Two New Beggiatoa Species Inhabiting Marine Mangrove Sediments in the Caribbean

Beggiatoaceae, giant sulphur-oxidizing bacteria, are well known to occur in cold and temperate waters, as well as hydrothermal vents, where they form dense mats on the floor. However, they have never been described in tropical marine mangroves. Here, we describe two new species of benthic Beggiatoaceae colonizing a marine mangrove adjacent to mangrove roots. We combined phylogenetic and lipid analysis with electron microscopy in order to describe these organisms. Furthermore, oxygen and sulphide measurements in and ex situ were performed in a mesocosm to characterize their environment. Based on this, two new species, Candidatus Maribeggiatoa sp. and Candidatus Isobeggiatoa sp. inhabiting tropical marine mangroves in Guadeloupe were identified. The species identified as Candidatus Maribeggiatoa group suggests that this genus could harbour a third cluster with organisms ranging from 60 to 120 μm in diameter. This is also the first description of an Isobeggiatoa species outside of Arctic and temperate waters. The multiphasic approach also gives information about the environment and indications for the metabolism of these bacteria. Our study shows the widespread occurrence of members of Beggiatoaceae family and provides new insight in their potential role in shallow-water marine sulphide-rich environments such as mangroves.


Introduction
Beggiatoa spp. are multicellular, filamentous colorless bacteria. Since their discovery by Vaucher in 1803, they are considered among the largest sulphur-oxidizing bacteria in nature [1]. permissions were required from these locations and activities. Our study did not involve endangered or protected species.

Fluorescence in situ Hybridization
Colorless filaments were prepared for FISH analysis according to previously described protocols [13]. After hybridization, samples were observed in MilliQ water with a drop of Vectashield using an epifluorescence Nikon microscope Eclipse 80i.
The specific probes (labelled with Cy3) were designed manually. Probes 16S ss-rRNA localization was optimized according to Fuchs et al. [35]. The probe's specificity was further tested with the online Probes Match tool provided by the Ribosomal Database Project [36].

Ultrastructural analysis
The ultrastructure of the colorless filaments was determined using a Scanning Electron Microscope (SEM Quanta 250, FEI). To this end, the bacterial filaments were fixed at 4°C in 2.5% glutaraldehyde in 0.1M cacodylate buffer (pH 7.2) which was made iso-osmotic to sea water by addition of sodium chloride and calcium chloride. Samples were then kept at 4°C until analysis. For conventional SEM analysis, samples were briefly rinsed, then dehydrated through a graded acetone series before drying with CO 2 using a critical point drier machine (EM CPD300, Leica). The samples were then sputter-coated with gold (Sputter Coater SC500, Biorad).
For EDX analysis, in order to avoid salt crystallization, samples were rinsed three times with deionized water, before observation with an ESEM Quanta 250 (FEI) operating from 10 to 20 kV under an environmental pressure of 7 Torrs at 5°C. EDX spectra were obtained using a M-max 500 mm 2 Oxford detector.
For Transmission Electron Microscopy (TEM) analysis, prefixed bacterial filaments were washed twice in 0.1M sodium cacodylate buffer in order to remove aldehydes before fixation for 45 min at room temperature in 1% osmium tetroxide in the same buffer. Then, samples were rinsed in distilled water, and post-fixed with 2% aqueous uranyl acetate for one hour more. After a rinse in distilled water, each sample was dehydrated through a graded acetone series and embedded in Epon-Araldite [37]. Thin sections (60 nm thick) were contrasted 30 min in 2% aqueous uranyl acetate and 10 min in 0.1% lead citrate before examination in a TEM LeO 912.

