Marine Biology 108, 277-291 (1991)

Marine Biology © Springer-Verlag 1991

Incorporation and utilization of bacterial lipids in the Solemya velum symbiosis *' ** N. Conway and J. McDowell Capuzzo Biology Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, USA Date of final manuscript acceptance: August 3, 1990. Communicated by J. Grassle, Woods Hole

Abstract. We undertook a detailed analysis of the lipid composition of Solemya velum (Say), a bivalve containing endosymbiotic chemoautotrophic bacteria, in order to determine the presence of lipid biomarkers of endosymbiont activity. The symbiont-free clam Mya arenaria (L.) and the sulfur-oxidizing bacterium Thiomicrospira crunogena (Jannasch et al.) were analyzed for comparative purposes. The 613C ratios of the fatty acids and sterols were also measured to elucidate potential carbon sources for the lipids of each bivalve species. Both fatty acid and sterol composition differed markedly between the two bivalves. The lipids of S. velum were characterized by large amounts of 18:lo97 (cis-vaccenic acid), 16:0, and 16 : 1(o7 fatty acids, and low concentrations of the highly unsaturated plant-derived fatty acids characteristic of most marine bivalves. Cholest-5-en-3/?-ol (cholesterol) accounted for greater than 95% of the sterols in S. velum. In contrast, M. arenaria had fatty acid and sterol compositions similar to typical marine bivalves and was characterized by large amounts of the highly unsaturated fatty acids 20 : 5co3 and 22 : 6co3 and a variety of plant-derived sterols. The fatty acids of T. erunogena were similar to those of S. velum and w~ere dominated by 18 : lco7, 16:0 and 16:1o97 fatty acids. The cis-vaccenic acid found in S. velum is almost certainly symbiontderived and serves as a potential biomarker for symbiontlipid incorporation by the host. The high concentrations of cis-vaccenic acid (up to 35% of the total fatty acid content) in both symbiont-containing and symbiont-free tissues of S. velum demonstrate the importance of the endosymbionts in the lipid metabolism of this bivalve. The presence of cis-vaccenic acid in all the major lipid classes of S. velum demonstrates both incorporation and utilization of this compound. The 613C ratios of the fatty acids and sterols of S. velum were significantly lighter ( - 38.4 to -45.3%0) than those ofM. arenaria ( - 2 3 . 8 to -24.2%0) and were similar to the values found for the * Woods Hole Oceanographic Institution Contribution No. 7356 ** Please address all correspondence and reprint requests to Dr Conway at her present address: Department of Biological Sciences, University of Pittsburgh, Pennsylvania 15260, USA

fatty acids of T. crunogena ( - 45%0); this suggests that the lipids of S. velum are either derived directly from the endosymbionts or are synthesized using endosymbiontderived carbon.

Introduction The small protobranch bivalve Solemya velum occurs in sulfide-rich sediments, has an extremely small gut, and has high concentrations of symbiotic bacteria within specific gill cells called bacteriocytes (Cavanaugh 1983, 1985). These bacterial symbionts appear to be chemoautotrophic, deriving their energy and reducing power from the respiration of inorganic substances, such as reduced sulfur compounds, and their cell carbon from the fixation of CO2 (Cavanaugh 1983, 1985). Biochemical and ultrastructural evidence gathered over the past decade suggest that similar invertebrate-chemoautotroph symbioses (hereafter called animal-bacteria symbioses) occur in at least four marine invertebrate phyla from both reducing sediments and deep-sea vents (Felbeck et al. 1981, Southward etal. 1981, Cavanaugh 1983, Giere et al. 1984). In many of the marine animal-bacteria symbioses studied to date the bacteria are contained in membrane-bound vacuoles within host cells (Southward et al. 1981, Cavanaugh 1983, Felbeck 1983, Fiala-Mtdioni 1984, DeBurgh et al. 1989). Most of the symbioses discovered involve sulfur-oxidizing bacteria; high activities of the enzymes involved in sulfur metabolism and CO2 fixation have been reported in the symbiont-containing tissues of some species (Cavanaugh 1983, Felbeck 1983). In addition, Cavanaugh et al. (1988) have confirmed that in the S. velum symbiosis, ribulose-biphosphate carboxylase, the enzyme involved in carbon fixation, is localized only in the bacterial cells. In many of these animal-bacteria symbioses, the digestive system of the host species is very small or absent, indicating atrophic function for the bacteria. In this context, stable isotope studies (Conway et al. 1989) suggest

278

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

that Solemya velum m a y obtain up to 100% of its carbon and nitrogen requirements from bacterial symbionts. Similarly, the detailed study of Anderson et al. (1987) suggests that the west coast species S. reidi m a y be able to meet its carbon needs through symbiont autotrophy. The distinct vacuolar m e m b r a n e separating the bacteria from the host cytoplasm m a y play a role in regulating nutrient or metabolite transfer between host and symbiont, or in protecting the bacteria f r o m host enzymes (DeBurgh et al. 1989). Understanding the details of the trophic relationships between marine invertebrates and chemoautotrophic bacterial symbionts is difficult using the standard physiological techniques of respirometry and enzyme kinetics. These methods require optimal animal and bacterial physiological condition, yet deep-sea and hydrothermal vent species are frequently damaged during collection. Moreover, optimal laboratory conditions for the maintenance of most species are not known, and bacterial productivity in animal-bacteria symbioses appears to decline as the period of laboratory maintenance lengthens (Anderson etal. 1987, N. Conway personal observations). Comparative biochemical and physiological studies m a y be useful in identifying chemical markers of endosymbiont activity (Conway and McDowell Capuzzo 1990). Symbiont-derived chemical markers could serve as valuable screening tools for the presence of symbionts in untested species. In addition, the identification of chemical markers, which are unaffected by physical damage to the organism during collection and short-term preservation, could be extremely useful in the study of nutrition in deep-sea symbioses. Lipids have been used in the past as biological markers for algae and bacteria in marine sediments (Perry et al. 1979, Volkman et al. 1980); to determine the biomass of bacterial symbionts in ascidians (Gillan et al. 1988); and as possible organic tracers for establishing benthic microbial community structure (Bobbie and White 1980). The lipid content of bivalve molluscs is generally dependent on dietary lipid intake (Moreno et al. 1980, Piretti et al. 1987). Consequently, the lipids of symbiont-containing bivalves will probably reflect a symbiont-based diet if the symbionts are being utilized for nutritional purposes. In addition, it m a y be possible to identify utilization of specific bacterial lipids by the host as certain lipids (e.g. hopanoids, eis-vaccenic acid, branched fatty acids) are generally associated with bacteria (Ourisson et al. 1979, Fulco 1983). Stable isotope ratios are important biochemical markers which have been used to estimate the contribution of bacterial symbionts to host nutrition (Rau 1985, Conway et al. 1989). The reduction of inorganic carbon during c h e m o a u t o t r o p h y generally results in a greater utilization of the 12C isotope relative to the 13C isotope (i.e. more negative 613C value) than occurs during photosynthetic carbon fixation (Degens 1969, Ruby et al. 1987). Consequently, chemoautotrophic bacteria m a y have more negative 613C values than photosynthetic organisms. These differences in 6~3C ratios are often transferred through the food-chain allowing determination of the origin of the carbon in an organism. The

results of stable isotope studies suggest that, in some species, bacterial endosymbionts m a y be vital in the provision of carbon and nitrogen to the host species (Conway et al. 1989). To date little information exists concerning the nature of the compounds being utilized. As part of an ongoing investigation of nutrition in animal-bacteria symbioses, we have undertaken a detailed comparative study of the biochemical composition of the Solemya velum symbiosis, the symbiont-free bivalve M y a arenaria, and Thiomicrospira crunogena, a sulfur-oxidizing bacterium. In particular, we have attempted to identify potential endosymbiont lipid biomarkers. We report here the results of a study of the fatty acid, sterol, hydrocarbon, hopanol, and lipid class composition of these species. In addition the lipid carbon stable isotope ratios were analyzed to allow the lipid carbon source for each species to be determined.

Materials and methods Specimen collection Specimens of Solemya velum (Say) and Mya arenaria (L.) were collected from the top 10 cm of organically-rich [0.5 to 8.0% total organic carbon (TOC)] sediments at Little Buttermilk Bay, a small tidal inlet in Cape Cod, Massachusetts, during November 1988. Animals were either dissected so that gills (which contain the endosymbiotic bacteria), foot (excluding gonads), and visceral mass (defined here as all remaining soft parts excluding adductors) were frozen in liquid nitrogen, or were frozen whole. All samples were maintained at - 7 0 °C prior to analysis. Three specimens of each species were used in analysis of tissues; eight specimens of each species were used in the determination of lipid compositions for whole bivalves. Thiomicrospira crunogena was isolated from 2600 m at 21°N from the hydrothermal vent area near the East Pacific Rise by Jannasch et al. (1985) and viable cultures were maintained at - 7 0 °C. This species is a Gram-negative vibrioid bacterium 0.4 to 0.5/zm in width and 0.5 to 3.0 #m in length. Eight liters of culture were grown using the methods of Jannasch et al. (1985). Cells were harvested by centrifugation of 200 ml aliquots of the culture at ca. 1000 x g for 20 min. The bacterial pellets were combined and frozen at - 7 0 °C for 1 wk until analysis.

