Environ. Sci. Technol. 2006, 40, 7077-7084
Effects of Temperature on Host-Pathogen-Drug Interactions in Red Abalone, Haliotis rufescens, Determined by 1H NMR Metabolomics E R I C S . R O S E N B L U M , * ,† RONALD S. TJEERDEMA,† AND MARK R. VIANT‡ Department of Environmental Toxicology, University of California Davis, Davis, California, 95616-8588, and School of Biosciences, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, U.K.
The antibiotic oxytetracycline (OTC) has shown immense promise for combating the causative agent of Withering syndrome (WS), a Rickettsia-like procaryote (WS-RLP) that has severely impacted California abalone (Haliotis spp.) populations. Using histology and nuclear magnetic resonance (NMR) spectroscopy based metabolomics, the effects of OTC treatments (10, 20, or 30 days) on WS-RLP infected abalone in seawater temperatures of 13.4 ( 1.2 and 17.3 ( 1.3 °C were investigated over 160 days. The highly efficacious nature of OTC in combating WS-RLP at both temperatures was demonstrated by histology. Metabolomics revealed, however, that the most significant metabolic changes in foot muscle depended upon posttreatment duration, irrespective of treatment and temperature. This was quite unexpected and would have been overlooked using histology alone. Metabolic changes in all animals at both temperatures included decreased levels of amino acids and carbohydrates and elevated taurine, glycinebetaine, and homarine. Subtle metabolic differences between OTC-treated and untreated abalone were observed at 17.3 °C only. These findings provide clear evidence that OTC eradicates WS-RLP which in turn reduces the metabolic decay associated with WS at elevated seawater temperature. Furthermore, this study documents the sequential metabolic changes that occur during pre-clinical WS, and demonstrates the application of metabolic phenotyping for understanding environmental effects on host-pathogendrug interactions.
Introduction Withering syndrome (WS) is a fatal disease affecting wild California abalone populations and the abalone aquaculture industry. It is characterized by a severely shrunken body mass and has been observed in black (Haliotis cracherodii), red (H. rufescens), pink (H. corrugata), green (H. fulgens), and white (H. sorenseni) abalone (1-5). While the diseasecausing pathogen is known to be the Rickettsiales-like * Corresponding author phone: 303-492-3591; fax: 303-492-5894; e-mail:
[email protected]. † University of California Davis. ‡ The University of Birmingham. 10.1021/es061354e CCC: $33.50 Published on Web 10/19/2006
2006 American Chemical Society
procaryote “Candidatus Xenohaliotis californiensis” (WSRLP, 2) studies have shown that infected California red abalone may also require a thermal stress to develop WS (1-3). Although reducing water temperatures may reduce parasite transmission and disease expression, abalone farms typically lack the ability to modulate seawater temperatures, limiting their ability to control the disease. Oxytetracycline (OTC) is approved by the U.S. Food and Drug Administration (FDA) to treat bacterial infections in aquatic species, and has proven effective against Piscirickettsia salmonis, a marine rickettsial pathogen (3). While orally administered OTC (for 14 consecutive days) in medicated feed has reduced the WS-RLP intensity and WS associated mortalities, almost nothing is known about posttreatment recovery times (3). The primary goals of this study were to investigate if metabolic recovery during OTC therapy coincides with WSRLP elimination. Using WS-RLP infected California red abalone we examined the metabolic constituents in foot muscle for 160 days post OTC-treatment at both ambient (13.4 °C) and elevated seawater temperatures (17.3 °C). Histological observations were used to quantify posttreatment levels of WS-RLP infection and atrophy of both foot muscle and digestive gland tissue, while metabolites in abalone foot muscle were evaluated using 1H NMR metabolomics. Earlier investigations using this technique found significant differences in the metabolic profiles obtained from the foot muscles of healthy California red abalone and those with overt signs of WS (4, 5). Studying a time course of the WS-RLP infection using NMR metabolomics complemented our secondary goal, to determine if changes in metabolite biomarkers can be used to screen for pre-clinical WS.
