronic leukemias.36 Adverse effect on neuro system was also reported.37 In this study, we investigated the toxic effects of benzene to C. elegans in an integrative manner, using computational behavior and toxicogenomics analyses. Behavior was analyzed using the line tracking system, which monitors line movements of the nematode. In addition head and tail movements as well as partial body movements were monitored in response to benzene treatment. Common apical end points, such as, survival and reproduction, were also investigated. Subsequently, in order to gain insight into the underlying mechanism of behavior toxicity, microarray and pathway analyses were also conducted. In order to identify genes involved in behavior response by benzene exposure, the three most significantly associated genes, namely cyp-35a2, pmk-1, and cep-1, were selected from the

microarray and pathway analysis. The behavior responses of the loss-of-function mutants of those genes as well as that of the wildtype were subsequently monitored using multiple parameters, such as, speed, acceleration, turning rate, and meander. Finally, a self-organizing map (SOM) was built to compare the behavior responses of the wildtype and mutants toward benzene exposure.

■

MATERIALS AND METHODS C. elegans and Exposure to Benzene. Wildtype and cyp-35a2(gk317), pmk-1(km25), and cep-1(gk138) mutant C. elegans were grown in Petri dishes on nematode growth medium (NGM) and fed with OP50 strain Escherichia coli. Worms were incubated at 20 °C, with young adults (3 days old) from an agesynchronized culture then used in all the experiments. For treatment of C. elegans, experimental benzene (Fluka, purity 99%) was dissolved in dimethyl sulfoxide (DMSO, Sigma-Aldrich). The solution was then added to K-media (0.032 M KCl and 0.051 M NaCl). C. elegans were exposed to 1 mg/L of benzene in K-media. Behavior Observation System. The behavior observation system consisted of an observation arena, optical microscope, camera, and personal computer, with the experiment taking place in an airtight chamber (Figure S1). The behavior experiment scheme is presented in Figure 1A. Briefly, observation was conducted on one age-synchronized young adult C. elegans for 2 h, 1 mg/L of benzene was then treated to the airtight chamber, and observation was continued for another 2 h using a CCD 8144

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

treatment, the worms tended to move around in more limited ranges within the observation arena (Figure 1C). Survival and reproduction were also investigated so as to compare the sensitivity of behavior response with the common apical end points. Exposure of young adults of C. elegans to 1 mg/L of benzene did not lead to any increase in mortality over 48 h, whereas exposure to the same concentration of benzene for 72 h resulted in a decrease of offspring numbers of about 20% (Figure S2). When comparing the responses of survival and reproduction, changes in movement behavior occurred at a shorter exposure period (i.e., 1 h) than other responses (48 and 72 h). We cannot rule out that the immediate behavior response might contribute to longer term effects, such as a decline in reproduction potential. Microarray and Pathway Analysis. Toxicogenomic analysis was further conducted on C. elegans exposed to the same concentration of benzene that caused behavior and reproduction alteration (1 mg/L) for 24 h, in order to determine the overall integrative mechanism of the toxic response. Microarray data indicated that 228 genes were differentially expressed by benzene exposure (194 up- and 34 down-regulated genes, Table S3; GSE No. 23013). Genes involved in xenobiotic metabolism (i.e., CYP450 monooxygenase (cyp-35a2, 3, 4, cyp35c1), flavin monooxygenase ( f mo-2), UDP-GT (ugt-21, -22, -46, -62), and GST (gst-21, -26, -36)) were significantly upregulated by benzene exposure. Collagens and major sperm protein families were also significantly up-regulated. DEGs mapped to the KEGG pathway suggested that the xenobiotic metabolism pathway of CYP450 and fatty acid metabolism was heavily involved in benzene toxicity (Table S4). Indeed, the metabolism of benzene has been widely studied in human and animal models due to its putative role in carcinogenesis (metabolic activation).40,41 Particularly, CYP2E1 has been implicated in the mechanism of benzene toxicity, which was evidenced by the absence of CYP2E1 that leads to a reduction of the cytotoxic and genotoxic effects in transgenic mice.42 Xenobiotic metabolism in C. elegans has been previously reviewed,43 and Leung et al. reported the absence of CYP1 family-like enzymatic activities in C. elegans by testing the genotoxicity of benzo[ap]yrene and aflatoxin.44 However, xenobiotic metabolism, including benzene metabolism, in C. elegans is still not fully understood. Our microarray result further revealed that cyp-35a2-4 and cyp-35c1 genes were increased upon exposure to benzene (Table S3). The cyp-35 family genes are known to be induced by various xenobiotics45−47 and to have a function in xenobiotic metabolism.48 Aarnio et al.49 observed that the cyp-35 family is very similar to human CYP-1 and CYP-2 families. As benzene is metabolized by CYP450 2E1 in humans,50,51 our DEG results of increased expression of the cyp-35 gene family indicate that CYP mediated benzene toxicity might be conserved in C. elegans. The pathway/network analyses provide an unbiased interpretation of gene array data based upon established interactions between genes/gene products and the identification of relevant biological processes. Therefore, we additionally performed Ingenuity Pathways Analysis (IPA) on the data set of DEGs to gain a comprehensive insight into the mechanism of toxicity. Proinflammatory transcription factor, NF-κB, which is absent in the C. elegans genome, appeared to be connected to the majority of molecules in the network generated with DEGs after benzene exposure (Figure S3). The activation of NF-κB, followed by proinflammatory cytokine production in response to volatile organic compound (VOC) exposure, has been previously reported in human cells,52,53 and, therefore, this result may insinuate that the functional counterpart of NF-κB genes and related networks in

