Changes in phosphatidylcholine molecular species in the shrimp Macrobrachium borellii in response to a water-soluble fraction of petroleum

The effect of the water-soluble fraction (WSF) of crude oil on lipid contents, lipid classes, FA, and PC molecular species was studied in high-phospholipid (hepatopancreas) and low-phospholipid (egg) tissues of a freshwater crustacean. After a 21-d exposure to a sublethal concentration of WSF, a significant decrease in shrimp total lipids was observed, although no alterations could be detected in the hepatopancreas or egg lipid contents. TAG/phospholipid ratios increased in the hepatopancreas and decreased in the eggs, suggesting alterations either in the mobilization of TAG to phospholipid pools or in the energy balance. The FA composition of phosphoglycerides in the hepatopancreas and eggs was dominated by PUFA, whereas the n−3/n−6 ratio was not affected by WSF exposure, although there was a significant increase in hepatopancreas 18∶1n−9. Analysis of the PC molecular species by HPLC-ELSD showed the presence of 15 species, with 16∶0/18∶1, 18∶1/18∶2, 16∶0/20∶5, and 16∶1/20∶5 being the major species in the hepatopancreas. The PC molecular species in the eggs showed a different pattern, dominated by 16∶0/18∶1 and 18∶1/18∶2. Of the PC molecular species, 10 contained 22∶6n−3, 20∶5n−3, and 20∶4n−6. Small amounts of di-PUFA species were also found. Exposure to WSF altered the PC molecular species in both tissues. The four major hepatopancreas molecular species and most of the ones containing PUFA decreased. This was compensated for by an increase in 16∶1/18∶1 (152%) and 18∶1/18∶1 (50%). The two major egg PC molecular species decreased, whereas the PUFA-containing ones increased. The contrasting responses of both tissues of WSF contamination suggests the presence of different homeostatic mechanisms.

The freshwater shrimp Macrobrachium borellii is a decapod crustacean widespread in South America that lives in turbid, tepid waters such as the ones in the Río de La Plata estuary (1,2).In several past studies the lipid composition and metabolism of M. borellii were investigated.Phospholipids (PL) and TAG were identified as the main lipid classes in adult tissues and eggs, respectively (3)(4)(5).Monounsaturated FA were the major group in egg TAG, whereas PUFA were the major group in PL after embryo organogenesis and in the total lipids of adult tissues (4,5).PL classes and their FA composition play a key role in the structure and function of most cellular membranes, including M. borellii membranes, whose main PL class is PC (6,7).Knowledge of the molecular species of PL is important for understanding the membrane properties and functions (8).However, unlike vertebrates, few reports are available on the molecular species of PL in invertebrates (9)(10)(11)(12)(13), and none are available on the effect of the water-soluble fraction (WSF) of petroleum hydrocarbons (HC) on them.
When lipophilic contaminants such as petroleum HC are taken up by organisms, they partition at some equilibrium level between water and the lipids of the organism and become bioconcentrated in lipid compartments such as membranes, lipoproteins, and lipid storage tissues (14).Benthic organisms are known to accumulate HC directly from the water column or from the interstitial water in the sediment (15).Thus, M. borellii was chosen because it is a species with ecological relevance in the Río de La Plata estuary in which we had previously studied the bioaccumulation and biodepuration rates of WSF; we had observed that lipid compartments retain polyaromatic HC for long periods of time (16).Oil spills occur periodically in the Río de La Plata estuary, which has become the most contaminated region in the country in terms of HC (17).
In the present study, we investigated the use of molecular species of PL as a potential biochemical marker for HC pollution.The effect of WSF on the lipid content and lipid class composition of M. borellii was examined.Changes in the pattern of PC molecular species after exposure to a sublethal concentration of HC are described.