DNA extraction and PCR amplification
DNA was extracted from colorless filaments using DNeasy Blood & Tissue kit (Qiagen) according to the manufacturer's instructions. 16S rDNA were amplified using primers 8F/ 907R (for morphotype 1) and 8F/1492R (for morphotype 2) as previously described [38,39]. PCR amplifications were performed as follows: 95°C for 5 min, 35 cycles of 94°C 30 s, 58°C 45 s, 72°C 1min 30 sec and finally 72°C 7 min. PCR products were purified using QIAquick PCR purification Kit (Qiagen) and cloned with pGEM-T cloning kit (Promega) according to manufacturer's instructions. Inserts from 20 positive clones of each construction were fully sequenced by Genoscreen (http://www.genoscreen.com) using vector primers T7 and SP6. The sequences obtained in this study were deposited in the GenBank database under accession no. KF892059 and KF892060.

Phylogenetic analysis
The 16S rDNA gene sequences obtained were compared with the National Center of Biotechnology information (NCBI) (http://www.ncbi.nlm.nih.gov) database using BLAST [40]. Best hits were included in phylogenetic analyses. The phylogenetic analyses were conducted using MEGA version 5 [41]. Sequences were aligned using SINA alignment service [42] of the SILVA web site (http://www.arb-silva.de) and alignments were checked manually. The phylogenetic tree was constructed from the multiple-aligned data using the Neighbor Joining (NJ) method with Tamurai-Nei as genetic distance model. Nodes robustness was assessed by performing 1000 bootstrap replicates, and only bootstrap values above 49% are indicated at the nodes of the tree. Leucothrix mucor, Thiothrix nivea, and Achromatium spp were used as outgroup.

Nomenclature
The electronic version of this article in Portable Document Format (PDF) in a work with an ISSN or ISBN will represent a published work according to the International Code of Nomenclature for algae, fungi, and plants, and hence the new names contained in the electronic publication of a PLOS ONE article are effectively published under that Code from the electronic edition alone, so there is no longer any need to provide printed copies.
The online version of this work is archived and available from the following digital repositories: PubMed Central, LOCKSS.

Lipid characterization
Lipids were extracted from freeze-dried biomass of the two morphotypes using a modified Bligh and Dyer extraction [43]. The extracts were subjected to acidic methanolysis (ibid.) in order to remove polar head groups and to obtain free fatty acids. An aliquot was methylated with BF 3 -MeOH, treated with BSTFA in pyridine and subsequently analyzed by gas chromatographymass spectrometry (GC-MS) using a TRACE GC with a DSQ-MS, using a fused silica capillary column (25 m, 0.32 mm internal diameter) coated with CP Sil-5 (film thickness 0.12 μm) and helium as a carrier gas. To determine the double bond position of the fatty acids, they were derivatized with dimethyldisulfide/I 2 and the resulting methyltioethers were analysed by GC/MS.

Sulphide measurements
In situ measurements. In an attempt to characterize the in situ conditions, 10 measurements were performed in sediment areas covered by the white bacterial mat with autonomous potentiometric captors. Sulphide and pH captors were both used in order to calculate the sulphide rates. The sulphide and the pH measuring system were the same as the one previously described [12] and have been used in various habitats [12,44]. The electrodes were calibrated in the laboratory before deployment.
A series of 10 short term measurements was performed in 1 cm sediment under several patches of white mat with tightly attached sulphide and pH electrodes. The average of these measurements was calculated with standard deviation.
Mesocosm measurements. Mangrove sediment was brought to the laboratory and installed in a glass recipient until the sediment was reorganized. Mat was collected from the field the day after and transferred immediately (within 1 hour) into the mesocosm on the sediment (see S1 Fig.).
Oxygen and sulphide profile measurements were carried out using Clark-style oxygen (Oxy100) and sulphide (H 2 S100) microsensors with a 10μm tip manufactured by Unisense (Aarhus, Denmark) connected to a four channel Unisense picoammeter. Calibrations were performed according to Unisense instructions. The pH was measured with autonomous probe similar to the one described [44] fixed to the micromanipulator.
Vertical profiles were determined by moving the microelectrodes using a micromanipulator into the mat and recording the electrical current with SensorBasic software. Total sulphide concentrations (S 2tot = H 2 S+HS -+S 2-) were calculated taking into account the measured pH and salinity [45] using a pK = 6.51.

