Four-year bacterial monitoring in the International Space Station - Japanese Experiment Module «Kibo» with culture-independent approach / Четырехлетний бактериальный мониторинг в японском экспериментальном модуле «Кибо» на Международной космической станции на основе культурно-независимого подхода

Ichijo Tomoaki PhD, Assistant Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan Ph.D., Assistant Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan nasu@phs.osaka-u.ac.jp Yamaguchi Nobuyasu PhD, Associated Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan Ph.D., Associated Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan yamaguti@phs.osaka-u.ac.jp Tanigaki Fumiaki Human Spaceflight Technology Directorate, Japan Aerospace Exploration Agency, Japan 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan Human Spaceflight Technology Directorate, Japan Aerospace Exploration Agency, Japan 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan nasu@phs.osaka-u.ac.jp Shirakawa Masaki PhD, Human Spaceflight Technology Directorate, Japan Aerospace Exploration Agency, Japan 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan Ph.D., Human Spaceflight Technology Directorate, Japan Aerospace Exploration Agency, Japan 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan nasu@phs.osaka-u.ac.jp

Nasu Masao PhD, Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan Ph.D., Professor, Environmental Science and Microbiology, Graduate School of Pharmaceutical Sciences, Osaka University, Japan 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan yamaguti@phs.osaka-u.ac.jp

Статья впервые опубликована: npj Microgravity 2, Article number: 16007 (2016); doi:10.1038/npjmgrav.2016.7

Research on microbial dynamics in crewed habitats in space is indispensable to achieve safe and healthy long-duration space habitation. According to the previous research, microgravity and spaceflight affect bacterial growth, virulence, biofilm formation, and so on (reviewed in Yamaguchi et al. ^[1]). For instance, Wilson et al. ^[2] reported that spaceflight-grown Salmonella enterica serovar Typhimurium increased their virulence when compared with control bacteria on Earth. Kim et al. ^[3] reported that Pseudomonas aeruginosa showed different structure and biomass of biofilms between normal and spaceflight conditions. Furthermore, immune dysfunction occurs associated with spaceflight ^[4]. Therefore, the relationship between humans and microbes may change in space habitation. For successful manned space missions, it is necessary to investigate the relationship between humans and microbes, and how microbes influence the systems and materials in such space environments. That is, we have to understand how and where microbes proliferate in a confined environment in space.

Since 2009, we have been continuously performing bacterial monitoring in Kibo, the Japanese Experiment Module of the International Space Station (ISS), in cooperation with the Japan Aerospace Exploration Agency (research title: “Microbe”:http://iss.jaxa.jp/en/kiboexp/news/101101_microbe-2_start.html). The objective of this research is to monitor microbes and analyze their dynamics in Kibo from environmental microbiological viewpoints. Previously, many studies on microbial contamination in space habitats have been based on culture-based approaches, despite the fact that the majority of microbes in natural environments are hard to culture under conventional culture methods ^[5-6]. In environmental microbiology, some culture-independent techniques are available, and these techniques have been introduced in recent research on microbial contamination in space habitats. Ichijo et al. ^[7] used total direct counting, quantitative PCR and PCR denaturant gradient gel electrophoresis for determination of the abundance and phylogenetic affiliation of bacteria on the interior surfaces in Kibo. Venkateswaran et al. ^[8] used quantitative PCR and pyrosequencing followed by propidium monoazide treatment to understand the distribution and diversity of viable microbes in debris collected from the ISS. These reports provided the useful information that the ISS environmental microbiota comprised human-related microbes.

For long space missions, understanding changes in microbial dynamics in a confined environment is essential, as described above. In this research, we collected samples from interior surfaces in Kibo every ca. 500 days after Kibo began operation. Bacterial abundance and phylogenetic affiliation were determined by fluorescent staining, 16S ribosomal RNA (rRNA) gene-targeted quantitative PCR and pyrosequencing. This is the first report on continuous monitoring of bacterial abundance and phylogenetic affiliation in a space habitat determined by culture-independent molecular microbiological methods.