EXPERIMENTAL PROCEDURES
Sample collection.Adult M. borellii were sampled during the spring and summer (October to February) in a watercourse free of petrogenic contamination close to the Rio de La Plata River in Argentina (20 km southwest of La Plata city), with temperatures ranging from 21 to 28°C.Ovigerous females with eggs were collected for the experiments in early summer, and adults were collected in the spring for another set of experiments.They were taken to the laboratory and kept in dechlorinated tap water at 24 ± 2°C and a 14-h light/10-h dark photoperiod for at least a week before the experiments (18).The eggs were removed from the pleopods of the ovigerous females and checked under a stereoscopic microscope to determine the stage of development (19).
Preparation of the WSF of crude oil.Light crude oil obtained from Punta Loyola (Santa Cruz, Argentina), stored at 4°C, was used to prepare the WSF.The oil was stirred in a 10-L stainless-steel mixing vessel equipped with a mechanical stirrer and a bottom drain, and was kept in a cold room at 4°C.Crude oil and fresh water in a ratio of 1:100 (vol/vol) were stirred at a low speed for 24 h and allowed to settle for an additional 48 h (shorter settling times produced a fraction with dispersed oil droplets).The WSF was collected daily using the bottom drain.During the experiments, fresh batches of WSF were prepared every 2 d, and a sample was extracted and analyzed by GLC following the method described by Lavarías et al. (16).
Exposure to WSF.Groups of 10 adults or 4 ovigerous females carrying eggs at developmental stage 3 (19) were placed into 750-mL glass flasks with air-tight screw-capped lids to avoid the loss of HC.The number of individuals per flask was selected to ensure the appropriate concentration of dissolved oxygen, which was checked daily.To evaluate the effect of WSF exposure on lipids in either the whole organism, hepatopancreas, or eggs, three groups were used as a control and another three were exposed to a sublethal concentration of WSF (0.6 ppm, chosen from the results of a previous study) (16).Shrimp were exposed for 21 d without being fed, and the WSF was replaced daily.After exposure, the eggs were removed from the female pleopods, the hepatopancreas was dissected from the adults, and the total weight of the samples was recorded.All animals per group were pooled at the end of the experiment.Individuals were dried on an absorbant tissue, weighed on a semi-analytical balance with 0.01-mg accuracy, and dissected; the size of the animals was homogeneous.Weight differences among the samples were minor, and mortality was always less than 10% of animals in the control and exposed groups.
Lipid extraction and analysis.Tissues were immediately homogenized in a Potter-type homogenizer (Thomas Scientific, Swedesboro, NJ) using 0.02 M Tris-HCl buffer, pH 7.5, and a subsample was taken for protein analysis for another experiment.For the analysis of whole organisms, homogenization was performed using an UltraTurrax tissue disrupter (Janke and Kunkel, Ika Werk, Germany).Lipids were extracted with a chloroform/methanol mixture following the method of Bligh and Dyer (20).Each group was analyzed separately.
The total lipid concentration was determined gravimetrically in an aliquot of chloroform, and the rest was kept in hexane under a nitrogen atmosphere at −70°C until further analysis.
A lipid class analysis was performed by TLC on silica gel Chromarods (type S-III) with quantification by FID using an Iatroscan TH-10, Mark III (Iatron Laboratories Inc., Tokyo, Japan), as described by Parrish and Ackman (21).The separation was conducted with a sequence of three different solvent systems according to Ackman et al. (22).The first development was carried out for 45 min in hexane/ethyl acetate/diethyl ether/formic acid (91:6:3:1, by vol).The Chromarods were dried, partially scanned to analyze the neutral lipids, and then developed in acetone for 15 min to quantify the carotenoid peak.Finally, the Chromarods were developed in chloroform/methanol/formic acid/water (50:30:4:2, by vol) for 60 min and completely scanned to reveal the different PL.Tetracosanol was used as an internal standard, and quantification was performed using the calibration curves of authentic standards run under the same conditions.Carotenoids from the shrimp were purified for use as standards to avoid error caused by differing responses of the FID to different carotenoids.The lipids were identified as described previously (5).
