Energy metabolism and metabolic depression during exercise in Callinectes sapidus, the Atlantic blue crab: effects of the bacterial pathogen Vibrio campbellii

The Atlantic blue crab, Callinectes sapidus (Rathbun), an important species for commercial and recreational fishing(Whitaker et al., 1998), lives in estuarine and other coastal environments along the Atlantic and Gulf coasts of North America. In these habitats, seasonal and diurnal fluctuations of salinity, temperature, oxygen and pH(deFur et al., 1990; Mikulski et al., 2000) can profoundly impact physiological processes. Such changes in environmental conditions not only alter fundamental processes of osmoregulation,respiration, circulation, digestion and acid–base balance (for a review,see Towle and Burnett, 2006),but also immune defenses (Newman and Feng,1982; Le Moullac and Haffner,2000; Holman et al.,2004; Burgents et al.,2005a; Macey et al.,2008). Interestingly, new evidence suggests that mounting an immune response against bacterial infection interferes with circulation, and consequently respiration, within crustaceans(Burnett et al., 2006; Scholnick et al., 2006).

Bacteria of the genus Vibrio occur naturally in marine environments and are frequently associated with the exoskeleton and hemolymph of blue crabs (Colwell et al.,1975; Tubiash et al.,1975; Sizemore and Davis,1985; Welsh and Sizemore,1985). Vibrio spp. are characterized as opportunistic crustacean pathogens and can cause mortality in the blue crab(Krantz et al., 1969; Johnson, 1976; Welsh and Sizemore, 1985). Experimentally introduced bacteria or other foreign particles are rapidly cleared (within 1 h) from crustacean hemolymph(McCumber and Clem, 1977; Merrill et al., 1979; White and Ratcliffe, 1982; Adams, 1991; Martin et al., 1993; van de Braak et al., 2002; Holman et al., 2004). The gills are particularly important sites for accumulation and subsequent removal of these bacteria or particles (Smith and Ratcliffe, 1980b; White and Ratcliffe, 1982; Martin et al., 1993; Martin et al.,2000).

Adult blue crabs, like many portunid crabs, are most noted for their proficient swimming capabilities but they are also capable of walking along the bottom, most often laterally, using three paired walking appendages(Spirito, 1972). Tagging studies have shown that blue crabs typically move 50–900 m within estuaries on a daily basis, but may migrate seasonally over larger distances of up to 800 km, using a combination of swimming high in the water column on flood tides and walking along the bottom on ebb tides(Hines, 2007). However, there are currently no data available on the energetics of walking in the blue crab,as studies of exercise physiology in blue crabs have been largely focused on swimming (Booth et al., 1982; Booth and McMahon, 1985).

Exercise by either swimming or walking in decapod crustaceans is, as in most animals, associated with an increase in the metabolic demand for oxygen(McMahon, 1981). There is anywhere from a two- to a sevenfold elevation of oxygen uptake during increased activity, made possible by increases in ventilation rate, heart rate and cardiac output (Herreid et al.,1979; McMahon et al.,1979; Wood and Randall,1981; Booth et al.,1982; Taylor,1982; Full and Herreid,1984; Houlihan and Innes,1984; Houlihan et al.,1984; Houlihan et al.,1985; Hamilton and Houlihan,1992). Elevated oxygen uptake persists for some time after activity is completed. This is characterized as excess post-exercise oxygen consumption (EPOC), which is used to recharge depleted energy stores and to dispose of anaerobic end products generated during activity(McMahon, 1981; Gaessar and Brooks, 1984; Gleeson and Hancock, 2001; Hancock and Gleeson, 2002). Highly aerobic crustaceans, such as the blue crab, typically have a small oxygen deficit recovery (e.g. Booth et al.,1982) and oxygen uptake often returns to pre-activity levels within 20–45 min (Herreid et al.,1979; Booth et al.,1982; Full and Herreid,1983; Houlihan and Innes,1984; Hamilton and Houlihan,1992). However, in organisms more reliant on anaerobic pathways to fuel energy demands during activity, this recovery may take as long as 5–24 h (McDonald et al.,1979; McMahon et al.,1979; Wood and Randall,1981). Of course, this also partially depends on the duration and intensity of the activity performed.

It is perhaps surprising that aerobic metabolism in shrimp(Scholnick et al., 2006) and in blue crabs (Burnett et al.,2006) decreases profoundly following injection of a sublethal dose of bacteria. This metabolic depression is especially interesting since the gills are implicated in the immune response(Smith and Ratcliffe, 1980a; White et al., 1985; Martin et al., 2000) during which time their respiratory function declines(Burnett et al., 2006). The decline is rapid and, in the case of the shrimp, it persists for 24 h. We were interested in investigating this phenomenon further in the blue crab and determining whether bacterial injection impaired the response of the blue crab to exercise.

Adult male blue crabs, Callinectes sapidus (60–140 g), were collected locally in Charleston, SC, USA and held at the Hollings Marine Laboratory in well-aerated, recirculating seawater (30‰;20–22°C; pH 7.5–8.5) on a 12 h:12 h dark:light cycle for a minimum of 3 days but no longer than 2 weeks before use in experiments. Crabs were fed daily with squid that were previously frozen, but food was withheld at least 1 day prior to and during each experiment. All experiments were performed in filtered seawater (30‰ salinity).

Crabs were prepared for injection as described by Holman et al.(Holman et al., 2004). At least 24 h before injecting bacteria (referred to as a bacterial challenge), a 1 mm hole was drilled through the carapace directly over the heart as a port for injection of bacteria or saline into the ventricle. Injection of bacteria directly into the ventricle ensured rapid and even distribution of the injected bacteria by the circulatory system as described by Macey et al.(Macey et al., 2008). Two similar holes were drilled over the pericardial sinus adjacent to the heart for sampling hemolymph from the pericardial space. A thin layer of latex rubber was secured over the drilled holes with cyanoacrylate glue and acted as a diaphragm to allow for injection or sampling of hemolymph with a needle. Each crab was placed individually in a small container with 1.5 l well-aerated 30‰ seawater and a thin layer of gravel.