camera. Recording was conducted for the whole observation period (total 4 h; 2 h for “before”; and 2 h for “after” treatment), and the trajectory patterns were determined based on 10 min monitoring; whereas multiparametric analysis was conducted based on 2 h monitoring, after 1 h recording of “before” and “after” treatment. Worms were not fed during the behavior observation, and a separate set of experiment was additionally conducted for the whole period of observation time (i.e., 4 h) without benzene treatment to investigate effect of starvation. The images of C. elegans were captured (30 frame per second) using a CCD camera (SAC-410ND CCD Color Camera, Samsung Inc.) mounted on an optical microscope (Olympus SZX 16 optical microscope, Olympus Corp., Tokyo, Japan). The images were transferred using the line-shape detection program according to Son et al.38 Four parameters (i.e., speed (μm/s), acceleration (μm/s2), turning rate (rad/s), and meander (rad/μm)) were determined in 1 min average values, consisting of 1800 subsegments of images. Self-Organizing Map (SOM). Self-organizing map (SOM) was used to classify behavioral responses of C. elegans to benzene exposure. The SOM is an efficient means of creating maps of multidimensional and complex data in order to approximate the probability density function of the input data in a more comprehensive fashion.39 In the SOM network, the output layer consisted of computation nodes (j) in low dimension (conveniently 2). Assuming input data consisting of four parameters, the value of a parameter, i, is expressed as a vector, xi, and was given to the input layer of the SOM. In the network each computation node, j, is connected to each node, i, of the input layer. The connectivity is represented as the weights, wij(t), between input and output nodes, adaptively changing in each iteration of calculation, t, until convergence is reached. Initially, the weight is randomly assigned in small values. Each neuron of the network computes the summed distance between the weights and the distance, dj(t), at output node, j, and the network is calculated as shown below: (N − 1)

dj(t ) =

∑

(xi − wij(t ))2

(i = 0)

The input data matrix consists of four behavior parameters (speed, acceleration, turning rate, and meander) as variables and 104 movement segments (i.e., 4 (strains) × 2 (“before” and “after” treatment) × 13 (body regions) = 104) as sample units. The details for the materials and methods are provided in the Supporting Information.

■

RESULTS AND DISCUSSIONS Behavior, Survival, and Reproduction. To investigate toxic potential of benzene to C. elegans, behavior, survival, and reproduction were monitored on C. elegans exposed to benzene. For behavior analysis, the line-movement tracks of the nematode were monitored for 4 h (2 h for “before” and 2 h for “after” treatment of benzene) and the trajectory patterns were determined based on 10 min monitoring after 1 h recording of “before” and “after” treatment (Figure 1A). The recorded video clips are available as Web-enhanced objects (video 1 and video 2), and each video clip showed distinct movement of a single C. elegans “before” (video 1) and “after” treatment (video 2). C. elegans was represented by a single line consisting of 13 body regions in every 0.03 s interval. Before the treatment, worms were distributed broadly, with more or less even dispersion throughout the observation arena (Figure 1B), whereas, after the benzene 8145

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

Figure 2. Expression of cyp-35a2, pmk-1, and cep-1 genes in wildtype (N2) (A), survival (B), and reproduction (C) of wildtype (N2), cyp-35a2 (gk317), pmk-1 (km25), and cep-1 (gk138) mutants. The expression of cyp-35a2, pmk-1, and cep-1 genes in wildtype (N2) exposed to 1 mg/L benzene using qRT-PCR, and the results are expressed as mean values compared to those of the control (control = 1, n = 3; mean ± standard error of the mean; **p < 0.01) (A). Survival and reproduction analysis was conducted on wildtype (N2), cyp-35a2 (gk317), pmk-1 (km25), and cep-1 (gk138) mutants exposed to 1 mg/L benzene, and the results were expressed as mean values compared to those of the control (control = 1, n = 3; mean ± standard error of the mean; *p < 0.05, **p < 0.01). The statistical difference between mutants and wildtype (N2) was also analyzed (#p < 0.05, ##p < 0.01) (B and C).

Figure 3. Average movement parameters of each strain of C. elegans “before” and “after” exposure to benzene. Speed (A), acceleration (B), turning rate (C), and meander (D). The results are expressed as mean values compared to “before” and “after” treatments based on the paired t −test (n = 5; mean ± standard error of the mean; *p < 0.05). Turkey’s test was conducted to compare statistical difference across the strains. Shared letters indicate no statistically significant difference (P ≥ 0.05), whereas different letters indicate that there is a significant difference (P < 0.05).

