Lytic archaeal viruses infect abundant primary producers in Earth’s crust

The continental subsurface houses a major portion of life’s abundance and diversity, yet little is known about viruses infecting microbes that reside there. Here, we use a combination of metagenomics and virus-targeted direct-geneFISH (virusFISH) to show that highly abundant carbon-fixing organisms of the uncultivated genus Candidatus Altiarchaeum are frequent targets of previously unrecognized viruses in the deep subsurface. Analysis of CRISPR spacer matches display resistances of Ca. Altiarchaea against eight predicted viral clades, which show genomic relatedness across continents but little similarity to previously identified viruses. Based on metagenomic information, we tag and image a putatively viral genome rich in protospacers using fluorescence microscopy. VirusFISH reveals a lytic lifestyle of the respective virus and challenges previous predictions that lysogeny prevails as the dominant viral lifestyle in the subsurface. CRISPR development over time and imaging of 18 samples from one subsurface ecosystem suggest a sophisticated interplay of viral diversification and adapting CRISPR-mediated resistances of Ca. Altiarchaeum. We conclude that infections of primary producers with lytic viruses followed by cell lysis potentially jump-start heterotrophic carbon cycling in these subsurface ecosystems.


nature research | reporting summary
April 2020

Data analysis
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A list of figures that have associated raw data -A description of any restrictions on data availability Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. This study was set out with the aim of elucidating viral infections of Ca. Altiarchaea across the globe using in silico (sequencing-based) methods as well as microscopy methods.
Public datasets were searched for Ca. Altiarchaea and chosen based on a hit with respective hamus proteins and manually selected based on available data structure. One sampling site was chosen to further elucidate the predicted host-virus relationships based on accessibility and biomass of the site: We collected microbial biofilm and water samples from an aquifer (sulfidic spring Mühlbacher Schwefelquelle Isling, MSI) that contained uncultivated archaea (mainly Altiarchaea) and a small proportion of bacteria. There was no further selection of samples for metagenomics or microscopy: For metagenomics, we used bulk samples that contained hundreds of biofilm flocks or dozens of liters of groundwater for filtration; for microscopy, biofilm flocks from the aquifer were randomly chosen after sampling.
For sampling the unfiltered planktonic microbial community for metagenomics, 70 L of groundwater were filtered onto a 0.1 µm pore-size PTFE membrane filter (Merck Millipore, Darmstadt, Germany). The flow-through was collected in a sterilized container, and a final concentration of 1 mg L-1 of iron (III) chloride (Carl Roth, Karlsruhe, Germany) was applied for chemical flocculation (  . Flocculates were filtered onto 5 x 0.2 µm membrane filters (<0.1 µm fraction, viral fraction). We chose to sample the archaeal biofilms, planktonic and viral fractions to determine viral infestations for planktonic and sessile Altiarchaeota. Metagenomes from whole community DNA reflected temperate and actively replicating viruses, whereas the viral fraction (virome) predominantly contained free viruses and lysed (Altiarchaeota) cells remnants. Sequencing viromes generally also increases the likelihood of finding rare viruses. For virusFISH we first established the method with respective controls and then randomly took 18 individual biofilm flocks from the ecosystem for determining infection categories of the 8.9kb long, dominant virus. There was no selection of biofilm flocks based on certain parameters.
Groundwater for metagenomic data of the Mühlbacher Schwefelquelle, Isling, Germany were collected 15th to 17th October 2018 in presence of Perla Abigail Figueroa Gonzalez, Janina Rahlff and Alexander Probst. Three types of samples were collected for metagenomics: i) biofilm flocks; ii) the planktonic community (>0.1 µm pore-size fraction); and iii) the viruses and lysed cells (<0.1 µm pore-size fraction) using a custom-made filtration system (serial number SN 0604_2018) by Sima-Tec (Schwalmtal, Germany) and 142 mm diameter filter membranes. Shotgun metagenome sequencing was performed using an Illumina HiSeq platform.
Biofilm material for virusFISH was collected 17th to 19th January 2019 by Victoria Turzynski, Indra Monsees and Alexander Probst. Detailed information on the sampling procedure for data collection can be found in the manuscript. Biofilm material was examined with an epifluorescence microscope Axio imager M2m equipped with an Axio Cam MRm (Carl Zeiss Microscopy GmbH, Germany). Groundwater for metagenomic data of the Mühlbacher Schwefelquelle, Isling, Germany were collected 15th to 17th October 2018. Biofilm material for TEM analysis was collected on 17th to 19th October 2018 and for virusFISH from 17th to 19th January 2019.
We do not show the data and analysis of non-matching spacers derived from viral mini CRISPR arrays that we checked for their role in interviral conflicts, because the spacers did not match any viral targets and there is consequently no data to show.
We initially identified the target virus for virusFISH in public metagenomes from 2012. To reproduce the identification and reconstruction of the virus and of the virus-host relationship from the metagenome, we did not perform technical replicates but biological replicates by focusing on size fractionation of the biological samples: We performed metagenomics of a biofilm sample, from the planktonic fraction in the community and from the viral fraction, for all of which we were able to recover the respective virus.
Detailed information can be found in the method section of the manuscript. FISH experiments were highly reproducible: Since we established virusFISH for this project in 2019, we have tested the chemically synthesized probe mix targeting Altivir_1_MSI 70 times in total and the Metallosphaera sp. virus probe (as negative control) 66 times in total. The testing was always carried out on Altiarchaea biofilm samples from MSI site.
During randomization experimental subjects are assigned to different treatments, but in this study we did not compare different treatments in that sense.
Blinding is usually performed in experiments involving living participants, which are kept unaware of the treatment group they get assigned to. This was not applicable in our study as human participants were not part of the experiments. Blinding during virusFISH experiments was not applicable, because the experimenter needs to know which biofilm flock was treated with positive and negative probes.
Groundwater was sampled from a ca. 10 degree celsius cold aquifer, no rainfall, pH ca. 7.2.
Samples were collected for microscopy and metagenomic analyzes from the sulfidic spring Mühlbacher Schwefelquelle Isling (MSI) in Regensburg, Germany (N 48°59.142, E 012°07.636). Data from other ecosystems were obtained from public databases or sampled and reported within another study as referenced in the manuscript.
No import or export of samples was conducted as the sampling site is in Germany and the executing lab was also located in Germany. Samples from the aquifer were taken after local authorities (city of Regensburg) had approved the sampling in the naturally protected area. Permission for sampling the Mühlbacher Schwefelquelle was granted on 29th June 2018 by Mr. Forster, Umweltamt Stadt Regensburg (Az. 31.4 Fo) and is valid for 5 years.
The sampling was non-invasive to the ecosystem as upwelling groundwater has been sampled.