Lipid analyses Bivalve tissues were dried to constant weight at 60°C and were ground to a fine powder. Comparison of the fatty acid profiles of Mya arenaria with previous studies of this species and other marine bivalves (see Joseph 1972) suggests that oxidation of double bonds is unlikely to have occurred during the drying procedure. The lipids of 5 to 50 mg of bivalve tissue were extracted, according to the methods of Farrington et al. (1988), by successive sonication with isopropyl alcohol, chloroform:methanol (1 : 1), and chloroform: methanol (3 : 1) after the addition of two internal recovery standards (a C19 free fatty acid and a C21 fatty acid methyl ester). The dry weights of Thiomicrospira crunogena extracted could not be determined because of the large amounts of sulfur produced externally by these bacteria. Removal of the sulfur prior to lipid extraction was not attempted because of the low culture yield. Tubes were centrifuged after each extraction and the supernatants decanted and combined. Lipids were partitioned from other material in a separatory funnel by the addition of 1/6 by volume saturated NaC1 solution. Lipid extracts were concentrated and split into two fractions. Half the lipid extract was used for total fatty acid, sterol, hydrocar-

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

279

Table 1. Solemya velum (a) and Mya arenaria (b). Sterol composition of whole bivalves and individual tissues." All values in ng m g - ~ dry wt Sterol

(a) Solemya velum Cholest-5-en-3/~-ol b 24-Ethylcholest-5en-3/3-ol c Total (b) Mya arenaria 26,27-Dinorergosta-5,22-dien-3/%ol d Cholesta-5,17(20)-dien-3fl-ol Cholest-5-en-3fl-ol 5-e-Cholestan-3/?-ol ~ U n k n o w n Cholestadienol Cholesta-5,24-dien-3]~-ol f 24-Methylcholesta-5,24-dien-3/~-ol 24-Methylcholesta-5,22-dien-3/3-ol g 24-Ethylcholesta-5,22-dien-3/~-olh U n k n o w n 24-Ethylcholestadienol 24-Ethylcholest-5-en-3/~-ol c 24-Ethylcholesta-5,24(28)-dien-3/~-ol i U n k n o w n Sterols Total

Gill (n = 3)

Foot (n = 3)

Visceral mass (n = 3)

Whole animal (n = 8)

Mean

sd

Mean

sd

Mean

sd

Mean

6120 860

1030 370

5130 130

790 130

5540 180

360 50

5420 190

990 140

6980

1400

5260

920

5730

410

5610

1130

100 30 3130 300 580 180 750 510 240 160 280 140 320

4 5 230 16 2 46 150 19 10 100 43 9 20

120 33 2560 410 490 290 810 490 100 200 300 110 320

11 7 320 320 110 15 39 28 25 17 60 1 10

68 7 1850 0J 260 270 490 330 111 80 440 190 200

3 13 180 0 64 155 67 84 83 29 450 200 7

72 18 2030 12 320 190 560 270 88 120 240 77 200

8 11 370 23 117 37 95 110 23 26 170 43 29

6700

660

6200

960

4290

1330

4190

1070

sd

a The trivial names of most of the sterols identified are: b cholesterol; c possibly sitosterol; d identification based on similarity of sterol mass spectra with standard; e cholestanol; f possibly desmosterol; g possibly brassicasterol; h possibly stigmasterol; i fucosterol or isofucosterol; J below detection limits of I ng mg - ~ dry wt. (We were unable to distinguish between the possible epimers of some of the sterols; in these cases the most likely epimer is listed)

bon, hopanoid, and sterol analysis and was treated with 0.5 N K O H in methanol to saponify esterified fatty acids. Fatty acids were converted to fatty acid methyl esters (FAMEs) by reaction with 10% BC13-methanol. Lipid groups were then separated by fractionation on a silica gel column (7 g Bio-Sil A, 100-200 mesh, 5% deactivated with water in 0.9 x 35 cm columns) and packed over activated copper (to remove elemental sulfur). The remaining lipid extract was separated into (1) polar lipids, (2) monoglycerides, (3) free sterols, (4) combined diglycerides, triglycerides and free fatty acids, and (5) steryl and methyl esters by thin layer chromatography (TLC) on W h a t m a n LK5D Linear K Silica Gel TLC plates, in order to determine the fatty acid composition of the constituent lipid classes. The solvent system consisted of dichloroethane : chloroform : acetic acid : acetone : isopropyl alcohol (92 : 8:0.1:0.03 : 0.03). Lipid bands were determined by running authentic lipid class standards on each plate followed by visualization with 2', 7'-dichlorofluorescein. Lipid classes were scraped off each plate and eluted from the silica gel with methanol, toluene, and hexane. The fatty acids of each lipid class were then saponified and free fatty acids converted to FAMEs as described above. All F A M E samples were analyzed by gas chromatography by coinjection on a J & W Scientific D u r a b o n d DB-5 30 m fused silica column (0.32 m m i.d., 0.25 m m film thickness) and a J & W Scientific D u r a b o n d DB-Wax 30 m fused silica column (0.25 m m i.d., 0.25 m m film thickness). Both columns were mounted in Carla Erba 4160 high resolution gas chromatographs equipped with flame ionization detectors. Identification of individual F A M E s was based on a comparison of the retention times of authentic standards on both columns. If a substance coeluted with a standard on both columns it was assumed to be the same as the standard. The results for individual fatty acids from the column yielding the lowest concentration for that fatty acid were used in data interpretation. F A M E identifications were confirmed by electron-impact gas chromatography mass spectrometry (GC-MS) using a J & W Scientific DB-5 fused silica column installed in a Carla Erba high resolution gas chromatograph interfaced to a Finnigan 4500 quadrupole mass

spectrometer and controlled by an INCOS data system. The eis configuration of the double bond in 18 : lcol 1 was verified by coinjection with cis- and trans-vaccenic acid standards; these compounds are resolved on the DB-5 column. Alkane and hopanol fractions were analyzed on the DB-5 fused silica column. Sterol composition was determined by GC-MS on the system described above. Prior to analysis, the sterols were acylated by reaction with pyridine and acetic anhydride overnight. Recovery of the C19 free fatty acid (after conversion to the corresponding C~9 FAME) and the C21 F A M E internal standards generally exceeded 95% after saponification, esterification, silica gel column chromatography, and gas chromatography, and exceeded 85% after TLC, saponification, esterification, and gas chromatography, demonstrating almost complete sample recovery and conversion of fatty acids to FAMES.

Stable isotope ratios For 613C measurements, the fatty acids of Thiomierospira crunogena and fatty acid and sterol samples pooled from two specimens of each bivalve species were blown to dryness under nitrogen and sealed under vacuum while frozen. Samples were combusted at 900°C, followed by cryogenic combustion of CO z. Isotope measurements were determined using a Finnigan 251 ratio mass spectrometer in the laboratory of Dr B. Fry at the Marine Biological Laboratory in Woods Hole. Standards were high purity gases from commerical cylinders calibrated against National Bureau of Standards reference materials and 613C ratios are reported relative to Pee Dee Belemnite (PDB) using the standard delta notation: (~X = [(Rsample/Rstandard) - 1] X 10 3 ,

(1)

where X = 13C and R = 1 3 C / 1 2 C . In order to determine the effects of methylation on the 1~13C ratios of the fatty acids, 6~3C ratios of the 10% BCla-methanol

280

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

Solemya velum 100

60

M y a arenaria 50

80 6O 40

40

20 0

30-

20"

Sterol 10"

==

Fig. 1. Solemya velum and Mya arenaria. Sterol composition of whole bivalves (mean_+sd)

Sterol derivitization reagents were determined, along with the 61aC ratios of a C19 commercial fatty acid standard measured before and after methylation to a C19 FAME. The C19 fatty acid standard was chosen, as the length of this carbon chain is representative of the fatty acid carbon lengths found in the animal and bacterial samples.

Results

Alkanes, hopanols, and sterols Concentrations of n-alkanes and hopanols greater than those observed in blanks were not observed in either species of bivalve or Thiomicrospira crunogena. Additionally, no sterols were present in the bacterial samples. Although both species of bivalve contained sterols (Table 1, Fig. 1), significant iuterspecific differences in sterol composition were apparent. In intact S. velum samples cholest-5-en-3/%ol (cholesterol) accounted for almost 97% of the total sterol content (Fig. 1). In contrast, whole Mya arenaria specimens contained at least 13 different sterols with cholest-5-en-3/?-ol (48%), 24-methylcholesta-5,24-dien-3/%ol (13%), 24-ethylcholest-5-en3/~-ol (possibly sitosterol, 5.7%), an unknown cholestadienol (7.6%), and 24-methylcholesta-5,22-dien-3/%ol

Table 2. Thiomicrospira crunogena. Fatty acid composition (% of total > 0.1%). Each sample consisted of ca 1/4 of the culture yield. Absolute amounts of lipids cannot be reported as the large amounts of sulfur in each sample prevented weight determination. Unsaturation index is calculated as the number of double bonds present in a fatty acid times the total percent of that fatty acid present

Fatty acid

Sample I

Sample2

Average

14 15 16 16 17 18 18 18

1.0 1.2 24.8 43.6 1.9 6.7 1.4 19.6

1.0 1.3 23.9 42.7 2.0 7.5 1.6 19.9

1.0 1.3 24.3 43.2 2.0 7.1 1.5 19.7

:0 :0 :0 : 1co7 :0 :0 : 1o)9 : lo~7

Unsaturation index

64.4

(possibly brassicasterol, 6.4%) occurring in the largest amounts (Table I b, Fig. 1). The gills of S. velum contained the highest concentrations of sterols and the largest relative amounts of 24-ethylcholest-5-en-3/~-ol (12.3% of the total), whereas the foot tissue had the lowest concentrations. In M. arenaria, the gills also had the highest

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

281

Table 3. Solemya velum. Fatty acid composition. All values in ng m g - 1 dry wt Fatty acid