Materials and Methods Abalone Treatments. Abalone used in the 13.4 and 17.3 °C experiments were obtained from The Abalone Farm (Cayucas, CA) and were received on 4/15/02 and 10/23/03 respectively. All abalone were offspring of farm-raised broodstock with an occasional wild parent and reared completely at The Abalone Farm. Upon arrival to the Pathogen Containment Facility at Bodega Marine Laboratory (Bodega, CA), abalone were placed in a 122-L receiving tank supplied with aerated, flowing seawater (ca. 13 °C). Pretreatment at 13.4 °C. Red abalone (n ) 500, ca. 2.5-5 cm long) were cohabitated with 25 RLP-infected animals for 8 weeks within the receiving tank and then all animals were transferred to 6 11 L “treatment” tanks. Water temperature was increased over one week to 17.3 ( 1.3 °C (to allow for increased WS-RLP shedding) and then held constant for 3 months. Donor animals that died during the infection period were replaced. Transfer of WS-RLP was evaluated using both histology and PCR (2, 3). Cohabitation continued until 80% infection prevalence was detected, at which point the WSRLP donors were removed from the system. Water temperatures were lowered to 13.4 ( 1.2 °C for 16 weeks prior to the initiation of the study and animals were acclimated to artificial unmedicated feed. Pretreatment at 17.3 °C. Red abalone (n ) 850, ca. 2.5 cm long) were allowed to acclimate for 2 weeks within the 122 L receiving tank after which 70 RLP-infected and visibly withered abalone were added. Water temperature within the receiving tank was increased to 17.3 ( 1.3 °C over a period of 6 weeks. Cohabitation continued until an infection prevalence of 80% was detected. WS-RLP donors were removed, and abalone were equally divided among 6 11 L VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7077
TABLE 1. Mean ((SD) Values for WS-RLP Infection, Mortality, and Indicators of Abalone Health after a 160-Day Post OTC-Treatment Period at Both 13.4 and 17.3 °C (**p e 0.001)a 13.4 °C
WS-RLP prevalence WS-RLP rating foot muscle atrophy rating digestive gland atrophy rating clinical symptoms of WS (total) mortalities (# mortalities/total) a
17.3 °C
untreated
OTC-treated
untreated
OTC-treated
75% (66) 1.0 ( 0.8** (66) 0.2 ( 0.4** (66) 0.1 ( 0.4** (66) 0 (225) 2.7% (6/225)
0% (75) 0.0 (75) 0.0 ( 0.1 (75) 0.0 ( 0.2 (75) 0 (225) 1.3% (3/225)
90% (81) 1.1 ( 0.6** (81) 0.2 ( 0.4** (81) 0.3 ( 0.6**(81) 1 (339) 5.0% (17/339)
2.5% (81) 0.0 ( 0.2 (81) 0.0 ( 0.2 (81) 0.0 ( 0.2(81) 0 (339) 0% (0/339)
Values in parentheses indicate the number of animals evaluated.
“treatment” tanks. These animals were acclimated for two weeks to the artificial unmedicated feed. OTC Treatment and Posttreatment Sampling. The experiments at each temperature were run independently of each other. In both experiments animals in three of the six tanks were maintained on medicated artificial feed (1.85% active OTC medicated feed at 104 mg OTC/kg biomass), while the other three tanks were maintained on control feed (identical amount, but lacking OTC). OTC concentrations in the medicated and control feed were confirmed prior to the onset of treatments. After 10 days, 1/3 of the animals from each of the 6 “treatment” tanks were moved to 6 new 4 L “posttreatment” tanks and maintained on kelp (Macrocystis pyrifera). Animals remaining in the 6 treatment tanks continued to be fed either the OTC or control diet for an additional 10 days, at which point an equal number as in the first transfer were removed to an additional 6 tanks and maintained on kelp. Animals in the original treatment tanks remained for 10 additional days of OTC treatment. After the 30-day period, all animals were maintained on kelp and housed in a total of 18 “posttreatment” tanks (3 replicates × 3 OTC treatment durations × 2 treatments). The day that animals were transferred out of the OTC-treatment tanks or control-feed tanks was designated as day 0 posttreatment. At days 3, 17, 23, 42, 61, 81, 102, 122, and 160 posttreatment, one abalone from each container was randomly selected and sacrificed. Digestive gland and foot muscle were excised and processed for histology. Foot muscle was also dissected, frozen, and stored at -80 °C for NMR analysis and dry/wet mass ratios. Histology. Digestive gland and foot muscle sections were preserved in Invertebrate Davidson’s Solution overnight and transferred to 70% ethanol. Paraffin-embedded tissue sections (5 µm) were deparaffinized, stained with hematoxylin and eosin, and viewed by light microscopy. Foot muscle atrophy was rated on muscle fiber content: 0 ) muscle fibers comprise >90% of tissue present, 1 ) 76-90%, 2 ) 50-75%, 3 e 50%. Digestive gland WS-RLP infection intensity was based on the number of WS-RLP inclusions in a 20× field of view: 0 ) absent, 1 ) 1-10, 2 ) 11-100, 3 > 100 inclusions. Digestive gland atrophy increases as the amount of connective tissue increases: 0 ) connective tissue comprises 25%. 