C. elegans may be involved in defense against benzene toxicity. It was also determined that p38 and ERK MAPK are strongly associated with benzene exposure (Figure S3). p38 MAPK is a well conserved pathway in C. elegans, which has prime importance in C. elegans immune and stress responses.54,55 Selection of Genes That Involved in Benzene Toxicity. Taking into account the microarray and pathway analysis results together, we selected cyp-35a2(gk317) and pmk-1(km25) mutants for deeper investigations of the investigation on genebehavior relationship. Among different cyp-35a genes (i.e., cyp35a2-4), cyp-35a2 was selected, as its expression was the most significantly increased after exposure to benzene, as confirmed by

individual PCR (Figure S4). We additionally selected the cep1(gk138) mutant, as benzene is known to be genotoxic and carcinogenic, and p53 (homologous C. elegans CEP-1) has a prime importance in genotoxic carcinogenesis. The expression levels of the selected genes (cyp-35a2, pmk-1, cep-1) were investigated in response to benzene exposure, and it was found that the expression of cyp-35a2 and pmk-1 genes were increased 3- and 1.5-fold compared to those of the control, respectively, whereas no alteration of expression was observed for cep-1 gene expression (Figure 2A). The survival and reproduction of mutants of these genes was also investigated after exposure to benzene, and was compared to that of the 8146

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

wildtype (Figure 2B, C). In terms of survival, no statistical difference was observed between the response of wildtype and that of mutants. On the other hand, benzene-induced reproduction failure was rescued in the cyp-35a2(gk317) mutant, whereas it was significantly exacerbated in the pmk-1(km25) mutant. Increased expression of cyp-35a2 gene (Figure 2A) and rescued toxicity by its loss-of-function mutant (Figure 2C) suggested that benzene-induced toxicity by toxic metabolites via CYP2 also occurred for C. elegans, and cyp-35a2 seems to be largely responsible for benzene metabolism in C. elegans. We previously found that the toxicity of fenitrothion, an organophosphorous pesticide, depends on metabolic activation via cyp35a2 in C. elegans,56 and cyp-35a2 was also found to be involved in nanotoxicity.57 The significantly increased sensitivity of the pmk-1(km25) mutant to benzene exposure reinforces the protective role of this gene in stress response. In our previous studies with silver nanoparticles, we reported the importance of PMK-1 p38 MAPK in oxidative stress related toxicity in C. elegans.57,58 Behavior Monitoring with Functional Mutants of Genes Involved in Benzene Toxicity. After confirming the potential role of selected genes in benzene toxicity, behavior analysis was investigated in the loss-of-function mutants of cyp35a2, pmk-1, and cep-1 genes as well as in wildtype C. elegans, “before” and “after” exposure to benzene. Four hours recording (2 h “before” and 2 h “after” treatment) on the line-movement tracks were thoroughly analyzed by dividing body regions into 13 parts from head to tail (1, head; 13, tail) (see Materials and Methods). The average values of movement parameters, such as, speed, acceleration, turning rate, and meander, of 13 body regions were evaluated. The wildtype and pmk-1(km25) mutant showed higher overall speed (60−70 μm/s) and acceleration (110−124.7 μm/s2) than cyp-35a2(gk317) and cep-1(gk138) mutants (Figure 3A, B). Results on speed and acceleration also indicated that cyp-35a2(gk317) and cep-1(gk138) mutants exhibited similar trends of lower activities, whereas wildtype and pmk-1(km25) mutants showed higher activities (Figure 3A, B). The turning rate was not much changed compared with the other parameters (Figure 3C).The results of meander were reversed, being higher in cyp-35a2(gk317) and cep-1(gk138) (0.45− 0.47 rad/μm) and lower in wildtype and pmk-1(km25) (0.33− 0.39 rad/μm) (Figure 3D). Though the worms were not fed during the whole observation period, starvation did not affect any of behavior parameters in response to benzene exposure (Figure S6). After benzene treatment, the values of parameters were characteristically different according to the strains tested. Marked increments of speed (14 μm/s) and acceleration (24.86 μm/s2) were observed for the cyp-35a2(gk317) mutant, and the increments were statistically significant (p < 0.05; t test, Figure 3A, B). It was also noteworthy that the parameters indicating direction change (turning rate and meander) appeared to decrease in all the strains after the treatment (Figure 3C, D). However, statistical difference was observed in wildtype and cyp-35a2(gk317) for turning rate and meander, respectively, and in cep-1(gk138) mutant for both parameters. The results confirmed that the activity of the nematodes was generally affected by benzene exposure, and, specifically, linear movement of cyp-35a2(gk317) mutant was enhanced by benzene; while pmk-1(km25) mutant showed no difference in any of parameters after benzene treatment (Figure 3). We further analyzed data from the measurements of 13 individual body regions (Figure 4). Differences in parameters characterizing 13 body regions in averages were overall similar to differences observed from whole body (Figure 3) except for a few