14 : 0 14 : lo)5 a-15 : 0 15 : 0 i-16:0 16 : 0 16 : lco7 a-17 : 0 17 : 0 18 : 0 i-18 : 0 18 : lo)9 18 : lco7 18 : 20)6 18 : 3d)3 18 : 3co6 18 : 40)3 20 : 0 20 : 1*a 20 : 109 20 : 10)7 20 : 2* 20 : 2O)6 20 : 3 * 20 : 3C03 20 : 4O)6 20 : 5co3 22:0 22 : 10)9 22 : 20)6 22 : 2* 22 : 3O)3 22: 3* 22 : 4C06 22 : 6C03 24:0 24 : 10)9 26:0 28:0 30:0 Total % _>1 unsaturation % >_2 unsaturations % > 3 unsaturations Unsaturation index

Whole animal

Gill

Mean ( n = 8 )

sd

1 450 22 27 64 2 3 630 4 690 37 58 1 530 29 950 8 070 1 430 52 12 32 27 0b 860 430 2 110 222 1 360 38 3 240 497 6 24 4 1 050 50 0 2 68 6 3 8 1 0

1 120 15 17 56 3 1 890 2 220 15 22 320 37 520 3 410 510 61 14 29 21 0 320 230 790 150 440 78 1 030 600 9 45 10 480 48 0 6 73 8 4 10 2 0 14 600

32 100

Mean(n=3)

Foot

Visceral mass

sd

Mean(n=3)

sd

Mean (n=3)

sd

840 35 41 73 6 4 720 7 020 36 80 2 180 130 860 10 800 1 540 120 12 58 38 0 670 940 2 950 202 1 630 86 3 620 22 18 47 12 1 190 46 0 100 19 4 7 0 0 0

330 20 20 23 10 1 300 2 290 13 23 900 110 410 4 730 360 210 17 70 40 0 100 850 1 660 290 950 68 1 270 31 13 81 20 470 64 0 92 17 4 12 0 0 0

580 36 31 260 0 2 360 2 360 43 64 1 250 19 700 5 000 1 310 230 11 47 61 0 910 200 1 770 62 1 510 0 1 770 270 26 41 30 1 120 86 0 67 200 21 23 74 0 0

220 10 7 280 0 500 750 16 26 400 34 440 1 540 420 400 11 19 57 0 330 140 820 110 62 0 670 250 23 39 53 550 100 0 120 350 36 24 43 5 0

1 190 9 8 24 0 2 350 2 900 34 43 1 190 15 1 110 6 200 1 460 0 11 49 16 0 800 300 1 690 180 1 240 130 2 770 580 9 47 0 1 160 32 0 0 91 6 0 4 0 0

770 11 9 21 0 570 630 17 15 140 26 940 2 570 660 0 19 47 15 0 310 150 620 160 520 220 1 150 360 8 82 0 590 60 0 0 63 11 0 0 0 0

40 100

16 900

22 500

8 850

25 600

10 700

78.8 32.5 17.2 142.9

78.6 28.3 13.9 130.3

79.0 38.3 18.9 149.7

80.4 36.6 19.1 153.8

Location of double bonds unknown b Below detection limit of 1 ng mg-1 dry wt

concentrations of sterols; the visceral mass had the lowest levels. The sterol content of intact S. velum samples (ca. 0.6% dry wt) was larger than that ofM. arenaria samples (ca. 0.4% dry wt) probably due to the large gills of S. velum. Fatty acids The fatty acids of Thiomicrospira crunogena (Table 2) were dominated by the three fatty acids 16" 0, 16" 1co7, and 18:lco7 (cis-vaccenic acid); 64.4% of the fatty acids

of this species had one double bond. These three dominant fatty acids in T. crunogena accounted for 87.3% of the total fatty acids present (Table 2) 1 and are the most commonly reported fatty acids in Gram-negative bacte1 The oJ classification of fatty acids refers to the position of the double bond closest to the methyl end of the fatty acid chain. For example, a C18 fatty acid with two double bonds, the closest of which is three carbon units from the methyl end of the lipid, is described as 18:2o)3. In the A-classification, the position of the double bonds are reported in relation to the carboxyl end of the fatty acid chain; thus the fatty acid just described would be 18:2 A12, 15

282

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

ria (Perry et al. 1979, Parkes and Taylor 1983, Gillan and Johns 1986). No fatty acids of greater than C~s in length were found and the iso and anteiso branched fatty acids characteristic of some species of Gram-negative bacteria (e.g. Legionella sp.) were not observed in T. erunogena. The fatty acids of Solemya velum and Mya arenaria differed significantly in terms of absolute and relative amounts. Differences within the tissues of each species were also apparent.

a

"~ ~, 20 "~ 10

Differences between bivalves The fatty acid concentrations of whole Solemya velum samples (Table 3) were greater than those of Mya arenaria (Table 4; mean values = 32.1#g m g - ~ dry wt and 16.1 yg rag- ~ dry wt, respectively). In S.velum, the high lipid content is primarily due to the presence of extremely large quantities of eis-vaccenic acid (18:1o)7) which accounted for between 0.5 and 1.1% of the tissue dry wt. In whole S. velum samples, the only three fatty acids occurring in concentrations exceeding 10% of the total are the three fatty acids 18 : 1o)7, 16: lco7, and 16:0, which account for > 50% of the entire fatty acid pool (Fig. 2). The fatty acid 18:2o)6 also occurred in relatively high concentrations in S. velum. The only polyunsaturated fatty acid (PUFA) found in high concentrations in S. velum was arachidonic acid (20 : 4o)6) which occurred in average concentrations of 3.2 yg m g - ~ dry wt in whole animal tissues (Table 3). In contrast, the major fatty acid present in M. arenaria was the PUFA 22 : 6~o3 (Table 4) which occurred in concentrations of 3.1 #g m g - ~ dry wt (Table4). Palmitic acid (16:0) and eicosapentaenoic acid (20 : 5co3) were the next most abundant fatty acids in M. arenaria, occurring in concentrations of 2.7 and 2.1 #g m g - ~ dry wt, respectively. These three fatty acids accounted for 48% of the total fatty acids in M. arenaria, although a 20 : 1 fatty acid also occurred in relatively high concentrations (Fig. 2 b). The ratio of 16 : 1 to 16 : 0 differed between the two bivalves. In both S. velum and the bacterial samples, there were larger amounts of 16 : lco7 than 16 : 0, whereas the opposite was true for M. arenaria. It is interesting to note the differences in the degree of unsaturation between the fatty acids of the two species. In both species, 70 to 80% of the fatty acids have at least one double bond (Tables 3 and 4) a typical pattern in marine bivalves (see Gardner and Riley 1972). However, greater than 40% of the fatty acids in Mya arenaria, but only about 17% of the fatty acids in Solemya velum, have at least three double bonds (Tables 3 and 4). This accounts for the large differences in unsaturation index 2 between the two species. Yet, the absolute concentrations of PUFAs present in both bivalve species are similar because of the higher overall concentrations of fatty acids in S. velum. The branched iso and anteiso fatty acids occurred in only small amounts in both species. 2 A measurement of the proportion of unsaturated fatty acids present, calculated as the sum of the number of unsaturations present per fatty acid times the percentage of the total fatty acids represented by that fatty acid

Solemya velum

30.

+Ill,++

, mmm m

Fatty Acid 30

b M y a arenaria

"~ 20

~. t0.

0

Fatty Acid Fig. 2. Solemya velum (a) and Mya arenar• (b). Fatty acid composition expressed as a percent of the total fatty acids found (mean_ sd). For clarity only whole specimen values are plotted. * : position of double bonds unknown

Differences between the total fatty acid content of individual animals often varied by as much as 50% resulting in the large standard deviation values calculated in Tables 3 and 4. It is important to note, however, that the relative proportions of the individual fatty acids of each species only varied by a factor of ca. 5 to 20%. This is seen when the fatty acids of each species are plotted as a percentage of the total fatty acids present (Fig. 2). Although only values for whole animals are plotted for clarity, the standard deviation values for relative fatty acid composition of individual tissues are also < 20%.

Differences between tissues In both species, the gills contained the highest fatty acid concentrations (Tables 3 and 4) with the mean values

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

283

Table 4. Mya arenaria. Fatty acid composition. All values in ng m g - 1 dry wt Fatty acid

14 : 0 14 : 1095 a-15 : 0 15 : 0 i-16 : 0 16 : 0 16 : lo97 a-17 : 0 17 : 0 18 : 0 i-18 : 0 18 : lo99 18 : lo97 18 : 2096 18 : 3o93 18 : 3096 18 : 4o93 20 : 0 20 : 1' a 20 : lo99 20 : lo97 20 : 2* 20 : 2o96 20 : 3 * 20 : 3o93 20 : 4096 20 : 5o93 22:0 22 : lo99 22 : 2o96 22 : 2* 22 : 3o)3 22 : 3 * 22 : 4096 22 : 6o93 24:0 24 : 1099 26:0 28:0 30:0 Total % > 1 unsaturation % _>2 unsaturations % > 3 unsaturations Unsaturation index

Whole animal

Gill

Foot

Visceral mass

Mean(n=8)

sd

Mean(n=3)

sd

Mean(n=3)

sd

Mean(n=3)

sd

240 10 16 100 42 2 730 790 110 220 800 70 1 010 390 160 51 9 210 32 1 320 530 420 170 230 120 9 780 2 060 7 14 5 4 28 300 60 3 100 3 1 3 1 0b

53 6 6 21 26 360 200 37 37 240 45 180 100 130 33 8 62 35 320 120 200 35 34 46 12 170 450 9 10 14 11 17 120 60 740 6 2 5 2 0

190 24 18 130 67 3 540 700 200 250 1 000 130 840 370 150 88 20 250 12 3 960 1 030 820 390 250 660 27 1 700 3 520 7 82 26 0 150 450 180 4 630 7 27 2 5 0