Morphology
The large colorless filamentous microorganisms were collected from extensive white mats ( Fig. 1a) located near the Rhizophora mangle roots in the tropical mangrove swamp in Guadeloupe. By light microscopy, it was noticed that the mat was composed mainly of 2 morphotypes with large colorless filaments (Fig. 1b).
The two colorless morphotypes described here are assemblages of cells constituting filaments of a total length of up to 30 mm ( Fig. 2 Fig. 2a-b) appeared, according to SEM observations, as pasted to the cell inner membranes. However, none of these vesicles were observed on the membranes separating two adjacent cells within the same filament (Fig. 3a). This observation was confirmed by TEM sections, which showed that these vesicles were linked to the inner membrane but not merged with it ( Fig. 3c). No free vesicles were observed in the cell cytoplasm, whatever section orientation was used ( Fig. 3c-d). Because sulphur is dissolved during dehydration processes, these empty vesicles observed in SEM

Phylogenetic analysis
The phylogenetic analysis was performed accordingly to the modern classification of large sulphur bacteria [41]. Neighbor-Joining (NJ) tree based on partial 16S rDNA sequences (925bp) revealed that the morphotype 1 forms a distinct clade with Uncultured Beggiatoa sp. clone WF120μm (Fig. 5) which falls into Candidatus Maribeggiatoa group [5]. The sister group was supported by the robust branch of the phylogenetic tree (100% bootstrap support from 1000 replicates). In contrast, phylogenetic analysis identified morphotype 2 as a sister group of Candidatus Isobeggiatoa spp. Thus, we proposed to name morphotype 1 strain as Candidatus Beggiatoa sp. Guadeloupe FWI and morphotype 2 strain as Candidatus Isobeggiatoa sp. Guadeloupe FWI in reference to the sister group they belong to and to the sampling site: Guadeloupe French West Indies.
The phylogenetic relationship of these two species was checked by in situ hybridization using specific probes (BEG572F for morphotype 1 and BEG282F for morphotype 2) designed from each bacterial sequence obtained in this study (Fig. 2b, e). The positive hybridization White arrows highlight the white sheath, the black arrows point out the membranes separating two bacterial cells, and the dotted arrows highlight the sulphur vesicles. The apex of the morphotype 1 filament is marked by a black star. The right identification of the two morphotypes is confirmed by the positive hybridization with the specific probes (BEG572F for morphotype 1 and BEG282F for morphotype 2) designed from each bacterial sequence obtained in this study (b and e are morphotypes 1 and 2, respectively). NONEUB probe was used as negative control (c and f for morphotypes 1 and 2, respectively). shown in Fig. 2b, e shows that the two morphotypes observed in the white mat correspond to the phylogenetic sequences previously obtained. A negative control was performed using NON338 probe (Fig. 2c, f). Lipid analysis showed that the fatty acids mainly consisted of C 16 and C 18 fatty acids with 0-1 double bonds and minor amounts of C 20 fatty acids in both morphotypes, with morphotype 2 containing significant amounts of a C 20 polyunsaturated fatty acid ( Table 1). The double bond position in the C 18:1 fatty acid in morphotype 2 was determined by DMDS adduction as ω-7, and, while concentrations of C 16:1 and C 18:1 fatty acids in morphotype 1 were too low for analysis after derivatization, retention times indicate an ω-7 position for those too.

Sulphur-oxidizing metabolism
In our study, EDXS analysis was performed using an environmental SEM (ESEM) allowing the observation of fully hydrated biological samples, and thus elemental sulphur was not dissolved during the preparation process of the samples. The EDX spectra showed that sulphur is the main element present within the organisms (Fig. 4a, b). Moreover, EDX cartography allowed to localize the elemental sulphur within granules that appeared as empty vesicles according to conventional SEM (Fig. 3b) and TEM pictures (Fig. 3c, d). Thus, both structural and EDXS These images highlighted small vesicles (white arrows) absent from the membranes separating two adjacent cells (dotted arrows). On higher magnification (b), some of these vesicles are fractured (white arrows), and appeared linked to the membranes. TEM microphotographs of the morphotype 1 (c) and morphotype 2 (d) highlight a large central empty space with all the cytoplasmic content postponed on the external membranes. The small vesicles (black arrows) also appear empty due to the loss of sulphur during dehydration process. They are absent from the membranes (dotted arrows) separating two adjacent cells. analyses demonstrate that the vesicles observed by light and electron microscope are sulphur storage granules (Fig. 2a, d; Fig. 3a-b).

