Study area and sampling
The L'Atalante deep hypersaline anoxic basin (DHAB) was discovered in the Mediterranean Sea in 1993 during an expedition that was a part of the European funded project "Mediterranean Ridge Fluid Flow". The bottom of the L'Atalante basin is a relatively flat area bounded to the southwest by the Cleft Basin and it is characterised by a morphological escarpment that is several hundreds of metres high, which is the sea-bottom expression of the main back thrust of the accretionary ridge. These characteristics originated from the dissolution of buried salt deposits (evaporitic deposits), which remained from the hypersaline waters of the Miocene period (5.5 My before present). The L'Atalante basin is characterised by the presence of a thick brine layer (ca. 40 m) with high density (1.23 g cm-3) and high contents of Na+ (4,674 mM), Cl- (5,289 mM) and Mg+ (410 mM) . This layer limits the mixing with the overlying oxic deep-waters to only the upper 1 m to 3 m of the brine, and it additionally acts as a physical barrier for particles settling to the bottom sediments. As a result, the inner part of the L'Atalante basin is completely anoxic since 53,000 yrs before present  and is characterized by elevated methane (0.52 mM) and hydrogen sulphide (2.9 mM) concentrations . Undisturbed sediment samples (down to a depth of 30 cm) were collected using a USNEL type box corer (surface ca. 0.2 m2), in 1998, 2005, 2006 and 2008. The samples from the DHAB sediment were collected in December 1998 (at 3,363 m depth, 35°18.20'N, 21°23.33'E), August 2005 (at 3,600 m depth, 35°18.23'N, 21°23.33'E), and June 2008 (at 3,450 m depth, 35°18.18'N, 21°23.35'E). In 1998 and 2008 additional sediment samples were collected outside the L'Atalante basin (ca. 10 miles from the DHAB; 35°11.84'N, 21°24.75'E) at ca. 3,250 m depth, for investigation of the characteristics of meiofaunal metazoans from the oxygenated adjacent systems (three sampling sites per period with three to five replicated deployments per site). In the northern-eastern Atlantic Ocean, oxygenated deep-sea sediment samples (55°29.87'N, 15°48.61'W at 600 m depth) were collected during the 2006 expedition. Loriciferans retrieved from these sediments were used for the comparison of their body composition with loriciferan specimens collected in the anoxic sediments of the L'Atalante basin. Sediments retrieved from the deep anoxic basin were immediately processed under strict anaerobic conditions.
Extraction and identification of benthic metazoans
For the extraction of metazoan fauna from the sediments, the samples (top 15 to 20 cm of the sediment cores) were pre-filtered through a 1,000-μm mesh (to remove larger debris), and a 20-μm mesh was used to retain all of the multi-cellular organisms. The fraction remaining on this latter sieve was re-suspended and centrifuged three times with Ludox HS40 (density 1.31 g cm-3) . All of the organisms isolated were counted and classified according to standard protocols [34, 35]. Only the organisms collected during the first expedition were stained with Rose Bengal (0.5 g L-1), a stain commonly used to highlight the body structures under light microscopy. On average of all collected samples metazoan abundance was 2,075 ind. m-2 in the L'Atalante sediments vs 21,548 ind. m-2 in the oxygenated sediments surrounding the basin. In the anoxic sediments of the L'Atalante basin, Loricifera accounted for 16.1% of the total metazoan abundance. No Loricifera were encountered in the oxygenated sediments surrounding the basin, where nematodes and copepods accounted for 95% and 4%, respectively, of the total metazoan abundance.
Identification of loriciferans to genus and species level with light and scanning electron microscopy
The extracted specimens were mounted on microslides in a drop of distilled water. The water was progressively replaced by increasing glycerol concentrations (5%, 10%, 25%, 50% and 100% vol water:vol glycerol). Then the specimens were sealed with Glyceel. The microslides were analyzed using a light microscope with phase contrast and Nomarski DIC optics. Micrographs of the specimens were taken on an Olympus BX51 microscope equipped with a digital Olympus C-3030 zoom camera and on a Leica DMRXA microscope with a digital Leica DC200 camera (Leica Camera AG, Solms, Germany). Morphological details of the loriciferans were obtained by scanning electron microscopy. Loriciferans extracted from sediments were carefully rinsed in distilled water and then dehydrated through a graded series of ethanol and acetone prior to critical-point drying. The dried specimens were mounted on aluminium stubs and coated with gold prior to observation under scanning electron microscopy (Philips XL20, Philips Electronics, Eindhoven, The Netherlands).
