Succession of the bacterial community from a spacecraft assembly clean room when enriched in brines relevant to Mars

Abstract

Interplanetary spacecraft are built in a spacecraft assembly facility (SAF), a clean room designed to reduce microbial contamination that could confound life detection missions or influence native ecosystems. The frigid hyperarid near-surface environment of Mars has ample hygroscopic Mg and Na salts of chloride, (per)chlorate and sulphate that may deliquesce to form dense brines, liquids with low water activity, and freezing points <0°C. The current study sought to define the climax microbial community after 6 mo of enrichment of SAF floor wipe samples in salt plains medium supplemented with 50% (w/v; ~2 M; a_w = 0.94) MgSO₄ or 20% (w/v; ~1.9 M; a_w = 0.91) NaClO₃. After 1 wk, 4 wk and 6 mo of incubation, metagenomic DNA extracts of the enriched SAF microbial community were used for high-throughput sequencing of 16S rRNA genes and subsequent phylogenetic analyses. Additionally, dozens of bacterial strains were isolated by repetitive streak-plating from the climax community after 6 mo of enrichment. Early in the enrichment, staphylococci greatly dominated and then remained abundant members of the community. However, actinobacteria succeeded the staphylococci as the dominant taxa as the cultures matured, including Arthrobacter, Brachybacterium and Brevibacterium. A diverse assemblage of bacilli was present, with Oceanobacillus being especially abundant. The SAF culture collection included representatives of Brachybacterium conglomeratum, Brevibacterium sediminis, Oceanobacillus picturae and Staphylococcus sciuri. These were characterized with biochemical and physiological tests, revealing their high salinotolerance. Shannon diversity indices were generally near 2, reflecting modest diversity at several levels of identity and the community structures were uneven throughout. However, minor members of the community seem capable of the ecosystem functions required for biogeochemical cycling. For instance, organisms capable of all the functions of the N cycle were detected. The microbial assemblage in SAFs is the most likely to be transported by spacecraft to another world. While individual microbial populations may exhibit the qualities needed for survival at the near-surface of Mars, certainly entire communities with the capacity for complete biogeochemical cycling, would have a greater chance of survival and proliferation.

Introduction

Robotic spacecraft that are destined to visit celestial bodies are built in clean rooms to protect against particles that could contaminate their components. Missions directed at worlds that have substantial astrobiological potential, where liquid water may exist, need to be free of levels of bioburden that are likely to cause planetary contamination. This is critical for life detection missions. Protecting the natural environment of locations that could support life increases the chance of detecting native microbial communities and recognizing these with certainty. It remains unclear whether forward contamination of Mars or the ocean worlds could lead to successful colonization by terrestrial microbes. We seek a better understanding of the microbial assemblages likely to be ensconced on robotic spacecraft and their tolerances to the extreme chemical and physical conditions of extraterrestrial environments.

The surface and near-surface of Mars is a challenging place for life to survive and proliferate (Mancinelli et al., 2004; Davila et al., 2010; Rummel et al., 2014). While ultraviolet radiation may be avoided in the shade of rocks and soils, the aridity of Mars is unavoidable in typical near-surface locales. Liquid water is expected to be scarce near the surface of Mars. Given the low surface temperatures, only dense brines have a real possibility of persisting as liquids under conditions similar to those found on Mars today. High concentrations of salts act to lower the freezing point of water, in some cases substantially. Certain chloride, (per)chlorate and sulphate salts relevant to Mars can lower the freezing point of water to near −70°C (Nuding et al., 2014; Rummel et al., 2014; Fischer et al., 2016; Jänchen et al., 2016; Primm et al., 2017; Nair and Unnikrishnan, 2020; Pál and Kereszturi, 2020; Rivera-Valentín et al., 2020). However, dense brines are so salty as to also lower the water activity (a_w) of the solution dramatically, to levels that can inhibit microbial growth (Grant, 2004; Schneegurt, 2012). Cells need liquid water to survive, but this water also must be bioavailable. Low a_w restricts growth to only those microbes physiologically adapted to these harsh chemical conditions. For instance, the vast majority of microbes cannot proliferate in seawater, with a modest salinity of 2% NaCl that lowers a_w to 0.97 (with a_w = 1.0 representing pure water). A saturated solution of NaCl lowers a_w to 0.75 and only halotolerant microbes can grow. LiCl and (per)chlorate salts can lower a_w to well below 0.6, which may be nonpermissive for microbial growth of any kind (Grant, 2004; Rummel et al., 2014; Hallsworth, 2019).