24 10 2 2 35 460 120 28 220 120 130 190 81 37 5 21 120 22 590 78 250 340 220 420 23 540 1 140 11 62 44 0 190 390 152 880 12 41 4 8 0

370 40 42 200 110 4 200 i 270 200 360 1 180 120 1 820 420 200 63 8 230 55 2 100 1 100 850 170 320 100 0 1 070 2 350 0 13 0 0 94 340 220 3 010 0 5 7 0 0

35 24 2 10 100 240 15 59 19 110 41 160 17 24 24 15 35 20 280 310 270 160 50 100 0 450 800 0 23 0 0 84 260 140 1 190 0 8 12 0 0

210 8 15 90 30 2 490 780 110 200 690 84 900 360 130 58 8 160 18 1 140 490 480 150 200 64 0 610 1 830 6 15 12 12 10 180 37 2 440 0 0 5 0 0

33 3 7 15 12 160 83 55 15 71 10 49 27 6 13 9 90 15 72 I00 260 22 7 60 0 180 710 10 13 21 20 17 160 64 1 060 0 0 9 0 0

3 940

25 900

6 990

22 600

5 080

14 000

3 460

16 100

72.8 42.2 41.3 247.8

78.4 47.8 41.8 259.4

69.5 35.8 32.7 204.6

71.5 41.1 37.5 232.1

a Location of double bonds unknown b Below detection limit of 1 ng mg-1 dry wt

f o r SoIemya velum a n d Mya arenaria gills b e i n g 40.1 a n d 25.9 # g m g 1 d r y wt, r e s p e c t i v e l y . T h i s is d u e t o t h e h i g h m e m b r a n e c o n t e n t o f b i v a l v e gills. T h e h i g h t o t a l f a t t y a c i d c o n c e n t r a t i o n s o f S. velum a r e p a r t l y d u e to t h e l a r g e c o n t r i b u t i o n o f t h e h y p e r t r o p h i e d gills to t h e s o f t tissue w e i g h t o f this species ( b e t w e e n 25 a n d 4 0 % o f t h e tissue w e i g h t ) a n d p a r t l y d u e to t h e l a r g e c o n c e n t r a t i o n s o f e n d o s y m b i o n t s in gill b a c t e r i o c y t e s . I n t h e gills o f S. velum, t h e c o n c e n t r a t i o n s o f 1 6 : 1 0 ) 7 a n d 18:1097 w e r e p a r t i c u l a r l y h i g h ( T a b l e 3); in fact, in s o m e s p e c i m e n s a l m o s t 1 . 5 % o f t h e d r y w e i g h t o f t h e gills w a s eis-vaccenic

acid. T h e f o o t tissue o f S. velum c o n t a i n e d t h e l o w e s t values, an expected finding considering the muscular n a t u r e o f this tissue. I n M. arenaria, t h e v i s c e r a l m a s s a n d not the foot had the lowest absolute fatty acid concentrat i o n s ; this m a y b e d u e to t h e p r e s e n c e o f n o t i c e a b l e g r a i n s o f s e d i m e n t in t h e gut, w h i c h c o u l d c o n t r i b u t e to t h e w e i g h t o f t h e g u t , b u t n o t to t h e f a t t y a c i d c o n c e n t r a t i o n . T h e gills o f M. arenaria c o n t a i n e d slightly g r e a t e r a m o u n t s o f 20 : 4096, 20 : 5(o3, 22 : 6093, 20 : 1, a n d 20 : 3 f a t t y a c i d s ( T a b l e 4). T h e gills o f S. velum h a d t h e l o w e s t tissue u n s a t u r a t i o n i n d e x o f t h e t h r e e S. velum tissue f r a c t i o n s ,

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

284

Table 5. Solemya velum. Mean fatty acid and sterol concentrations of the major lipid classes analyzed a. All values in ng mg 1 dry wt; n=4

Table 6. Mya arenaria. Mean fatty acid and sterol concentrations of the major lipid classes analyzed ~. All values in ng rag-1 dry wt; n=4

Fatty acid

Fatty acid

14 : 0 16 : 0 16 : 1097 18 : 0 18 : 1099 18 : 1097 18 : 2093 18 : 4o93 20 : 1099 20 : 1097 20 : 2 *c 20 : 20)6 20 : 4co6 20 : 5093 22 : 2096 22 : 2* Total

Phospholipids

Monoglycerides

Triglycerides Sterylesters diglycerides and methyl and free esters b fatty acids

Mean

Mean

Mean

sd

840 550 1 970 200 2 180 350 720 120 480 300 4 320 1300 720 440 18 11 330 100 98 56 550 430 160 240 1 320 870 220 290 140 240 530 470

150 440 230 220 120 490 84 19 61 45 56 50 160 50 13

sd

sd

97 410 310 170 1 520 1 140 300 660 550 74 700 510 58 560 420 654 1380 1370 93 180 140 42 46 50 55 230 130 42 120 95 120 130 170 71 36 66 160 250 210 110 42 83 29 56 120 44 98

14 600 5980 2190 2060 6340 5460

% of total 39.4-t-4.5 animal fatty acids

6.5_+3.2

23.2-t-4.7

Phospholipids

Monoglycerides

Triglycerides Steryl esters diglycerides and methyl and free esters b fatty acids

Mean

sd

Mean

sd

Mean

sd

14 : 0 16:0 16 : 1097 18:0 18 : 1099 18 : lo97 18 : 2093 20 : 1 .c 20 : lo99 20 : 1o97 20 : 2096 20 : 4o96 20 : 5093 22 : 6093

120 1840 280 420 610 180 71 650 130 230 62 230 390 680

32 43 31 300 2 72 85 130 36 130 37 19 49 39 420 13 31 40 72 20 38 39 170 6 160 39 300 86

15 52 40 25 49 17 24 23 23 2 67 11 12 7

200 1520 250 460 440 200 91 84 140 120 12 180 680 630

63 430 91 84 120 53 35 120 55 46 16 150 260 230

Total

5877 1450 975

367

4990

Mean sd 250 730 310 400 300 670 80 17 130 78 11 27 220 57

200 530 410 150 160 590 77 30 150 95 24 31 160 100

23

51

33202770

% of total 42.5-1-0.5 animal fatty acids

6.6-t-1.7

Mean sd 130 730 200 270 360 64 51 61 57 130 9 27 96 164

150 560 290 140 360 54 40 120 45 150 6 26 97 140

1 7 5 0 2350 2180

32.9_+14.6

15.3_+11.3

12.2_+11.3

a Only fatty acids exceeding 1% of the total are reported b Some fatty acids are methylated during the lipid extraction procedure; these methyl esters coelute with the steryl esters during TLC Location of double bonds unknown

w h e r e a s in M. arenaria the u n s a t u r a t i o n i n d e x o f the gills was the highest o f the tissue m e a s u r e m e n t s (Tables 3 a n d 4).

F a t t y acid c o m p o s i t i o n o f the m a j o r lipid classes T h e r e were n o significant differences b e t w e e n the distrib u t i o n o f f a t t y acids a m o n g the lipid classes o f b o t h b i v a l v e species (Tables 5 a n d 6). In b o t h species o f bivalve, ca. 4 0 % o f the f a t t y acids were f o u n d in the p h o s p h o l i p i d p o o l , 30 to 4 0 % in the c o m b i n e d triglyceride, diglyceride a n d free f a t t y a c i d fraction, 12 to 15% in the steryl a n d m e t h y l ester p o o l , a n d a b o u t 6 % in the m o n o g l y c e r i d e s (Tables 5 a n d 6; Fig. 3). T h e r e were slightly lower a m o u n t s o f f a t t y acids in the tri- a n d diglyceride f r a c t i o n s o f Solemya velum s a m p l e s t h a n in Mya arenaria. A p r o p o r t i o n o f the f a t t y acids in the steryl ester f r a c t i o n is p r o b a b l y due to the f o r m a t i o n o f F A M E s b y r e a c t i o n w i t h m e t h a n o l d u r i n g the lipid e x t r a c t i o n process, a n a r t i f a c t t h a t c a n n o t be a v o i d e d if c h l o r o f o r m : m e t h a n o l e x t r a c t i o n solvents are used. F A M E s c o e l u t e with steryl esters using the T L C m e t h o d o l o g i e s u s e d here. I n S. velum, the p h o s p h o l i p i d f r a c t i o n g e n e r a l l y h a d relatively larger a m o u n t s o f the f a t t y acids 20:4096, 16:1o97 a n d 18 : lo97 a n d less 14: 0, 1 6 : 0 a n d 18 : lo99 t h a n the o t h e r lipid classes (Fig. 3 a). T h e f a t t y acid c o m p o s i t i o n

a Only fatty acids exceeding 1% of the total are reported b Some fatty acids are methylated during the lipid extraction procedure; these methyl esters coelute with the steryl esters during TLC c Location of double bonds unknown

o f the o t h e r lipid classes o f S. velum were very similar (Fig. 3 a). In M. arenaria the p h o s p h o l i p i d s h a d relatively l a r g e r a m o u n t s o f 20: 1, w h e r e a s the tri-, diglyceride, a n d free f a t t y acid f r a c t i o n s h a d a large p e r c e n t a g e o f the fatty acids 20:5co3 a n d 22:6co3 (Fig. 3b). I n M. arenaria, P U F A s are f o u n d d i s t r i b u t e d a m o n g all the lipid classes, w h e r e a s in S. velum, the p r e d o m i n a n t P U F A 20 : 4096, is m o r e localized in the p h o s p h o l i p i d p o o l . Sterol esters a c c o u n t e d for ca. 8 % o f the sterols o f S. velum a n d 4 to 5 % o f the sterols o f M. arenaria (Table 7). T h e r e was c o n s i d e r a b l e v a r i a b i l i t y b e t w e e n i n d i v i d u a l s o f each species in b o t h a b s o l u t e a n d relative lipid class f a t t y acid c o m p o s i t i o n (Fig. 3, Tables 5 a n d 6).