Figure 4. Differences between “before” and “after” treatment in the movement parameters of 13 body regions of C. elegance. Speed (A), acceleration (B), turning rate (C), and meander (D). The number on the x axis indicates the body region from head to tail (1, head; 13, tail). Age synchronized young adults were individually observed with 5 replications of each strain and treatment. The results are expressed as mean values, and standard errors of the mean are omitted for the clarity of the data. The symbol “*” presents statistical significance between “before” and “after” treatment based on the paired t-test (p < 0.05).

cases of partial difference in body regions (e.g., pmk-1(km25)) (Figure 4). The cyp-35a2(gk317) mutant showed a similar difference across the body regions “before” and “after” treatment (Figure 4). All points of body regions increased in speed and acceleration in cyp-35a2(gk317) “after” treatment (Figure 4A, B). It was worth noting that the different body regions responded differently to the chemical in the pmk-1(km25) mutant although 8147

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

Figure 5. Clustering of strains of C. elegans on the SOM according to four movement parameters (speed, acceleration, turning rate, and meander) “before” and “after” treatment. Clustering (A) (alphabets, the number and character (a/b) listed units in the node of SOM indicate strains, body region and treatment (b: “before” treatment, a: “after” treatment), respectively. (i.e., “pmk-13a” means the 13th region of pmk-1 after benzene treatment)). Dendrogram for clustering (B) and profile of parameters superimposed on the SOM (C).

tail part of the body (Figure 4B). Decrease in meander was also notable in the mid body of cyp-35a2(gk317) and cep-1(gk138) mutants with statistical significance (Figure 4D). Body shape responses of the pmk-1(km25) mutant were somewhat fundamentally different from the other strains tested

no significant changes were observed in the average values of the parameters for whole body (Figure 3). An increase in the speed and acceleration at the head part of pmk-1(km25) mutant was observed compared to that of the wildtype (Figure 4A), whereas the same strain showed a decrease in acceleration at mid to the 8148

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

“before” and “after” treatment (Figure 4). The pmk-1 has been intensively investigated in immune and stress response of C. elegans,59,60 and it was also found to be involved in responses to chemicals exposure by our group57,58 as well as by other research groups.61,62 However, no study has previously reported on the neurobehavior aspect of pmk-1, and the current study provided no clear explanation on how and why partial body shape was specifically affected in this strain. Further investigation of neurobehavior related stress response will be required in the pmk-1(km25) strain. Self-Organizing Map (SOM) Analysis. We further built a SOM by using the data for body parameters to create a comprehensive review of the behavior responses of different strains (Figure 5). A vertical gradient was observed for all of the strains “before” treatment; the pmk-1(km25) mutant was located in the upper part of the SOM, whereas the cep-1(gk138) and cyp35a2(gk317) mutants were located at the bottom part, across the wildtype, which was located in the middle of the map (Figure 5A). The data obtained “before” and “after” treatment were additionally grouped horizontally on the map for each of the different strains. The location of the pmk-1(km25) mutant was further divided into two parts: the top right for “before” treatment and top left for “after” treatment. Similarly, wildtype and cyp-35a2(gk317) mutant strains were also divided horizontally (Figure 5A). Although the pmk-1(km25) and cyp-35a2(gk317) mutants and wildtype were densely grouped, the movements of the cep-1(gk138) mutant were widely distributed throughout the map after treatment compared to the other strains. This result, along with gene expression and toxicity results (Figure 2), collectively suggests limited involvement of cep-1 in response to benzene exposure. Each sample unit was classified into 8 clusters based on the dendrogram of the hierarchical cluster analysis using Ward’s linkage method with the Euclidian distance measure (Figure 5B). Clusters were fairly well grouped according to the strains “before” and “after” treatment. Cluster V was mostly represented by the wildtype “before” treatment, whereas cluster IV was grouped by the wildtype “after” treatment. Body regions of pmk1(km25) “before” treatment were mostly placed in cluster II, whereas body regions of pmk-1(km25) “after” treatment were mainly observed in cluster I. Clusters VI and VII represented cyp-35a2(gk317) “before” and “after” treatment, respectively. The pmk-1(km25) and cyp-35a2(gk317) mutants were located in the opposite sides, crossing over the wildtype (Figure 5A, B). This suggests that benzene affected the two strains of C. elegans in starkly different manners. The grouping of the cep-1(gk138) mutant was relatively similar to that of cyp-35a2(gk317) “before” treatment (Figure 5B, clusters VI and VIII), indicating that the impact of chemical treatment was relatively similar between these two mutants, as compared to pmk-1(km25). When the movement parameters were compared among different clusters based on the SOM (Figure 5C, S5), the parameters exhibited a similar pattern “before” and “after” treatment, as shown by the average values of parameters (Figure 3). All the parameters for the wildtype tended to decrease slightly “after” treatment (Figure 5C, S4, clusters IV and V). Speed and acceleration of pmk-1(km25) mutant were higher than other strains “before” treatment, followed by further increase “after” treatment (Figure S5A, B, clusters I and II). It was noteworthy that clusters VI representing cyp-35a2(gk317) “before” treatment showed the lowest values in speed and acceleration, whereas increases of the same parameters were observed “after” the treatment (Figure 5C, S5A, B, Clusters VII), confirming the

