This study has demonstrated that re-examination of a fairly simple, well studied environmental habitat using a combination of strategic environmental sampling, deep sequencing, and de novo metagenomic assembly can reveal significant new information. We have discovered and characterized nearly complete genomes representing a novel archaeal lineage prevalent in hypersaline systems worldwide, yet very different from all previously described members of class Halobacteria.
We propose the creation of a new class ‘Nanohaloarchaea' within phylum Euryarchaeota to accommodate this new lineage. We further propose partitioning class Nanohaloarchaea to place J07AB43 and J07AB56 into distinct genera, Candidatus ‘Nanosalina sp. J07AB43' and ‘Candidatus Nanosalinarum sp. J07AB56'. Evidence supporting these proposals includes: (i) comprehensive euryarchaeotal phylogenetic analyses based on 16S rRNA genes and ribosomal proteins; (ii) lineage-specific features, including numerous genes without previously described close relatives; and (iii) significant intra-lineage diversity and abundance within geographically distinct hypersaline habitats worldwide. Evolutionary distinctness of J07AB43 and J07AB56 from other halophilic archaea is reinforced by taxonomic patterns of BLASTP matches for their predicted proteomes against GenBank nr, as well as amino acid composition-based clustering. The sister-grouping of class Halobacteria and class Nanohaloarchaea reflects probable derivation from an ancient common halophilic ancestor with a ‘high salt-in' osmotic regulation strategy, followed by subsequent divergence along separate evolutionary paths.
Lineage-specific characteristics that distinguish ‘Candidatus
Nanosalina sp.' and ‘Candidatus
Nanosalinarum sp.' from most other extreme halophiles include their small physical size, compact genomes, single-copy rRNA operon, low G+C composition, unique proteome amino acid composition, absence of conserved gas vesicle genes and atypical predicted pathways associated with carbohydrate metabolism. Small compact genomes, as well as single-copy rRNA operons, have been proposed to minimize metabolic costs in habitats where neither broad metabolic repertoire nor high numbers of paralogous proteins are needed to accommodate rapid growth under fluctuating environmental conditions (Klappenbach et al., 2000
). Small cell size, which increases surface to volume ratio, could be an adaptation for optimizing nutrient uptake capacity. Alternatively it is possible that small physical size allows Nanohaloarchaea to remain suspended in oxygenated surface waters to support aerobic metabolism, thus eliminating the need for gas vesicles to provide buoyancy.
The low G+C compositions of the two Nanohaloarchaea genomes, especially J07AB43 (43%), are surprising considering their prevalence in high light habitats. In the absence of compensatory mechanisms, lower G+C would be expected to increase susceptibility to ultraviolet-induced DNA damage. One possible explanation is that the low G+C composition of J07AB43 is related to ecological lifestyle. Low G+C composition and genomic streamlining have been associated with decreased nitrogen requirements and a slow-growing, energy-conservative lifestyle in marine bacteria (Giovannoni et al., 2005
). However, the habitats from which these Nanohaloarchaea were isolated are not generally considered to be nutrient-limited (Oren, 2002b
). Alternatively, it has been proposed that the low G+C composition of H. walsbyi
(48%) compared with other halophiles is a specific adaptation to counteract the over-stabilizing effect of high magnesium concentrations on DNA structure (Bolhuis et al., 2006
). If extremely high environmental magnesium cannot be adequately excluded from the cell, lower genomic G+C helps maintain DNA structural flexibility and avoids difficulties in strand separation caused by elevated melting temperatures. These same principles could apply to J07AB43, providing a possible selective advantage under high magnesium conditions expected in evaporative high salt environments.
Nanohaloarchaea are estimated to represent at least 10–25% of the total archaeal community in surface water samples from LT, Australia and CV, California, USA. We believe these values are robust, based on agreement of three independent analysis techniques: amplification of environmental 16S rRNA gene sequences; statistical analysis of metagenomic sequencing reads assembled into near-complete draft genomes; and quantitative FISH of cells from natural water samples labeled with lineage-specific probes. Microscopic counts reveal that Nanohaloarchaea are present at cell concentrations exceeding 106
in hypersaline habitats of Australia and North America. The sporadic identification of Nanohaloarchaea in other surveys of hypersaline communities worldwide suggests that Nanohaloarchaea represent a significant yet neglected fraction of the biomass and diversity in these habitats.
The inability of earlier studies to recognize the significant contribution of Nanohaloarchaea to hypersaline community composition is likely due to limitations of the tools routinely used to assess environmental microbial diversity, including laboratory culture, microscopy, amplification of 16S rRNA gene fragments, and sequence database similarity searches for unassembled metagenomic reads. The isolation of cultured strains from environmental habitats is known to exclude many organisms that are highly successful in their native habitats. It is therefore not surprising the 96 hypersaline archaeal isolates described to date do not include any Nanohaloarchaea. Repeated efforts to culture these microorganisms in our own laboratory have also been unsuccessful. Furthermore, cultivation-independent microbial diversity studies based on 16S rRNA gene amplification are known to suffer from primer bias (Sipos et al., 2007
). Mismatches between Nanohaloarchaea and many commonly used universal primers may have impeded detection in earlier studies. Primers likely to have been particularly problematic are highlighted in (Amann et al., 1990
; Lane, 1991
; DeLong, 1992
; DasSarma and Fleischman, 1995
; Ihara et al., 1997
; Brunk and Eis, 1998
; Daims et al., 1999
; Grant et al., 1999
; Baker et al., 2003
; Raes and Bork, 2008
). The exceptionally small size of Nanohaloarchaea compared with other halophilic microorganisms makes them difficult to visualize by microscopy in the absence of selective enrichment techniques or group-specific probes, and can prevent recovery during sample concentration procedures targeting larger microorganisms or smaller viruses (Rodriguez-Brito et al., 2010
). Similar issues have been noted for other nano-sized archaea, identified solely by 16S rRNA gene sequencing (Casanueva et al., 2008
; Gareeb and Setani, 2009
The presence of ultrasmall, uncultivated novel archaeal lineages in natural environments may be a common occurrence. Nanohaloarchaea represent the third nano-sized archaeal lineage to be described. However, unlike the thermophilic Nanoarchaeum equitans
(Huber et al., 2002
) or the acidophilic ARMAN lineages (Baker et al., 2006
), members of the Nanohaloarchaea appear to be free-living based on microscopic observations. The larger genomes of Nanohaloarchaea (approximately 1.2
Mbp) relative to other symbiotic/parasitic nano-sized archaea (ARMAN, <1
Mbp; Nanoarchaeum equitans
Mbp) are consistent with a possible non-host associated lifestyle for this group. It is interesting to contemplate the environmental pressures selecting for the evolution of ultrasmall microorganisms with small genomes, and to consider the extent of an ultrasmall microbial biosphere. The realization that ultrasmall populations can comprise a significant fraction of the total microbial community, yet have eluded previous detection, suggests that historical estimates of microbial biomass and numerical abundance in natural environments may be substantially underestimated. This is particularly relevant in non-extreme habitats where the existence of ultrasmall microbial populations have not yet been described or investigated.
Routine metagenomic analysis methods currently rely on the expectation that undiscovered microorganisms will have a certain degree of similarity to those already known, creating a potential bias against novel discoveries. Although this study exposes limitations of commonly used microbial diversity assessment tools in the context of detecting novel archaea in hypersaline lakes, these limitations apply even more emphatically to other more complex microbial communities, which often contain elaborate mixed consortia of Bacteria, Archaea, Eukarya and viruses. This study reinforces the utility of community genomics and de novo sequence assembly as important methods for the detection and analysis of biological diversity.