To assess crabs for a pre-existing systemic infection with culturable bacteria, 150 μl of hemolymph was sampled from the pericardial sinus,diluted 1:10 in sterile 10 mmol l^–1 Hepes-buffered 2.5% NaCl(hereafter referred to as saline), suspended in marine agar and overlaid onto a sterile tryptic soy agar (TSA; Difco, Becton Dickenson and Co., Sparks, MD,USA) microbial culture plate supplemented with 2% NaCl. These plates were incubated for 24 h at 25°C and then examined for colony forming units(CFU) of bacteria. Only those crabs with no culturable bacteria present in the hemolymph, as detected by this assay, were used for experimentation.

The bacterial pathogen used was V. campbellii 90-69B3, stably transfected with the Vibrio-derived plasmid EVS146 that expresses green fluorescent protein and resistance to antibiotics kanamycin and chloramphenicol (Dr E. Stabb, University of Georgia). The parental strain of V. campbellii 90-69B3 was isolated from diseased shrimp Litopenaeus vannamei by D. Lightner and L. Mahone (University of Arizona). V. campbellii is an opportunistic pathogen, and its pathogenicity in L. vannamei is comparable to that of other marine Vibrio spp. that are commonly isolated from marine waters(Mikulski et al., 2000). V. campbellii was selected for the present study because its distribution, inactivation and elimination from tissues of shrimp and crabs have been extensively documented (Holman et al., 2004; Burgents et al.,2005b). Stability of the plasmid and growth characteristics of the transfected strain have been described previously(Burgents et al., 2005b).

One day prior to injection, V. campbellii was streaked from a frozen glycerol working stock onto a TSA plate containing 2.5% NaCl, 100 μg ml^–1 kanamycin A (Sigma-Aldrich, St Louis, MO, USA) and 5μg ml^–1 chloramphenicol (Sigma-Aldrich) and incubated at 25°C for 24 h. Bacteria from this plate were then transferred to a tube containing sterile 10 mmol l^–1 saline using a sterile inoculating loop and the bacterial concentration was adjusted to an optical density of 0.1±0.005 at 540 nm, which equals a bacterial concentration of 10⁸ CFU ml^–1(Mikulski et al., 2000; Burgents et al., 2005b). The prepared bacterial suspension was adjusted with sterile saline to obtain the desired concentration for injection. Injection volumes of saline or saline containing Vibrio never exceeded 300 μl.

Lethality of V. campbellii 90-69B3 in the blue crab was assessed by performing three 48-h LD₅₀ bacterial challenges. In each replicate experiment, crabs (60–125 g; N=7 for each dose) were injected with saline (control) or one of three doses(2.5×10⁴, 2.5×10⁵ or 2.5×10⁶ CFU g^–1 crab) of V. campbellii. The inoculum of bacteria for each injection was prepared as described above to obtain a concentration of 10⁸ CFU ml^–1 saline. To prepare the lowest dose, this bacterial suspension was diluted 1:3 with sterile saline to obtain a concentration of 2.5×10⁷ CFU ml^–1. This concentration, when injected as 1 μl g^–1 crab, delivered a dose of 2.5×10⁴Vibrio g^–1 crab. Animals receiving the mid-range dosage were injected with 2.5 μl g^–1 crab of the original bacterial suspension (10⁸CFU ml^–1 saline) which delivered the desired dose of 2.5×10⁵ CFU g^–1 crab. To prepare the highest bacterial challenge dose, 10 ml of the original bacterial suspension of 10⁸ CFU ml^–1 saline was pelleted by centrifugation(2000 g for 5 min), the supernatant discarded, and the pellet resuspended in 1 ml of saline to obtain a concentration of 10⁹ CFU ml^–1. Each animal was injected with 2.5 μl g^–1 crab of the concentrated bacteria to deliver a final dose of 2.5×10⁶ CFU g^–1 crab. Control animals received an injection of 1 μl g^–1 crab of sterile saline.

In each replicate trial, crabs were injected directly into the ventricle with the appropriate dose of saline containing Vibrio or saline(control) within their individual containers ensuring that the gill chambers stayed fully submerged while doing so. The water within each container was changed every 24 h during the experiment. Mortality was recorded at 0, 2, 4,8, 12, 24 and 48 h following bacterial challenge.

Based on our determination of an LD₅₀ dose of V. campbellii (methods described above and reported in Results), we used a sublethal dose of 2.5×10⁴ CFU g^–1 crab for subsequent experiments. This is equivalent to a circulating dose of 1×10⁵ CFU ml^–1 hemolymph, assuming a hemolymph volume of 25 ml 100 g^–1 body mass(Gleeson and Zubkoff,1977).

Oxygen uptake following the injection of a sub-lethal dose of V. campbellii was measured at 25°C using a flow-through respirometry system. The P_O2 of seawater entering and leaving a sealed 1.8 l circular, plexiglass respirometry chamber (17 cm wide,8 cm high) lined with sterile gravel was measured every 10 s using an O₂ electrode (Yellow Springs Instruments, Yellow Springs, OH, USA)and a Sable System (Las Vegas, NV, USA) data acquisition system. Seawater entering the chamber was gassed with 30% O₂ (31 kPa) to ensure that the partial pressure within the chamber did not drop below 80% air saturation(16.5 kPa O₂) when a crab was present in the chamber. Flow rates through the chamber usually ranged between 80 and 150 ml min^–1.

Typically, a crab was placed in the respiratory chamber, then the chamber was sealed, covered and the animal left undisturbed for 2 h. After this time,the chamber was opened and the crab was injected with saline (control) or saline containing V. campbellii (2.5×10⁴ CFU g^–1 crab). The chamber was then resealed and O₂uptake was measured continuously for 4 h. Afterwards, the crab was removed from the respiratory chamber and held overnight in an individual container of well-aerated 30‰ seawater at 25°C. Approximately 22 h after injection, the crab was returned to the sealed respiratory chamber and left undisturbed for 2 h. A final measurement of O₂ uptake over 30 min was recorded.