Using clean rooms as spacecraft assembly facilities (SAFs) greatly reduces the bioburden on spacecraft but does not eliminate biocontamination. Clean rooms can be contaminated by infiltration of outside air, the introduction of spacecraft parts, tools and test equipment, and by workers entering the SAF, despite personal protective equipment and airlocks. Previous studies of the microbial communities in SAFs during the assembly of several spacecraft have demonstrated that a diverse collection of bacteria and archaea can be cultivated from swabs of SAF surfaces (Foster and Winans, 1975; Puleo et al., 1977; La Duc et al., 2003; Moissl et al., 2008). Molecular analyses have uncovered an even broader assemblage of microbes than cultivation campaigns (Moissl et al., 2007; La Duc et al., 2009, 2012; Weinmaier et al., 2015; Bashir et al., 2016; Danko et al., 2021; Hendrickson et al., 2021; Highlander et al., 2023). Clean rooms are relatively low in humidity, so successful microbial colonizers may demonstrate salinotolerance, since dry environments tend to deposit evaporite minerals. Substantial salinotolerance has been demonstrated for individual microbial isolates from SAFs and more broadly across large isolate collections that appear to be enriched for salinotolerant representatives (Moissl-Eichinger et al., 2013; Venkateswaran et al., 2014; Smith et al., 2017; Zanmuto et al., 2018). This observation increases the likelihood that microbes contaminating SAFs might gain a foothold on Mars or a salty ocean world. However, previous studies have not examined a broad range of salts and were limited to individual microbial strains in isolation. The current study followed changes in the microbial communities derived from swabs of SAF surfaces, when enriched for months in dense brines of MgSO₄ and NaClO₃. The results show that certain genera rise to dominance during ecological succession under these extreme chemical conditions. The climax community that persists seems more likely to survive and proliferate than individual microbial strains in potential brines on Mars or the ocean worlds.

Methods

Sampling of SAF

Sterile polyester wipes (Texwipe; Kernersville, NC), moistened with 15 ml of sterile water, were used to swab 1-m² surfaces of high-traffic floors of the aseptic assembly facility at Jet Propulsion Laboatory (JPL) during the assembly of the Mars 2020 Sample Caching System hardware (Fig. 1). The aseptic assembly facility is a certified ISO 5 clean room. All entrants into the ISO 5 clean room donned sterile gowning and gloves. The environment was monitored for biological cleanliness by surface sampling, air sampling and utilization of an instantaneous detection system for airborne particles (microbial and inert).

Figure 1. Map of the aseptic assembly facility in the JPL SAF showing the locations of three wipe samples of high-traffic areas of the floor.

Three wipe samples were taken and a fourth wipe was used as a procedural control. A fresh pair of sterile gloves was worn for each sample collection. The wipes were packaged in sterile polypropylene tubes with screw caps and shipped overnight in a cool container from JPL to Wichita State University. Upon arrival, the wipes were wetted with 30 ml of a sterile chaotropic solution (0.1% Na pyrophosphate) to dislodge microbes. After 10 min, the liquid was squeezed from the wipes in the tubes with a sterile syringe plunger. The extracts were used to inoculate enrichment cultures and for direct DNA extractions.

Enrichment cultures

Selective media with high concentrations of salts were used to enrich for salinotolerant microbes. Enrichment cultures were performed in Salt Plains (SP) medium containing (per liter): NaCl, 1 g; KCl, 2.0 g; MgSO₄⋅7H₂O, 1.0 g; CaCl₂⋅H₂O, 0.36 g; NaBr, 0.23 g; FeCl₃⋅6H₂O, 1.0 mg; trace minerals, 0.5 ml; yeast extract, 10.0 g; tryptone, 5.0 g; glucose, 1.0 g; and brought to a final pH of 7.0 (Caton et al., 2004), supplemented with either 50% (w/v) MgSO₄ (2.0 M; a_w = 0.94) or 20% (w/v) NaClO₃ (1.9 M; a_w = 0.91). Flasks (100 ml) were inoculated with aliquots (1 ml) of the fluids extracted from SAF wipes and maintained on an Innova rotary shaker (125 rpm; 1-in stroke dia; New Brunswick Scientific, Edison, NJ) at room temperature for a month. Samples (6 ml) were taken weekly during this incubation for metagenomic analyses. After a month, the enrichment cultures were removed from the shaker, wrapped with parafilm to limit evaporation and stored static for an additional 5 mo, before sampling for metagenomic analyses and bacterial cultivation and isolation.

Bacterial isolation and characterization

After 6 mo of enrichment, viable bacteria remaining in the cultures were isolated through serial dilution plating on SP medium supplemented to 10% NaCl. Colonies for characterization were selected by morphological characteristics and abundance. Five consecutive streak-plates of isolated colonies were used for isolation. The isolates were maintained as agar slants at room temperature and also stored at −70°C as 50% glycerol stocks. Isolates were identified by 16S rRNA gene sequencing and characterized by cell and colony morphology and a variety of biochemical tests. Gram stain was performed using Harleco reagents (Sigma–Aldrich) following the manufacturer's instructions. The endospore stain (Thermo Scientific) was carried out using the manufacturer's instructions. The SIM tests for motility, sulphide production and indole production were performed with agar stabs at 37°C using SIM medium (BBL) following the manufacturer's instructions. The presence of catalase was determined by applying 3% hydrogen peroxide solution to smears of culture on microscope slides. The presence of oxidase was determined using DrySlides (BBL). Starch agar plates (Difco) were used at 37°C and flooded with iodine solution once grown, to observe hydrolysis by amylase. Lipase enzyme was detected using Spirit Blue Agar (Difco) plates at 37°C. Gelatinase was detected using Nutrient Gelatin (Thermo Scientific) deeps that were stab inoculated and incubated at 37°C. Mannitol fermentation was observed on Mannitol salt agar plates (BBL) at 37°C. Lactose fermentation was similarly detected using MacConkey agar plates (BBL) at 37°C. Glucose and sucrose fermentation to acid and gas were observed in an assay medium (per l; 100 g NaCl, 10 g tryptone, and 0.018 g phenol red, pH 7.3) to which 0.5% (w/v) substrate and an inverted Durham tube were added.