1H NMR Spectroscopy. Foot muscle was analyzed using 1H NMR spectroscopy, as described previously (4). Briefly, muscle samples were extracted using perchloric acid, lyophilized, and then resuspended in 0.2 M sodium phosphate buffer (in D2O; 1 mM sodium 3-(trimethylsilyl) propionate2,2,3,3-d4 (TMSP)). Extracts were analyzed at 500.11 MHz using an Avance DRX-500 spectrometer (Bruker, Fremont, CA) at 295 K using a 1 D sequence with presaturation of the water resonance. The spectra were phased, baseline corrected, and calibrated (TMSP at 0 ppm) using XWINNMR (Version 3.1; Bruker). Differences between the metabolic profiles of the treatment groups were identified using pattern recognition methods, as described below. 7078
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
Analysis of NMR Data. ProMetab software (Viant 2003) in MATLAB (Version 6.1; The MathWorks, Natick, MA) was used to segment each 1-D spectrum into 0.005-ppm chemical shift bins between 0.2 and 10.0 ppm, and to integrate the total spectral area within each bin (4). Bins from 4.70 to 5.05 ppm (residual water) were excluded. Bin sizes were increased in the regions between 2.715 and 2.73, 2.784-2.794, 2.9242.936, 2.94-2.96, 6.135-6.17, 7.71-7.76, 8.02-8.18, 8.2658.285, 8.57-8.61, and 8.705-8.735 ppm (pH-sensitive resonances), resulting in data matrices of 162 spectra × 1728 bins for each of the 17.3 and 13.4 °C experiments. The total area of each spectrum was normalized to dry tissue mass enabling interpretation of relative metabolite levels on a dry mass basis (4). All elements of the matrices were log transformed and the columns were mean-centered. Principal component analyses (PCA) of the pre-processed NMR datasets (13.4 and 17.3 °C) were conducted using PLS-Toolbox (Version 2.1; Eigenvector Research, Manson, WA) within MATLAB. Statistical Analysis. Relative metabolite concentrations were obtained by integrating the binned data (prior to log transformation), allowing two-way ANOVAs to detect for significant responses to both OTC-treatment and to changes occurring over the posttreatment time course. Significant changes in metabolite levels were determined using the highly conservative Bonferonni corrected value of p ) (0.05/14) ) 0.0036 (Number Cruncher Statistical System 2001 Edition; Kaysville, UT). Metabolite peak areas found to have nonnormal distributions were re-evaluated after log transformation. Foot muscle and digestive gland atrophy as well as WSRLP ratings were evaluated by Kruskal-Wallis one-way ANOVAs after rank transformation; wet-to-dry mass ratios were tested by two-way ANOVAs.
Results RLP Infection and Mortality. Infection ratings increased significantly over the time course only in the untreated animals at 17.3 °C (p < 0.05). In the un-medicated abalone sampled for NMR analysis, 90% of those at 17.3 °C hosted the WS-RLP and 75% at 13.4 °C were WS-RLP infected (Table 1); of these only one animal sampled from the 17.3 °C untreated tanks showed overt signs of late-stage WS. This occurred at 102 days, which is typical for the time period between infection and advanced disease. Within the 13.4 and 17.3 °C untreated tanks 2.7% and 5.0% of the animals, respectively, died during the posttreatment period (Table 1). Histological observations of the digestive gland and postesophagus tissue showed complete clearance of WS-RLP in all OTC-treated animals at 13.4 °C (81/81), and in 97.5% of abalone (79/81) at 17.3 °C. Only 3 mortalities occurred over the entire posttreatment time course within the treated tanks; all occurred in the 13.4 °C experiment. Indicators of Abalone Health. OTC-treated abalone had a lower incidence of foot muscle and digestive gland atrophy than untreated animals (p < 0.001; Table 1), in addition to