In a separate group of crabs, changes in anaerobic metabolism were assessed by measuring hemolymph lactate levels over time after injection of saline or V. campbellii. Hemolymph (30 μl) was sampled from the pericardial space of each crab at 30 min, 2 h, 4 h and 24 h following injection and immediately deproteinized in ice-cold 12% perchloric acid (1:10 dilution). The samples were then centrifuged and lactate concentrations measured enzymatically using a NADH-linked colorimetric assay (Sigma Technical Bulletin No 862-UV; lactate dehydrogenase, Sigma #L2500) and modified according to Graham et al. (Graham et al.,1983).

Walking was initiated and sustained on a variable speed treadmill equipped with a flow-through respirometry system. The treadmill is enclosed in a 4.3 l clear, watertight plexiglass chamber and powered with an external DC motor. The crabs were able to move freely about in the 18 cm×17 cm×7 cm high compartment above the tread. In each experiment, crabs were injected with saline (control) or saline containing V. campbellii, placed within the treadmill chamber, and allowed to rest for 1 h. Animals then walked on the treadmill at a speed of 8 m min^–1 for 30 min. After this activity was completed, each crab was allowed to recover for 3 h.

Three different sets of experiments employed this apparatus and activity regime. In the first set of experiments, O₂ uptake was measured in 14 crabs (N=7 each for saline-injected and Vibrio-injected crabs) using flow-through respirometry. In the second, hemolymph was sampled for lactate analysis from a different set of 18 crabs (N=9 for each injection treatment) at designated timepoints during the activity regime. In the final set of experiments, a total of 84 crabs were sampled at designated times to quantify whole body lactate and muscle metabolites.

O₂ uptake was measured at 25°C in the treadmill respirometer as described above. Flow rates through the chamber ranged between 80 and 200 ml min^–1. Instantaneous measurements of O₂ uptake can be obtained by the difference in O₂ pressures in the water flowing into and out of the chamber at a known water flow rate when the rate of O₂ uptake is constant, as observed under resting conditions. However, when the oxygen uptake rate of the animal changes to a new steady-state level, there is a delay in the establishment of a new steady-state P_O2 of the water exiting the respirometer. The approach to a new steady-state level is an exponential function of time. This delay or `washout time' has been described by Frappell et al. (Frappell et al., 1989)and was accounted for in the present study by applying the Z transformation(Bartholomew et al., 1981). In order to minimize any washout effects associated with the transition from resting O₂ uptake to that which occurs during walking and the early stages of recovery, water flow through the respirometer was increased during activity. Increasing the flow rate during the activity period decreases the washout time and insures that the O₂ pressure within the chamber stays reasonably constant and does not drop below 80% air saturation. In fact, P_O2 of the water exiting the treadmill chamber during walking changed very little from resting values as a result of increasing the water flow rate.

Each crab was placed in the treadmill chamber and left undisturbed for 2 h,during which time well-oxygenated seawater passed through the chamber. The lid of the chamber was then removed and without removing the crab from the water,saline (control) or saline containing V. campbellii was injected directly into the ventricle though the injection port. The chamber was resealed and the animal was allowed to rest in the treadmill chamber for 1 h. Each crab then walked on the treadmill for 30 min as described above. Subsequently, the treadmill was stopped and O₂ uptake for each animal was recorded during a 3 h recovery period. Excess post-exercise oxygen consumption (EPOC) was calculated by integrating the area described by the elevation of O₂ uptake above the base-line value from 0 to 80 min after exercise, by which time O₂ uptake had returned to base-line levels.

Hemolymph lactate concentration was measured on a separate group of crabs injected with saline (control) or saline containing V. campbellii and exercised according to the same regimen. Each crab was injected with saline(control) or saline containing V. campbellii into the ventricle while remaining submerged in its individual container. Hemolymph (30 μl) was sampled from the pericardial sinus after 1 h at rest in the treadmill chamber,immediately after 30 min of walking, and at 1, 2 and 3 h during the recovery period. Hemolymph samples were immediate deproteinized and lactate was measured as described above.

An additional 84 crabs were injected, exercised and rested following the same regimen described above. To measure metabolites in the whole crab and separately in muscle tissues at different times before, during and after the exercise period, crabs were removed rapidly from the apparatus at the appropriate time and plunged into liquid nitrogen to freeze all tissues. Crabs were frozen at the following times: 1 h after injection and prior to walking,immediately after 30 min of walking on the treadmill, and 5, 30, 60, 120 and 180 min after walking. Frozen crabs were removed from the liquid nitrogen and cut (anterior to posterior) into two equal halves and stored at–80°C until analysis. Whole body lactate and succinate were measured in one half of the crab. Lactate, succinate, phosphagens and adenylates were measured in the leg muscle (described below).

To measure whole body lactate and succinate one half of each crab was cryogenically homogenized using a Teflon disk mill (W. P. Law, Johns Island,SC, USA). The intact frozen crab tissue was broken up into smaller pieces with a sledgehammer and placed in a Teflon disk mill that was pre-cooled overnight in a XLC1211 liquid N₂ vapor-phase freezer (Chart Industries-MVE,Marietta, GA, USA) and powdered using a TS-250X Teflon Disk Mill Shaker(Siebtechnik GMBH, Mülheim an der Ruhr, Germany). The powdering process took less than 2 min and transformed the frozen tissue into a uniformly homogeneous particulate powder (Pugh et al., 2007). The powder was transferred to a small container for storage at –80°C until analysis. A known mass (∼0.4 g) was transferred to a tube containing three parts ice-cold 0.9 mmol l^–1 perchloric acid (HClO₄) and homogenized while on ice for 30 s at 30,000 r.p.m. using a Polytron PT 2100 homogenizer system(Westbury, NY, USA). Samples were transferred to 2 ml microcentrifuge tubes and centrifuged for 10 min at 11,000 g at 4°C. The supernatant was removed from the microcentrifuge tube and the remaining pellet was washed with 0.2 mmol l^–1 HClO₄ using one-third the volume of the initial extraction, and centrifuged as stated above. The pooled supernatant from the two previous steps was neutralized with a volume of 3.75 mmol l^–1 potassium carbonate(K₂CO₃) equal to 10% of the total amount of perchloric acid used in both extraction steps. The samples were shaken vigorously and centrifuged a final time for 10 min at 11,000 g at 4°C and the resulting supernatants were analyzed for lactate and succinate. Lactate was assayed as described above for hemolymph. Succinate concentrations were measured in a subset of samples using a commercially available NADH-linked colorimetric assay kit (R-Biopharm, Marshall, MI, USA).

