Elemental characterization of microorganism granules by EFTEM in the tube wall of a deep-sea vent invertebrate

Biology of the Cell 94 (2002) 243–249 www.elsevier.com/locate/biocell

Original article

Elemental characterization of microorganism granules by EFTEM in the tube wall of a deep-sea vent invertebrate Jean-Pierre Lechaire *, Bruce Shillito, Ghislaine Frébourg, Françoise Gaill Équipe Adaptations aux Milieux Extrêmes, UMR 7622 CNRS, Université Pierre et Marie Curie, Bâtiment A, 7, quai St-Bernard, 75252 Paris cedex 5, France Received 10 December 2001; accepted 6 June 2002

Abstract Microorganisms colonizing the exoskeletons of the tube worm Riftia pachyptila are described at the ultrastructural level. The prokaryotic cells from the worm tube wall differ from those colonizing the exoskeleton outer surface in the presence of an electron dense granule. The morphology and distribution of these bacteria-like cells are described. Prokaryotic organisms are assembled in nodules which increased in size in the oldest part of the exoskeleton. The aspect, location and elemental composition of the intracellular granules are determined. Most of them (100 nm in diameter) are located close to the cell membrane and exhibit a homogeneous and amorphous content. EDX and EFTEM microanalyses show that these structures contain phosphorus, oxygen and iron. All together these data suggest that these granules are iron polyphosphates. These structures may act as energy sources for making ATP during anoxic conditions as existing in hydrothermal environments. © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Keywords: Polyphosphate; EFTEM; Microorganism; Exoskeleton; Hydrothermal vent

1. Introduction The hydrothermal vent fluids from the East Pacific Rise (13° N) are known to contain high metal concentrations compared to the environmental sea water (reviewed in Childress and Fisher, 1992). Among the hydrothermal vent communities, the tube worm Riftia pachyptila has been studied for its symbiotic metabolism (Felbeck, 1981; Gaill, 1993; Arndt et al., 2001), for the bioaccumulation of several metals by detoxication processes (Cosson, 1996; Truchet et al., 1998) and the morphogenesis of its tube (Gaill and Hunt, 1986; Shillito et al., 1995; Shillito et al., 1997; Ravaux et al., 1998). Some prokaryotic organisms were observed embedded in the Riftia tube wall (Gaill and Hunt, 1986; Gaill and Hunt, 1991) and preliminary results (Lechaire et al, 1998) have shown that these microorganisms were quite different from the filamentous bacteria colonizing the outer

* Corresponding author. Fax: +33-1-44-27-52-50. E-mail address: [email protected] (J.P. Lechaire). © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. PII: S 0 2 4 8 - 4 9 0 0 ( 0 2 ) 0 1 1 9 9 - 1

surface of the tube. Unlike the latter, the main characteristic of these microorganisms was the presence of an intracellular electron dense granule. Intracellular granules are often present in prokaryotic organisms (Shively, 1974) and the presence of polyphosphate has been reported in more than 100 different species of bacteria (Bode et al., 1993). A variety of microorganisms accumulate large reserves of inorganic phosphate granules that are classified as amorphous minerals, considering the lack of cristalline order (Lins and Farina, 1999). Inorganic polyphosphates (poly P) are found in volcanic condensates and deep-oceanic vents (Kornberg, 1995). Poly P are excellent ligands for certain cations (Ehrlich, 1999). Intracellular poly P must therefore vary somewhat in elemental composition, depending on the cations that are bound to the poly P. The aim of this work was to study the morphology and the distribution of these bacteria inside the exoskeleton of R. pachyptila, the aspect and the location of the granules in the intracellular compartment and identify the elemental composition of the granules using different approaches: EDX and EFTEM microanalyses.