Stable i s o t o p e values F a t t y acid a n d sterol 613C r a t i o s for all the species e x a m ined are s h o w n in Table 8, a l o n g w i t h w h o l e a n i m a l isot o p e values, a n d 613C r a t i o s o f the m e t h y l a t i o n r e a g e n t s a n d the C19 f a t t y a c i d a n d d e r i v a t i z e d C19 F A M E . I n all cases, lipid 6 1 a c r a t i o s were c o n s i d e r a b l y m o r e negative t h a n c o r r e s p o n d i n g w h o l e tissue values, p a r t i c u l a r l y in Solemya velum a n d Thiomicrospira erunogena. Differences in the lipid i s o t o p e r a t i o s b e t w e e n the two bivalves were e x t r e m e l y large (Table 8). T h e lipid i s o t o p e r a t i o s in S. velum were b e t w e e n 14 a n d 22%0 m o r e negative t h a n

285

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum 40"

a

Table 7. Solemya velum and Mya arenaria. Free and esterified sterol composition

Solemya velum

30

• [] l~ []

[-

Composition (ng rag- 1 wet wt) (n = 4)

Phosphotipids Monoglycerides Tfi-, Diglycerides + FFAs Steryl Esters

"~ 20-

% of total sterols

Mean

sd

6004 520

520 170

92.0 8.0

2360 103

380 75

94.9 5.1

2480 97

640 82

96.1 3.9

4840 220

920 150

95.6 4.4

Solemya velum Cholesterol Free Steryl esters

10'

Mya arenaria Cholesterol Free Steryl esters Other Sterols Free Steryl esters Total sterols Free Steryl esters

Fatty Acid 40

b

Mya arenaria

30

• Phospholipids [ ] Monoglycerides [ ] Tri-, Diglycerides + FFAs [ ] Steryl Esters

~,~

Table 8. Solemya velum, Mya arenaria and Thiomicrospira erunogena. (a) Lipid 613C ratios from Conway et al. (1989) (parts per thousand relative to PDB) and (b) 6~aC ratios of the BC13-methanol used in fatty acid methylation and a commercial C19 fatty acid standard before and after mcthylation to a Cz9 FAME

20

(a)

Intact tissue 613C*

Fatty acids 613C

Sterols 313C

S. velum M. arenaria T. crunogena

-31.5 to -33.9%0 - 17.1 to - 17.8%o --27.3%0

-45.4%0 -23.8%0 --45%0

-38.5%0 -24.2%0

0

0

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

¢q

.

.

¢-I

.

.

¢q

.

¢-q

Fatty Acid Fig. 3. Solemya velum (a) and Mya arenaria (b). Fatty acid composition of the lipid classes of each species expressed as a percent of the total fatty acids found in each lipid class (mean _+sd). FFA: free fatty acids; *: position of double bonds unknown

corresponding values for M y a arenaria. Furthermore, the extremely negative 313C ratios for S. velum fatty acids (-45.4%0) were very similar to those found in the sulfuroxidizing bacterium T. crunogena (-45.0%0). Although the b~3C ratios of the 10% BCla-methanol derivitization reagents were extremely negative (-44.7%o), methylation of the animal and bacterial fatty acids is unlikely to have affected the 3~3C ratios of the resulting FAMEs by more than -1%o as evidenced by the small effect of methylation on the 6A3C ratio of the commerical C19 fatty acid standard (Table 8).

.

* Conway et al. 1989 (b)

613C

BC13-methanol C19 Fatty acid C19 FAME after BC13-methanol esterification

- 44.7%o -26.8%o -27.5%o

animal can often be inferred from an examination of its lipid composition. The lipid composition of an organism may be particularly informative in cases where only a few animals are available for study, and laboratory feeding studies are not possible, as is the case with many animalbacteria symbioses. The lipid composition of Solemya velum is unusual, more closely resembling that of bacteria than typical marine bivalves with regard to both fatty acid composition and 613C ratios.

Sterols Discussion

The lipid content of many marine bivalves is related to dietary intake (Moreno et al. 1980, Piretti et al. 1987), particularly in the case of PUFAs and sterols, as many of these lipids are only synthesized de novo by algae and higher plants. Consequently, aspects of the diet of an

Reviews of the sterol composition of molluscs (e.g. Voogt 1972, G o a d 1976) demonstrate that marine molluscs contain complex mixtures of C26 to C3o sterols, with cholesterol (C27) generally occurring in the highest concentrations. Many of the sterols found in marine molluscs are commonly synthesized by plants and algae and are presumed to be derived from the diet. Data on sterol

286

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

synthesis in bivalves are controversial. Although it is generally stated that bivalves are incapable of synthesizing sterols, Fagerland and Idler (1960) reported the incorporation of radiolabelled acetate in the sterols of Mytilus californianus and Saxidomus giganteus, and Teshima and Patterson (1981) demonstrated the synthesis of cholesterol, desmosterol, isofucosterol, and 24-methylenecholesterol from acetate in Crassostrea virginica. Teshima and Kanazawa (1974) reported the synthesis of cholesterol, 22-dehydrocholesterol, desmosterol and 24-methylenecholesterol from mevalonate in Mytilus edulis. Voogt (1975), however, did not recover any radioactivity in the sterols of M. edulis after the addition of labelled acetate, and the majority of bivalves studied appear to be capable of only limited de novo sterol synthesis. Thus, the majority of the sterols in Mya arenaria probably reflect an algal-based diet, with perhaps some in vivo sterol interconversion. The extremely limited range of sterols in Solemya velum is unusual and suggests a diet deficient in plant and algal material. Little information exists in the literature concerning the sterol composition of protobranchs; cholesterol was the only sterol found in S. velum by Idler and Wiseman (1971), although Nucula sp. had eight different sterols. The 24-ethylcholest-5-en-3fi-ol found in S. velum in our study may reflect the ingestion of small amounts of plant-derived particulate organic matter. Overall, however, S. velum does not appear to be incorporating significant amounts of the sterols of exogenous particulate organic matter. This is further suggested by the extremely low (-38.5%o) 613C ratios found for the sterols of S. velum which would appear to have been synthesized using isotopically depleted carbon derived from bacterial chemosynthesis. The utilization of exogenous particulate matter would most probably result in stable isotope ratios closer to those of Mya arenaria (-24.2%0) and other marine animals and algae ( - 1 1 . 2 to -22.8%0; Kokko et al. 1984). The origin of the high cholesterol concentrations in S. velum is uncertain. S. velum or the endosymbionts may be capable of cholesterol synthesis. If in vivo sterol biosynthesis occurs in the animal tissues of the S. velum symbiosis, then it is likely that the carbon is derived from compounds initially produced by the bacterial endosymbionts. This would explain the unusually low 613C ratios found in the sterols of S. velum. There have been several reports since 1967 of sterolcontaining bacteria (McCaffrey et al. 1989 and references therein). Sterols and hopanoids are both synthesized from squalene but steroid synthesis is considered to be evolutionarily more advanced and associated with eukaryotes. The low sterol carbon isotope ratios of Solemya velum may be the result of sterol synthesis by the endosymbionts, or synthesis of sterol precursors by the symbionts followed by conversion of these precursors to sterols by the animal host. S. velum and Mya arenaria could also derive some of their sterol requirements by ingestion and direct epithelial uptake of dissolved sterols. Dissolved sterols (predominantly cholesterol and 24ethylcholestrol, mostly derived from plant material) have been reported in seawater samples (Gagosian 1975, Goad 1976 and references therein). Uptake of dissolved sterols

has been reported for some bivalves (Piretti et al. 1989). This phenomenon is unlikely to occur to a large extent in S. velum, since the very negative 613C ratios for S. velum sterols are inconsistent with photosynthetically derived carbon sources, although small amounts of sterols could be provided in this way. Fatty acids The fatty acid composition of Thiomicrospira crunogena, an obligate chemoautotroph (Jannasch et al. 1985), was similar to those of the Group III obligate Thiobacilli subdivision of Katayama-Fujimura et al. (1982) and Thioploca sp. (McCaffrey et al. 1989, Table 9). Bacteria can synthesize monounsaturated fatty acids (MUFAs) by using one of two biochemical pathways, which generally appear to be mutually exclusive (Goldfine 1972, Fulco 1983) although a few species are reported to utilize both pathways (Wada et al. 1989). The anaerobic pathway, found only in bacteria, proceeds in the absence of oxygen and produces long chain MUFAs by the elongation of medium chain length cis-3-unsaturated intermediates. Generally, the anaerobic pathway leads to the formation of cis-vaccenic acid (18 : 1o)7) as the major end-product, with 16:1o)7 produced as a secondary product. In contrast, the aerobic or O2-dependent MUFA synthesis pathway, which is found in some bacteria and all eukaryores, results in a large variety of fatty acids and utilizes 16 : 0 or 18 : 0 as preferred substrates with double bond insertion occurring in the A9 position (in some species of bacteria, A5, 8, 9 and 10 isomers can also be found). Because the two pathways appear to be mutually exclusive, we can infer from the high concentrations of eis-vaccenic acid in T. crunogena and other sulfur-oxidizers examined to date (e.g. Thioploca spp., McCaffrey etal. 1989, Table9; Thiobaeillus sp., Katayama-Fujimura et al. 1982) that the anaerobic MUFA synthesis pathway operates in these, and perhaps all, sulfur-oxidizing bacteria. Three major differences are apparent when the fatty acids of both bivalve species are examined. (1) Solemya velum, like Thiomierospira erunogena, contained extremely large amounts of the MUFAs 16 : lo)7 and 18 : io)7 (the fatty acid predominantly produced by bacterial anaerobic MUFA synthesis), while Mya arenaria had only small amounts of these lipids. (2) The 613C ratios of the fatty acids of S. velum (-45.4%o) were similar to those of Thiomicrospira crunogena ( - 45%0) but differed markedly from the fatty acid 613C ratios ofM. arenaria (-23.8%o). (3) M. arenaria contained high concentrations of PUFAs while the PUFA content of S. velum was low. The cis-vaccenic acid (18: 1o)7) found in S. velum is almost certainly bacterial in origin; when the morphology of S. velum is considered (very small digestive system, large hypertrophied gills containing high concentrations of chemoautotrophic bacteria) it is highly likely that the cis-vaccenic acid is derived from the endosymbionts of this species. Similarly, the greater amounts of 16:1o)7 relative to 16:0 found in S. velum probably reflect the incorporation of endosymbiont lipids. Comparison of