Statistical analyses were performed using SigmaStat 3.0. All data are given as means ± s.e.m. Significance level is α=0.05 for all tests.

The effect of bacterial challenge dose on the percentage survival of blue crabs over 96 h was analyzed by two-way ANOVA. Post-hoc comparisons between crabs challenged with bacteria and crabs injected with saline (control group) were performed using the Holm–Sidak (HS) method. 48 h LD₅₀ values and confidence intervals for each replicate experiment were calculated by the trimmed Spearman–Karber method(Hamilton et al., 1977). This particular method is a non-parametric statistical procedure that estimates the trimmed mean LD₅₀ and associated 95% confidence intervals from a distribution of the log₁₀ of percentage mortality over the range of exposure concentrations after a fixed exposure time.

For experiments with resting crabs, oxygen uptake and hemolymph lactate data were analyzed using one-way or two-way repeated measures ANOVA (RM-ANOVA)to allow comparisons between treatment groups (saline vs Vibrio) and with time. The oxygen uptake data were not normally distributed so the data were ranked and the ANOVA analysis performed on the ranked data. Hemolymph lactate data were transformed using the square root to achieve normality. Post-hoc comparisons between treatments (saline vs Vibrio)at individual timepoints were performed using the HS method.

For experiments with exercised crabs, O₂ uptake data were compared between treatments (saline vs Vibrio) and over time using repeated measures (RM) ANOVA. Data were log₁₀ transformed to achieve normality. If a statistically significant difference between treatments was found, post-hoc comparisons of individual timepoints between treatments were performed using the HS method. A significant difference in time within treatment resulted in a post-hoc comparison of resting oxygen uptake values with values measured during and after activity within each treatment using the HS method.

All lactate, succinate and muscle metabolite data were analyzed using two-way ANOVA to compare between treatments (saline vs Vibrio) and between timepoints within each treatment. Values for whole body lactate and hemolymph lactate were log₁₀ transformed to achieve normality. Post-hoc comparisons using the HS method were employed when a significant difference between treatments was found. When a significant difference was indicated between timepoints, post-hoc analysis using the HS method was used to compare pre-activity levels with values measured during recovery within each treatment.

V. campbellii was lethal to C. sapidus in a dose-dependent manner (Fig. 1). There was a significant effect of injection dose on the survival of blue crabs over time (two-way ANOVA; P<0.001). Only 1 out of 21 saline-injected crabs died during experimental trials. All mortality occurred within 48 h following injections. No significant difference was found in percentage survival between crabs receiving saline injections and those receiving Vibrio injections of 2.5×10⁴ CFU g^–1 crab (Fig. 1). However, percentage survival was significantly lower in crabs receiving an injection of 2.5×10⁵ CFU g^–1crab after 24 h compared with control crabs (HS; P<0.02). Crabs injected with 2.5×10⁶ CFU g^–1 crab exhibited a significantly lower percentage survival after only 2 h compared with crabs receiving saline injections (HS; P<0.001). 48 h LD₅₀values were analyzed using the trimmed Spearman–Karber method(Hamilton et al., 1977) and ranged from 5.1×10⁵ CFU g^–1 crab to 7.9×10⁵ CFU g^–1 crab with an average value of 6.2×10⁵ CFU g^–1 crab. The 95% confidence interval for these data ranged from 1.2×10⁵ CFU g^–1 crab to 2.9×10⁶ CFU g^–1crab across all three replicates.

Prior to injection, oxygen uptake rates in saline-injected crabs(3.52±0.46 μmol g^–1 h^–1) and Vibrio-injected crabs (2.89± 0.11 μmol g^–1h^–1) were not significantly different from one another(one-way ANOVA; P=0.233). Injection of V. campbellii caused a rapid and sustained reduction in oxygen uptake over a 24 h period(Fig. 2; two-way RM ANOVA on ranked data; P=0.012). A 24% reduction in oxygen uptake occurred within the first 20 min after injection compared with saline-injected crabs(HS; P<0.05). Oxygen uptake in Vibrio-injected crabs was reduced by 30% after 4 h and was further reduced to 42% after 24 h compared with oxygen uptake measured in saline-injected crabs (HS; P<0.05).

Although hemolymph lactate levels were generally low (<1 mmol l^–1), there was a significant effect of injection treatment on hemolymph lactate over a 24 h period (two-way RM ANOVA; P=0.004). Accumulation of lactate in the hemolymph was significantly greater in Vibrio-injected crabs compared with saline-injected crabs 2 h following injection; this difference in hemolymph lactate persisted for 24 h(Fig. 3; HS; P<0.03). Hemolymph lactate levels in saline-injected crabs were different only between the 0.5 h and the 24 h timepoints(Fig. 3; one-way RM ANOVA with HS; P<0.02) whereas levels in Vibrio-injected crabs did not significantly change with time.

Blue crabs are capable of walking on a treadmill at a moderate pace of 8 m min^–1 for 30 min. Blue crabs mainly walked laterally, but occasionally walked backwards, using three pairs of walking legs while the chelipeds and swimming legs were mainly used for stabilization. Animals also occasionally rotated 180° during activity, a behavior commonly observed in blue crabs (Spirito, 1972). Crabs subjected to either treatment (saline or Vibrio) were able to perform this level of activity, so a direct comparison of measurements between treatments could be made.

Oxygen uptake significantly changed with activity in both saline-injected crabs and Vibrio-injected crabs (one-way RM ANOVA; P<0.001 for both treatments). Oxygen uptake increased by 2.1–2.5 times resting levels during activity in both treatment groups(HS; P<0.03 for both treatments) and reached steady state within the first 10 min (Fig. 4). Animals quickly recovered from activity and oxygen uptake levels remained elevated only during the first 40 min of recovery from walking (HS; P<0.01 for both treatments).