244

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

2. Materials and methods 2.1. Animal collection Specimens of R. pachyptila were collected at 2600 m in depth during the French American HOT 96 cruise, at the 13° N site (Genesis) on the East Pacific Rise. The collection was achieved with the French submersible, Nautile. 2.2. Fixation and tissue processing The upper part (15 cm length) of a Riftia tube (90 cm total length) was fixed in 10% neutralized formalin when animals arrived on shipboard. Small samples (1 mm3) were dissected at the base of the fixed specimen for further morphological and elemental study. For morphological studies by photonic and electronic microscopies, the samples were post-fixed in a 1% OsO4 buffered solution. For electron microscopy microanalysis, the samples were not osmicated in order to avoid artefacts in the chemical electron microanalysis. All the samples were dehydrated in graded alcohol and propylen oxide solutions and further embedded in Epon resin. Semi-thin sections were obtained from a Leica ultramicrotome (Ultracut E), and stained with a toluidine blue and azure II mixture (Richardson et al., 1960) with 1% saccharose. Ultrathin sections were contrasted with uranyl acetate and lead citrate for the morphological study, and not contrasted for microanalysis (EFTEM–EDX). 2.3. Energy dispersive X-ray microanalysis (EDX) EDX microanalysis was carried out using a JEOL 2000 EX transmission electron microscope, operating at 200 kV, and acquired with an energy dispersive X-ray detection system (Tracor 5400 FX) equipped with an Si–Li diode and beryllium window allowing detection of elements heavier than fluorine. Microanalysis was collected in scanning transmission mode using a 100 nm probe. The samples were analysed for Fe/P ratio (at.%), Fe- and P-K-factors were determined experimentally from iron and phosphorus bearing minerals (orthopyroxene and apatite). Absorption corrections corresponding to the thickness of the ultrathin sections were carried out. The ultrathin sections (70 nm) are deposited onto 400 square mesh nickel grids. 2.4. Energy filtered transmission electron microscopy (EFTEM) The EFTEM observations were performed using an LEO 912 electron microscope (LEO Electron Optics GmbH, Oberkochen, Germany) equipped with an LaB6 source, and operated at 120 kV. The LEO 912 features a Koehler-type illumination system (Benner and Probst, 1994) and an in-column omega-type electron energy filter (Lanio, 1986; Jeanguillaume, 1987).

In order to reduce drift upon irradiation, the ultrathin sections (50 nm) were laid on a 700-mesh hexagonal copper grid with thin bars (Agar Scientific, Oxford Instruments, Orsay, France). Parallel electron energy loss spectra (PEELS) and electron spectroscopic images (ESI) were recorded with a cooled slow-scan CCD camera (Proscan, Penzing, Germany) equipped with a 1024 × 1024 pixelsized chip and operating in a 14-bit mode. Acquisition was accomplished with the ESIvision program (versions 2.11 and 3.0 Soft-Imaging Software, SIS, GmbH, D-48153 Münster). 2.4.1. ESI procedure Due to the Koehler-type illumination system of the microscope, the irradiation can reproducibly be adjusted. Setting the emission current to 4 µA and the condenser system to 2.5 mrad resulted in a dose rate of 6.0 × 104 e– (nm2 s)–1. Exposure time was 1 s for each image. The entrance aperture of the spectrometer was set at 1.5 mm. The spectrometer slit width was set to 15 eV, and the primary magnification to 20 000×. All images were corrected for the camera offset and gain variations. After averaging 2 × 2 pixels, the effective pixel size on the resulting 512 × 512 images was 1.4 nm. For ESI acquisition, we used the three-window power-law method. 2.4.2. PEELS procedure The parameters for illumination were the same as for ESI acquisition. The primary magnification was set to 100 000× in order to select only one granule in the area delimited by the entrance aperture of the spectrometer which was set at 100 µm. In these conditions, the specimen was exposed to a dose rate of 4.6 × 104 e– (nm2 s)–1. The acquisition time was 1 s and the integration maintained up to 40 s. For the phosphorus spectrum, a back-ground subtraction was performed according to the Egerton power-law model (Egerton, 1986).

3. Results A schematic view of the animal in its tube is shown in Fig. 1A. The Riftia tube wall is made of a fibrillar system of chitin bundles embedded in an amorphous proteic matrix (Gaill and Hunt, 1986; Shillito et al., 1995; Shillito et al., 1997). Fig. 1B illustrates the distribution of microorganism nodules inside the tube wall. The term nodule corresponds to a cluster of several bacteria. These nodules, as shown in Fig. 2C, bear an elliptic shape and are located between the chitin bundles. 3.1. Morphological characteristics of the microorganism nodules Considering the staining intensity of the toluidine blue (Fig. 2A), we can define two zones in the Riftia tube. An