Sulphide measurements
In order to characterize the mat environment in its natural biotope, in situ sulphide measurements were performed. The values obtained in mangrove from the ten profiles ranged from 189 μM to 2396 μM, with an average of 1187 μM (±728). Profiles from mesocosm experiments in the laboratory, in presence or absence of a bacterial mat, are shown in Fig. 6. In absence of bacterial mats (Fig. 6a), oxygen penetrated 0.5 mm into the sediment while sulphides were detected (detection level~1 μM) below a depth of 0.2 mm. Sulphide concentrations reached 988 μM (± 627) at 0.5 mm depth and increased continuously with depth. Therefore, the anaerobic sulphate-reducing bacteria (SBR) contained in the sediment were functional and produced sulphides by degradation of the organic matter by sulphate reduction.
In contrast, when the bacterial mat was present on the sediment (Fig. 6b), the oxygen concentration in the water column gradually decreased from 153 μM (± 45) to zero 3 mm above the sediment while sulphides were detected before entering the sediment. It was also observed that sulphide rates increased with depth until 3 mm with a maximum of 8197 μM (± 6030) and then decreased.
In the presence of the mat, a slope rupture of the sulphide concentration curve can be noticed, which means that the sulphide concentrations decreased quicker in the mat than in the sediment while it diffused to the surface. This data suggests that the bacterial mat consumed the sulphides coming from the sediment (due to SBR activity) quicker than natural sulphide oxidation by oxygen present in the seawater.

Phylogenetic placement of the new Beggiatoaceae species
In this study we described two new species of Beggiatoaceae family, which are forming microbial mats in marine mangroves from the Caribbean. According to the recently updated large sulphur bacteria phylogeny by Salman et al [46,47] our sequences studied here belong to two distinct taxa, Candidatus Isobeggiatoa sp. and Candidatus Maribeggiatoa sp., and share many characteristics with Beggiatoa alba, the type specie of Beggiatoaceae. In fact, for the two bacteria described here, the same morphology can be observed: multicellular filaments harbouring discoid cells with sulphur granules visible into incident light. Furthermore, both filaments can move by gliding on solid surface and possess a sheath.
Morphotype 1 is phylogenetically close to another bacterial species of 120 μm diameter identified as Maribeggiatoa [5] suggesting a division based on cell diameter within the Maribeggiatoa genus: one cluster harbouring species with a diameter between 12 to 18 μm, and a second one with diameter between 25 to 37 μm. Our data suggest that morphotype 1 and Uncultured Beggiatoa sp. clone 120 μm reported by Mckay et al. could form a third cluster with diameters ranging from 60 μm to 120 μm [5]. Additional phylotypes and new taxa identification are needed in order to clarify this cluster. In contrast, morphotype 2 clearly belongs to Isobeggiatoa group, which only gathers filaments with diameters between 10 and 40 μm. Interestingly, the sequence obtained here only shared 94% identity with available sequences from  Genbank. According to some authors, percentages lower than 95% could indicate a new genus [48,49]. Furthermore, all the Isobeggiatoa spp. sequences available are from cold temperate water species (Denmark and Germany), and Antarctic environments [46]. The results of the lipid analysis (Table 1) confirmed the genetic results, as the main fatty acids detected contained 16 and 18 carbon atoms with 0-1 ω-7 double bonds, concurrent with previously published results for Beggiatoa [50,51] and other sulphur-oxidizing bacteria Thioploca and Thiomargarita [51,52]. Interestingly, Jacq and co-authors [53], who characterized two types of filamentous bacteria retrieved from subtidal hydrothermal vents in southern California, were the only ones to also report small, but significant amounts of polyunsaturated C 20 fatty acids in both phenotypes. A C 20 polyunsaturated fatty acid was only detected in morphotype 2, phylogenetically characterized here as Isobeggiatoa, but was absent in morphotype 1 (i.e. Beggiatoa), suggesting that the Beggiatoa-like mats observed in subtidal hydrothermal vents may have been Isobeggiatoa [53]. C 15 and C 17 fatty acids, which are characteristic for sulphate-reducing bacteria, were absent.
This result shows that information based on 16S rRNA gene sequences is insufficient to identify new species and how it is necessary to use multiphasic approach to classify them. Nevertheless, further molecular investigations involving additional marker genes (i.e. 23S rDNA, ITS) and other multiphasic approach (e.g., physiological traits) could be used in order to resolve in depth the phylogeny of these species [46].