increase in these parameters due to chemical treatment as stated above (Figure 3). Overall, the general trends observed in average values (Figure 3) were confirmed by the SOM. It was worth noting that the difference in parameters regarding direction change (turning rate and meander) across different clusters were hardly observed (Figure S5 C, D) compared to the parameters showing linear movement (speed and acceleration) (Figure S5 A, B). Benzeneinduced increase in speed and acceleration in the cyp-35a2(gk317) mutant (Figures 3, 4) may indicate that, in the absence of cyp-35a2 gene, toxic metabolites cannot be produced. Thus, worms have an increased resistance to benzene toxicity. Increased locomotive activity in the cyp-35a2(gk317) mutant reflected this phenomenon. It is interesting to relate these behavior alterations with changes in gene expression and reproduction (Figure 2). As described above, increased expression by benzene was observed in cyp-35a2 and pmk-1 genes but not in cep-1 gene (Figure 2A). Similarly, the reproduction decline in wildtype was changed in cyp-35a2(gk317) and pmk-1(km25) mutants (i.e., rescued in cyp-35a2(gk317) whereas exacerbated in pmk-1(km25)) but not in cep-1(gk138) mutant (Figure 2C). These results collectively suggest that the alterations of benzene-induced cyp-35a2 and pmk-1 gene expression may have functional concurrence, such as, reproduction decline and behavioral alterations in C. elegans; however, the exact mechanism on how the expression change of these genes alters those parameters needs further investigation. The different responses of each body part of pmk-1(km25) mutant merits further investigation of the neurobiological mechanisms at play.

■

CONCLUSIONS Benzene exposure led to changes in gene expression, reproduction, and behavior. The multiparametric behavior analysis on selected mutants of genes indicated that linear movement was enhanced by benzene exposure in the cyp-35a2(gk317) mutant, whereas turning activity was declined in all strains tested. It is remarkable that the pmk-1(km25) mutant showed changes in partial body movement. According to the SOM analysis, the benzene treated pmk-1(km25) mutant group was the most densely clustered, and the cyp35a2(gk317) mutant strain was clustered at the opposite side of the map, suggesting the distinct role of those two genes in toxic mechanisms of benzene in C. elegans, which was also supported by gene expression and mutant reproduction test. This study will contribute to our understanding of the relationship between gene and behavioral responses toward benzene exposure in C. elegans. Taking into account the considerable number of conserved genes and pathways between C. elegans and mammalian species, this information has the potential of more general application in broadening our understanding of behavioral toxicity.

■

ASSOCIATED CONTENT

* Supporting Information S

The details for materials and methods, Tables S1−S4, and Figures S1−S6. This material is available free of charge via the Internet at http://pubs.acs.org. W Web-Enhanced Features *

The recorded video clips are available as Web-enhanced objects (videos 1 and 2).