Significant differences in oxygen uptake were found between treatment groups (two way RM ANOVA; P=0.012). Oxygen uptake in saline-injected crabs at rest before exercise (4.59±0.37 μmol g^–1h^–1) was similar to that found in Vibrio-injected crabs (3.65±0.4 μmol g^–1 h^–1). However, oxygen uptake was significantly reduced during walking in Vibrio-injected crabs (7.74±0.88 μmol g^–1h^–1) compared with saline-injected crabs (11.52±0.89μmol g^–1 h^–1; HS; P<0.03). Vibrio-injected crabs continued to consume significantly lower amounts of oxygen during 3 h of recovery from walking(Fig. 4; HS; P<0.03).

Hemolymph lactate levels increased with activity (one-way ANOVA; P<0.02 for both treatments) and were significantly different between saline-injected and Vibrio-injected crabs(Fig. 5A; two-way ANOVA; P<0.001). Hemolymph lactate was only slightly elevated in saline-injected crabs after activity (1.00±0.19 mmol l^–1; HS; P=0.004) and quickly returned to pre-activity levels within 1 h during recovery(Fig. 5A). Hemolymph lactate in Vibrio-injected crabs increased considerably during activity (HS; P=0.001), and was significantly higher (3.63±0.67 mmol l^–1) immediately after activity (HS; P<0.001);this elevation in hemolymph lactate above control animals persisted during 2 h of recovery (Fig. 5A).

Whole body lactate (Fig. 5B)changed over time in both treatment groups (one-way ANOVA; P=0.002 for both treatments). Resting whole body lactate averaged 7.12±0.66μmol g^–1 in Vibrio-injected crabs and 5.86±0.46 μmol g^–1 in saline-injected crabs(Fig. 5B). Whole body lactate significantly increased ∼1.8-fold during activity to 11.77±1.88μmol g^–1 in Vibrio-injected crabs and 10.53±1.15 μmol g^–1 in saline-injected crabs (HS; P<0.01 for both treatments). Elimination of whole body lactate was slow, with values reaching pre-activity levels within 2 h of recovery (150 min after the onset of walking; Fig. 5B). Although whole body lactate in saline-injected crabs declined immediately after exercise, whole body lactate in Vibrio-injected crabs increased to 12.58±1.00 μmol g^–1 at 30 min recovery. As a result, there was significantly more whole body lactate present in Vibrio-injected crabs at 30 min recovery than in saline-injected crabs (Fig. 5B; two-way ANOVA with HS, P=0.005).

Leg muscle (from the merus of the second walking leg) lactate significantly increased with activity overall (one-way ANOVA; P<0.01 for both treatments) and there were significant differences between treatment groups(two-way ANOVA; P=0.007). Vibrio-injected crabs had significantly more muscle lactate (6.29±0.29 μmol g^–1) that saline-injected crabs (3.48±0.61 μmol g^–1) at rest (Fig. 5C; HS; P=0.007). Muscle lactate increased significantly during activity to 10.53±1.16 μmol g^–1 in Vibrio-injected crabs and 7.65±1.08 μmol g^–1 in saline-injected crabs; the difference between injection treatments persisted through 30 min of recovery(Fig. 5C; HS; P<0.03). Similarly to that observed in the whole body, muscle lactate levels dropped more rapidly in saline-injected crabs, resulting in significantly more muscle lactate present in Vibrio-injected crabs(9.47±1.06 μmol g^–1) at 30 min recovery than in saline-injected crabs (6.31±0.35 μmol g^–1)(Fig. 5C; HS; P=0.017).

Selected samples of whole body and leg muscle tissues were analyzed for succinate, an alternative anaerobic end product generated by many invertebrates (Livingstone,1991). Succinate concentrations in the whole body and leg muscle were very similar and ranged from 0.6 to 1 μmol g^–1 in both tissues (Table 1). Succinate levels did not significantly change with activity in either tissue type (two-way ANOVA; P>0.05). However, there was a significant treatment effect on leg muscle succinate (two-way ANOVA; P=0.012),where in Vibrio-injected crabs succinate was significantly higher immediately after walking than in saline-injected crabs (HS; P<0.05; Table 1).

Within leg muscles at rest, concentrations of arginine phosphate in both treatment groups were higher than those for free arginine. This corresponds to an arginine charge of 0.66 for saline-injected crabs and 0.57 for Vibrio-injected crabs (Table 1). During activity, arginine phosphate was readily dephosphorylated into arginine, causing arginine phosphate concentrations to drop significantly in saline- and Vibrio-injected crabs (one-way ANOVA with HS; P<0.01 for both treatments) with a corresponding significant increase in free arginine (HS; P<0.01; Table 1). After activity,arginine levels exceeded arginine phosphate levels, causing the arginine charge to significantly drop to approximately 0.4 in both treatments(Table 1; one-way ANOVA with HS; P<0.001 for both treatments). Both arginine and arginine phosphate levels returned to pre-activity levels within 1 h of recovery,reflected by an increase in arginine charge to approximately 0.7 for both treatments, and were indistinguishable from pre-activity levels(Table 1). There were no significant differences between saline-injected and Vibrio-injected crabs in phosphagen store measurements over time.

Although there is an apparent fluctuation of ATP and ADP levels in response to activity in both treatment groups, the concentrations were not significantly different from resting values at any time during or after activity (Table 2). However,there was a significant change in AMP levels over time in saline-injected crabs (one-way ANOVA; P=0.027). AMP levels significantly increased with activity in saline-injected crabs and returned to pre-activity levels within 1 h of recovery (HS; P<0.01; Table 2). There was no difference in ATP or ADP concentrations between treatment groups at any timepoint. However, significantly higher AMP levels were found in Vibrio-injected crabs at rest (two-way ANOVA with HS; P<0.02). The adenylate energy charge, representing the overall energy status of the muscle, was high at rest (0.89±0.01) but significantly decreased with exercise to 0.77±0.03(Table 2; two-way ANOVA with HS; P<0.001). Energy charge quickly returned to pre-activity levels within 1 h following exercise. There were no treatment differences in adenylate energy charge values before or after exercise.