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

245

3.3. Microorganism granules At higher magnification (Fig. 2B), only one electron dense granule is observed in these microorganisms. Most of these granules are electron dense and round shaped. In section, their diameter can reach up to 100 nm. In some rare cases, two or three granules were detected (not shown). These additional granules appeared polygonal and lighter in their content. We decided to focus our analysis on the 100 nm in diameter granule population. These granules are located close to the cell membrane (Fig. 2B). Their content is homogenous and seems amorphous as suggested by electron diffraction experiments (not shown). 3.4. Chemical composition of granules

Fig. 1. Location of the microorganisms in the tube wall of R. pachyptila. (A) Riftia in its tube. (B) Schematic view of the microorganism nodules in the tube wall thickness. Longitudinal section (LS), inside (I) and outside (O) of the tube wall regions are indicated.

inner zone, the lighter one, corresponds to the fresh secretion of the exoskeleton. A more heavily stained zone, at the outer face with delaminated layers, corresponds to the mature zone. The density of the microorganism’s nodules in the tube wall is >100 nodules/mm2 (see Fig. 2A), occupying 2.5% of the section’s area measured in longitudinal tube sections (parallel to the larger dimension of the tube, see Fig. 1B). This density is lower in the fresh zone (about 70 nodules/mm2 of the tube section) than in the oldest one. In the mature zone, the distribution appears to be more homogeneous (130 nodules/mm2 of tube section). The size of these nodules may be up to 80 µm in height and up to 20 µm in thickness. The geometry of the cross-section of both longitudinal and transverse (not shown) sections is “elliptic”, suggesting that they are symmetrical as discs. In fact their size increases from the internal side to the external side. 3.2. Microorganisms After conventional staining (Fig. 2C), we focused our analysis on individual nodules, embedded between successive layers of chitin microfibrils. These nodules contain numerous polymorphous bacteria-like organisms (4 µm–2) embedded in an electron dense matrix. In section, these microorganisms have a length ranging up to 1.8 µm and a diameter up to 300 nm. The content of these microorganisms is electron dense: lighter in the central part and more dense in the outer part where granules are observed.

EDX microanalysis was performed in order to determine the elemental composition of the large granule content. The control spectrum (Fig. 3A) from an epon region at the outside of the tissue showed a nickel peak due to the nickel support grid and an artefactual copper peak due to the sample holder. Residual traces of chloride contamination were due to the epoxy resin. When we compared the EDX spectra (only one was shown in Fig. 3B) from different large granules (100 nm in diameter), we observed mainly phosphorus and iron peaks in each granule. The Fe/P ratio (at.%) was about 0.75. The Ca peak was not observed in each granule (not shown). The PEELS method allowed us to precise the elemental composition of a large granule (100 nm). As previously shown in EDX, we confirmed the presence of phosphorus (P-L2,3) in Fig. 4A and iron (Fe-L2,3) in Fig. 4B. Moreover oxygen (O-K) was now detected (Fig. 4B) which could not be detected with the EDX system used. The ESI method was performed in order to image the location of these elements within a granule. With this method, we confirmed the presence of P, O and Fe in the same granule (Fig. 5). The area occupied by oxygen was similar to the phosphorus and to the iron ones. These areas were rather homogeneously distributed in the granule.

4. Discussion The hydrothermal vent fluids from the East Pacific Rise have high levels of mineral compounds (reviewed in Childress and Fisher, 1992). For example at 13° N the hydrothermal fluid has a high level of Fe comprised between 100 000 and 600 000 mg/l instead of <0.06 mg/l encountered in sea water. Cosson (1996) has shown a bioaccumulation of several elements in the tissues of the vestimentiferan worm R. pachyptila. In the environment, microbial mineral formation is a form of immobilization of chemical elements and can be a form of detoxication (Ehrlich, 1999). In this report, we describe microorganism nodules in the Riftia tube wall. These microorganisms are quite different

246

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

Fig. 2. Morphological aspects of the microorganism nodules ((A) optical micrograph; (B,C) transmission electron micrographs). (A) Distribution of the microorganism nodules in the tube wall thickness. In longitudinal semi-thin section, the larger nodules (up to 80 µm in height) are located near the outer (o) part and the smaller one (less than 0.5 µm in height) close to the internal side (i). A lighter stained zone is visible at the inner side (i) of the tube wall (arrowhead). Scale bar: 100 µm. (B) Higher magnification view of the microorganisms observed in a nodule. These microorganisms exhibit electron dense granules which are generally close to the cell envelope at one pole of the microorganism. When the microorganisms are tangentially sectioned, the dense granule may appear in the central part of the microorganism (arrowhead). Most of these granules are round-shaped. In section, their diameter ranges from 70 to 100 nm. Scale bar: 240 nm. (C) Nodule embedded in the tube wall matrix (M). This matrix consists of chitin microfibrils organized in parallel bundles, with various orientations. Polymorphous (rounded, elongated, polygonal or digitated) microorganisms are observed in a homogeneous matrix (m). According to the section plane through the microorganism nodule, electron dense granules are present or not in the microorganism (24% of the microorganisms observed contain a granule). Generally these microorganisms have only one granule, and occasionally two (arrowhead). The white holes (arrows) observed correspond to the granules eliminated during the thin sectioning. Scale bar: 1 µm.