FIGURE 1. PCA scores plots of (a) the entire 13.4 °C foot muscle dataset and (b) the entire 17.3 °C foot muscle dataset. Ellipses represent the mean scores of nine animals ( SD (along PC1 and PC2). Gray ellipses represent no OTC treatment and unfilled ellipses represent OTC-treated abalone; numbers indicate the day samples were taken posttreatment. Movements along both PC1 (c and d) and PC2 (e and f) are presented separately. Solid lines represent the movement in OTC-treated animals, dashed lines represent untreated animals. Error bars represent one SD. Both OTC-treated and untreated groups show the major cause of spectral variation within the datasets to be dominated by time. a lower severity of digestive gland atrophy (p < 0.05; Table 1). The severity and prevalence of tissue atrophy were similar in untreated animals at both temperatures, and did not increase over the posttreatment time course (p > 0.05; Table 1). Dry/wet tissue mass ratios were evaluated throughout the 122 days posttreatment as a decrease in this ratio has previously been associated with tissue breakdown during WS (4, 5). Significant decreases occurred for all groups at both temperatures; however, significant differences between OTC-treated and untreated groups, and the cross product of the OTC-treatment and posttreatment sample time were found only in animals sampled from the 17.3 °C tanks (p < 0.001; Figure A in Supporting Information). Effect of Duration of OTC Treatment. Within the OTCtreated animals used for NMR analysis, only two abalone were detected with low-level WS-RLP infections. Differences between spectra from the 10, 20, and 30 day treatment groups therefore reflect the potential metabolic impact arising from OTC treatment duration and do not arise from the pathogen itself. PCAs were initially conducted on OTC-treated animals only, and the PC1 and PC2 scores at both ambient and
elevated temperatures showed no evidence of significant metabolic difference between the three treatment durations (p e 0.05; data not shown). Therefore the three treatment groups within each temperature were pooled, as were the three untreated groups. Subsequent analyses compared OTCtreated and untreated animals at both 13.4 and 17.3 °C, with a 160-day posttreatment time course component. Metabolites in Tissue Extracts. Several of the foot muscle polar metabolites within the 1-D 1H NMR spectrum have been identified previously (4). These include amino acids (Ala, Asp, Gly, Leu, Ile, Lys, Val, Tyr, Trp, Phe, taurine), Rand β-glucose, glycogen, carnitine, a phosphagen, and various glycolytic products and osmoregulators (see Figure B, Supporting Information). PC analyses were conducted separately on the 13.4 and 17.3 °C datasets of muscle extracts (Figure 1a and b). Groupings within these scores plots focus on changes occurring in OTC-treated and untreated animals over time. Each time point is a composite of nine spectra (three replicates per posttreatment time point × three pooled OTC-treatment durations). VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7079
TABLE 2. Factors of Treatment and Time on Individual Metabolite Concentrations (p < 0.0036); Only Metabolites Showing Significant Changes during the Experiment Are Displayed. days posttreatmenta metabolite homarine glycine-betaine hypotaurine taurine phosphoarginine + (arginine) ATP + (ADP) glycogen glucose carnitine tryptophan phenylalanine tyrosine glycine lysine aspartate alanine valine isoleucine leucine
13.4 °C v 62 v 42 v 42 v 62 v 17 V 17 V 17 v 42 v 23 V160 V 17 V 17 V 17 v 23 V160 V 23 V 17 V 42 V 17 V 17
17.3 °C v 42 v 17 V102 v 81 v 17 V122 v 23 V102 v 17 V102 V 17 V 17 v160 v 17 V122 V 160 V 42 V 17 V 17 V 62 V 17 V 17 V 17 V 17 V 17
days × treatment
treatmentb 13.4 °C
17.3 °C
13.4 °C
17.3 °C
U 18% *c OTC 13% OTC 10%
*
OTC 15% OTC 20% OTC 25%
* *
OTC 30%
OTC 19% * OTC 10%
OTC 17% OTC 12%
OTC 16% OTC 12%
a Each arrow and number indicates the direction and day that significant metabolic changes were detected. b Letters indicate whether OTCtreated (OTC) or untreated (U) abalone had larger mean concentrations, and percentage represents the size of this difference (n ) 81). c * Metabolites in which two-way interactions occurred.