Large numbers of Vibrio spp. are often detected in dead or moribund blue crabs (Krantz et al.,1969; Johnson,1976; Welsh and Sizemore,1985). However, Vibrio are often cultured from the hemolymph of apparently healthy blue crabs (e.g. Davis and Sizemore, 1982) and shrimp (e.g. Saulnier et al.,2000). These observations suggest that most Vibrio spp. act as opportunistic pathogens. The pathogenicity of V. campbellii is similar in C. sapidus (LD₅₀=6.2× 10⁵ CFU g^–1 crab) and the Pacific whiteleg shrimp, L. vannamei (LD₅₀=3.06×10⁵ CFU g^–1 animal), but somewhat higher (LD₅₀=6.08×10⁴ CFU g^–1 animal) in the grass shrimp Palaemonetes pugio. This difference in virulence may be attributed to species-specific differences among hosts. Importantly, the LD₅₀data reported here for the blue crab document that the dose of 2.5×10⁴ CFU g^–1 animal used for immunological challenge experiments(Holman et al., 2004; Burgents et al., 2005b) (and the present study) is well below the LD₅₀ value of V. campbellii in crabs and shrimp.

A sub-lethal dose of V. campbellii has a significant impact on resting aerobic metabolism, causing a 30% reduction in oxygen uptake 4 h after injection, and a more substantial 42% reduction in oxygen uptake at 24 h(Fig. 2). These results are similar to those of Scholnick et al.(Scholnick et al., 2006) who found that the shrimp L. vannamei experienced a 27% reduction in oxygen uptake 4 h after injection of V. campbellii, which declined 24 h later to 40%. Reduction of aerobic metabolism is accompanied by a small, but significant, increase in anaerobic metabolism as indicated by elevated hemolymph lactate levels (Fig. 3). The significant elevation of lactate in the hemolymph was not evident until 2 h following bacterial injection. It is possible that these small lactate changes could be indirect results of changes in hemolymph pH(not measured) which may alter the dynamic equilibrium between lactate and pyruvate. The result is an overall metabolic depression similar to the one observed in the Pacific white shrimp(Scholnick et al., 2006).

The metabolic depression appears to persist during walking exercise(Fig. 4) when oxygen demand rises. In these experiments, crabs walked 240 m at a speed of 8 m min^–1 and in all cases did this without difficulty. This is not surprising since this species is capable of swimming at speeds of 7 m min^–1 for at least 30 min(Houlihan et al., 1985) and in the wild displays rapid directional movements over long distances(Spirito, 1972; Wolcott and Hines, 1989; Hines et al., 1995).

The factorial metabolic scope of activity, defined as the ratio of active to resting oxygen uptake (Full and Herreid, 1983) is rather small, ranging from two to seven in most decapod crustaceans (McMahon et al.,1979; Herreid,1981; McMahon,1981; Booth et al.,1982; Full and Herreid,1984; Houlihan and Innes,1984). During walking, O₂ uptake significantly increased ∼2.5-fold and reached steady state within the first 10 min of activity (Fig. 4). This particular aerobic pattern is similar to that found in swimming blue crabs,indicating that blue crabs largely rely on aerobic ATP production during these activities (Booth et al.,1982). An aerobic scope of ∼2.5 is surprisingly small and probably does not represent the maximal oxygen uptake for blue crabs(McMahon, 1981; Booth and McMahon, 1992). Indeed we chose a level of activity that blue crabs in both treatment groups could complete, but that clearly challenged the O₂ delivery system. Despite having efficient mechanisms to enhance oxygen uptake during increased activity, crabs infected with Vibrio did not increase oxygen uptake to levels measured in control animals (Fig. 4). These results are consistent with the idea that bacterial exposure impairs respiratory function in the blue crab because of the formation of hemocyte aggregates or nodules that become trapped in the gills(Johnson, 1976; Smith and Ratcliffe, 1980a; Martin et al., 2000; Burnett et al., 2006; Scholnick et al., 2006).

The excess post-exercise oxygen consumption (EPOC) during recovery in both treatments is partially due to energetic costs incurred during activity. EPOC is used to re-cycle or remove activity-induced buildup of lactate and replenish energy and oxygen stores used during sustained activity but could also be used for energetic demands of elevated ventilation and circulation that continues for some time after activity is completed(Herreid, 1981; Gaessar and Brooks, 1984; Gleeson and Hancock, 2001; Hancock and Gleeson, 2002). A relatively small O₂ deficit associated with a rapid recovery (<1 h) after walking in blue crabs is consistent with results found for swimming blue crabs and other running terrestrial crabs(Houlihan and Innes, 1984; Houlihan et al., 1984; Houlihan et al., 1985; Hamilton and Houlihan, 1992). The EPOC associated with walking was relatively small for all crabs but Vibrio-injected crabs (4.11 μmol O₂g^–1 crab) had a comparatively smaller EPOC than control crabs(6.51 μmol O₂ g^–1 crab).

Although an aerobic steady state was reached in walking crabs, anaerobic metabolism was clearly used to generate some energy(Fig. 5A–C). This could be a consequence of a limitation of oxygen supply to the gills or of mild muscle hypoxia induced by a limitation of oxygen delivery at the tissues(McMahon, 1981; Burnett et al., 2006). Resting whole body lactate was high (∼6 μmol g^–1) compared with resting levels of 1.2 μequiv g^–1 previously found in blue crabs (Booth and McMahon,1985). A twofold increase in whole body lactate during 30 min of walking (Fig. 5B) is modest compared with a 6.1-fold increase in blue crabs that swam for 30 min(Booth and McMahon, 1985). Blue crabs were able to eliminate excess whole body lactate in less than 2 h after walking (Fig. 5B), which is much shorter than the recovery period (∼8 h) for swimming blue crabs(Booth and McMahon, 1985). Levels of lactate in leg muscles were similar in value and pattern to whole body lactate during and after walking, suggesting that whole body lactate measurements are representative of lactate levels present in muscle tissue(Fig. 5B,C).