Salinotolerance was measured in SP medium supplemented with various concentrations (all w/v) of NaCl (0.1, 10, 20 and 30%), MgSO₄ (30, 40, 50 and 60%), and NaClO₃ (5, 10, 20 and 30%). Shake-tubes (2 ml in 13 × 100-mm tubes) were lightly inoculated (to below 0.05 OD units at 600 nm) and incubated at room temperature for 4 wk. Growth was measured by absorbance spectrophotometry at 600 nm using a Genesys 10S instrument (Thermo Fisher) at 1, 3, 7, 14, 21 and 28 d after inoculation. The threshold for positive growth was 0.2 OD units.

DNA extraction and molecular analyses

Crude DNA extracts were made from aliquots (6 ml) of isolate and enrichment cultures using a freeze–thaw technique (Caton et al., 2004). Cells were collected by serial microcentrifugation for 5 min at 14 000 × g. Pellets were resuspended in 300 μl of sterile water before six cycles of freezing in liquid N₂ and thawing at 80°C, with vigorous vortex mixing every other cycle. Homogenates were clarified by microcentrifugation for 10 min at 14 000 × g and the final supernatant heated for 5 min at 80°C. Extracts were stored at −20°C before PCR amplification. Direct extracts from wipe samples before enrichment yielded insufficient DNA for reliable PCR amplification and community analyses as expected, since samples with extremely low biomass require specialized extraction methods, as previously reported (Highlander et al., 2023).

Gene sequences from bacterial isolates encoding 16S rRNAs were amplified using universal bacterial primers (EUBpA: 5′-AGAGTTTGATCCTGGCTCAF-3′ and EUBpH: 5′AAGGAGGTGATCCAGCCGCA-3′) (Edwards et al., 1989). Each of the 25-μl reactions contained 2.5 μl of each primer (0.2 μM), 1 U of DreamTaq DNA polymerase in master mix (Thermo Scientific) and 5 μl of DNA extract. A thermal cycler (Eppendorf Mastercycler) denatured the DNA at 95°C for 2 min, followed by 40 cycles of 95°C for 1 min, 50°C for 1 min, and 72°C for 1 min, with a final 5-min extension at 72°C. Positive controls using Halomonas sp. str. HL12 (Kilmer et al., 2014) and negative controls with no added DNA were included with each run. PCR amplicons were visualized under ultraviolet light with ethidium bromide stain after electrophoresis on a 1.5% agarose gel to confirm amplicon size and purity. Single-pass Sanger sequencing was performed by a commercial vendor (Eurofins Genomics, Louisville, KY) using the EUBpA primer. Isolate sequences appear in GenBank with accession numbers OP440608 to OP440641. Phylogenetic trees were constructed by maximum-likelihood analyses in MEGA v7.0 (Kumar et al., 2016), with control sequences selected from GenBank using BLAST, from alignments made using the SILVA v.138 database.

Crude metagenomic DNA extracts from enrichment culture samples were used for Illumina sequencing (miSeq v2 Nano PE-250bp) of 16S rRNA genes (v3/v4) by a commercial vendor (University of Kansas Center for Genomics). Each sample in the multiplex reactions produced ~ 30 000 reads and the FASTQ files were demultiplexed for forward and reverse reads. Metagenomic 16S sequence libraries were analysed on the Galaxy platform (Afgan et al., 2018) following the 16S microbial analysis package workflow, incorporating tools designed for mothur (Schloss et al., 2009). Forward and reverse reads were combined to create contigs, generating an average read length of 465 nucleotides. Chimaeras were found using chimera.vsearch and removed using remove.seqs in mothur on the Galaxy platform. These were aligned using SILVA v.138 and clustered into OTUs with a (97%) species threshold with reference to SILVA taxonomy v.138.

Results

Succession of SAF bacterial communities enriched in 50% MgSO₄

Microbes from three wipes of SAF floor surfaces were enriched in SP medium containing 50% (w/v) MgSO₄. Aliquots were withdrawn after 1 wk, 4 wk and 6 mo. Direct metagenomic DNA extracts were made from these samples and used for PCR amplification, high-throughput sequencing of 16S rRNA genes and phylogenetic analyses to describe the bacterial community. A succession of bacterial populations was observed over time within the SAF microbial community during these high-salt enrichments (Figs. 2 and 3). The harsh chemical conditions of the enrichment media were expected to greatly limit the populations of microbes present, even after a 1-wk exposure. Staphylococcus was the dominant taxa observed after 1 wk of MgSO₄ enrichment, comprising >90% of the sequences observed (grey columns of Fig. 2). After 4 wk of enrichment, the proportion of Staphylococcus decreased substantially in wipes 2 and 3. This trend continued, such that Staphylococcus was succeeded by actinobacteria in the climax community observed after 6 mo of enrichment for wipes 2 and 3 and Staphylococcus was greatly reduced in the wipe 1 community.

Figure 2. Relative abundance of bacterial classes observed by high throughput sequencing and phylogenetic analyses of 16S rRNA genes from metagenomic extracts of SAF wipe communities after 1 wk, 4 wk, and 6 mo of enrichment in SP medium supplemented with 50% (w/v; ~2.0 M) MgSO₄. The grey bars show the abundance of Staphylococcus relative to other members of the community. The coloured bars show the relative abundance of the bacterial classes observed without the inclusion of Staphylococcus.

Figure 3. Relative abundance of bacterial genera observed by high-throughput sequencing and phylogenetic analyses of 16S rRNA genes from metagenomic extracts of three SAF wipe communities after 6 mo of enrichment in SP medium supplemented with 50% (w/v; ~2.0 M) MgSO₄ or 20% (w/v; ~1.9 M) NaClO₃.