Treated vs Untreated at 13.4 °C, Over 160 Days Posttreatment. At 13.4 °C, no significant differences exist between OTC-treated and untreated mean PC1 or PC2 scores (Figure 1c and e). However, significant movement occurs in both PC1 and PC2 scores over the 160-day posttreatment period characterized along PC1 by a progression from positive to negative scores (p e 0.0001) and by a transient movement into positive PC2 space (p e 0.0001; Figure 1c and e). This movement in PC space arises from changes in metabolite concentrations. All identified metabolites, except for hypotaurine, were significantly affected by posttreatment sample time (p < 0.0036; Table 2). Within both OTC-treated and untreated animals significant increases occur in homarine, glycine-betaine, taurine, phosphoarginine (co-incident with arginine, which is at lower concentration), ATP (co-incident with ADP, which is at lower concentration), carnitine, tryptophan, and lysine. These metabolites either increase concentration early on and then level off (e.g., ATP (Figure 2a) and phosphoarginine); increase throughout the entire time course (e.g., homarine (Figure 2c), glycine-betaine, and carnitine); or increase early on then decrease in late stage (e.g., tryptophan and lysine). The remainder of the identified metabolites showed overall decreases during the posttreatment time course, including slow continuous decreases (e.g., glycine (Figure 2e) and the other amino acids) and abrupt decreases (e.g., glycogen (Figure 2g) and glucose). In addition, glycine, valine, and isoleucine show significantly lower levels in untreated versus treated abalone (p e 0.0036; Table 2). No two-way interactions were present between OTC-treatment and posttreatment sample time in either PC score or individual metabolite levels. Treated vs Untreated at 17.3 °C, Over 160 Days Posttreatment. At 17.3 °C, significant differences exist between OTC-treated and untreated mean PC1 (p e 0.01; Figure 1b and d) and PC2 scores (p e 0.0001; Figure 1d and f). For PC2, this occurs on day 102 largely due to movement of infected abalone into negative PC2 space (Figure 1f), which is associated with increasing homarine levels in infected abalone (PC loadings data not shown). The same movement along PC1 observed at 13.4 °C also occurred throughout the posttreatment time at 17.3 °C (Figure 1b). Both treated and untreated abalone exist in positive PC1 during days 3-23 and then move into negative PC1 over days 42-160 (p e 7080
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
0.0001; Figure 1d). Unlike the consistently decreasing trend observed in PC1 scores, PC2 scores tend to migrate around the axis. The PC1 loads plot (Figure 3) indicates the aromatic and aliphatic amino acids, glycogen, carbohydrates, and phosphoarginine are at higher concentrations at earlier time points, while later time points are associated with increases in homarine, glycine-betaine, taurine, and hypotaurine. Significantly lower levels of the following metabolites were detected in untreated versus OTC-treated animals: valine, isoleucine, glycine, lysine, phenylalanine, tryptophan, aspartate, carnitine, ATP, and phosphoarginine. Homarine was the only metabolite at significantly higher concentration in the untreated animals at 17.3 °C (p < 0.0036; Table 2; Figure 2d). Post-hoc tests revealed, however, that this elevation was transient, with higher homarine levels first detected on day 102 but gone by day 160 (Figure 2d). This loss of difference in homarine resulted from late-stage decreases (on day 160) in untreated abalone, which was not observed in the OTCtreated animals, suggesting an effect induced by WS. Significant effects arising from the posttreatment sample time are observed in all identified metabolite levels at 17.3 °C (p e 0.0036; Table 2). Within both OTC-treated and untreated animals significant increases occurred in homarine, glycine-betaine, hypotaurine, taurine, phosphoarginine, ATP, and carnitine. Increases in ATP (Figure 2b) and phosphoarginine were observed through day 62; however, unlike the leveling off observed in the 13.4 °C abalone, both metabolites showed significant decreases on day 102. The same trend was observed in glycine-betaine and taurine, in which levels initially increased followed by significant decreases on day 122. Increased homarine (Figure 2d) and carnitine levels persisted throughout the entire time course. While many metabolites at 13.4 and 17.3 °C are affected by both the single factors of posttreatment sample time and OTC treatment, post-hoc tests revealed that the direction of change in these metabolites during the posttreatment period is similar in both treated and untreated animals. At 17.3 °C, significant two-way interactions were observed in hypotaurine, phosphoarginine, carnitine, tyrosine, and lysine (Table 2). These effects are the result of ordinal changes (ones in which the relative differences that exist between infected and uninfected significantly increase over time).