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

287

Table 9. Solemya velum and Mya arenaria. Total lipid fatty acid composition compared with sulfur-oxidizingbacteria and four other bivalves. (Fatty acids presented as % of total fatty acids> 1%) Fatty acid a

Thioploca sp. b

14 : I 14 : 0 15 : 0 0.2 16 : 1 41.4 16 : 0 17.7 17 : 0 0.1 18 : 4 18 : 3 18 : 2 18 : lco9 0.4 18 : lco7 36.9 18 : 0 0.7 20 : 5 20 : 4 20 : 3 20 : 2 20 : 1 0.1 22 : 6 22 : 5 22 : 4 22 : 3 22 : 2 22 : 0 " b c a

Thiomicrospira crunogena e

Solemya velum c

Mya arenaria c

1.0 1.2 43.6 24.8 1.9

4.1 0.3 14.4 11.0 0.2 0.1 0.2 4.6 2.9 24.8 5.2 1.4 10.6 4.5 7.2 4.0 0.2

1.5 0.1 0.6 4.9 17.0 1.4 1.3 0.4 1.0 6.2 2.4 5.0 12.7 4.8 0.8 2.5 14.3 19.0

0.2 3.4

0.3 2.0 0.1

1.4 19.6 6.7

Mya arenaria d

Mytilus edulis d

Solen strictus d

Mercenaria mercenaria a

Crassostrea virginica d

2.2 0.1

3.2

1.8

2.0

3.5

4.5 12.5

7.5 15.9

1.2 16.3

4.8 12.3

4.2 28.9

1.8 1.4 1.9 6.2

1.9 1.1 1.1 11.1

2.5 1.4 3.2 10.6

1.1 0.9 0.2 8.4

3.0 11.2 5.5 0.8 1.2 1.1 13.4 0.5 3.8

3.7 10.2 3.8 0.2 0.3 7.7 13.7 0.2

3.5 7.0 4.6 0.3 1.5 4.4 14.0 2.1 1.0

5.5 18.3 3.6 0.1 0.2 8.6 15.0 2.0 0.4 0.1

2.6 3.3 2.0 3.9 4.3 3.6 11.2

7.1 9.7

4.1 1.4

Positional isomers are not reported for the majority of fatty acids as they were not determined in many of the studies cited From McCaffrey et al. (1989) This study See references in Joseph (1982)

the fatty acids of S. velum with those of other bivalves and sulfur-oxidizing bacteria reveals the presence of these strong "bacterial" signatures in S. velum (Table 9). The small amounts of cis-vaccenic acid found in M y a arenaria taken concurrently from the same sediments indicates that the large amounts in S. velum are unlikely to be the result of external sediment contamination. Cis-vaccenic acid may be an important biomarker for animal-bacteria symbiosis and may be useful in screening other marine organisms for the presence of bacterial symbionts. In this context, other invertebrate-chemoautotroph symbioses recently examined in this laboratory have been found to contain large amounts of cis-vaccenic acid, demonstrating the potential of this endosymbiont biomarker as a screening technique (Conway 1990, Giere et al. 1991). The low 6a3C ratios found in the fatty acid pool of S o l e m y a velum, and the similarity between the lipid 513C ratios in S. velum and Thiomicrospira crunogena, also demonstrate the utilization of bacterial lipids by S. velum. Lipids are generally depleted in 13C relative to the bulk organic carbon of most marine species (Prahl and Meuhlhausen 1989). In marine plankton 613C ratios of total lipids differ from c~3C ratios of plankton total organic carbon by ca. -9.5%o (Prahl and Meuhlhausen 1989), whereas in Escherichia eoli, total lipid 613C ratios differ from total organic carbon values by ca. -7.1%o (Monson and Hayes 1982). Thus, the very negative 613C ratios found in the fatty acids of S. velum suggest that the fatty acids of this species are synthesized using isotopicatly-light carbon (ca. - 3 0 to -40%o), such as that pro-

duced by chemoautotrophic bacteria (Degens 1969, R u b y et al. 1987). These lipids may either be synthesized by the chemoautotrophic symbionts of S. velum and transported to the host (as would appear to be the case for cis-vaccenic acid), or may be synthesized by S. velum using carbon derived from the bacterial symbionts. Gillan et al. (1988) used the presence of bacterial fatty acids in sponges to determine the contribution of symbiotic bacteria to total sponge biomass. In our study, however, we found large concentrations of 18 : 1co7 in the foot tissue of S o l e m y a velum, where endosymbionts have not been found (Cavanaugh 1983). This suggests that this bacterial lipid is being transported to, and utilized by, non-bacterial cells, and concentrations of bacterial lipids are not indicative of bacterial population size. In fact, if we use the techniques of Gillan et al. (1988) to calculate the contribution of eis-vaccenic acid to the biomass of S. velum, we estimate that the bacteria account for ca. 26% of the animal's biomass. This is a conservative estimate as it assumes that the bacteria only contain one fatty acid, yet it is orders of magnitude greater than the actual biomass of bacteria in S. velum which can be estimated using the bacterial counts of Cavanaugh (1983). Thus it is likely that the host tissue is incorporating lipids produced by the bacterial endosymbionts. In addition, the presence of 18 : lo)7 in all the major lipid classes examined, particularly the phospholipid and triglyceride pools (Table 5), suggests that this lipid plays a significant role in all aspects of the lipid metabolism of S. velum.

288

N. Conway and J. McDowellCapuzzo: Bacterial lipids in Solemya velum

In an ideal mutualistic symbiosis, both members of the association must evolve the means to utilize the metabolic products of each other. The incorporation of bacterial lipids into all the major lipid classes and all the animal tissues of the S. velum symbiosis suggests that a fight metabolic coupling exists between the two members of this symbiosis. These results suggest that specific bacterial lipids, such as cis-vaccenic acid, are translocated from endosymbiont to host and utilized intact. It is likely that in S. velum, cis-vaccenic acid performs the physiological roles normally filled by oleic acid (18:1039) in bivalves, including energy storage and production and membrane structural functions. Transfer of bacterial membrane material from endosymbiont to host is probably the source of the bacterial lipids found in Solemya velum. DeBurgh et al. (1989) suggest that the membrane surrounding the bacteria in invertebrate-chemoautotroph symbioses may play a role in regulating nutrient transfer. Portions of the bacterial membrane appear to be pinched off in the Lucinoma aequizonata symbiosis (R. Vetter personal communication). Furthermore, the accumulation of lipids under the bacterial zone of gill cells has been suggested by transmission electron microscopy (Fiala-M6dioni et al. 1984); abundant lipid inclusions have been seen in the gill cells of both the Bathymodiolus thermophilus and Calyptogena phaseoliformis symbioses, and Fiala-M6dioni and LePennec (1987) suggest that these lipid inclusions might be related to the endosymbiont metabolism and represent some type of reserve for the host. In addition, FialaM6dioni et al. (1986) demonstrated the incorporation of 7 to 15% of the radiolabel in the lipidic fraction of B. thermophilus after incubation in seawater containing radiolabelled bicarbonate. Our study provides evidence to suggest the accumulation of endosymbiont lipids by the bivalve host and incorporation of these lipids into all the lipid class pools of this species. We can deduce from the lipid data for S. velum that the bacterial lipids do not appear to undergo desaturation and elongation to other fatty acids, as only low concentrations of other fatty acids of the co7-series were present 3. Solemya velum tissues contained smaller amounts of the PUFAs commonly found in marine bivalves (20 : 5033 and 22:6co3, Table 3), although the PUFA 20:4co6 occurred in high concentrations. Monounsaturation of fatty acids in animal cells is achieved with a A9 desaturase, which has a maximum affinity for stearic acid (18 : 0) and results in 18 : 1A9 (18 : 1039). When PUFAs are formed, additional double bonds can only be introduced between the carboxyl group and the nearest existing double bond. Therefore, de novo animal PUFAs are unlikely to have an co number less than 7. Yet animals require 033 and 036 PUFAs, and these essential fatty acids (EFAs) must be included in the diet. EFAs are vital components of membranes, where they are important 3 Animalfatty acid synthetasescannot insert a double bond into a fatty acid closer to the methylend of the carbon chain than the co7 position; consequently,althoughcis-vaccenicacid could be elongated to 20:1, further desaturations would not occur closer to the methyl end of the chain, and the new fatty acid would still be a member of the co7 series

