The sampling campaign was carried out in January 2017 and samples were collected from a fumarole’s wall (14°14′21″N; 40°17′55″E) located at the main Dallol hydrothermal outcrop in the Danakil depression and a blue pool surrounding the small fumarole. Samples were collected aseptically, using sterilized spatulas and plastic aseptic materials in 12 mL sterile glass vials. The vials were completely filled to prevent head space. To ensure that the samples did not oxidize during transportation and storage, the vials were sealed with a septum tap, covered with parafilm tape and kept under anaerobic conditions. Samples were transported at room temperature.

In Situ” physico-chemical parameters measurement

Physico-chemical parameters (T: °C, Eh: mV; conductivity: mS/cm2) were measured “in situ” using a multi-parametric probe, YSI 556 MPS. Gas composition by gas chromatography and elemental composition using ICP-MS.

Electron microscopy

Samples were fixed in 4% paraformaldehyde and 2% glutaraldehyde, in 0.1 M phosphate buffer (pH 7.2), for 2 h at room temperature. Fixed samples were washed three times in the same buffer and post-fixed with 1% OsO4 in water for 60 min at room temperature in the dark. After three washes in distillate water, samples were incubated with 2% aqueous uranil acetate for 1 h at room temperature, washed again, and dehydrated in increasing concentrations of ethanol 30, 50, and 70% 20 min each, 90% 2 × 20 min, and 100% 2 × 30 min at room temperature. Dehydration was completed with a mixture of ethanol/propylene oxide (1:1) for 10 min and pure propylene oxide 3 × 10 min. Infiltration of the resin was accomplished with propylene oxide/Epon (1:1) for 45 min and pure LR White resin (London Resin Company limited, England), overnight at room temperature. Polymerization of infiltrated samples was done at 60 °C for 2 days. Ultrathin sections of the samples were stained with uranyl acetate and lead citrate by standard procedures.

Four types of electronic microscopy techniques were used for this study: Scanning Electron Microscope (SEM) (JEOL-5600 coupled to an EDX, INCA) with an Energy Dispersive X-Ray Analyzer (EDX) and SEM-FI (Philips XL30-FEG); Scanning Electron Microscopy-Field Emission Gun (SEM-FEG) (Philips XL30-FEG); Transmission Electronic Microscope (TEM) (JEM-1010); TEM/STEM electronic microscope (JEOL 2100 K) with FEG. Samples were mounted on conductive graphite stubs and sputter and gold-coated in a Quorum, Q150T-S apparatus to ensure electrical conductivity and prevent charging under electron beams.

For SEM analysis, samples were also analyzed with EDX to obtain semiquantitative chemical data. A STEM unit was coupled to the microscope with acquisition of contrast images z (HAADF). The qualitative element composition of samples was determined using an INCA-X-SIGHT with a Si-Li detector (Oxford, England) with a detection limit of 10% for the main element. The operating energy was 200 kv. For TEM analyses, the instrument was operated at 200 kv (Catalysis Institute and PETROQUIMICA-CSIC), equipped with an energy dispersive X-RAY microanalysis instrument INCA (Oxford, England). The TEM/STEM was operated at 200 kv with an EDX, which was coupled to a STEM unit, with acquisition of contrast images z (HAADF). The same grids were used for both TEM and TEM-STEM.

Fluorescence microscopy

Samples were fixed in 4% formaldehyde in phosphate-buffered saline (PBS) for 2 h and stored at 4 °C until further processing. Fixed samples were dispersed by 3 cycles of sonication of 30 sec, with a 30 sec break in between with 1 pulse per second (intensity of 20%). The samples were filtered sequentially through a 0.4 μm pore size polycarbonate filters, a 0.2 μm pore size filter and a 0.1 μm pore size filter (Millipore, Germany). Filters were pretreatment as previously described18. FISH was performed as described by Glöckner, et al.19 using Cy3 single-labeled Narc1214 probe20 (Biomers, Ulm, Germany). Stringencies were regulated adjusting formamide and NaCl concentration in hybridization and washing buffer at 30% (vol/vol) and 0.112 M respectively. Filters were counterstaining by incubation with SYBR® Gold (Molecular Probes, Eugene, OR, USA) 1X diluted in milliQ water for 15 min, before being mounted on glass a slide using Vectashield (Vector Laboratories, Burlingame, CA, USA): Citifluor (Citifluor, London, United Kingdom) (1:4). Samples were imaged using a confocal laser scanning microscope LSM710 (Carl Zeiss, Jena, Germany) equipped with diodo (405 nm), argon (458/488/514 nm) and helium and neon (543 and 633 nm) lasers. Fiji software21 was used to process images.