FA analysis.Preparative high-performance TLC (HP-TLC) (Merck, Darmstadt, Germany) was used to isolate the neutral lipids from the polar lipids, as described previously (5).FAME from PL were prepared using a base-catalyzed transesterification microscale procedure according to Christie (23).Amidebound FA, like those in sphingomyelin, are not affected by alkaline transesterification; therefore, the PL FA analyzed came mainly from the PC and PE in this species.FAME were analyzed by GLC on an Omegawax 250 (30 m × 0.25 mm, 0.25 µm film; Supelco, Bellefonte, PA) capillary column in a Hewlett-Packard HP-6890 equipped with a FID.The column temperature was programmed for a linear increase of 3°C/min from 175 to 230°C.FA were identified by their characteristic retention times and by co-injecting the sample with authentic standards (Supelco, St. Louis, MO) or a FAME mixture of established composition run under the same conditions.
PC molecular species.PL fractions were isolated from lipid extracts by HPLC using a Merck-Hitachi L-6200 Intelligent pump (Hitachi Ltd., Tokyo, Japan).An Evaporative Light-Scattering Detector 500 (Alltech Associates Inc., Deerfield, IL) operating at a nitrogen flow rate of 2.2 mL/min and a drift tube temperature of 90°C was used for detection.To purify the PC, an Econosil Silica column (250 × 4.6 mm × 10 µm; Alltech Associates Inc.) and a guard column packed with the same material were used.Elution was performed as described elsewhere (24).PC was collected manually, the eluate was evaporated to dryness under a stream of nitrogen, and the PC was redissolved in chloroform/methanol (1:1, vol/vol).
Resolution of the PC molecular species was performed on two end-capped Lichrosphere 100 RP18 columns in series (250 × 4 mm; Merck, Darmstadt, Germany).Isocratic elution was applied with a mobile phase composed of acetonitrile/ methanol/triethylamine (40:58:2, by vol) (25).Detection and quantification were done by ELSD using nitrogen as the nebulizer gas at a flow of 1.8 L/min and a temperature of 100ºC (24).
The column eluate was collected every 30 s using an LKB 2212 Helirac fraction collector (Bromma, Sweden).The individual PC molecular species were identified by determining the FA composition of each peak by GLC as described, and its relative abundance was calculated from the areas of the peaks.
Statistical analyses.Data collected from all experiments were analyzed by ANOVA using Instat, v. 2.0 (GraphPad, San Diego, CA).Data were transformed prior to performing the statistical procedures.Results were considered significant at a level of 5%.

Changes in lipid content and composition by WSF exposure.
After a 21-d exposure to WSF, the total lipid concentrations of exposed animals decreased significantly compared with control organisms; however, the total lipid concentrations of the hepatopancreas and eggs did not vary significantly (Table 1).
To determine the effect of WSF on the lipids, each lipid class was studied separately in the adult hepatopancreas and eggs of control and exposed organisms.We also checked the muscle, but no changes were observed; therefore, it was not included in the present study.PL were the major lipids in the hepatopancreas, accounting for 74.4%, by weight of the total lipids, whereas egg lipids were dominated by TAG, representing 69.5% of the total lipids.This markedly different lipid composition provided us with two contrasting models within the same species to study the effect of HC contamination.After a 21-day exposure to WSF, both the hepatopancreas and eggs showed significant changes in their lipid class composition.The hepatopancreas showed a significant decrease in PL as well as a decrease in the carotenoid pigment (Table 2).PC in particular, representing the major lipid class in this tissue, showed a significant reduction from 62 to 53% by wet weight (P < 0.05) in WSF-exposed animals.However, the average sphingomyelin concentration in the WSF-exposed hepatopancreas was almost two times higher than the hepatopancreas control (Table 2).The eggs, in contrast, showed a significant decrease (more than 11%) in TAG reserves in WSF-exposed embryos compared with the percentage in controls.In addition, there was an overall increase in PL, whereas the carotenoid pig-ment did not change significantly in WSF-exposed eggs (Table 2).The PL/TAG ratio decreased in the hepatopancreas of WSFexposed animals because of the decrease in PL, whereas in the eggs the ratio increased from 0.4 to 0.6, reflecting both the decrease in TAG and the increase in PL (Table 2).