Sulphur metabolism of the new Beggiatoa species
Beggiatoa mats, as all microbial mats, are self-sustaining communities that support all major biogeochemical cycles [54]. The characterization of their chemical environments, either in situ or in mesocosms, by sensor measurements can provide information about their contributions to the ecosystem [55,56,57]. Mesocosm measurements were similar to those observed in previous studies undertaken in marine mangroves. Sulphide concentration increased with depth in the sediment in absence of mat [12,58]. Moreover, in our study, under the Beggiatoa mat, a decrease of sulphide concentration was observed after 4 mm depth. This phenomenon was already noticed in an ultramafic hydrothermal vent field [59], and in a sulfidic cave [60] where the sulphide rate did not only increase with depth, as shown in numerous studies [19,61]. No explanation for this phenomenon was given in the hydrothermal vent. However, Macalady et al. showed that in the sulfidic caves, it could be explained by diffusion-controlled transport and also by the fact that in sulfidic caves, sulphides diffused both from water above and from sediment below [60].
Measurements in a mangrove under a Beggiatoa mat showed that oxygen was fully consumed 2 cm above the mat [19]. In our study, an anoxic zone was present a few mm above the mat. Furthermore, under the filament network, sulphide concentrations were more important in the mesocosms than previously reported in literature. Indeed, a concentration of 1489 μM (± 1328) of sulphide was reached at 2 mm depth into the sediment in mesocosm, whereas in our in situ measurements, at 1 cm, we measured an average concentration of sulphide of 1193 μM (±728), similar to the measurements performed in the Twin Cays mangrove, where 1400 μM (± 1000) is reached at 1 cm depth [19]. Although concentrations are higher in mesocosm at this depth, at 1 cm, they are lower in the mesocosm than in situ. A significant heterogeneity existed within the sediment, in situ, and in the mesocosm as evident by the large standard deviations. This heterogeneity allowed us, to date, to consider the mesocosm as similar to in situ conditions. To our knowledge, no study has been conducted in mangrove habitats on Beggiatoa mats using microprobes. The concentrations measured by Lee et al. were obtained by colorimetry [19], with less precision than the measurements performed here using probes.
In two previous studies done in mangrove environments, sulphide concentrations were always lower than 1000 μM for 10 cm depth [12,58]. However, the locations were not next to or under Beggiatoa mats which are expected to be present at places where higher sulphide concentrations are available. Indeed, members of the Beggiatoaceae family are known to migrate in order to find the best gradient sulphur/oxygen for their development [62,63]. High currents or other mechanisms could also explain the mats' localization into the mangroves.
Our results are similar to those obtained in other environments. In hydrothermal vents and in the Santa Barbara Basin, it was shown that all oxygen was consumed within the first millimeters of the sediment while sulphide concentrations increased with depth [59,64]. However, in these two environments the sulphide concentrations were 25 to 150 times lower than those observed during our mesocosm experiment. In the hydrothermal vents, the maximum of sulphide concentration observed was 250 μM at 30 mm sediment depth [59], and in the Santa Barbara basin, a maximum of 50 μM was observed at 12 mm sediment depth [64].
The morphological study of Candidatus Beggiatoa sp. Guadeloupe FWI and Candidatus Isobeggiatoa sp. Guadeloupe FWI, highlighted that the sulphur inclusions visible in SEM and TEM images and identified by EDX, are joined to the plasmic membrane and are absent from membranes separating two adjacent cells. However, it was impossible to distinguish whether the sulphur granules were surrounded by a single membrane against the outer membrane, in invaginations of the cytoplasmic membrane, as previously suggested [28].