The colour columns of Fig. 2 detail the classes of bacteria observed in the MgSO₄ enrichments, but separately from the Staphylococcus results for clarity. The non-Staphylococcus populations after 1 or 6 mo of incubation were rich in actinobacteria. However, populations of Alphaproteobacteria, bacilli, clostridia and Gammaproteobacteria were detected in low abundance in the communities observed after 1 wk of enrichment. Only a limited number of genera were found to persist after 6 mo of enrichment in 50% MgSO₄. Staphylococcus remained a major constituent of the SAF climax communities. Brachybacterium and Brevibacterium were common and numerous across all wipes. Glutamicibacter (Arthobacter) was only a substantial portion of the community from wipe 2. The taxonomic and metabolic diversity of the microbial populations detected in the climax microbial communities after 6 mo of enrichment are discussed below.

Diversity indices were calculated for each of the bacterial communities at four levels of sequence identity (Table 1). Common identity thresholds were used to describe diversity at the taxonomic levels of division (88%), genus (94%), species (97%) and strain (99%). More than 2000 different OTUs were recorded at the species level, with ~1000 genera represented. Coverage overall was high, indicating that a substantial proportion of the diversity of the microbial community was observed, despite relatively small sequence libraries. Chao estimators suggest that the enriched microbial community contained ~10⁴ species of bacteria. Non-parametric Shannon diversity and Inverse Simpson indices were generally low (~2) and increased over time, being highest after 6 mo. This likely reflects the lessening predominance of Staphylococcus, as evidenced by increases in the evenness of the communities by the Shannon Equitability Index. Slower growing K-strategists seemed to establish richer communities during the 6 mo of enrichment. However, the bacterial community remained uneven across all wipes, levels of identity and time points.

Table 1. Diversity indices for the enrichment cultures in 50% MgSO₄

Succession of SAF bacterial communities enriched in 20% NaClO₃

In a similar fashion, direct metagenomic DNA extracts were used for PCR amplification, high-throughput sequencing and phylogenetic analyses of 16S rRNA genes to describe the bacterial community developing from three wipes of SAF floor surfaces when enriched in SP medium containing 20% NaClO₃ (w/v). A succession of bacterial populations was observed over time within the microbial community during these harsh enrichments (Figs. 3 and 4). The chemical reactivity and lower a_w of chlorate solutions were expected to limit the populations of microbes present even more than enrichments in MgSO₄ (Al Soudi et al., 2017). Again, Staphylococcus was dominant at early time points, but showed signs of actinobacteria succession by 6 mo, especially apparent for wipe 3 (grey columns of Fig. 4). The final bacterial communities enriched from wipes 1 and 2 were nearly entirely actinobacteria (colour columns of Fig. 4). The wipe 3 community retained large populations of bacilli, along with actinobacteria.

Figure 4. Relative abundance of bacterial classes observed by high-throughput sequencing and phylogenetic analyses of 16S rRNA genes from metagenomic extracts of SAF wipe communities after 1 wk, 4 wk, and 6 mo of enrichment in SP medium supplemented with 20% (w/v; ~1.9 M) NaClO₃. The grey bars show the abundance of Staphylococcus relative to other members of the community. The coloured bars show the relative abundance of the bacterial classes observed without the inclusion of Staphylococcus.

Alphaproteobacteria were more prominent in communities after 1 wk and 1 mo of enrichment than in the climax communities at 6 mo. Staphylococcus and Brachybacterium dominated the bacterial communities of wipes 1 and 2 after 6 mo of enrichment in 20% NaClO₃ (Fig. 3). Wipe 3 appeared to retain a different bacterial community, one dominated by bacilli with Arthrobacter and Brachybacterium. The diversity indices of the NaClO₃ enrichments followed the same trends as the MgSO₄ enrichments (Table 2). The number of OTUs observed and predicted for the NaClO₃ enrichments were similar to those of the MgSO₄ enrichments. Apparent diversity increased during the NaClO₃ enrichment across all levels of sequence identity. The evenness of the bacterial community remained low but increased substantially during the enrichment.

Table 2. Diversity indices for the enrichment cultures in 20% NaClO₃

Major and minor genera detected in SAF community enrichments

Considering the genera detected across all wipes, enrichments and time points, the dominant members of the community were all Gram-positive bacteria (Fig. 3). Staphylococcus was the most abundant genus detected, comprised of 24 species, with Staphylococcus saprophyticus (~93%) far outnumbering other species, followed by S. xylosus (~6%) and S. warneri (~0.3%). Staphylococci such as Staphylococcus saprophyticus are halotolerant (grow in 15% NaCl), saprophytic, non-spore-forming, fermentative, facultative anaerobes that are common in foods, marine and soil environments and in the human microbiome (Scott, 1953; Kloos et al., 1976; La Duc et al., 2003; Wilson, 2005; Probst et al., 2010; Garza-González et al., 2011; Medved'ová et al., 2019).