Incorporation of (3H)-leucine
For investigating the vitality of the meiofaunal metazoans, the top 5 cm of intact sediment cores were incubated with (3H)-leucine . Replicate sediment samples (n = 3, internal diameter 5.5 cm, approximately 120 cm3 of sediment per replicate sample) were kept in the dark at in-situ temperature and under anoxic conditions (a N2 atmosphere); these were injected with 10 mL (3H)-leucine dissolved in 0.2 μm filtered, autoclaved and degassed deep-sea water (final concentration 0.2 mCi mL-1). Controls for the incubation experiments were obtained as follows: additional sediment cores were frozen immediately after collection at -80°C, to kill all metazoans within the samples. After thawing, when the samples reached the in-situ temperature, the sediments were incubated with an aqueous solution of (3H)-leucine and then processed as described above. We used deep freezing to kill animals, as previous studies have demonstrated that meiofauna fixed using chemical compounds (that is, formaldehyde, glutaraldehyde and ethanol) show a significant loss in the incorporated radioactivity . All samples were incubated on deck (101,325 Pa) under anoxic conditions (N2 atmosphere) for four hours in the dark, and at the in-situ temperature (about 14°C). At the end of the incubations, the samples were deep-frozen in liquid N2 to stop any additional substrate uptake. In the laboratory, the organisms were extracted from the sediment as previously described. Due to the relatively low numbers of loriciferans in the sediment cores (n = 3 both in the control and treated samples) the organisms were analyzed individually. Meiofaunal organisms were rinsed with 0.2-μm pre-filtered seawater (to minimise interference due to radioactivity incorporated by prokaryotes that were potentially present on the metazoan surface)  and transferred to scintillation vials. The samples were digested at 50°C for 24 h using 1 mL tissue solubiliser (Soluene-350, Packard Inc., Meriden, Connecticut, US). After addition of 10 mL scintillation cocktail, the radioactivity (as disintegration per minute; DPM) in the loriciferans was determined in a liquid scintillation counter (Packard, Tri-Carb 2100 TR). DPM data were normalised per individual.
To test the accuracy and consistency of the radiotracer experiments carried out on sediments collected in the L'Atalante basin, additional experiments were performed on coastal sediments of the Mediterranean Sea. Loriferans were not present in these samples; therefore nematodes were used as model organisms. After incubation with the radiolabelled substrate, the nematodes (diameter: 20 to 30 μm and length: 200 to 900 μm) were extracted from the sediments and analyzed individually or pooled together (from 2 to 10 individuals). These experiments demonstrated that the radioactivity incorporated into the nematodes is significantly higher than that found in organisms used as controls, even when a single individual is analyzed (Table 1). Moreover, radioactivity measured from the nematodes incubated with radioactive substrates increased linearly with the increasing number of individuals analyzed.