The Beggiatoa species described here are the predominant species in the filament network and thus probably the main microorganisms responsible for the sulphur consumption observed in the mat. Nevertheless, other sulphur-oxidizing bacteria could also participate in sulphide oxidation. The presence and activity of bacteria other than the giant Beggiatoa spp. were not determined in this study. Mesocosm measurements showed that, while oxygen was absent from the first millimeter of the mat, Beggiatoa cells were still present. These were probably cells from the anaerobic layer using dissimilatory nitrate reduction to ammonium in order to oxidize sulphur. Their need to oxidize sulphur and/or ammonia would cause migration to the oxic sediment layer. Indeed, SEM and TEM images highlighted a large free space in the cell with all the cytoplasmic content positioned against the outer membrane of the cell. These large vacuoles could be the nitrate vacuoles already encountered in previous large marine Beggiatoa spp. [27,29,63,65,66], and observed in Isobeggiatoa and Marithioploca strains [41]. These nitrate vacuoles allow the bacteria to survive anaerobically, oxidizing sulphides through nitrate reduction into dihydrogen and ammonia [1,30,67,68,69].
In our study, the internal component of the central space was not identified but TEM images showed that the empty area has no intracytoplasmic membrane. This is in accordance with De Albuquerque et al., who showed that the vacuoles have no internal membranes into marine and hypersalines studied mats [27], as observed also in Thioploca [69]. However, some marine sulphur-oxidizing bacteria from Thiothrix genera showed such vacuoles with no nitrate accumulation [70]. Thus, in absence of more information about the nature of the vacuoles and the nitrification rates of the mat, it is impossible to draw conclusions on the metabolism of nitrogen in these two new species of Beggiatoaceae.
It could be interesting to study the ammonium consumption of the Beggiatoa mat in marine mangrove in order to estimate their contribution to the nitrogen cycle regarding the mat composition. Furthermore, a recent study has shown that some non-marine Beggiatoa spp. from sulfidic caves are able to fix nitrogen [31]. This suggests that is possible that also Candidatus Beggiatoa sp. Guadeloupe FWI and Candidatus Isobeggiatoa sp. Guadeloupe FWI could fix nitrogen.
The Beggiatoa mats are also known to provide food for benthic foraminifera in temperate tidal flats and Antarctic shallow waters [64], but also for meiofauna and macrofauna of the Denmark cold waters [61]. In mangroves, the interactions between meiofauna and microbial mats have shown that some nematods and annelids feed on these mats, so the mat could be the source of a complex food web [71]. Thus, a detailed study of the interactions between these compartments will help to understand how the Beggiatoa mats contribute to the mangrove ecosystem.
This multidisciplinary study has revealed two new species of Maribeggiatoa and Isobeggiatoa, inhabiting the marine mangrove. This study is the first evidence for the presence of Isobeggiatoa spp. outside of northern Europe or Arctic waters. The multiphasic approach with use of microprobes, electron microscopy, lipid and phylogenetic analysis, has provided detailed information on species, and their sulphidic environment. Furthermore, the mesocosm study addresses some issues of the metabolism of these two species; and the results indicate that the role of the central vacuole is related to the dissimilatory nitrate reduction to ammonium. Our results are a first approach to ultimately understand the contribution of Beggiatoaceaedominated microbial mats to the biochemical cycles and food web of mangroves. They could constitute a base for further studies dealing with marine mangrove microbial mats.