FIGURE 2. Changes in metabolite concentrations during the posttreatment period in: (a) 13.4 °C ATP (+ADP); (b) 17.3 °C ATP (+ADP); (c) 13.4 °C homarine; (d) 17.3 °C homarine; (e) 13.4 °C glycine; (f) 17.3 °C glycine; (g) 13.4 °C glycogen; and (h) 17.3 °C glycogen, over the entire time course. Solid lines represent the movement in OTC-treated animals, dashed lines represent untreated animals, and error bars are one SD. Asterisks indicate specific posttreatment sample days where OTC-treated and untreated metabolite concentrations differed significantly (p e 0.0036). PCA was also performed on the NMR spectra from treated and untreated abalone (at 17.3 °C) from the last three posttreatment sampling times (days 102, 122, and 160), allowing the biochemical differences between treated and untreated animals to be examined more extensively (Figure 4). Significant differences exist between treated and untreated spectra along PC1 (p < 0.0001). The PC1 loads plot confirms that OTC-treated abalone exhibit higher levels of amino acids, ATP, and phosphoarginine, while untreated animals have higher homarine and glycine-betaine (Figure 4).
Discussion Disease Progression in Untreated Abalone. Numerous studies have shown that elevated water temperature hastens WS development and increases mortality in infected abalone (1-5). Furthermore researchers have shown that WS-RLP free red abalone could be maintained at 18 °C for >1 year with no mortalities or clinical signs of WS (2). The NMR metabolic findings here are consistent with these observations. Metabolic changes were observed in both treated and untreated animals at 13.4 °C; however, they were of similar VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7081
FIGURE 3. PCA loads plots for the entire 17.3 °C dataset that identifies which metabolites exhibit the largest concentration changes through the study. Metabolites with positive PC1 loadings occur at higher concentration in the samples with positive PC1 scores (Figure 1b), namely those at earlier time points.
FIGURE 4. (a) PCA scores plots of 17.3 °C OTC-treated and untreated abalone from posttreatment sample days 102-160 only. Ellipses represent the mean scores of 9 animals ( SD (along PC1 and PC2). Gray ellipses represent no OTC treatment and unfilled ellipses represent OTC-treated abalone. (b) PC1 loads plot for 17.3 °C posttreatment days 102-160 shows that positive PC1 space is highly weighted by all amino acids, ATP (+ADP), and phosphoarginine (+arginine), while negative PC1 loadings contain homarine and glycine-betaine. magnitude. While changes also occurred in both treated and untreated abalone at 17.3 °C, those in untreated animals were significantly more pronounced. The reductions in carbohydrates and amino acids, and increases in secondary osmoregulators in untreated animals at both temperatures are similar to the metabolic changes pre7082
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
viously identified in both food-limited and WS-affected animals (4, 5). Metabolic Changes in OTC Treated Animals. OTC effectively reduced the WS-RLP burdens and WS-induced mortalities at a temperature known to favor its development (17.3 °C). Successful treatment of WS requires both eliminating the pathogen and metabolic recovery during posttreatment grow-out. In fact, had this study and the success of treatment been based solely on histology, the unexpected finding of “adverse” metabolic changes in OTC-treated animals would have been overlooked. The metabolic changes observed in OTC-treated abalone are similar to those reported during starvation (6-9). For example observations in starved disk abalone found significant losses of glycogen and glycine (9), while starved red abalone showed decreases in glycogen (99%) and glycine (98%) (4, 7). In this study we found 97% and 90% reductions in treated animals’ glycogen and glycine levels at 17.3 °C, and 90% and 75% reductions at 13.4 °C. Furthermore, precipitous drops occurred in glucose and amino acid levels of both treated and untreated animals. Roberts et al. observed that starved abalone survive for several weeks by utilizing protein for energy (7). During the 160-day period we observed decreasing amino acids and carbohydrates, as muscle carnitine increased, an effect previously associated with increased protein turnover (4). Overall, this suggests that all abalone in the current experiment showed metabolic changes that are consistent with dietary limitation, discussed further below. Reductions in carbohydrates and amino acids observed in abalone held at 17.3 °C were characterized by dramatic declines up to day 50, while the decreases at 13.4 °C were more gradual over the entire time course (Figure 4e and f). Furthermore, at 13.4 °C, ATP, phosphoarginine, glycinebetaine, and taurine increased over the entire study, but at 17.3 °C late-stage declines were observed (Table 2) suggesting that a depletion of fuel for cellular energy occurred after 102 days at 17.3 °C (but not 13.4 °C). This is commensurate with the onset of advanced WS, and supports previous findings that metabolic rates respond to temperature (10). Therefore, one explanation for the increased severity of WS at elevated temperatures is increased metabolic rate leading to increased nutritional requirements during a period when animals suffer from a loss of digestive gland function. Changes in Osmoregulators and Secondary Metabolites. Although the main role of free amino acids in marine osmoconformers is in osmoregulation (11, 12), they are also