As actinobacteria succeeded staphylococci in the enrichment cultures over 6 mo, Arthrobacter, Brachybacterium and Brevibacterium were the main genera observed. Among the 13 species of Brevibacterium detected, nearly all (~90%) were Brevibacterium casei, with lower abundances of B. oceani (~7%) and B. permense (~1%). Brevibacterium casei are non-fermentative strict anaerobes that live as saprophytes on skin and in spoiled foods (Trujillo and Goodfellow, 2012). Brachybacterium paraconglomeratum (~26%), B. sacelli (~13%) and B. saurashtrense (~16%) were the most common of 11 Brachybacterium species detected. Brachybacterium are non-spore-forming halotolerant bacteria (growing up to 15% NaCl) found in aerobic or microaerobic habitats in seawater, sediments, cheeses and poultry litter (Collins et al., 1988; Park et al., 2011; Buczolits and Busse, 2012). Another actinobacterium, Arthrobacter, observed in several enrichment samples, comprised 14 species, with the most abundant being A. uratoxydans (~24%). Glutamicibacter spp. are currently described as Arthrobacter, with the A. uratoxydans species involved in soil N cycles through ammonification and nitrate respiration (van Waasbergen et al., 2000; Eschbach et al., 2003). Of the 14 species of Oceanobacillus observed in certain enrichment cultures, nearly all (>99%) were Oceanobacillus picturae, a halotolerant (grows in 10% NaCl), fermentative, facultative anaerobe from the human gut (Lu et al., 2001; Lagier et al., 2015; Mondal et al., 2017). More than 25 species of Bacillus were detected but these were not major constituents of these enriched communities.

Minor but relatively abundant members of the communities included the actinobacteria Plesiocystis, Pseudonocardia and Zhihengliuella, the latter being a halotolerant micrococci found in marine, sediment and soil habitats (Zhang et al., 2007). Plesiocystis is a marine myxobacterium that is reported to be halophilic, requiring >1% NaCl to grow (Iizuka et al., 2003). Genera detected in lower abundance included Acinetobacter, Cellulomonas, Clostridium, Compostibacillus, Corynebacterium (seven species), Curtobacterium, Curvibacter, Dietzia, Geodermatophilus, Geothrix, Kocuria, Limisphaera, Micropruina, Nonomuraea, Ornithinibacillus, Paenisporosarcina, Planococcus, Quadrisphaera, Saccharomonospora, Skermanella, Streptococcus (six species), Thermodesulfobium and Virgibacillus (six species). Our sequence libraries were relatively small, so there were likely species in low abundance that were not detected. However, take note that several of the minor genera detected are known for metabolic activities central to biogeochemical cycles (v.i.). Even minor members of a microbial community may have important roles such as Azorhizobium (N fixation), Geothrix (Fe respiration), Methylobrevis (1-C metabolism) and Synechococcus (photosynthesis). Many of the species detected are typically associated with human microflora or soil communities, including several pathogens detected at low abundance (v.i.). Representatives also were detected from clearly halotolerant groups such as Halobacillus, Haloechinothrix, Oceanicola, Salinicoccus, Salimicrobium and Virgibacillus (v.i.).

Characterization of SAF bacterial isolates

After the enrichment cultures had incubated for 6 mo, bacterial isolates were obtained from spread-plates and collected based on colony morphology and colour. Phenetic characterization was performed on 38 isolates; 23 and 15 isolates were from the 20% NaClO₃ and 50% MgSO₄ enrichments, respectively. All but three isolates were identified by 16S rRNA gene sequencing and analyzed phylogenetically (Fig. 5). Isolates SAF 1 to 23 derive from NaClO₃ enrichment cultures, while the rest of the isolates derive from MgSO₄ enrichment cultures. Biochemical and physiological characteristics of the isolates are given in Table 3.

Figure 5. Phylogenetic tree based on 16S rRNA gene sequences for SAF bacterial isolates obtained by repetitive streak-plating from the climax microbial community after 6 mo of enrichment in SP medium supplemented with 50% (w/v; ~2.0 M) MgSO₄ or 20% (w/v; ~1.9 M) NaClO₃. Source of isolate: *, Wipe 1; †, Wipe 2; ‡, Wipe 3.

Table 3. Characterization of SAF bacterial isolates

^a Cell shapes: b, bacillus; c, coccus.

^b Colony colours: c, cream; p, pink; w, white.

^c No gas observed, except for SAF27 on lactose.

^d All isolates grew at low salt concentrations (0.1%).

The cultivable community was not diverse, comprising strains from only four genera, namely, Brachybacterium, Brevibacterium, Oceanobacillus and Staphylococcus (Fig. 5). Fortunately, these represent the dominant taxa detected in the enriched communities by 16S rRNA sequencing of direct metagenomic extracts, with the exception that representatives of Arthrobacter (Glutamicibacter) were not isolated. Staphylococcus was the most abundant genus recovered in both enrichment brines, with all but one isolate (SAF18) clustering with S. sciuri. Oceanobacillus picturae was recovered only from the NaClO₃ enrichment cultures. One isolate of Brevibacterium sediminis was obtained from MgSO₄ enrichment cultures. The Brachybacterium isolates recovered were in the B. conglomeratum cluster and all derived from the MgSO₄ enrichment cultures.

The isolates were a mixture of bacilli and cocci and their colonies were not highly coloured, appearing mainly cream and white. All the isolates were within Gram-positive genera, although some isolates did not retain the Gram stain well. Oceanobacillus were the only motile isolates and the only genus among the isolates known to form endospores. All the isolates were catalase-positive and all but three (Brachybacterium sp. str. SAF 30, 37 and 43) were oxidase-positive. More than half of the isolates exhibited lipase and/or gelatinase activity, but amylase activity was absent. Nearly all the isolates fermented glucose and mannitol to acid.