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

Fig. 3. EDX spectrum from a 100 nm sectioned granule. (A) Control from a region outside of the tissue on the resin. The Ni and Cu peaks come from the support grid and from the sample holder, respectively. Residual traces of Cl contamination due to the resin are discernible (counting time: 120 s). (B) EDX spectrum from a granule showing P, Ca and Fe peaks (counting time: 90 s).

from the endosymbionts localized in the Riftia trophosomal tissue which have no granule (Gaill, 1993; Truchet et al, 1998). When we observed freshly secreted tube at the surface of the internal side of the tube, we never saw microorganism nodules (data not shown). Since the secretion of the fresh tube material is deposited at the inner tube wall part (Gaill et al., 1997), and because the size of the microorganism nodules increases from the inner to the outer part of the tube, we suggest a development of the nodules in the thickness of the tube. These nodules might act as biological markers of the tube ageing. The elemental composition obtained by EFTEM microanalysis confirms the presence of the elements P and Fe detected by X-ray microanalysis and moreover the presence of oxygen. These elements are homogeneously distributed in the microorganism granules. Bacteria play a role in some cases of mineral formation (Ehrlich, 1999). The minerals may be extra- or intracellular, or in the cell envelope. The granules we observed are intracellular. In bacteria, the main intracellular mineral depositions are sulphur (S°), calcium carbonate (CaCO3),

247

Fig. 4. EFTEM microanalysis (PEELS method) from a sectioned granule (100 nm in diameter). (A) Phosphorus (P-L2,3) spectrum obtained after back-ground subtraction. The first peak observed at the left of the P spectrum is a parasite signal integrated within the defined region of interest (ROI) during the acquisition of the phosphorus spectrum (not shown) before the back-ground subtraction. (B) Spectrum showing the presence of oxygen (O-K) and iron (Fe-L2,3).

magnetite (Fe3O4), greigite (Fe3S4), apatite (Ca10(PO4)6(OH)2) and polyphosphate (PO32–-O-[PO3–]nPO32–) (Ehrlich, 1999). Other nonmineral types of inclusion bodies have been described in bacteria as polyhedral bodies (carboxysomes) (Shively, 1974). These intracellular inclusions have been divided into two major groups based on the presence or absence of a surrounding membrane (Shively, 1974). Magnetosomes (magnetite and greigite bodies; Mann et al., 1990) and carboxysomes (Kaplan et al., 1994) are considered as membrane-enclosed inclusions, while polyphosphates are in the group of nonmembrane-enclosed intracellular deposits. The fact that we observed nonmembrane-enclosed granules containing P, O and Fe indicates that these structures are probably polyphosphate granules. Bode et al. (1993) have detected by electron energy loss spectrum (EELS) poly P granules in Helicobacter pylori. When we compared the specific phosphorus spectrum obtained by EELS in our study with the spectrum of phosphorus obtained for poly P granules in H. pylori, the

248

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

Fig. 5. EFTEM microanalysis (ESI method) from sectioned granules. (A) Zero loss filtered image of microorganism showing three granules (80–100 nm in diameter). (B) Phosphorus map. (C) Oxygen map. (D) Iron map. Scale bar: 100 nm.