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. 8149

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

Author Contributions

(16) Hart, A. C., Ed. Behavior (July 3, 2006), WormBook, ed. The C. elegans Research Community, WormBook, doi/10.1895/wormbook.1.87.1 http://www.wormbook.org (accessed May 30, 2014). (17) Loer, C. M.; Kenyon, C. J. Serotonin-deficient mutants and male mating behavior in the nematode Caenorhabditis elegans. J. Neurosci. 1993, 13 (12), 5407−5417. (18) Ségalat, L.; Elkes, D. A.; Kaplan, J. M. Modulation of serotonincontrolled behaviors by Go in Caenorhabditis elegans. Science 1995, 267 (5204), 1648−1651. (19) Hardaker, L. A.; Singer, E.; Kerr, R.; Zhou, G.; Schafer, W. R. Serotonin modulates locomotory behavior and coordinates egg-laying and movement in Caenorhabditis elegans. J. Neurobiol. 2001, 49 (4), 303−313. (20) Zheng, Y.; Brockie, P. J.; Mellem, J. E.; Madsen, D. M.; Maricq, A. V. Neuronal control of locomotion in C. elegans is modified by a dominant mutation in the GLR-1 ionotropic glutamate receptor. Neuron 1999, 24 (2), 347−361. (21) Sawin, E. R.; Ranganathan, R.; Horvitz, H. R. C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 2000, 26 (3), 619−631. (22) Ma, D. K.; Rothe, M.; Zheng, S.; Bhatla, N.; Pender, C. L.; Menzel, R.; Horvitz, H. R. Cytochrome P450 drives a HIF-regulated behavioral response to reoxygenation by C. elegans. Science 2013, 341 (6145), 554− 558. (23) Yu, H.; Aleman-Meza, B.; Gharib, S.; Labocha, M. K.; Cronin, C. J.; Sternberg, P. W.; Zhong, W. Systematic profiling of Caenorhabditis elegans locomotive behaviors reveals additional components in Gprotein Gαq signaling. Proc. Natl. Acad. Sci. U.S.A. 2013, 110 (29), 11940−11945. (24) Husson, S. J. et al. Keeping track of worm trackers (September 10, 2012), WormBook, ed. The C. elegans Research Community, WormBook, doi/10.1895/wormbook.1.156.1. http://www.wormbook.org (accessed May 30, 2014). (25) Leung, M. C.; Williams, P. L.; Benedetto, A.; Au, C.; Helmcke, K. J.; Aschner, M.; Meyer, J. N. Caenorhabditis elegans: an emerging model in biomedical and environmental toxicology. Toxicol. Sci. 2008, 106 (1), 5−28. (26) Zhao, Y.; Wu, Q.; Tang, M.; Wang, D. The in vivo underlying mechanism for recovery response formation in nano-titanium dioxide exposed Caenorhabditis elegans after transfer to the normal condition. Nanomedicine 2013, 10 (1), 89−98. (27) Anderson, G. L.; Boyd, W. A.; Williams, P. L. Assessment of sublethal endpoints for toxicity testing with the nematode Caenorhabditis elegans. Environ. Toxicol. Chem. 2001, 20 (4), 833−838. (28) Dhawan, R.; Dusenbery, D. B.; Williams, P. L. Comparison of lethality, reproduction, and behavior as toxicological endpoints in the nematode Caenorhabditis elegans. J. Toxicol. Environ. Health, Part A 1999, 58 (7), 451−462. (29) Yu, Z.; Chen, X.; Zhang, J.; Wang, R.; Yin, D. Transgenerational effects of heavy metals on L3 larva of Caenorhabditis elegans with greater behavior and growth inhibitions in the progeny. Ecotoxicol. Environ. Saf. 2013, 88, 178−184. (30) Zhang, Y.; Ye, B.; Wang, D. Effects of metal exposure on associative learning behavior in nematode Caenorhabditis elegans. Arch. Environ. Contam. Toxicol. 2010, 59 (1), 129−136. (31) Ali, S. J.; Rajini, P. S. Elicitation of dopaminergic features of Parkinson’s disease in C. elegans by monocrotophos, an organophosphorous insecticide. CNS Neurol. Disord.: Drug Targets 2012, 11 (8), 993−1000. (32) Anderson, G. L.; Cole, R. D.; Williams, P. L. Assessing behavioral toxicity with Caenorhabditis elegans. Environ. Toxicol. Chem. 2004, 23 (5), 1235−1240. (33) Ruan, Q. L.; Ju, J. J.; Li, Y. H.; Liu, R.; Pu, Y. P.; Yin, L. H.; Wang, D. Y. Evaluation of pesticide toxicities with differing mechanisms using Caenorhabditis elegans. J. Toxicol. Environ. Health, Part A 2009, 72 (11− 12), 746−751. (34) Saldanha, J. N.; Parashar, A.; Pandey, S.; Powell-Coffman, J. A. Multiparameter behavioral analyses provide insights to mechanisms of

#

These authors contributed equally.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

■

REFERENCES

This work was supported by the grant from the Korea Ministry of Environment as “Environmental Health R&D Program″ (2012001370009) and ‘‘Eco-technopia 21 Project’’ and also by the grant from the Midcareer Researcher Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (2013R1A2A2A03010980).