High salinity tolerance was observed for the isolates across three different salts. All the isolates exhibited growth tolerance to 20% NaCl (3.4 M; a_w = 0.85) (Table 3). All but seven isolates (82%) grew at 30% NaCl (5.1 M; a_w = 0.76), near saturation, showing extraordinary halotolerance. None of the isolates appeared halophilic (requiring high NaCl for growth). Epsotolerance also was high for the SAF isolates, with all growing at ≥40% MgSO₄ (1.6 M; a_w = 0.95) and nearly all (79%) showing growth at 50% MgSO₄ (Table 3). Seven isolates grew at 60% MgSO₄ (2.4 M; a_w = 0.91), near saturation (~67%). All but four of the isolates grew well at 20% NaClO₃ (Table 3). Three of the isolates (Brachybacterium sp. str. SAF37 and Staphylococcus sp. str. SAF 2 and 3) exhibited growth at 30% NaClO₃ (2.8 M; a_w = 0.89).

Discussion

Planetary protection protocols seek to prevent the contamination of extraterrestrial bodies by terrestrial life. For this reason, spacecraft are constructed in clean rooms certified for flight assemblies and are subject to rigorous cleaning protocols to reduce bioburden before launch. Controlled environments in SAFs exert selective pressures on indigenous microbes, where stress-tolerant species may adapt to conditions of low nutrient availability, relatively low humidity and low overall biomass. For instance, SAFs with low humidity (40 ± 5%) may select for salinotolerant microbes given that dry environments tend to retain salt evaporites. The SAF isolates appear to be relatively tolerant of high salinity, radiation and oxidants (Venkateswaran et al., 2001, 2003a, 2003b, 2014; La Duc et al., 2003; Link et al., 2003; Kempf et al., 2005; Smith et al., 2017; Zanmuto et al., 2018). The microbes isolated from SAFs closely match those isolated from spacecraft (Favero et al., 1966; Favero, 1971; Puleo et al., 1977). This suggests that SAF environments may enrich for organisms with great potential for successful survival or even colonization following forward contamination by spacecraft.

The microbial populations of SAFs have been described previously by classical cultivation and by molecular analyses to reveal complex communities (Favero et al., 1966; Favero, 1971; Foster and Winans, 1975; Puleo et al., 1977; Moissl et al., 2008; La Duc et al., 2009; Stieglmeier et al., 2009; Ghosh et al., 2010; Probst et al., 2010; Bashir et al., 2016; Hendrickson et al., 2017, 2021; Probst and Vaishampayan, 2020; Danko et al., 2021; Smith et al., 2022; Highlander et al., 2023; Lu et al., 2023). The microbial communities appear to be dominated by bacteria, with fewer fungi and archaea. Previously isolated representatives of Arthrobacter, Bacillus, Exiguobacterium, Filibacter, Oceanobacillus, Sporosarcina, Staphylococcus and Streptococcus, are bacterial genera typically associated with soils and human microbiomes (La Duc et al., 2003; Link et al., 2003; Venkateswaran et al., 2003a, 2003b; Kempf et al., 2005; Satomi et al., 2006; La Duc et al., 2007; La Duc et al., 2012). In molecular libraries, representatives of anaerobes and facultative strains of Firmicutes, actinobacteria and Gammaproteobacteria were observed. The most common Firmicutes, included the genera Bacillus, Clostridium, Enterococcus, Lactobacillus, Paenibacillus, and Staphylococcus (Moissl et al., 2008; Stieglmeier et al., 2009; La Duc et al., 2014; Smith et al., 2022; Lu et al., 2023). Moraxellaceae dominated in a cleanroom study of floor wipes that focused on human pathogens (Bashir et al., 2016) and were dominant, along with actinobacteria, in a recent metagenomic study (Highlander et al., 2023). A wide variety of media and conditions, both common and extreme, aerobic and anaerobic, were used to isolate microbes from Herschel SAF wipe samples, producing a culture collection with greater diversity than other studies (Moissl-Eichinger et al., 2013). Similarly, isolates from the SAF in French Guiana came from a diverse set of media enrichments, comprising nearly 50 genera (Schwendner et al., 2013). Several of the genera detected in fresh SAF samples were found in the microbial assemblage remaining after our extended SAF enrichments in dense brines, including actinobacteria, Arthrobacter, Bacillus, Oceanobacillus and Staphylococcus. Selective pressures appear to play a role in forming the SAF community and its enrichment of salinotolerant microbes. The microbial communities found in surgical suites and electronics clean rooms may be like those moulded by the conditions of SAFs, given the similarities of these environments.

