Computational Genome and Pathway Analysis of Halophilic Archaea

Halophilic archaea inhabit hypersaline environments and share
common physiological features such as acidic protein machineries in
order to adapt to high internal salt concentrations as well as
electron transport chains for oxidative respiration. Surprisingly,
nutritional demands were found to differ considerably amongst
haloarchaeal species, though, and in this project several complete
genomes of halophilic archaea were analysed to predict their
metabolic capabilities. Comparative analysis of gene equipments
showed that haloarchaea adopted several strategies to utilize
abundant cell material available in brines such as the acquisition
of catabolic enzymes, secretion of hydrolytic enzymes, and
elimination of biosynthesis gene clusters. For example, metabolic
genes of the well-studied Halobacterium salinarum were found to be
consistent with the known degradation of glycerol and amino acids.
Further, the complex requirement of H. salinarum for various amino
acids and vitamins in comparison with other halophiles was
explained by the lack of several genes and gene clusters, e.g. for
the biosynthesis of methionine, lysine, and thiamine. Nitrogen
metabolism varied also among halophilic archaea, and the
haloalkaliphile Natronomonas pharaonis was predicted to apply
several modes of N-assimilation to cope with severe ammonium
deficiencies in its highly alkaline habitat. This species was
experimentally shown to possess a functional respiratory chain, but
comparative analysis with several archaea suggests a yet unknown
complex III analogue in N. pharaonis. Respiratory chains of
halophilic and other respiratory archaea were found to share
similar genes for pre-quinone electron transfer steps but show
great diversity in post-quinone electron transfer steps indicating
adaptation to changing environmental conditions in extreme
habitats. Finally, secretomes of halophilic and non-halophilic
archaea were predicted proposing that haloarchaea secretion
proteins are predominantly exported via the twin-arginine pathway
and commonly exhibit a lipobox motif for N-terminal lipid
anchoring. In N. pharaonis, lipoboxcontaining proteins were most
frequent suggesting that lipid anchoring might prevent protein
extraction under alkaline conditions. By contrast, non-halophilic
archaea seem to prefer the general secretion pathway for protein
translocation and to retain only few secretion proteins by
N-terminal lipid anchors. Membrane attachment was preferentially
observed for interacting components of ABC transporters and
respiratory chains and might further occur via postulated
C-terminal anchors in archaea. Within this project, the complete
genome of the newly sequenced N. pharaonis was analysed with focus
on curation of automatically generated data in order to retrieve
reliable gene prediction and protein function assignment results as
a basis for additional studies. Through the development of a
post-processing routine and expert validation as well as by
integration of proteomics data, a highly reliable gene set was
created for N. pharaonis which was subsequently used to assess
various microbial gene finders. This showed that all automatic gene
tools predicted a rather correct gene set for the GC-rich N.
pharaonis genome but produced insufficient results in respect to
their start codon assignments. Available proteomics results for N.
pharaonis and H. salinarum were further analysed for
posttranslational modifications, and N-terminal peptides of
haloarchaeal proteins were found to be commonly processed by
N-terminal methionine cleavage and to some extent further modified
by N-acetylation. For general function assignment of predicted N.
pharaonis proteins and for enzyme assignment in H. salinarum,
similarity-based searches, genecontext methods such as
neighbourhood analysis but also manual curation were applied in
order to reduce the number of hypothetical proteins and to avoid
cross-species transfer of misassigned functions. This permitted to
reliably reconstruct the metabolism of H. salinarum and N.
pharaonis. Generated metabolic data were stored in a newly
developed metabolic database that also integrates experimental data
retrieved from the literature. The pathway data can be assessed as
coloured KEGG maps and were combined with data resulting from
transcriptomics and proteomics techniques. In future,
expert-curated reaction entries of the created metabolic database
will be a valuable source for the design of metabolic experiments
and will deliver a reliable input for metabolic models of
halophilic archaea.