DNA extraction and amplification of ribosomal genes

Samples were filtered through polycarbonate membranes (0.45, 0.2 and 0.1 µm diameter; Millipore, USA). Total genomic DNA was extracted from those membranes using the DNeasy PowerSoil Kit (Qiagen, Germany) as described in the manufacture’s instructions, and concentrated using a SpeedVac Concentrator. The quantity of the extracted DNA was analysed by fluorimetry using a Qubit 2.0 fluorometer (Thermo Fhiser Scientific, USA). Total genomic DNA were stored at −20 °C for sequencing.

Ribosomal genes were amplified using a primers set specific to amplify V2-V3 region of 16S rRNA in Archaea domain (Arch1F 5′-CGGRAAACTGGGGATAAT-3′ and Arch1R 5′-TRTTACCGCGGCGGCTGBCA-3′). PCR reactions were performed as described22,23. Gel electrophoreses (1% agarose Conda, Spain in 0.5X TBE buffer, 90 mV during 30 min and staining with GreenSafe Premium NZYTech) were carried out to check the size and quality of PCR products.

Ribosomal genes library preparation and sequencing

Library preparation and 2 × 300 pair-end sequencing by Illumina MiSeq were made by Genomic Unit in Parque Científico de Madrid Foundation/FPCM (Madrid, Spain).

Total genomic DNAs was quantified by Picogreen. Then, an input of 16 pg of DNA and 27 cycles were used in a first PCR with Q5® Hot Start High-Fidelity DNA Polymerase (New England Biolabs, USA) in the presence of 100 nM primers for 16S amplification (5′-ACACTGACGACATGGTTCTACACCTACGGGNGGCWGCAG-3′ and 5′- TACGGTAGCAGAGACTTGGTCTGACTACHVGGGTATCTAATCC-3′, these primers amplify the V3-V4 region of 16S), 200 nM primers for Archaea amplification (5′-ACACTGACGACATGGTTCTACACGGRAAACTGGGGATAAT-3′ and 5′-TACGGTAGCAGAGACTTGGTCTTRTTACCGCGGCGGCTGBCA-3′),

After the first PCR, a second PCR of 15 cycles was perfomed with Q5® Hot Start High-Fidelity DNA Polymerase (New England Biolabs, USA) in the presence of 400 nM of primers (5′-AATGATACGGCGACCACCGAGATCTACACTGACGACATGGTTCTACA-3′ and 5′-CAAGCAGAAGACGGCATACGAGAT-[10 nucleotides barcode]-TACGGTAGCAGAGACTTGGTCT-3′) of the Access Array Barcode Library for Illumina Sequencers (Fluidigm, USA).

The amplicons were validated and quantified by a Bioanalyzer. An equimolecular pool was purified by gen extraction and titrated by quantitative PCR using the “Kapa-SYBR FAST qPCR kit for LightCycler480” with a reference standard for quantification. The pooled amplicons were denatured and added to the flowcell at a density of 9 pM. Clusters formed, which were sequenced using a “MiSeq Reagent Kit v3”, in a 2 × 300 pair-end sequencing run on an Illumina MiSeq sequencer.

Detection of DPANN OTUs by microbiome analyse

Quality of reads was evaluated by means of FastQC software. PANDAseq Assembler was using for assembling forward and reverse reads and convert in a fasta file24. Sequencing data were processed using Qiime software package version 1.9.025. High quality contigs were clustered into OTUs based on 94% sequence similarity with UCLUST. The first sequence for each OTU as the representative OTU, which were aligned using PYNAST26. The taxonomic identity of each phylotype was determined using the SILVA_132_QIIME database ( Since Ultra small microorganisms are the target of this study, their representative sequences were filtered from biom table in order to analysis only these taxa using and scripts in Qiime. Finally, the ARB software package28 was used to reconstruct the phylogenetic tree. The neighbour-joining with Felsenstein correction included in the ARB package was used for phylogenetic inference. The robustness of the reconstructed trees was evaluated by bootstrap analysis of 1000 resampled datasets.