Bacterial abundance

Bacterial abundance on the interior surfaces in Kibo is shown in Table 1. We determined bacterial abundance with different approaches, fluorescent microscopy and quantitative PCR targeting the bacterial 16S rRNA gene, to confirm the reliability of the results. Quantitative PCR outputs the copy number of the 16S rRNA gene in a sample. To convert the copy number to a cell number, we need to know the copy number of the target gene in a single cell. However, bacterial cells carry 1–15 copies of the 16S rRNA gene in their genomes ^[9]. We therefore calculated the average copy number of 16S rRNA genes in the bacterial community in Kibo as 5 copies/cell, based on the community structure determined by pyrosequencing as described below and the ribosomal RNA operon copy number database rrnDB ^[9]. As shown in Table 1, ca. 10³ cells/cm² of bacteria were detected on the interior surfaces in Kibo in Microbe-I (Table 2). In Microbe-II and Microbe-III (Table 2), bacterial abundance was ca. 10² cells/cm², or less than the quantification limit. Overall, bacterial abundance did not exceed 10⁴ cells/cm².

Table 1: Bacterial abundance on interior surfaces in Kibo determined by fluorescent microscopy and quantitative PCR

Abbreviations: NT, not tested; qPCR, quantitative PCR; TDC, total direct counting with fluorescent microscopy.

Table 2: Sampling and shipping to our laboratory

* Kibo operation was started on 4 June 2008

Bacterial community structure

Bacterial community structure on the interior surfaces in Kibo was analyzed by 16S rRNA gene-targeted pyrosequencing. Among 99,967 raw reads, 71,190 were high-quality reads after processing with the Quantitative Insights into Microbial Ecology (QIIME) pipeline. These reads were distributed in each sample (13 samples) with an average of 5,472 reads. At first, we estimated sufficient read number for bacterial community analysis using chao1 index (Supplementary Table S1) and a rarefaction curve (Supplementary Figure S1). As shown in Supplementary Table S1, we found that observed operational taxonomic units (OTUs) revealed 73–100% of bacterial family-level OTUs (equivalent to OTU at 90% similarity) that were present in our sample. We therefore confirmed the number of high-quality reads obtained was sufficient to reveal the bacterial community structure in the Kibo sample at the family level.

Figure 1a shows the bacterial community structure at the phylum level on the interior surfaces in Kibo. Bacteria of the phyla Proteobacteria (beta- and gamma-subclasses), Firmicutes, and Actinobacteria were frequently detected. The families Enterobacteriaceae and Staphylococcaceae were dominant in each sample (Figure 1b).

Figure 1

Bacterial community structure on the interior surfaces in Kibo. (a) At the phylum level. (b) Expanding beta- and gamma-Proteobacteria and Firmicutes to the family level.

Studies on the relationship between humans and microbes in space habitation environments are critical for success in long-duration space missions. To reduce potential hazards to the crew and the spacecraft infrastructure, indoor quality control is important, as defined by the roadmap or scenario of each agency. In this study, we performed microbial monitoring in the ISS—“Kibo” for 4 years after its completion, and analyzed samples with culture-independent techniques.

The majority of bacteria in environments are hard to culture under conventional culture condition. Therefore, bacterial number is usually underestimated with culture methods. To obtain bacterial numbers more accurately, bacterial abundance was determined with culture-independent techniques. We used both total direct counting and quantitative PCR to obtain more reliable results. Total direct counting potentially gives us the absolute cell numbers. In case the number of contaminants (e.g., non-biological fluorescent particles or detritus) is relatively large in the sample, the accuracy of total direct counting tends to decrease. In this study, we set up a criterion when we enumerate bacterial cells by total direct counting; tiny particles whose size are smaller than bacteria generally distributed in environments are omitted. Under this criterion, cell number might be underestimated. The results show that the bacterial number was below 10⁴ cells/cm². Therefore, we conclude that the bacterial abundance in Kibo was well controlled during the 1,596-day operation covered by our study. However, Kibo is a module for performing experiments and not for living activities. It is therefore important to compare results obtained in Kibo with those in modules for living activities to evaluate the whole microbial world in the ISS.