At least some of the lactate that builds up in active muscle during activity is released into the hemolymph(Fig. 5A). Resting hemolymph lactate (<2 mmol l^–1) was similar to levels measured previously in the blue crab (Booth et al.,1982; Booth and McMahon,1985; Gannon and Wheatly,1995; Johnson et al.,2004). A modest 2.2- to 2.5-fold increase in hemolymph lactate in crabs that had walked for 30 min was comparable to two- to fourfold increases reported for blue crabs after swimming bouts of 15 min(Gannon and Wheatly, 1995) but less than the 14- to 15-fold increase reported after swimming for 25 min(Booth et al., 1982) and 30 min(Booth and McMahon, 1985; Milligan et al., 1989) and 1 h(7-fold increase) (Booth et al.,1984).

Differences in O₂ uptake during walking between control and Vibrio-challenged crabs were not offset by the differences in lactate production. However, blue crabs could also be excreting lactate into the environment (de Zwaan and Skjoldal,1979; Head and Baldwin,1986) or producing other anaerobic end products, such as succinate. Generation of succinate is a high efficiency–low energy output pathway that is most often used by invertebrates during periods of environmental anaerobiosis (Livingstone,1991). However, succinate levels in whole body and muscle were considerably lower than those of lactate in both tissue types and changed very little with injection treatment and with strenuous activity, suggesting that succinate production and, probably, anaerobic ATP-generating pathways overall,do not play an important role in Vibrio-associated changes in metabolism (Table 1).

Only a small number of studies have documented adenylate and phosphagen usage within crustacean muscles. In the walking leg muscle, arginine phosphate was found in high concentrations (25–30 μmol g^–1)along with comparatively lower concentrations of arginine (15–20 μmol g^–1; Table 1). This corresponds to a resting arginine charge of 0.66, which is comparable to 0.69 found in blue crab levator muscle(Hardy et al., 2006) and 0.59 in crayfish leg muscle (Speed et al.,2001) but lower than 0.8 in red crab leg muscle(Morris and Adamczewska,2002). These slight differences are mostly due to varied concentrations of free arginine present in the muscle. Arginine phosphate in blue crab leg muscle (29.6 μmol g^–1; Table 1) was similar to that found in blue crab levator muscle (29.9 μmol g^–1) Hardy et al., 2006) and in spiny lobster leg muscle (31.35 μmol g^–1)(Speed et al., 2001), but higher than concentrations found in highly aerobic ghost crabs (21.5 μmol g^–1) (Full and Weinstein,1992). ATP content (5.7 μmol g^–1; Table 2), indicating muscle energetic condition, was comparable to concentrations in red crab leg muscle(4.76 μmol g^–1)(Morris and Adamczewska, 2002)and in spiny lobster leg muscle (5.14 μmol g^–1)(Speed et al., 2001) but lower than ATP content measured in crayfish tail muscle (7.43 μmol g^–1) (Head and Baldwin,1986) and in blue crab levator muscle (7.3 μmol g^–1) (Hardy et al.,2006). The resting adenylate energy charge in blue crab leg muscle(0.89) was high (Table 2) but very comparable to the energy charge found in other crustacean muscles(England and Baldwin, 1983; Speed et al., 2001; Morris and Adamczewska,2002).

High concentrations of arginine phosphate and ATP are generally associated with muscle fibers that predominantly use anaerobic energy production(England and Baldwin, 1983; Head and Baldwin, 1986; Speed et al., 2001). Although the muscle tissue analyzed here most likely contained a combination of aerobic and anaerobic muscle fibers and no attempt was made to distinguish between the two fiber types, phosphagen energy stores are clearly important during walking as shown by a pronounced 57% depletion in arginine phosphate associated with a 55% increase in free arginine, causing a dramatic drop in arginine charge from 0.66 to 0.35 (Table 1). This large drop in the phosphagen energy stores appears to protect ATP content,which did not significantly change with increased activity(Table 2). This response seems like a general strategy used during muscle oxygen deficiency in crustaceans(Hill et al., 1991). Although only a small fluctuation in ATP levels occurred in response to 30 min walking,the adenylate energy charge (AEC) decreased by 13% during activity mostly due to an increase in AMP levels (Table 2). This indicated that the energetic demands of walking exceeded the buffering capacity of arginine phosphate stores. However, the energetic demands on blue crab walking leg muscle do not seem to be as extreme as the requirements for more intense muscular activity, such as those for crayfish tail muscle after simulated escape behavior, which caused a 28% drop in AEC after exhaustive tail flipping (England and Baldwin, 1983).

The steady state O₂ uptake at a particular work load, converted into energy ATP equivalents, is usually taken as the total energetic cost of that particular activity in highly aerobic vertebrates (reviewed by Herreid, 1981), but a mixture of both aerobic and anaerobic responses are generally reported for crustaceans. We estimated the energetic cost of 30 min of walking by converting the measurement of each metabolic component to ATP energy equivalents (Table 3). Energy production during activity is associated with oxygen consumption (1 mol O₂=6.33 mol ATP), lactate accumulation (1 mol lactate=1.5 mol ATP)and net arginine phosphate and ATP use (1 mol arginine phosphate=1 mol ATP);these are assumptions based on glycogen as the main fuel source(Hohnke and Scheer, 1970; England and Baldwin, 1983). Although aerobic ATP production accounted for the majority (74–80%) of the calculated energetic cost during 30 min of walking, only 15–20% was supported by anaerobic metabolism (Table 3). Although we conservatively assumed that a steady accumulation of lactate was occurring over the 30 min activity period, this contribution is probably most important during the first few minutes of activity. Less than 5%of the total cost was attributed to net arginine phosphate and net ATP use(Table 3). The total energetic cost of exercise in Vibrio-injected crabs (68.0 μmol ATP g^–1 h^–1) was 74% of the total cost estimated for saline-injected crabs (91.6 μmol ATP g^–1h^–1). This difference in energetic cost is almost entirely due to a reduced contribution of aerobic ATP production to the total energy cost.

