Population structure and phylogenetic characterization of marine benthic Archaea in deep-sea sediments

Abstract

During the past few years Archaea have been recognized as a widespread and significant component of marine picoplankton assemblages and, more recently, the presence of novel archaeal phylogenetic lineages has been reported in coastal marine benthic environments. We investigated the relative abundance, vertical distribution, phylogenetic composition, and spatial variability of Archaea in deep-sea sediments collected from several stations in the Atlantic Ocean. Quantitative oligonucleotide hybridization experiments indicated that the relative abundance of archaeal 16S rRNA in deep-sea sediments (1500 m deep) ranged from about 2.5 to 8% of the total prokaryotic rRNA. Clone libraries of PCR-amplified archaeal rRNA genes (rDNA) were constructed from 10 depth intervals obtained from sediment cores collected at depths of 1,500, 2,600, and 4,500 m. Phylogenetic analysis of rDNA sequences revealed the presence of a complex archaeal population structure, whose members could be grouped into discrete phylogenetic lineages within the two kingdoms, Crenarchaeota and Euryarchaeota. Comparative denaturing gradient gel electrophoresis profile analysis of archaeal 16S rDNA V3 fragments revealed a significant depth-related variability in the composition of the archaeal population.

FIG. 1

Depth-related abundance of archaeal 16S rRNA as a fraction of total prokaryotic rRNA from an AC sediment core. Bars are means of two determinations, with error bars representing the range.

FIG. 2

RFLP pattern distribution for 61 16S rDNA clones isolated from different vertical depth intervals from the AP and CR-1 sediments (4,500- and 2,616-m depths, respectively).

FIG. 3

Phylogenetic analysis of deep-sea benthic Archaea. A maximum likelihood tree was constructed for the archaeal sequences by using fastDNAml (17). The scale represents the expected number of changes per sequence position. Abbreviations for the benthic Archaea (in boldface) are formed by a combination of the sampling station abbreviation, followed by the letter A (Archaea), the phylotype number, and the vertical depth interval of the sediment core from which the specific phylotype was obtained. Abbreviations for as-yet-uncultivated phylotypes from various environments: BBA (AF004343 to AF004348), from continental shelf sediments (56); JM8 (L24201), from the gut of a deep-sea sediment feeder (38); Mariana15 (D87350), from Mariana Trench sediments (31); SBAR5 (M88075), SBAR16 (M88077), OARB (U11040), WHARN (M88078), ANTARCTIC5 (U11044), pN1-2 (U86455), pN1-73 (U86462), and p712-3 (U81540), from marine picoplankton (9, 20); FFSB1 (X96688) and SCA1145 (U62811), from soil (6, 30); pGrfA4 (U59968), pGrfC26 (U59986), pGrfB286 (U59984), LMA134 (U87515), LMA238 (U87517), pLAW11 (U77569), and pLAW12 (U77568), from freshwater sediments (26, 35, 50); Arc.98 (AF005760) and Arc.168 (AF005764), from a deep-subsurface paleosol (8); pJP33 (L25300), pJP41 (L25301), pJP89 (L25305), pSL17 (U63339), pSL22 (U63340), and pSL123 (U63445), from a Yellowstone National Park hot spring (5); 2MT1 (AF015981) and 2MT8 (AF015992), from salt marsh sediments (42); and WCHD3 (AF050616), from a contaminated aquifer (14).

FIG. 4

Group distribution for 41 16S rDNA sequences obtained from six different vertical depth intervals from AC, CR, and AP sediments.

FIG. 5

Comparative DGGE profile analysis of archaeal 16S rDNA V3 fragments from two sampling stations and five vertical depth intervals. Lanes: 1, CR-1, 0 to 2 cm; 2, CR-1, 11 to 13 cm; 3, CR-1, 25 to 27 cm; 4, CR-2, 0 to 2 cm; 5, CR-2, 21 to 23 cm. Letters indicate sequenced 16S rDNA fragments; putative Crenarchaeota are indicated in boldface type, and Euryarchaeota are indicated in regular type (see text and Table 3).

FIG. 6

Maximum-likelihood unrooted phylogenetic tree showing the position of the deep-sea benthic Archaea relative to members of the three main domains of life. Branches in boldface type indicate extremely thermophilic organisms. The scale represents the expected number of changes per sequence position. The numbers give the bootstrap values obtained for a bootstrap sampling of 100.