two spectra are superposed. Thus we can assume that these granules are polyphosphate granules. Furthermore the fact that we obtained P and O spectra by EELS and that the location of P and O by ESI were similar in the same granule also suggests that the granules are indeed phosphate granules. The spectra we obtained with our biological samples did not allow to resolve the fine structures of the P peak. Higher spectral resolution would permit to confirm or infirm the existence of poly P. Possibly the XANES (X-ray absorption near-edge structure) method could answer this question in analysing the fine structures of this peak (Varlot et al., 2000). However, the percentage for each element detected in the granules by EDX analyses yields an Fe/P (at.%) ratio of about 0.75 ± 0.07. This ratio was found identical for six granules analysed and suggests a polyphosphate with a short chain. Further semi-quantitative EDX studies would be necessary to precise the type of poly P. The granules described in our study have a diameter of 100 nm which is in agreement with the diameters comprised between 50 and 500 nm for poly P granules described by authors in different species of bacteria (Bode et al. (1993) for H. pylori; Lins and Farina (1999) for magnetotactic bacteria). These authors observed that the poly P granules showed evaporation typical of polyphosphate during exposure to high intensity electron beam. We did not observe this phenomenon. Bode et al. (1993) precised that this phenomenon is dependent on the time of exposure and was not so

obvious in smaller granules. The iron contained in the poly P granules could stabilize the granule under the electron beam in the same way that staining with heavy-metal salts enhances the contrast and makes the specimen less sensitive to radiation (Luther, 1992). It is known that numerous and varied biological functions are performed by poly P depending on the need (Kornberg, 1995). This author suggested that poly P may represent energy sources, which would serve ATP synthesis. Kornberg (1995) also mentioned a possible role of poly P as a regulator for stress and survival. As poly P are known to fluctuate in response to nutritional and other parameters, it seems possible that they act as a reservoir of oxygen in the case of environmental anoxia. This could be the case for microorganism granules in Riftia tube taking into account that the hydrothermal vent fluids are anoxic. The fact that we always observe a granule at a pole of the microorganism certainly corresponds to a biological function which remains to be determined. Among other possible functions, poly P may act as a chelator of metals ions (e.g., Mn2+ and Ca2+) in physiologic adjustments to growth and development (Kornberg, 1995). The chelation of other metals (e.g., Zn, Fe, Cu and Cd) may reduce their toxicity (Kornberg, 1995). Poly P granules described by Lins and Farina (1999) in magnetotactic bacteria can incorporate different elements (Al, Fe and Zn). These authors propose that granules are involved in accumulation of phosphates and accumulation of “unwanted”

J.P. Lechaire et al. / Biology of the Cell 94 (2002) 243–249

(toxic) metals. The processes of Fe deposition in Riftia granules, and whether these granules are involved in detoxication, accumulation of iron or if iron could further be used in metabolism as suggested by Lins and Farina (1999) remain to be determined.

Acknowledgements We wish to thank Drs. Christian Colliex and Claudie Mory, Laboratoire de Physique des Solides, Université Paris-Sud, 91 Orsay, France and Carmen Quintana, Instituto de Microelectronica de Madrid–CNM-CSIC, Madrid, Spain for their helpful discussion in this work concerning the EFTEM. The authors are grateful to Dr. François Guyot, Laboratoire de Minéralogie, LMCP et IPGP, 75005 Paris, France for the X-ray microanalysis and Nadine Le Bris, Département Environnement Profond, IFREMER, 29 Plouzané, France for her precious discussion. The EFTEM observations were carried out at the Service de Microscopie Electronique de l’IFR–Biologie Intégrative 2062 CNRS.

References Arndt, C., Gaill, F., Felbeck, H., 2001. Anaerobic sulfur metabolism in thiotrophic symbioses. J. Exp. Biol. 204, 741–750. Benner, G., Probst, W., 1994. Köhler illumination in the TEM: fundamentals and advantages. J. Microsc. 174, 133–142. Bode, G., Mauch, F., Ditschuneit, H., Malfertheiner, P., 1993. Identification of structures containing polyphosphate in Helicobacter pylori. J. Gen. Microbiol. 139, 3029–3033. Childress, J.J., Fisher, C.R., 1992. The biology of hydrothermal vent animals: physiology, biochemistry and autotrophic symbioses. Oceanogr. Mar. Biol. Annu. Rev 30, 337–341. Cosson, R.P., 1996. La bioaccumulation des éléments minéraux chez le vestimentifère Riftia pachyptila (Jones): bilan des connaissances. Oceanologica Acta 19 (2), 163–176. Egerton, R.F., 1986. Electron Energy Loss Spectroscopy in the Electron Microscope. Plenum Press, New York. Ehrlich, H.L., 1999. Microbes as geologic agents: their role in mineral formation. Geomicrobiol. J. 16, 135–153. Felbeck, H., 1981. Chemoautotrophic potential of the hydrothermal vent tubeworm, Riftia pachyptila Jones (Vestimentifera). Science 213, 336–338. Gaill, F., 1993. Aspects of life development at deep sea hydrothermal vents. FASEB J. 7, 558–565. Gaill, F., Hunt, S., 1986. Tubes of deep sea hydrothermal vent worms Riftia pachyptila (Vestimentifera) and Alvinella pompejana(Annelida). Mar. Ecol. Prog. Ser. 34, 267–274.