FA. Table 3 shows the FA profiles of phosphoglycerides from the hepatopancreas and eggs of control and WSF-exposed organisms.Regardless of treatment, EPA (20:5n-3) was the major FA in the hepatopancreas, followed by arachidonic (20:4n-6) and oleic acids (18:1n-9).The exposure of shrimp to WSF did not significantly change the FA profile of the PL, except for an increase in 18:1, which made it the second most important FA in the PL.The FA profile of the eggs was also dominated by 20:5n-3, followed by similar amounts of 18:1n-9, 16:0, and 18:0.After exposure to WSF, there were no statistically significant differences in the relative FA composition between the WSF and control groups.PUFA was the major group in the PL, regardless of the tissue or treatment analyzed.
PC molecular species.After studying the effect of WSF on total lipids, the lipid composition, and PL FA, we identified PC as the only PL whose concentration was significantly altered by the contaminant in both the hepatopancreas and eggs (Table 2).We therefore studied PC in detail, analyzing its molecular species.The eggs and hepatopancreas were chosen, as in preliminary experiments these tissues were found to be the ones most sensitive to HC pollution.

DISCUSSION
Total lipid content as a pollution indicator.In this work, we analyzed the effect of a sublethal exposure to WSF on lipid contents, lipid classes, and the PL FA composition of adult M. borellii tissues and eggs and characterized the PC molecular species as well as the variations triggered by this contaminant.When we analyzed the total lipid contents, we observed a significant decrease in the exposed adults, probably because the shrimp catabolized more of their lipid reserves to obtain energy for the detoxification processes triggered by the contaminant.Wang and Stickle (26), studying the crab Callinectes sapidus, also observed a decrease in the total lipid concentration after a 15-d exposure to crude oil.Nevertheless, when we determined total lipids in the hepatopancreas and eggs of M. borellii-tissues metabolically active in biosynthesis, degradation, and depuration-we found no differences between exposed and control organisms.Nowadays, it has become clear that the traditional use of total lipid content or lipid-normalized pollutant data as a pollution indicator is not always a safe approach because of variations in lipid extraction and analysis, differences in contaminant partitioning among lipid pools, and seasonal/physiological changes in the lipid concentration and composition (27).In addition, the homeostatic mechanisms of the organisms may not be overcome, thus avoiding the appearance of significant changes in lipid concentration.It is therefore important to consider the variations in lipid classes or, like the novel approach explored in the present study, changes in the PL molecular species induced by exposure to WSF.
Variations in lipid and FA classes in response to WSF.We observed a significant decrease in PC concentration in the hepatopancreas of exposed shrimp, whereas the sphingomyelin increased significantly and the TAG showed an increasing trend.Similar changes in PC and TAG were also found in crabs (28) and mussels (29) in response to other pollutants.Those authors suggested this might result from membrane damage, a fact that could also be the case with M. borellii, whose membranes contain PC as the major component (30); PC changes could be affecting the structure and properties of M. borellii membranes (4).Moreover, there are indications that xenobiotics increase membrane lipid peroxidation (31), and because the hepatopancreas is in charge of xenobiotic detoxification in adult crustaceans (32), it was not surprising to observe that the concentration of astaxanthin, a very potent antioxidant, decreased in WSF-exposed organisms, suggesting its consumption to prevent oxidative damage.The significant decrease in the PL/TAG ratio of the hepatopancreas of exposed organisms would also be reflective of a mechanism to raise TAG levels, either by increasing TAG synthesis or by decreasing TAG mobilization into PL pools, thereby raising the number of hydrophobic reservoirs for WSF and diminishing the contaminant bioavailability by dilution.This mechanism also has been observed in other aquatic organisms (29).Previous studies in M. borellii have shown that the bioconcentration of aromatic HC from the environment and their subsequent depuration rate is related to the lipid contents (16).Capuzzo and Lancaster (33) demonstrated the impairment of lipid utilization in lobsters exposed to oil and suggested that the disruption in energetics was caused by alterations in their lipid metabolism.