(1) Liu, Y.; Lee, S. H.; Chon, T. S. Analysis of behavioral changes of zebrafish (Danio rerio) in response to formaldehyde using Selforganizing map and a hidden Markov model. Ecol. Modell. 2011, 222 (14), 2191−2201. (2) Breckels, R. D.; Neff, B. D. Pollution-induced behavioural effects in the brown bullhead (Ameiurus nebulosus). Ecotoxicology 2010, 19 (7), 1337−1346. (3) De Lange, H. J.; Peeters, E. T. H. M.; Lurling, M. Changes in ventilation and locomotion of Gammarus pulex (Crustacea, Amphipoda) in response to low concentrations of pharmaceuticals. Hum. Ecol. Risk Assess. 2009, 15 (1), 111−120. (4) Duquesne, S.; Liess, M.; Bird, D. J. Sub-lethal effects of metal exposure: physiological and behavioural responses of the estuarine bivalve Macoma balthica. Mar. Environ. Res. 2004, 58 (2−5), 245−250. (5) Gaworecki, K. M.; Klaine, S. J. Behavioral and biochemical responses of hybrid striped bass during and after fluoxetine exposure. Aquat. Toxicol. 2008, 88 (4), 207−213. (6) Painter, M. M.; Buerkley, M. A.; Julius, M. L.; Vajda, A. M.; Norris, D. O.; Barber, L. B.; Furlong, E. T.; Schultz, M. M.; Schoenfuss, H. L. Antidepressants at environmentally relevant concentrations affect predator avoidance behavior of larval fathead minnows (Pimephales promelas). Environ. Toxicol. Chem. 2009, 28 (12), 2677−2684. (7) Little, E. E. Behavioral measures of environmental stressors in fish. In Biological Indicators of Aquatic Ecosystem Stress; Adams, S. M., Ed.; American Fisheries Society: Bethesda, MD, 2002; pp 431−472. (8) Scott, G. R.; Sloman, K. A. The effects of environmental pollutants on complex fish behaviour: integrating behavioural and physiological indicators of toxicity. Aquat. Toxicol. 2004, 68, 369−392. (9) Lemly, A. D.; Smith, R. J. A behavioral assay for assessing effects of pollutants of fish chemoreception. Ecotox. Environ. Saf. 1986, 11 (2), 210−218. (10) Li, Y.; Lee, J. M.; Chon, T. S.; Liu, Y.; Kim, H.; Bae, M. J.; Park, Y. S. Analysis of movement behavior of zebrafish (Danio rerio) under chemical stress using hidden Markov model. Mod. Phys. Lett. B 2013, 27 (02), 1350014−1350027. (11) Lee, W. Y.; Macko, S. A.; Nicol, J. A. C. Changes in nesting behavior and lipid content of a marine amphipod (Amphithoe valida) to the toxicity of no. 2 fuel oil. Water, Air, Soil Pollut. 1981, 15, 185−195. (12) Gerhardt, A. Biomonitoring for the 21st Century. In Biomonitoring of Polluted Water. Reviews on Actual Topics, Environmental Research Forum Vol 9; Gerhardt, A., Ed.; TransTech Publ.: Switzerland, 1990; pp 301. (13) Boyd, W. A.; Smith, M. V.; Freedman, J. H. Caenorhabditis elegans as a model in developmental toxicology. Methods Mol. Biol. 2012, 889, 15−24. (14) Martinez-Finley, E. J.; Aschner, M. Revelations from the Nematode Caenorhabditis elegans on the Complex Interplay of Metal Toxicological Mechanisms. J. Toxicol. 2011, 2011, 895236. (15) Williams, P. L.; Dusenbery, D. B. Using the nematode Caenorhabditis elegans to predict mammalian acute lethality to metallic salts. Toxicol. Ind. Health. 1988, 4 (4), 469−478. 8150

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151

Environmental Science & Technology

Article

cyanide resistance in Caenorhabditis elegans. Toxicol. Sci. 2013, 135 (1), 156−168. (35) Agency for Toxic Substances and Disease Registry (ATSDR). Toxicological Profile for Benzene; U.S. Department of Health and Human Services, Public Health Service: Atlanta, GA. 2007. (36) Schnatter, A. R.; Rosamilia, K.; Wojcik, N. C. Review of the literature on benzene exposure and leukemia subtypes. Chem. Biol. Interact. 2005, 153−4, 9−21. (37) Lo Pumo, R.; Bellia, M.; Nicosia, A.; Micale, V.; Drago, F. Longlasting neurotoxicity of prenatal benzene acute exposure in rats. Toxicology 2006, 223 (3), 227−234. (38) Son, K. H.; Ji, C. W.; Park, Y. M.; Cui, Y.; Wang, H. Z.; Chon, T. S.; Cha, E. Y. Recurrent Self-Organizing Map implemented to detection of temporal line-movement patterns of Lumbriculus variegatus (Oligochaeta: Lumbriculidae) in response to the treatments of heavy metal. In Environmental Toxicology; Kungolos, A. G., Brebbia, C. A., Samaras, C. P., Popov, V., Eds.; Southampton and Mykonos, pp 77−92. (39) Kohonen, T. The self-organizing map. Proc. IEEE 1990, 78, 1464−1480. (40) Ross, D. Metabolic basis of benzene toxicity. Eur. J. Haematol. 1996, 57 (60), 111−118. (41) Snyder, R. Xenobiotic metabolism and the mechanism(s) of benzene toxicity. Drug Metab. Rev. 2004, 36 (3−4), 531−547. (42) Valentine, J. L.; Lee, S. S.; Seaton, M. J.; Asgharian, B.; Farris, G.; Corton, J. C.; Gonzalez, F. J.; Medinsky, M. A. Reduction of benzene metabolism and toxicity in mice that lack CYP2E1 expression. Toxicol. Appl. Pharmacol. 1996, 141 (1), 205−213. (43) Lindblom, T. H.; Dodd, A. K. Xenobiotic detoxification in the nematode Caenorhabditis elegans. J. Exp. Zool., Part A 2006, 305 (9), 720−730. (44) Leung, M. C.; Goldstone, J. V.; Boyd, W. A.; Freedman, J. H.; Meyer, J. N. Caenorhabditis elegans generates biologically relevant levels of genotoxic metabolites from aflatoxin B1 but not benzo[a]pyrene in vivo. Toxicol. Sci. 2010, 118 (2), 444−53. (45) Menzel, R.; Rödel, M.; Kulas, J.; Steinberg, C. E. W. CYP35: Xenobiotically induced gene expression in the nematode Caenorhabditis elegans. Arch. Biochem. Biophys. 2005, 438 (1), 93−102. (46) Reichert, K.; Menzel, R. Expression profiling of five different xenobiotics using a Caenorhabditis elegans whole-genome microarray. Chemosphere 2005, 61 (2), 229−237. (47) Roh, J. Y.; Jung, I. H.; Lee, J. Y.; Choi, J. Toxic effects of di(2ethylhexyl) phthalate on mortality, growth, reproduction and stressrelated gene expression in the soil nematode Caenorhabditis elegans. Toxicology 2007, 237 (1−3), 126−133. (48) Thomas, J. H. Rapid birth-death evolution specific to xenobiotic cytochrome P450 genes in vertebrates. PLoS Genet. 2007, 3 (5), e67. (49) Aarnio, V.; Lehtonen, M.; Storvik, M.; Callaway, J. C.; Lakso, M.; Wong, G. Caenorhabditis elegans Mutants Predict Regulation of Fatty Acids and Endocannabinoids by the CYP-35A Gene Family. Front. Pharmacol. 2011, 2, 12. (50) Koop, D. R.; Laethem, C. L.; Schnier, G. G. Identification of ethanol-inducible p450 isozyme 3a (p450iie1) as a benzene and phenol hydroxylase. Toxicol. Appl. Pharmacol. 1989, 98 (2), 278−288. (51) Snyder, R.; Witz, G.; Goldstein, B. D. The toxicology of benzene. Environ. Health Perspect. 1993, 100, 293−306. (52) Roder-Stolinski, C.; Fischader, G.; Oostingh, G. J.; Eder, K.; Duschl, A.; Lehmann, I. Chlorobenzene induces the NF-kappa B and p38 MAP kinase pathways in lung epithelial cells. Inhal. Toxicol. 2008, 20 (9), 813−820. (53) Roder-Stolinski, C.; Fischader, G.; Oostingh, G. J.; Feltens, R.; Kohse, F.; von Bergen, M.; Morbt, N.; Eder, K.; Duschl, A.; Lehmann, I. Styrene induces an inflammatory response in human lung epithelial cells via oxidative stress and NF-kappaB activation. Toxicol. Appl. Pharmacol. 2008, 231 (2), 241−247. (54) Kim, D. H.; Feinbaum, R.; Alloing, G.; Emerson, F. E.; Garsin, D. A.; Inoue, H.; Tanaka-Hino, M.; Hisamoto, N.; Matsumoto, K.; Tan, M. W.; Ausubel, F. M. A conserved p38 MAP kinase pathway in Caenorhabditis elegans innate immunity. Science 2002, 297, 623−626.