249

Gaill, F., Hunt, S., 1991. The Biology of Annelid Worms from High Temperature Hydrothermal vent Regions. Rev. Aquat. Sci 4 (2-3), 107–137. Gaill, F., Shillito, B., Ménard, F., Goffinet, G., Childress, J.J., 1997. Rate and process of tube production by the deep-sea hydrothermal vent tubeworm Riftia pachyptila. Mar. Ecol. Prog. Ser 148, 135–143. Jeanguillaume, C., 1987. Electron energy-loss spectroscopy and biology. Scanning Microsc. 1, 437–450. Kaplan, A., Schwarz, R., Lieman-Hurwitz, J., Ronen-Tarazi, M., Reinhold, L., 1994. Physiological and molecular studies on the response of cyanobacteria to changes in the ambient inorganic carbon concentration. In: Bryant, D.A. (Ed.), The Molecular Biology of Cyanobacteria. Kluwer Academic Publishers, Netherlands, pp. 469–485. Kornberg, A., 1995. Inorganic polyphosphate: toward making a forgotten polymer unforgettable. J Bacteriol. 177, 491–496. Lanio, S., 1986. High-resolution imaging magnetic energy filters with simple structure. Optik 73, 99–107. Lechaire J.-P., Frebourg G., Gaill F. Morphogenesis and aging of the tube wall from a deep-sea worm (Riftia pachyptila): an EFTEM study, in electron microscopy 1998, ICEM 14, Cancun, Mexico, 31 August–4 September 1998, 1, 695–696. Lins, U., Farina, M., 1999. Phosphorus-rich granules in uncultured magnetotactic bacteria. FEMS Microbiol. Lett. 172, 23–28. Luther, P.K., 1992. Sample shrinkage and radiation damage, in Electron Tomography: Three-Dimensional Imaging with the Transmission Electron Microscope. In: Frank, J. (Ed.). Plenum Press, New York, pp. 39–60. Mann, S., Sparks, N.H.C., Frankel, R.B., Bazylinski, D.A., Jannash, H.W., 1990. Biomineralization of ferrimagnetic greigite (Fe3S4) and iron pyrite (FeS2) in a magnetotactic bacterium. Nature 343, 258–261. Ravaux, J., Gay, L., Voss-Foucart, M.F., Gaill, F., 1998. Tube growth process in the deep-sea hydrothermal vent tube-worm Riftia pachyptila: synthesis and degradation of chitin. Cah. Biol. Mar. 39, 99–107. Richardson, K.C., Jarret, L., Finke, E.H., 1960. Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Technol 35 (313). Shillito, B., Lechaire, J.-P., Goffinet, G., Gaill, F., 1995. Composition and morphogenesis of the tubes of vestimentiferan worms. In: Parson, L.M., Walker, C.L., Dixon, D.R. (Eds.), Hydrothermal Vents and Processes, Geological Society Special Publication 87, pp. 295–302. Shillito, B., Lechaire, J.-P., Childress, J., Gaill, F., 1997. Diffraction contrast imaging of extracellular matrix components using zero-loss filtering. J. Struct. Biol 120, 85–92. Shively, J.M., 1974. Inclusion bodies in prokaryotes. Annu. Rev. Microbiol. 28, 167–187. Truchet, M., Ballan-Dufrançais, C., Jeantet, A.Y., Lechaire, J.P., Cosson, R., 1998. Le trophosome des vestimentifères, Riftia pachyptila et Tevnia jerichonana : Bioaccumulations métalliques et métabolisme du soufre. Cah. Biol. Mar. 39, 129–141. Varlot, K., Kasrai, M., Martin, J.M., Vacher, B., Bancroft, G.M., Yamaguchi, E.S., Ryason, P.R., 2000. Antiwear film formation of neutral and basic ZDDP: influence of the reaction temperature and of the concentration. Tribol. Lett. 8, 9–16.