Incorporation of Cell-Tracker™ Green CMFDA
After sediment retrieval from the anoxic basin, the top 5 cm of the sediment cores and its anoxic overlying water were maintained under strict anaerobic conditions (N2 atmosphere) and incubated on deck (101,325 Pa) in the dark and at the in-situ temperature (ca 14°C). The samples were used for incorporation experiments with Cell-Tracker™ Green CMFDA, fluorescent probe (5-chloromethylfluorescein diacetate; Molecular Probes, Inc., Eugene, Oregon, US; 10 μM final concentration). The Cell-Tracker™ Green fluorescent CMFDA probe penetrates the cells and reacts with the intracellular enzymes, generating fluorescence . This molecular probe is specifically designed for testing the presence of metabolic activity and is therefore used here to support the evidence of viability of the metazoans present within the anoxic deep-sea sediments. The sediment samples were incubated for four hours. Controls for the incubation experiments were obtained as follows: additional sediment cores were frozen immediately after collection at -80°C to kill all metazoans within the samples. After thawing, when the samples reached the in-situ temperature, the sediments were incubated with an aqueous solution Cell-Tracker™ Green CMFDA, and then processed as described above. At the end of the incubation, the samples were deep-frozen in liquid N2 to stop any metabolic reactions, and the recovered loriciferans were placed on concave slides containing a drop of 0.9% NaCl solution (previously autoclaved). The fluorescence of the organisms was examined using a confocal microscope equipped with Kr/Ar mixed gas laser (Bio-Rad MRC 1024 UV; Bio-Rad, Hercules, California, US) using excitation wavelengths 488 nm and the emission has been detected after passing a bandpass filter of 522/35 nm. The confocal laser images were acquired (using the same laser emission power, iris and electronic gain for all acquisitions) in the Bio-Rad PIC format using the Bio-Rad Lasersharp Acquisition software (Release 2.1). The organisms were investigated using exactly the same magnification (× 40) in order to allow data comparison. Images were taken at depths of 3 μm for a total of 21 sections per animal and analyzed using the Bio-Rad Lasersharp processing tool. This enabled merging all of the sections (without any contrast manipulation) and measuring the mean scale colour (0 to 255) of the entire animal body. Images were sequentially acquired and stored as TIFF files. The reliability of the control used in the experiment was previously tested by means of repeated (n = 5) incubation experiments with Cell-Tracker™ Green CMFDA performed on two nematode species cultured in the laboratory (Diplolamelloides myily and Diplolaimella diewgatentis). All of the specimens were analyzed by confocal laser microscopy, as described above.
X-ray micro-analysis of the elemental composition of Loricifera
After extraction from the sediment, loriciferans from both the L'Atalante basin (undescribed species of the genus Spinoloricus, only adults) and the deep NE Atlantic Ocean (Rugiloricus cauliculus cfr) underwent quantitative X-ray micro-analysis, after coating with graphite. Specimens collected in the oxygenated sediments were used as a reference. The selected parts were: abdomen, the posterior lorica and the whole organism (Additional file 4).
Spectroscopic infra-red determinations
Fourier transformed Infra-Red (FT-IR) spectroscopic determinations were carried out on loriciferans collected both from the anoxic sediments of the L'Atalante basin and from oxic sediments of the NE Atlantic Ocean. Spectral data were obtained with a Perkin-Elmer Spectrum One FT-IR equipped with a Perkin-Elmer Autoimage microscope (PerkinElmer Life and Analytical Sciences, Shelton, Connecticut, US). Spectra were measured from 4,000 to 400 cm-1 at a spectral resolution of 4 cm-1 with 128 scans. The spatial resolution was 30 × 30 μm. Background scans were obtained from a region of no sample and rationed against the sample spectrum. The samples were deposited first on a steel support to collect reflectance spectra and on the centre of a BaF2 plate for transmittance spectral acquisition. Specific areas of interest were identified by means of the microscope television camera. Baseline (polynomial line fit) was performed in all cases while Second Derivative, Fourier Self Deconvolution and Curve Fitting (Gaussian character) procedures were used to determine the absorbance ratio between the bands of interest. All spectra were scaled for equal intensity in the Amide I band. For data handling, the Spectrum v.303 (Perkin-Elmer) software package was used.
Analysis of the ultra-structure of loriciferans by transmission electron microscopy
For ultrastructural studies, loriciferans (undescribed species of the genus Rugiloricus) extracted from sediments were carefully rinsed in distilled water and then stored in glutaraldehyde (2% final solution) for transmission electron microscopy examinations. After treatment with osmium (one hour incubation) and acetone dehydration (two times at 60% for one minute, and three times at 100% for one minute), loriciferans were embedded in epoxy resin. Ultrathin sections (78 nm) were obtained using a microtome (Model RMC MTX, Boeckeler Instruments Inc., Tucson, Arizona, USA) equipped with a diamond knife. Sections were collected on carbon-coated formvar supports, stained with lead citrate, and examined by transmission electron microscopy (Philips EM 208).