While the enriched SAF communities of the current study included microbes typically associated with the human microbiome, the assemblage did not resemble the community found on human skin (Byrd et al., 2018). This supports the conclusion that our SAF samples were not simply contaminants introduced during the handling and cultivation of the samples. The absence of microbial growth in the process blank, a wipe handled as the others but never having swabbed a surface, further demonstrates that handling and cultivation did not introduce skin contaminants. Although Staphylococcus, common skin microbes, were observed in high abundance in enrichment cultures, Streptococcus and Propionibacterium, also common skin microbes, were in low abundance, even early in the enrichment cultures. Furthermore, the Staphylococcus detected did not include Staphylococcus aureus or Staphylococcus epidermidis, the most abundant Staphylococcus species on human skin. The predominant isolate from the 6-mo enrichment cultures, Staphylococcus sciuri, is a minor pathogen causing urinary tract infections (Kloos et al., 1976; Garza-González et al., 2011). It forms a separate taxonomic cluster (one of six Staphylococcus clusters) that is distinguished by being oxidase positive (Shaw et al., 1951). Staphylococcus saprophyticus, prominent in the sequence libraries, is a minor pathogen and found in foods and marine environments, often forming biofilms (Schleifer and Bell, 2009). Oceanobacillus, while widespread in marine and saline environments, is commonly found in the human gut and the skin-associated Kocuria that were detected have been found at the Atacama Desert (Lagier et al., 2015; Azua-Bustos et al., 2020). It is interesting to note that Mycobacterium leprae, the causative agent of leprosy, also appeared in the sequence libraries. Human pathogens detected at low abundance in previous metagenomic studies of SAF floor wipe samples included Acinetobacter, Bacillus, Enterobacter, Enterococcus, Escherichia, Legionella, Pseudomonas and Staphylococcus, with their associated virulence factors (Bashir et al., 2016).

Several of the bacterial genera detected in the sequence libraries from SAF enrichment cultures are well known to be halotolerant such as Anoxybacter, Brachybacterium, Gracibacillus, Halobacillus, Haloechinothrix, Kitasatospora, Lentibacillus, Oceanobacillus, Oceanicola, Ornithinibacillus, Plesiocystis, Salinicoccus, Salimicrobium, Skermanella, Staphylococcus, Tetragenococcus, Virgibacillus (V. halophilus and V. salexigens) and Zhihengliuella. Members of these genera tend to exhibit growth tolerance to 5–15% NaCl, however, certain strains may show even greater salinotolerance. For instance, Staphylococcus generally tolerate up to 15% NaCl, but isolates from soils of the Great Salt Plains of Oklahoma were shown to grow at >20% NaCl in culture (Maitland and Martyn, 1948; Caton et al., 2004; Litzner et al., 2006). Halobacillus isolates from these salt plains grew at 25% NaCl and an Oceanobacillus isolate appeared halophilic, growing from 10% to 25% NaCl. It is surprising that no salinotolerant Gram-negative bacterial isolates were recovered from SAF samples such as Halomonads, fast-growing polyextremophiles that are common in natural hypersaline environments (Mata et al., 2002; Caton et al., 2004; Caton and Schneegurt, 2012; Kilmer et al., 2014). Notably, the halotolerant bacteria isolated from a variety of common oligosaline soils were entirely Gram-positive bacilli, including Bacillus, Halobacillus, Oceanobacillus, Staphylococcus and Virgibacillus, mainly exhibiting growth tolerances of 25–30% NaCl (Howell et al., 2022). This supports the reasonable expectation that many of the microbes found in SAFs may derive from local soils. Archaea present in SAF enrichment cultures would not be detected by the bacterial 16S rRNA gene primers used here and the SP medium and growth conditions used do not target archaea (Edwards et al., 1989; Caton and Schneegurt, 2012).

The SAF enrichment cultures and the SAFs themselves are moderate indoor environments. The vast majority of the genera observed were not expected to proliferate well under environmental extremes. However, microbes known for their tolerances to temperature, pH and radiation were detected. Thermophilic bacteria were observed such as Anoxybacter and Thermodsulfobium (Mori et al., 2003; Zeng et al., 2015). Psychrophilic bacteria with growth tolerance to low temperatures (<10°C) included Paenisporosarcina (related to Planococcus) known from glaciers and soils and Deinococcus frigens, an isolate from Antarctica (Hirsch et al., 2004; Reddy et al., 2013). It is interesting to note that D. frigens is highly resistant to UV radiation, as Deinococcus are remarkably radiation tolerant due to superior DNA repair mechanisms (Hirsch et al., 2004). Bacillus alkalinitrilicus is a halotolerant alkaliphile that tolerates high pH conditions (Sorokin et al., 2008). Overall, fermentative organisms are tolerant to the low pH conditions created by acidic fermentation products such as lactic acid. Thus, microbes in SAFs have the potential to proliferate in habitats over a wide range of extreme conditions as suggested previously (Venkateswaran et al., 2014; Bashir et al., 2016; Smith et al., 2017; Zanmuto et al., 2018).

The vast majority of bacterial genera detected in sequence libraries or recovered as isolates from SAF enrichment cultures in the current study are unremarkable metabolically, being heterotrophic aerobes or facultative anaerobes. A notable aspect of the climax microbial community after 6 mo of enrichment was the variety of metabolic activities associated with certain populations that were detected. Although these genera may be in low abundance in this microbial community and in the natural communities of soils and waters, their biogeochemical roles are critical to ecosystem functioning. Quite a few of the genera detected are known to ferment small molecules under anaerobic conditions, often central to C recycling and the acetogen guild. Two classes of photosynthetic organisms were detected in the SAF enrichment cultures that can fix C, converting atmospheric CO₂ into bioavailable sugars. Synechococcus is a unicellular cyanobacterium found in marine environments (Castenholz, 2012). Thiolamprovum is a green sulphur bacterium in the Chromatiaceae that can perform anoxygenic photosynthesis in anoxic environments (Imhoff, 2014). Methylobrevis is a methylotrophic Rhizobiales that does not fix N but exhibits 1-C metabolism (Poroshina et al., 2015). All methanogens are strictly anaerobic archaea and hence would not have been detected here. However, the key processes of the C cycle appear to be functional in the climax SAF microbial community.

