Genome sequence of Haloarcula marismortui: a halophilic archaeon from the Dead Sea

Abstract

We report the complete sequence of the 4,274,642-bp genome of Haloarcula marismortui, a halophilic archaeal isolate from the Dead Sea. The genome is organized into nine circular replicons of varying G+C compositions ranging from 54% to 62%. Comparison of the genome architectures of Halobacterium sp. NRC-1 and H. marismortui suggests a common ancestor for the two organisms and a genome of significantly reduced size in the former. Both of these halophilic archaea use the same strategy of high surface negative charge of folded proteins as means to circumvent the salting-out phenomenon in a hypersaline cytoplasm. A multitiered annotation approach, including primary sequence similarities, protein family signatures, structure prediction, and a protein function association network, has assigned putative functions for at least 58% of the 4242 predicted proteins, a far larger number than is usually achieved in most newly sequenced microorganisms. Among these assigned functions were genes encoding six opsins, 19 MCP and/or HAMP domain signal transducers, and an unusually large number of environmental response regulators-nearly five times as many as those encoded in Halobacterium sp. NRC-1--suggesting H. marismortui is significantly more physiologically capable of exploiting diverse environments. In comparing the physiologies of the two halophilic archaea, in addition to the expected extensive similarity, we discovered several differences in their metabolic strategies and physiological responses such as distinct pathways for arginine breakdown in each halophile. Finally, as expected from the larger genome, H. marismortui encodes many more functions and seems to have fewer nutritional requirements for survival than does Halobacterium sp. NRC-1.

Figure 1.

Evolutionary comparison between the main chromosomes of H. marismortui (vertical axis) and Halobacterium sp. NRC-1 (horizontal axis). Scales are equal and in kilobasepairs (kb). Blue points in the dot-plot indicate the location of putative orthologous genes. Green boxes and lines indicate membership in ADHoRe clusters; larger boxes correspond to more populous clusters. The location of the cdc6d/orc7 orthologous pair is indicated in red. The graphs above and to the right depict the distribution of inversion boundaries as identified by GRIMM. Regions of significant nucleotide skews (absolute Z-score higher than three) are depicted in the graphs below and to the left of the dot-plot. Dotted lines are included as an aid for visualization.

Figure 2.

Functional associations among proteins implicated in the denitrification process in H. marismortui. (Inset) Key indicates functional relevance of node and edge color attributes (for other details, see text). A detailed list of dentrification proteins is included in Supplemental Table C.

Figure 3.

Comparative metabolic reconstruction in H. marismortui and Halobacterium sp. NRC-1. An example of metabolic reconstruction using the Web intermediary software and KEGG: arginine metabolism and the urea cycle in H. marismortui compared with Halobacterium sp. NRC-1. Enzymes identified in Halobacterium sp. NRC-1 are shaded in green, and those identified in H. marismortui are shown with a red border and red font. The ADI pathway of arginine fermentation in Halobacterium sp. NRC-1 is indicated with solid blue arrows. Each enzyme identified in H. marismortui is hyperlinked to its corresponding entry in the SBEAMS annotation database. This view of KEGG metabolic pathways for the two halophiles is available on http://halo.systemsbiology.net.

Figure 4.

Distribution in H. marismortui and Halobacterium sp. NRC-1of sensory domains among unique proteins having at least one such domain. The distribution of light-sensing (GAF [PF01590]) and redox-sensing domains (PAS [PF00989]), (PAC [PF00785]) in H. marismortui (A) and in Halobacterium sp.NRC-1 (B). (C) The distribution of methyl accepting chemotaxis (MCP–PF00015), HAMP (PF00672), histidine kinase (PF00512 and PF02518), and response regulator (PF00015) domains in H. marismortui. The blue circles represent the kinase domains. The red shaded area indicates the number of proteins that also contain PAS, PAC, or GAF domains.