Macrobrachium borellii egg lipids are composed primarily of TAG, which serve as the main source of energy during development (5).In this model we observed a significant decrease of this reserve in embryos exposed to WSF.This contrasting effect, compared with the hepatopancreas, is probably related to the fact that, being a closed system, egg embryos are faced with a different pressure: They have no source of energy other than the vitellus and must expend more energy coping with the increase in free radicals and detoxification mechanisms induced by WSF, as shown by the highly significant increase in the PL/TAG ratio.Although the concentration of the antioxidant astaxanthin did not change, the enzymatic defense systems are well developed in these shrimp embryos (see ensuing discussion).We have previously shown that developing embryos have a high metabolic rate and an important consumption of lipids from the vitellus at later stages (5); thus, an alteration of the major energy source attributable to the presence of WSF was expected.
The FA profile of PL from the hepatopancreas of M. borellii did not differ much from the one previously reported for egg PL (5), the major FA being 20:5n-3, followed by 20:4n-6, the 18:1 isomers, 16:0, and 18:0.The n-3/n-6 ratio was below one, as is usual in freshwater environments.In both the hepatopancreas and eggs, this FA composition was not greatly affected by the WSF.Nevertheless, the significant increase in 18:1 in the hepatopancreas after WSF exposure is noteworthy, as it became the second most important FA in the tissue, and this in-crease was coincident with the increase in PC molecular species containing this FA (see ensuing discussion).Embryos are enclosed in the eggshell, and as closed systems, they cannot afford to lose the PUFA they need for the active membrane synthesis that takes place during organogenesis.The embryos of the crustacean Gammarus locusta are known to have a more developed antioxidant defense system than that of adults (34).A study in progress shows that M. borellii embryos have background levels of antioxidant defense enzymes such as catalase and glutathion transferase over twice as high as levels in the adult hepatopancreas.This probably allows the embryos to reduce oxidative damage, thus accounting for the unchanged PUFA concentration in embryos caused by the stressor.In contrast to results for the hepatopancreas, the astaxanthin concentration did not change by WSF exposure, suggesting that this antioxidant pigment is conserved by the embryo, as was shown in a previous study (5).
Variations in PC molecular species in response to WSF.After studying the effect of WSF on total lipids, lipid classes, and the phosphoglyceride FA composition, we identified PC as the only lipid in both the hepatopancreas and eggs whose concentration was varied significantly by the contaminant.We therefore focused on this lipid class and analyzed its molecular species to investigate whether there was a remodeling of the PC molecular species in response to WSF.In the hepatopancreas and eggs, we were able to identify 15 PC molecular species, dominated by 16:0/18:1 (24 and 31% of the total, respectively).Currently, very few reports are available on the PC molecular species of aquatic arthropods.The PC molecular species of the amphipod Gammarus sp. were also dominated by this molecular species (12).Unlike those results and the ones reported for Limulus polyphemus amebocytes (13), no fully saturated species were detected in the caridean decapod M. borellii (present results), Pandulus sp., or Parapandulus sp.(9).Moreover, we found minor amounts of di-polyunsaturated molecular species.Nevertheless, the vast majority of PC molecular species in these species consisted of sn-1 saturated or monoenoic 16-or 18-carbon FA and sn-2 monoenoic and polyenoic PC molecular species.This was expected, as it is a feature also found in the composition of most PC molecular species of freshwater (35,36) and terrestrial (37)(38)(39) vertebrates.In general, the molecular species composition reflected the high level of PUFA found in the PL FA profile of M. borellii.