Bacterial community structure was determined by amplicon sequencing with a high-throughput sequencer. In this study, we used a two-step PCR protocol to generate amplicons. Amplicon sequencing with two-step PCR has several advantages ^[10]. For example, this method increases reproducibility and recovers higher genetic diversity. Selection of the amplified region can influence the profile of the bacterial community generated by pyrosequencing ^[11]. Various regions such as V1–V2 ^[12-13], V3–V5 ^[14-15], V4–V5 ^[16-17], and V6–V8 ^[18-19] have been analyzed for bacterial community structure by pyrosequencing. In this study, before the experiment, we evaluated several primer sets (targeting V1–V3, V3–V5, V7–V8, and V6–V8) by analyzing bacterial diversity on the incubator surface of Microbe-I by denaturing gradient gel electrophoresis. We selected a primer set of 968f and 1401r, which can amplify the V6–V8 region of the 16S rRNA gene based on greater diversity (data not shown).

The results of bacterial community analysis show that Actinobacteria, Firmicutes, and Proteobacteria were frequently detected on the surfaces of equipment in Kibo. In particular, Staphylococcaceae belonging to the phylum Firmicutes, Enterobacteriaceae belonging to the Proteobacteria (gamma-subclass), and Neisseriaceae belonging to the Proteobacteria (beta-subclass) were dominant taxonomic groups on the equipment surfaces (Figure 1b). Actinobacteria are commonly found in both terrestrial and aquatic ecosystems ^[20]. This phylum is known as the dominant species of human gut microbiota ^[21-22]. Enterobacteriaceae, Staphylococcaceae, and Neisseriaceae are also human related. Enterobacteriaceae, a large group of the gamma-Proteobacteria, are commonly found in the human gut microbiota ^[23]. Fierer et al. ^[12] reported that Actinobacteria, Firmicutes, and Proteobacteria contain the majority of the sequences retrieved from human palm surfaces. According to this report ^[12], Staphylococcaceae, one of the most abundant taxonomic groups collected from Kibo, are relatively highly abundant on human palm surfaces; Neisseriaceae made up ca. 3% of the sequences from human palm surfaces; Enterobacteriales were also commonly found there. As described above, most of the bacteria detected on the Kibo equipment surfaces belong to human microbiota; thus, we suggest that bacterial cells are transferred to the surfaces in Kibo from the astronauts. Kibo was disinfected with isopropyl alcohol before launch. We measured bacterial numbers on the surfaces in Kibo before launch by fluorescent microscopy, but the bacterial number was below the quantification limit (<2×10² cells/cm²). Our results show that immediately after beginning the operation of Kibo, few bacteria were present there. The bacterial community in Kibo formed from human-related bacteria through astronaut activities in the early stage of operation.

Microbial community structure on the equipments in Kibo varied during the first 4-year operation (Figure 1b). For example, on the surfaces of diffuser, handrail, and inside of incubator, Neiseriaceae occupied 3.4–48% of total bacterial communities on Microbe-I and Microbe-II; however, on Microbe-III, Neiseriaceae was rarely detected there (<1%). Relative abundance of Enterobacteriaceae seemed to increase in Microbe-III. Although human-related bacteria occupied majority of bacterial community in Kibo in the early stage of operation, their bacterial community was not stable during this period (Figure 1b). Because bacterial community in the space habitat was not established, it is considered that the community is easy to change affected by external factors such as astronauts’ activities and experiments in Kibo.

Legionella were dominant in the sample collected from the outer surface of the incubator on Microbe-I (Figure 1b). Legionella is normally found in aquatic and terrestrial environments ^[23], but not among human microbiota. Therefore, it seems unlikely that Legionella cells were transferred to the surface via the astronauts. In the monitoring of space habitat, Legionella species was only found in the Russian Space Station Mir ^[24]. As environmental bacteria have be