Similarly, microbes known to perform all the major processes of the N cycle appear to be present in the climax SAF microbial community. Certain organisms exhibit nitrate reduction such as Arthrobacter, Micropruina, Oceanicola, and Oceanobacillus (Hirsch et al., 1961; Shintani et al., 2000; van Waasbergen et al., 2000; Eschbach et al., 2003; Zheng et al., 2010; Lagier et al., 2015; Zeng et al., 2015). For instance, Arthrobacter uratoxydans (Glutamicibacter uratoxydans) from humic soils and the deep subsurface can perform denitrification through nitrate respiration and additionally produces uricase for nitrate ammonification (van Waasbergen et al., 2000; Eschbach et al., 2003). Denitrification also has been attributed to Streptomyces spp. (Hirsch et al., 1961; Shoun et al., 1998). Certain members of Mycobacterium, Nocardia and Streptomyces can perform nitrification by oxidizing ammonia to nitrite or nitrate (Hirsch et al., 1961). Azorhizobium can fix N, converting dinitrogen gas into bioavailable forms, while in plant root nodules or when free-living in soils (Dreyfus et al., 1988; Ryu et al., 2020). Arthrobacter, Corynebacterium, Herbospirillum and Mycobacterium species also are known to fix N (Gtari et al., 2012). Thus, a complete N cycle may be operating in the climax microbial community of the SAF enrichments of the current study, including N fixation, nitrification, denitrification and ammonification.

While microbes that perform all the processes of the S cycle were not detected, organisms known to perform the key process of dissimilatory sulphate reduction included Desulfotomaculum, Desulfovibrio, Desufuribacillus and Thermodesulfobium (Mori et al., 2003; Kuever et al., 2012). Certain actinobacteria species also are known to perform anaerobic sulphate respiration (Zeng et al., 2015). Finally, Anoxybacter, Geothrix and Pelobacter are chemoorganotrophs that can grow anaerobically by fermentation or by using Fe respiration, where oxidized Fe acts as the terminal electron acceptor (Coates et al., 1999; Tang et al., 2010; Zeng et al., 2015).

The succession of actinobacteria over staphylococci was the most notable shift in community structure observed during hypersaline enrichment culturing of SAF wipe samples over time. Early in the enrichment, fast-growing bacteria, with limited generalist metabolisms, dominated the microbial community. As the batch cultures matured, actinobacteria known for their metabolic versatility became dominant. This is an example of a classic secondary succession process commonly observed in natural microbial communities (Atlas and Bartha, 1987). Initially the SAF enrichment cultures were replete with the nutrients supplied by the eutrophic SP medium. Certain microbes will exploit those readily metabolized nutrients (sugars, amino acids) and rapidly proliferate. These r-strategists have the evolutionary advantage of rapid growth, but they typically have limited metabolic capabilities and are not particularly adaptive to environmental changes. The bloom of r-strategists will bust when the readily available nutrients are depleted, and their abundance then falls. The K-strategists in the community live near their carrying capacity and are more adaptive, tolerant and versatile members of the community that often form permanent biofilms. These organisms are evolutionarily successful because their versatile metabolic capabilities allow them to utilize recalcitrant nutrients of lower quality, the complex components of dead cells such as cellulose and chitin. The SAF enrichment cultures also included species that proliferate in specific niches, using electron sources and sinks associated with lithotrophy and anaerobic respiration. These tend to outlast r-strategists in stable environments, as do K-strategists in general.

The microbes present in SAFs are the most likely to be carried by spacecraft to a planetary body. The extremely hypersaline media used for our SAF enrichment cultures are analogs for the harsh chemical conditions proposed for liquid water near the surface of Mars or in discrete locations in the icy worlds (Chevrier et al., 2009; Chin et al., 2010; Chevrier and Rivera-Valentin, 2012; Hanley et al., 2012; Toner et al., 2014). While sulphates are important on Mars, (per)chlorate brines may be the most likely sources of liquid water near the surface (Vaniman et al., 2004; Hecht et al., 2009; Kounaves et al., 2010; Kminek et al., 2017). The extraordinary chlorate tolerances observed for certain SAF isolates and enrichment cultures supports our previous conclusions that (per)chlorate tolerance is more widespread among microbes than might be expected given their chemical reactivity (Wilks et al., 2019). The microbial communities which develop over months in these brines from SAF surface wipes appear to comprise the functionalities central for maintaining biogeochemical cycles. In addition, many of the end members of the enrichment cultures were extremophiles and anaerobic respirers, even using Fe, which is a common oxidant on Mars.

Previous studies of microbes from SAFs mainly focused on the capabilities of individual isolates as pure cultures. However, studying microbes in the context of complex communities can enhance our understanding of natural ecosystems. When considering the forward contamination of celestial bodies, we suggest that it may be better to study the survivability and proliferation of microbial communities rather than individual microbial isolates. The climax microbial communities after enrichment of SAF samples may support fully functioning biogeochemical cycles. Subsequent experiments might begin with climax communities like these, developed over months or years, which then can be followed over long incubation periods to better mimic what may occur should the SAF microbial community be carried to another world. Sufficient biomass should be present to form biofilms, whereby microbes can retain water, recycle nutrients and find protection from harsh chemical conditions and radiation. Entire microbial communities, with every key functional niche filled, are far more likely to survive and proliferate on another world than any single remarkably resilient organism. Thus, hypertolerant microbial communities are a greater threat than microbial isolates to successfully colonize another world following forward contamination by a visiting spacecraft.