used as an energy source during periods of prolonged starvation (12). The initial response to depletion of free amino acids observed here was a simultaneous increase in less metabolically active osmoregulators such as taurine, glycinebetaine, and homarine. These initial increases were followed by significant declines in glycine-betaine and taurine in untreated abalone at 17.3 °C, and subsequently by a decrease in homarine, which suggests that even these compounds are metabolized during prolonged starvation. Viant et al. observed a similar trend in red abalone. In early stages of WS both homarine and glycine-betaine were elevated in red abalone, but in later stages of the disease glycine-betaine decreased while high levels of homarine persisted (4). This led to the idea that homarine could serve as a preclinical biomarker of disease. However, until now, little was known about the temporal changes in homarine over the time course of WS. Our observation that homarine increases in treated (WS-RLP negative) animals indicates that changes in homarine are not specific to pathogen. Both treated and untreated abalone continued to show metabolic changes that resembled starvation and WS, a possible response to damaged digestive capabilities that occurred prior to OTC-treatment. This supports the hypothesis that WS arises due to compromised digestive gland function and that clinical symptoms of WS are simply physiological starvation. WS-RLP infects secretory and absorptive cells within the digestive gland resulting in a loss of enzyme production and nutrient uptake. Researchers have shown that abalone with WS experience these digestive gland alterations prior to the depletion of cellular fuels (2, 13). Thus, the previously observed increase in homarine in WS-RLP infected abalone and decrease in food-limited abalone most likely represent distinct phases during the same physiological response, the balance between energy metabolism and the preservation of osmoregulation. Furthermore, the initial transient increase in homarine observed in WS-RLP infected abalone maintained at 17.3 °C, along with increased homarine levels in both the infected and uninfected animals at 13.4 °C, raises questions about the specificity of homarine as a biomarker to differentiate WS from other environmentally relevant stressors such as food limitation. Digestive Gland Function in Treated and Untreated Animals. Our histological data did not show evidence of digestive gland atrophy in the OTC-treated animals beyond day 3. This, along with reduced mortalities in medicated abalone, indicates that treatment occurred prior to the terminal stages of WS, which suggests that metabolic recovery from WS-RLP-induced digestive gland damage could potentially occur after a period of time longer than that investigated here. However, diet can affect both the chemical composition of marine invertebrate tissues (8, 14) and the activities of digestive enzymes (16); e.g., kelp-fed abalone adjusted their polysaccharide-digesting enzymes to utilize this substrate (15). Therefore a more plausible explanation for the metabolic decay that we observed in both treated and untreated abalone throughout the study is due to the change in diet from an artificial food to kelp. A second control group, in which non WS-RLP exposed abalone were conditioned on non-medicated feed, would be needed to test this hypothesis. Metabolomics as a Complementary Tool to Histology. 1H NMR metabolomics allowed both the molecular documentation of the increased severity of WS in elevated water temperatures and the sequential metabolic changes that occurred throughout the time course of pre-clinical WS. In addition, our documentation of the metabolic differences between OTC-treated and untreated abalone at 17.3 °C confirms that WS was progressing in untreated animals, and that the changes in these untreated abalone were not only due to elevated temperature. The highly significant difference in the muscle metabolic phenotypes of treated and untreated
abalone from 102 to 160 days, along with histological observations, gives clear evidence that OTC eradicates WSRLP which in turn reduces the metabolic decay associated with WS. Furthermore this study demonstrates the use of metabolic phenotyping in furthering the understanding of temperature on drug-pathogen-host interactions using the OTC, WS-RLP, and abalone model system. The detection of metabolic differences between treated and untreated animals at 17.3 °C prior to the onset of clinical WS suggests the use of metabolomics as not only a sensitive tool for disease detection in marine organisms but also as a complementary tool to histology. These conclusions support the further development and integration of metabolomics as a complementary tool for identifying and characterizing pathological events in aquatic species, and suggest the usefulness of such methodology in the aquaculture industry for understanding drug-host-pathogen interactions and the optimization of drug treatments.
Acknowledgments We thank the Abalone Farm for providing abalone. This research was funded by the National Sea Grant College Program, National Oceanic and Atmospheric Administration, U.S. Department of Commerce, under grants NA06RG0142 (project R/A-115) and NA04OAR4170038 (project R/A-122B), and by the California Department of Fish and Game. M.R.V. thanks the Natural Environment Research Council, U.K., for an Advanced Fellowship (NER/J/S/2002/00618). The views expressed herein are those of the authors and do not necessarily reflect the views of NOAA or any of its subagencies.