regulators of membrane fluidity and play a role in membrane transport processes. Additionally, they are precursors for prostaglandin synthesis and act as energy sources in a manner similar to saturated and monounsaturated fatty acids (Guarnieri and Johnson 1970 and references therein). EFAs are abundant in marine phytoplankton and higher plants, yet they are absent in the vast majority of bacteria studied to date (Goldfine 1972, Phillips 1984). These PUFAs have been recorded in some bacteria (Johns and Perry 1977, DeLong and Yayanos 1986); however, they can only be synthesized using the aerobic fatty acid synthesis system described previously. The high concentrations of cis-vaccenic acid in Solemya velum suggest that the bacterial endosymbionts of this species utilize the anaerobic MUFA synthesis pathway and are thus unlikely to synthesize 6o3 and 036 PUFAs. Thus, while Mya arenaria and other typical bivalves will ingest adequate amounts of these fatty acids in the diet, species such as S. velum, with reduced digestive capabilities, would appear to be limited in their ability to obtain EFAs in the diet. In most marine bivalves, the PUFAs 20 : 5033 and 22:6033 predominate (Table 9) and the fatty acid profiles reported here for M. arenaria are in remarkably good agreement with those of other bivalve molluscs (Table 9). S. velum had only trace amounts of the 033 fatty acids, an observation consistent with the hypothesis that the endosymbionts serve as the primary nutritional source for the symbiosis. The PUFA composition of Solemya velum raises two questions. How does S. velum function with a limited supply of EFAs, and what is the source of the EFAs found in S. velum tissues? The high concentrations of 033 PUFAs in most marine bivalves studied indicate that these PUFAs are very important in these organisms. The relatively large amounts of 20 : 4036 in S. velum suggest that this fatty acid must fulfill some of the functions normally associated with 20:5033 and 22:6033 in most other bivalves, including maintenance of membrane fluidity. Localization of 20:40)6 in the phospholipid fraction of S. velum may be a PUFA conservation strategy to ensure proper membrane functioning, as the lipids of S. velum have a relatively smaller proportion of PUFAs than M. arenaria and other bivalves. The high concentrations of eis-vaccenic acid in the phospholipid pool of S. velum suggests that this lipid is also important in membrane structure. In many bacteria cis-vaccenic acid is the major MUFA present and is an important component of bacterial membranes. Concentrations of cis-vaccenic acid may be increased in bacteria during acclimation to low temperatures, ensuring maintenance of membrane fluidity (Okuyama et al. 1977). In animal systems two different strategies are used to ensure membrane fluidity during low temperature exposure (Hazel and Sellner 1979 and references therein). In most cases, the ratios of PUFAs to saturated fatty acids are increased and the proportion of MUFAs remains fairly constant. In some instances, however, MUFA concentrations are increased while PUFA and saturated fatty acid concentrations remain stable. It would be interesting to see if the cis-vaccenic acid concentrations of S. velum

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum increase during cold exposure, since this lipid is almost certainly important in maintaining membrane integrity in S. velum tissues, as evidenced by the high concentrations of eis-vaccenic acid in the phospholipid pool of S. velum. Changes in the concentrations of bacterial lipids in the bivalve tissues during periods of cold stress would demonstrate the existence of mutual metabolic control mechanisms. This lipid may function with 20 : 40)6 in lieu of the 0)3 PUFAs of typical marine bivalves. The bivalves examined in this study were collected in November; thus the high concentrations of cis-vaccenic acid may be the result of cold exposure. If this is the case, one would expect levels to be even higher in December and January, when the annual temperatures at Little Buttermilk Bay reach a minimum. The source of the 20:40)6 P U F A found in Solemya velum remains to be determined. Possibilities include in vivo synthesis by either the animal or bacterial components of the symbiosis or the ingestion of small amounts of exogenous food sources. The presence of large amounts of fatty acids associated with the anaerobic bacterial M U F A synthesis pathway suggests that synthesis of PUFAs by the endosymbionts is unlikely. Additionally, animal fatty acid synthesis pathways are considered incapable of 0)6 fatty acid synthesis. Exogenous sources may provide the high concentrations of 20 : 40)6 found in S. velum, yet large amounts of other lipids would be concomitantly taken in and the 613C ratios of S. velum fatty acids have a strong bacterial signature. Perhaps ingestion of small amounts of 20 : 40)6 and other members of the 0)6 fatty acid series such as 18:20)6 and 20:20)6 and subsequent elongation by the animal may be important in contributing to the large 20 : 4co6 concentrations found in S. velum. Low 20 : 40)6 turnover rates could also explain the high concentrations of this fatty acid. To summarize, the fatty acid, sterol, and lipid stable isotope compositions of Solemya velum strongly suggest that these compounds are directly or indirectly derived from endosymbiotic chemoautotrophic bacteria, while marine photosynthetic primary production plays at best a small role in the nutrition of this species. The major fatty acid in S. velum was the monounsaturated fatty acid cis-vaccenic acid, a lipid normally found in large concentrations only in bacteria. This lipid may serve as a useful screening tool for the investigation of other invertebratechemoautotroph symbioses. Cis-vaccenic acid occurs in high concentrations in all the lipid classes of Solemya velum, particularly the phospholipid and triglyceride pools, demonstrating its importance in both membrane structure and lipid storage processes. The sterols and 0)3 PUFAs derived from phytoplankton and normally found in high concentrations in marine bivalves were present in only trace amounts in S. velum, demonstrating a limited input of plant-derived material. The 0)6 PUFAs present in S. velum appear to be conserved for use in the phospholipid pool of this species; the source of these lipids remains to be determined. The 613C ratios of the lipids of S. velum strongly suggest an endosymbiont source for the lipid carbon. This work clearly shows the usefulness of a biomarker approach for the study of animal-bacteria symbioses. As-

289

sumptions based on the lipid composition of the Solemya velum symbiosis support the results of earlier work with this species indicating that the bacterial endosymbionts form the major nutritional source for this species (Conway et al. 1989). S. velum is highly suitable for general studies of nutrition in bivalve-chemoautotroph symbioses, as large numbers may be collected and examined for endosymbiont biomarkers. Consequently, new techniques can be calibrated on S. velum, and other relatively accessible symbioses, and applied to rarer specimens when the opportunities arise. The techniques described here may be very useful in the study of other symbioses. While we are not suggesting that the study of preserved specimens should replace traditional laboratory studies of living specimens, examination of biochemical parameters may be the most appropriate course of study when specimens are damaged, sample sizes are small, or laboratory studies are not feasible.

Acknowledgements. We would like to thank M. McCaffrey for his patience and advice throughout the lipid analyses, B. Fry and C. Johnson for mass spectrometry analyses, D. Leavitt, B. Lancaster, and H. Schempf for technical assistance throughout these studies, and J. Farrington for advice, supplies, and laboratory space during the lipid composition studies. H. Jannasch kindly provided viable cultures of Thiomierospira crunogena and S. Molyneaux and C. Wirsen assisted in the culturing and harvesting of T. erunogena cells. C. Cavanaugh, M. McCaffrey, J. Farrington, and E. DeLong provided helpful comments on earlier versions of this manuscript. G. Rau and an anonymous reviewer provided thorough and helpful reviews of this manuscript. This work was supported by NOAA Grant No. NA87AA-O-OMO93 (to J.M.C.), an Ocean Ventures Fund Award (to N. C.) and the education office of the Woods Hole Oceanographic Institution.

Literature cited

Anderson, A. E., Childress, J. J., Favuzzi, J. A. (1987). Net uptake of CO 2 driven by sulphide and thiosulphate oxidation in the bacterial symbiont-containing clam Solemya reidi. J. exp. Biol. 133:1-31 Bobbie, R. J., White, D. C. (1980). Characterization of benthic microbial community structure by high-resolution gas chromatography of fatty acid methyl esters. Appl envirl. Microbiol. 39:1212 1222 Cavanaugh, C. M. (1983). Symbiosis of chemoautotrophic bacteria and marine invertebrates from sulphide-rich habitats. Nature, Lond. 302:58 61 Cavanaugh, C. M. (1985). Symbiosis ofchemoautotrophic bacteria and marine invertebrates. Ph.D. Thesis, Harvard University Cavanaugh, C. M., Abbott, M. S., Veenhuis, M. (1988). Immunochemical localization of ribulose-l,5-biphosphate carboxylase in the symbiont-containing gills of Solemya velum (Bivalvia: Mollusca). Proc. natn. Acad. Sci. USA 85:7786-7789 Conway, N. (1990). The nutritional role of endosymbiotic bacteria in animal-bacteria symbioses: Solemya velum; a case study. MIT/WHOI Ph.D. thesis Conway, N., McDowell Capuzzo, J. (1990). The use of biochemical indicators in the study of trophic interactions in animal-bacteria symbioses: Solemya velum,a case study. In: Barnes, M., Gibson, R. N. (eds.) Trophic relationships in the marine environment. Proc. 24th Eur. mar. biol. Syrup. Aberdeen University Press, Aberdeen, p. 553-564