The general lack of a large recruitment of anaerobic metabolism and stability of AEC during stress are defining characteristics of metabolic depression (Guppy et al.,1994; Reipschlager and Pörtner, 1996). We found only a 5% increase in anaerobic metabolism in response to Vibrio injection to compensate for the large reduction in aerobic metabolism(Table 3). Additionally, there were no differences in arginine and adenylate energy charge (Tables 1 and 2). These results together support the hypothesis of a metabolic depression induced by injection of bacteria that persists during increased activity.

The phenomenon of metabolic depression has been documented in the majority of major animal phyla (reviewed by Guppy and Withers, 1999; Storey and Storey, 2004). Many invertebrates are capable of decreasing overall metabolism in response to environmental stressors such as hypoxia,desiccation and extreme cold (Guppy et al., 1994; Burnett,1997; Guppy,2004). Metabolic depression may also be a stress response in crustaceans exposed to bacteria. Although the actual systemic bacterial infection itself may cause a metabolic depression to occur, this possibility is unlikely since culturable bacteria are rapidly cleared from the hemolymph and other major tissues in blue crabs and penaeid shrimp(Holman et al., 2004; Burgents et al., 2005b). Also,we cannot rule out the possibility that there may be direct or indirect effects of the bacteria itself, its associated endotoxins and exotoxins or hormones that are released in response to bacterial stress. For example,bacterial lipopolysaccharide not only induces the normal hemocytopenic response but also the release of crustacean hyperglycemic hormone, a hormone that is considered the primary regulator of metabolic fuel availability in crustaceans and can cause multiple physiological effects in moderately stressed organisms (Lorenzon et al.,1997; Lorenzon et al.,1999; Morris,1999; Fanjul-Moles,2006). Therefore, it is possible that reduction of oxygen uptake in response to bacterial exposure may be hormonally controlled, as has been found in respiration and ionic regulation in the hololimnetic shrimp Macrobrachium potiuna in response to salinity stress(de Souza and Moreira,1994).

Animals experiencing a metabolic depression generally show decreased overall activity (Storey and Storey,1990). However, Vibrio-injected crabs along with their controls were clearly able to accomplish walking exercise. The difference in metabolic rate between the control and Vibrio-treated crabs during exercise is nearly identical to the metabolic depression observed in resting crabs. This result was unexpected. Vibrio-injected crabs were able to walk for 30 min and they were clearly able to do this metabolically in a manner similar to control crabs (Table 3). We had expected that Vibrio-injected crabs would find it more difficult to deliver the increased oxygen required during exercise if their oxygen transport system were truly impaired(Burnett et al., 2006);however, these experiments were conducted in well-aerated water and we suspect that the results would be quite different in hypoxic conditions.

Although metabolic depression could indicate a reduction in all cellular processes, we hypothesize that a reduction in metabolism is probably due to decreases in protein synthesis rate and the associated ATP costs of protein turnover (Storey and Storey,1990; Storey and Storey,2004). Protein synthesis is an energy-expensive process and can account for as much as 37% of aerobic metabolism in crustaceans(Houlihan et al., 1990). Indeed, a significant reduction in protein synthesis during metabolic depression has been reported in numerous organisms and is frequently associated with the liver and hepatopancreas(Hochachka et al., 1996; Fuery et al., 1998; Pakay et al., 2002).

The mechanisms by which priorities of energy expenditure are re-arranged to sustain necessary functions while under stress are unclear. Interestingly,activation of AMP-activated protein kinase (AMPK), a `metabolic master switch'in mammals that acts to preserve ATP during stress by increasing ATP generating pathways (glycolysis, fatty acid oxidation, glucose uptake) and decreasing ATP consuming pathways (synthesis of fatty acids, cholesterol and protein), occurs in response to depleted cellular energy pools that occur during prolonged exercise and hypoxic stress(Winder and Hardie, 1999; Kemp et al., 2002). AMPK has recently been identified as a central mediator of the translational control of protein synthesis during hypoxic stress in rat hepatocytes(Horman et al., 2002; Kemp et al., 2002). It has now been confirmed that AMPK is present in crustaceans and is activated in response to both temperature and hypoxic stress(Pinz et al., 2005; Frederich et al., 2009). Therefore, AMPK, if present in the blue crab, may be playing a role in regulating energy metabolism during bacterial stress.

Despite the fact that the total energy expenditure during walking in Vibrio-exposed crabs accounted for only 74% of the energy expenditure in saline-injected blue crabs, the decrease in the amount of ATP produced by aerobic means does not appear to affect the overall performance of crabs during moderate walking activity. A crab's ability to walk continuously for 30 min at a moderate pace is not greatly affected by Vibrio injection. However, crabs infected with bacteria are less aerobic while doing so, as are resting crabs. While movement in optimal environmental conditions may not pose much of a problem, should that crab encounter hypoxic waters or some other stressful environmental condition, reduced metabolism associated with bacterial infection may render the animal less fit to meet the demands of normal activity necessary for survival in the wild.

Regardless of the mechanisms by which aerobic metabolism is reduced,significant metabolic adjustments that occur after bacterial stress would surely be intensified by the effects of environmental hypoxia or high levels of animal activity. Hypoxic conditions occur in coastal and estuarine environments and are increasing on a global level(Diaz and Rosenberg, 2008). Blue crabs are able to tolerate external hypoxic conditions by compensating with behavioral, metabolic and circulatory adjustments(Batterton and Cameron, 1978; deFur et al., 1990; Burnett, 1997; Mangum, 1997; Burnett and Stickle, 2001) but environmental hypoxia itself negatively impacts immune function(Holman et al., 2004; Burgents et al., 2005a; Macey et al., 2008). Blue crabs are also able to perform high levels of activity necessary for survival in the field (Spirito, 1972; Booth et al., 1982). In cases of both hypoxia and increased activity, the oxygen delivery system fails to completely supply the necessary energy needed, and is supplemented by anaerobic pathways producing lactic acid(Booth et al., 1982; Booth and McMahon, 1985; Burnett and Stickle, 2001), in a manner similar to that when a blue crab is experimentally infected with bacteria. Dealing with either of these additional stressful situations while simultaneously fighting a bacterial infection would only further affect the already-compromised metabolic status of an animal and possibly its survival in the field.