(55) Troemel, E. R.; Chu, S. W.; Reinke, V.; Lee, S. S.; Ausubel, F. M.; Kim, D. H. p38 MAPK regulates expression of immune response genes and contributes to longevity in C. elegans. PLoS Genet. 2006, 2, e183. (56) Roh, J. Y.; Choi, J. Cyp35a2 gene expression is involved in toxicity of fenitrothion in the soil nematode Caenorhabditis elegans. Chemosphere 2011, 84 (10), 1356−1361. (57) Eom, H. J.; Ahn, J. M.; Kim, Y.; Choi, J. Hypoxia inducible factor-1 (HIF-1)-flavin containing monooxygenase-2 (FMO-2) signaling acts in silver nanoparticles and silver ion toxicity in the nematode Caenorhabditis elegans. Toxicol. Appl. Pharmacol. 2013, 270 (2), 106− 113. (58) Lim, D.; Roh, J. Y.; Eom, H. J.; Choi, J. Y.; Hyun, J.; Choi, J. Oxidative stress-related PMK-1 P38 MAPK activation as a mechanism for toxicity of silver nanoparticles to reproduction in the nematode Caenorhabditis elegans. Environ. Toxicol. Chem. 2012, 31 (3), 585−592. (59) Kim, D. H.; Feinbaum, R.; Alloing, G.; Emerson, F. E.; Garsin, D. A.; Inoue, H.; Tanaka-Hino, M.; Hisamoto, N.; Matsumoto, K.; Tan, M. W.; Ausubel, F. M. A conserved p38 MAP kinase pathway in Caenorhabditis elegans innate immunity. Science 2002, 297 (5581), 623−626. (60) Troemel, E. R.; Chu, S. W.; Reinke, V.; Lee, S. S.; Ausubel, F. M.; Kim, D. H. p38 MAPK regulates expression of immune response genes and contributes to longevity in C. elegans. PLoS Genet. 2006, 2 (11), e183. (61) Wang, S.; Tang, M.; Pei, B.; Xiao, X.; Wang, J.; Hang, H.; Wu, L. Cadmium-induced germline apoptosis in Caenorhabditis elegans: the roles of HUS1, p53, and MAPK signaling pathways. Toxicol. Sci. 2008, 102 (2), 345−351. (62) Kezhou, C.; Chong, R.; Zengliang, Y. Nickel-induced apoptosis and relevant signal transduction pathways in Caenorhabditis elegans. Toxicol. Ind. Health 2010, 26 (4), 249−256.

8151

dx.doi.org/10.1021/es500608e | Environ. Sci. Technol. 2014, 48, 8143−8151