290

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum

Conway, N., McDowe11 Capuzzo, J., Fry, B. (1989). The role of endosymbiotic bacteria in the nutrition of Solemya velum: Evidence from a stable isotope analysis of endosymbionts and host. Limnol. Oceanogr. 34:149 155 DeBurgh, M. E., Juniper, S. K., Singla, C. L. (1989). Bacterial symbiosis in Northeast Pacific Vestimentifera: a TEM study. Mar. Biol 101:97-105 Degens, E. T. (1969). Biogeochemistry of stable carbon isotopes. In: Egtinton, E., Murphy, M. J. T. (eds.) Organic geochemistry. Springer-Verlag, Berlin, p. 304-329 DeLong, E. F., Yayanos, A. A. (1986). Biochemical function and ecological significance of novel bacterial lipids in deep-sea procaryotes. Appl. envirl Microbiol. 51:730-737 Fagerland, U. H. M., Idler, D. R. (1960). Marine sterols. VI. Sterol biosynthesis in molluscs and echinoderms. Can. J. Biochem. Physiol. 38:997-1002 Farrington, J. W, Davis, A. C., Sulanowski, J., McCaffrey, M. A., Clifford, C. H., Dickenson, P., Volkman, J. K. (1988). Biogeochemistry of lipids in surface sediments of the Peru upwelling area - 15°S. In: Mattivelli, L., Novelli, L. (eds.) Advances in organic geochemistry 1987. Org. Geochem. 13:607-617 Felbeck, H. (1983). Sulfide oxidation and carbon fixation by the gutless clam Solemya reidi: an animal-bacteria symbiosis. J. comp. Physiol. 152:3-11 Felbeck, H., Childress, J. J., Somero, G. N. (1981). Calvin-Benson cycle and sulfide oxidation enzymes in animals from sulfide-rich habitats. Nature, Lond. 293:291-293 Fiala-M~dioni, A. (1984). Ultrastructural evidence of abundance of intracellular symbiotic bacteria in the gills of bivalve molluscs of deep hydrothermal vents. Cr. Lebd. S6anc. Acad. Sci., Paris (Set. III) 298:487-492 Fiala-M6dioni, A., Alayse, A. M., Cahet, G. (1986). Evidence of in situ uptake and incorporation of bicarbonate and amino acids by a hydrothermal vent mussel. J. exp. mar. Biol. Ecol. 96: 191-198 Fiala-M6dioni, A., LePennec, M. (1987). Trophic structural adaptations in relation to the bacterial association of bivalve molluscs from hydrothermal vents and subduction zones. Symbiosis 4: 63 74 Fulco, A. J. (1983). Fatty acid metabolism in bacteria. Prog. Lipid Res. 22:133-160 Gagosian, R. B. (1975). Sterots in the western North Atlantic Ocean. Geochim. Cosmochim. Acta 39:1443 1454 Gardner, D., Riley, J. P. (1972). The component fatty acids of the Iipids of some species of marine and freshwater molluscs. J. mar. biol. Ass. U.K. 52:827-838 Giere, O., Conway, N. M., Gastrock, G., Schmidt, C. (1991). 'Regulation' of gutless annelid ecology by endosymbiotic bacteria. Mar. Ecol. Prog. Set. 68:287-299 Giere, O., Felbeck, H., Dawson, R., Liebezeit, G. (1984). The gutless marine oligochaete Phallodrilus leukodermatus Giere, a tubificid of structural, ecological and physiological significance. Hydrobiologica 115 : 83 - 89 Gillan, F. T., Johns, R. B. (1986). Chemical markers for marine bacteria: Fatty acids and pigments. In: Johns, R. B. (ed.) Biological markers in the sedimentary environment. Elsevier, Amsterdam, p. 291-309 Gillan, F. T., Stoilov, I. L., Thompson, J. E., Hogg, R. W., Wilkinson, C. R., Djerassi, C. (1988). Fatty acids as biological markers for bacterial symbionts in sponges. Lipids 23 : 1139-1145 Goad, L. J. (1976). The steroids of marine algae and invertebrate animals. In: Malins, D. C., Sargent, J. R. (eds.) Biochemical and biophysical perspectives in marine biology, vol. 3. Academic Press, New York, p. 213-318 Goldfine, H. (1972). Comparative aspects of bacterial lipids. Adv. microb. Physiol. 8 : 1 - 5 8 Guarnieri, M., Johnson, R. M. (1970). The essential fatty acids. Adv. Lipid Res 8 : 115-174 Hazel, J. R., Sellner, P. A. (1979). The regulation of membrane lipid composition in thermally acclimated poikilotherms. In: Gilles, R. (ed.) Animals and environmental fitness. Physiological and

biochemical aspects of adaptation and ecology. Proc. 1st Conf. Eur. Soc. Comp. Physiol. Biochem. Pergamon Press, New York, p. 541-559 Idler, D. R., Wiseman, P. (1971). Sterols of Molluscs. Int. J. Biochem. 2:516 Jannasch, H. W., Wirsen, C. O., Nelson, D. C., Robertson, L. A. (1985). Thiomierospira crunogena sp. nov., a colorless, sulfuroxidizing bacterium from a deep-sea hydrothermal vent. Int. J. system Bact. 35:422-424 Johns, R. B., Perry, G. J. (1977). Lipids of the marine bacterium Flexibacter polymorphus. Archs Microbiol. 114: 267 - 271 Joseph, J. D. (1982). Lipid composition of marine and estuarine invertebrates. Part II: Mollusca. Prog. Lipid Res. 22:109-153 Katayama-Fujimura, Y., Tsuzaki, N., Kuraishi, H. (1982). Ubiquinone, fatty acid and DNA base composition determination as a guide to the taxonomy of the genus Thiobacillus. J. gen. Microbiol. 128:1599-1611 Kokko, W. C. M. C., Epstein, S., Look, S. A., Rau, G., Fenieal, W., Djerassi, C. (1984). Culture studies and an application of 13C/ 12C isotope ratio mass spectrometry. J. biol. Chem. 259: 81688173 McCaffrey, M. A., Farrington, J. W., Repeta, D. J. (1989). Geochemical implications of the lipid composition of Thioploea sp. from the Peru upwelling region - 15°S. Org. Geochem. 14: 61 68 Monson, K. D., Hayes, J. M. (1982). Carbon isotope fractionation in the biosynthesis of bacterial fatty acids. Ozonolysis of unsaturated fatty acids as a means of determining the intramolecular distribution of carbon isotopes. Geochim. Cosmochim. Acta. 46:139-149 Moreno, J., Pollero, A. E., Moreno, V. J., Brenner, R. R. (1980). Lipids and fatty acids of the mussel (Mytilus platensis d'Orbigny) from South Atlantic Waters. J. exp. mar. Biol. Ecol. 48: 263 -276 Okuyama, H., Yamada, K., Kameyama, Y., Ikezawa, H., Akamatzu, Y., Nojima, S. (1977). Regulation of membrane lipid synthesis in Escherichia coli after shifts in temperature. Biochem. 16:2668-2673 Ourisson, G., Albrecht, P., Rohmer, M. (1979). The hopanoids. Palaeochemistry and biochemistry of a group of natural products. Pure appl. Chem. 51:709 729 Parkes, R. J., Taylor, J. (1983). The relationship between fatty acid distributions and bacterial respiratory types in contemporary marine sediments. Estuar. cstl mar. Sci. 16 : 173-189 Perry, G. J., Volkman, J. K., Johns, R. B, Bavor, H. J. Jr. (1979). Fatty acids of bacterial origin in contemporary marine sediments. Geochim. Cosmochim. Acta 43:1715-1725 Phillips, N. W. (1984). The role of different microbes and substrates as potential suppliers of specific, essential nutrients to marine ~letritivores. Bull. mar. Sci. 35:283-298 Piretti, M. V., Pistore, R., Pagliuca, G. (1989). Uptake and utilization of lipid constituents dispersed in culture water by the bivalve mollusc Scapharca inaequivalvis (Bruguiere). Comp. Biochem. Physiol. 92B: 755-758 Piretti, M. V., Tioli, F., Pagliuca, G. (1987). Investigation of the seasonal variations of sterol and fatty acid constituents in the bivalve molluscs Venus gallina and Seapharca inaequivalvis (Brugui6re). Comp. Biochem. Physiol. 88B: 1201-1208 Prahl, F. G., Meuhlhausen, L. A. (1989). Lipid biomarkers as geochemical tools for paleoceanographic study. Berger, W H., Smetacek, V. S., Wefer, G. (eds.) Productivity of the ocean: present and past. T. Wiley and Sons limited, New York, p. 271 289 Rau, G. H. (1985). 13C/12Cand aSN/14N in hydrothermal vent organisms: ecological and biogeochemical implications. Bull. biol. Soc. Wash. 6:243-247 Ruby, E. G., Jannasch, H. W, Deuser, W G. (1987). Fractionation of stable carbon isotopes during chemoautotrophie growth of sulfur-oxidizing bacteria. Appl. envirl. Microbiol. 53:1940-1943 Southward, A. J., Southward, E. C., Dando, P. R., Rau, G. H., Fehlbeck, H., Flugel, H. (1981). Bacterial symbionts and low

N. Conway and J. McDowell Capuzzo: Bacterial lipids in Solemya velum 13C/12C ratios in tissues of Pogonophora suggest unusual nutrition and metabolism. Nature, Lond. 293:616-620 Teshima, S.-I., Kanazawa, A. (1974). Biosynthesis of sterols in abalone, Haliotis gurneri, and mussel, Mytilus edulis. Comp. Biochem. Physiol 47B: 555-561 Teshima, S.-I., Patterson, S. W. (1981). Sterol biosynthesis in the oyster Crassostrea virginica. Lipids 16:234-239 Volkman, J. K. (1986). A review of sterol markers for marine and terrigenous organic matter. Org. Geochem 9 : 8 3 - 9 9 Volkman, J. K., Johns, R. B., Gillan, E T., Perry, G. J., Bavor, H. J. Jr (1980). Microbial lipids of an intertidal sediment. Fatty acids and hydrocarbons. Geochem. Cosmochim. Acta 44: 1133 - 1143

291

Voogt, P. A. (1975). Investigations of the capacity of synthesizing fl-sterols in Moltusca. XII. Biosynthesis and composition of sterols in some bivalves (Anisomyasia). Comp. Biochem. Physiol. 50B: 499-504 Voogt, P. A. (1972). Lipid and sterol components and metabolism in Mollusca. In: Florkin, M., Sheer, B. T. (eds.) Chemical zoology, Vol. 7. Academic Press, London and New York, p. 245 Wada, M., Fukunaga, N., Sasaki, S. (1989). Mechanism of biosynthesis of unsaturated fatty acids in Pseudomonas sp. strain E-3, a psychotrophic bacterium. J. Bacteriol. 171:4267-4271