Supporting Information Available Figure A: dry-to-wet mass ratios of OTC-treated and untreated abalone foot muscle tissue over 122 days post OTC-treatment. Figure B: representative one-dimensional 1H NMR spectrum of a healthy abalone foot muscle.
Literature Cited (1) Moore, J. D.; Robbins, T. T.; Friedman, C. S. Withering syndrome in farmed red abalone, Haliotis rufescens: Thermal induction and association with a gastrointestinal Rickettsiales-like prokaryote. J. Aquat. Anim. Health 2000, 12, 26-34. (2) Braid, B.; Moore, J. D.; Robbins, T. T.; Hedrick, R. P.; Tjeerdema, R. S.; Friedman, C. S. Health and survival of red abalone, Haliotis rufescens, under varying temperature, food supply and exposure to the agent of withering syndrome. J. Invert. Pathol. 2005, 89, 219. (3) Friedman, C. S.; Trevelyan, G.; Robbins, T. T. Development of an oral administration of oxytetracycline to control losses due to withering syndrome in cultured red abalone Haliotis rufescens. Aquaculture 2003, 224, 1-23. (4) Viant, M. R.; Rosenblum, E. S.; Tjeerdema, R. T. NMR-based Metabolomics: A powerful approach for characterizing the effects of environmental stressors on Organism Health. Environ. Sci. Technol. 2003, 37, 4982-4989. (5) Rosenblum, E. S.; Viant, M. R.; Braid, B. M.; Moore, J. D.; Friedman, C. S.; Tjeerdema, R. S. Characterizing the metabolic actions of natural stresses in the California red abalone, Haliotis rufescens using 1H NMR metabolomics. Metabolomics 2005, 1, 199-209. (6) Carefoot, T. H.; Quian, P. Y.; Taylor, B. E.; West, T.; Osborne, J. M. Effect of starvation on energy reserves and metabolism in the Northern abalone, Haliotis kamtschatkana. Aquaculture 1993, 118, 15-325. (7) Roberts, R. D.; Lapworth, C.; Barker, R. J. Effect of starvation on the growth and survival of post-larval abalone (Haliotis iris). Aquaculture 2001, 200, 323-338. (8) Watanabe, H.; Yamanaka, H.; Yamakawa, H. Changes in the content of extractive components in disk abalone fed with marine algae and starved. Nippon Suisan Gakk 1993, 59, 20312036. (9) Takami, H.; Yamanaka, H.; Nakano, H. Survival and physiological stress of juvenile disk abalone Haliotis discus discus during longterm starvation. Fish. Sci. 1995, 61, 111-115. VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7083
(10) Paul, A. J.; Paul, J. M. Respiration rate and thermal tolerances of pinto abalone Haliotis kamtschatkana. J. Shellfish Res. 1998, 17, 743-745. (11) Schoffeniels, E. Adaptation with respect to salinity. Biochem. Soc. Symp. 1976, 41, 179-204. (12) McCoid, M.; Miget, R.; Finne, G. Effect of environmental salinity on the free amino acid composition and concentration in penaeid shrimp. J. Food Sci. 1984, 49, 327-330. (13) Gardner, G. R.; Harshbarger, J. C.; Lake, J.; Sawyer, T. K.; Price, K. L.; Stephenson, M. D.; Haaker, P. L.; Togstad, H. A. Association of prokaryotes with symptomatic appearance of withering syndrome in black abalone Haliotis cracherodii. J. Invert. Pathol. 1995, 66, 111-120.
7084
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
(14) Mai, K.; Mercer, J. P.; Donlon, J. Comparative studies on the nutrition of two species of abalone, Haliotis tuberculata and Haliotis discus. Aquaculture 1994, 128, 115-130. (15) Cook, P.; Coyne, V. The role of bacteria in the digestion of seaweed by the abalone Haliotis midae Erasmus. Aquaculture 1997, 155, 381-390. (16) Harris, R.; Samain, J.; Martin-Jezequel, V. Effects of algal diet on digestive enzyme activity in Calanus helgolandicus. Mar. Biol. 1986, 90, 353-361.
Received for review June 6, 2006. Revised manuscript received August 29, 2006. Accepted September 6, 2006. ES061354E