After 3 wk of exposure to WSF, the composition of the molecular species of hepatopancreatic PC was modified such that it probably compensated for any changes triggered by the stressor.Tissues with a lower TAG/PL ratio were more greatly affected by the WSF, as suggested by Bergen et al. (27) for mussels; with the hepatopancreas there was a significant decrease in most of the PUFA-containing PC molecular species, probably due to the high susceptibility of PUFA to lipoperoxidation induced by aromatic HC detoxification (32).
One inherent property of the cells in poikilothermic animals is their ability to adjust the physicochemical characteristics of their membranes to the prevailing temperatures.Saturated FA produce less fluid PL with increasing chain lengths (38), but PUFA make the membranes only slightly more fluid than do monounsaturated FA.Therefore, the decrease in the fluidizing highly PUFA-containing species triggered by WSF exposure could be compensated for by the increase in di-monounsaturated 18:1-containing PC species as well as the decrease in 16:0/18:1.Interestingly, these results are similar to those found for shrimp exposed to different temperatures in which changes in the PC molecular species were produced in such a way that the fluidity was roughly maintained (35).Reshuffling and restructuring of the molecular species might be an important factor in lipid membrane adaptation, minimizing alterations caused by exposure to the WSF, although the present data do not allow us to differentiate between an adaptation to the WSF or a direct effect of WSF on the membranes.
The lack of change caused by WSF in the FA composition of the eggs, together with significant changes in the PC molecular species (Fig. 2), indicates that the type and proportion of the FA at both the sn-1 and sn-2 positions had altered.This could be a consequence of the availability of different FA (for example, from the TAG pool), but also of changes found in the specific activity of enzymes involved in PC synthesis.A retailoring process was evidently induced, and a deacylation-reacylation at carbons sn-1 and sn-2 was also manifest, although not much information is available on these processes in crustaceans.The profile of the PC molecular species might also have been influenced by different types of pathways, such as PE N-methylation, which has been found to be very active in crustaceans (40), although this was not checked on these shrimp.
In contrast to the hepatopancreas, the eggs showed a significant decrease in molecular species containing oleic acid and an increase in the PUFA-containing molecular species in WSFexposed embryos.The increase in PUFA-containing molecular species may be related to the fact that the embryos must conserve these EFA for proper development, and in this sense, we previously observed that developing time and hatching are not significantly altered in WSF-exposed embryos (16).Ongoing experiments are focusing on the changes in membrane properties, e.g., chemical reactivity, fluidity, and interaction with proteins, induced by WSF.
The PC species 16:0/18:1 was the most affected by exposure to WSF, and its marked decrease in the hepatopancreas or eggs could be used as a complementary early biochemical biomarker, indicating the presence of HC pollution.Nevertheless, work is needed to validate the suitability of PC species as biomarkers in the environment.The shift in molecular composition of PC is noteworthy since it illustrates that relatively small changes on a total scale (lipids and FA composition) are translated into major changes in the PUFA-and monounsaturated FA-containing PC fractions.The different responses observed in the adults and eggs of PC species reinforce the need to include different life history stages of species in toxicological tests, as well as an appropriate biomarker that integrates chemical data with biological responses, when aiming for a better evaluation of hazardous compounds.

TABLE 1 Total Lipid Concentration of Whole Organisms, Hepatopancreas, and Eggs of Macrobrachium borellii from Control and Water-Soluble Fraction (WSF)-Exposed Organisms a
Values are the means of triplicate analyses ± 1 SD and are expressed as mg/g wet weight.*P < 0.05. a

TABLE 2 Lipid Class Composition a of the Hepatopancreas and Eggs of M. borellii from Control and WSF-Exposed Organisms
a Values (mg lipid/100 mg tissue wet wt) are the means of triplicate analyses ± 1 SD.*Significant changes between control and exposed tissue (P < 0.05); **very significant (P < 0.01).Hp, hepatopancreas; ES, esterified sterols; ST, free sterols; ASX, carotenoids; SM, sphingomyelin; PL, phospholipids; for other abbreviation see Table1.

TABLE 3 Major FA of the Phosphoglyceride Control and WSF-Exposed Hepatopancreas and Eggs of M. borellii a
a