The Open Bioinformatics Journal, 2008, 2, 11-19 11
1875-03 62/08 2008 Be ntham Science Publ ishers Ltd.
Nucleotide Composition and Amino Acid Usage in AT-Rich
Hyperthermophilic Species
Subhash Mohan Agarwal1,* and Atul Grover2
1Bioinformatics Center, School of Information Technology, Jawaharlal Nehru University, New Delhi 110067, India and
2Department of Bioscience and Biotechnology, Banasthali University, Banasthali 304022, India
Abstract: Nucleotide composition, codon usage and amino acid content are important molecular signatures that vary in
different groups of organisms. AT-rich (or GC poor) hyperthermophiles have relatively been unexplored in these aspects.
In this study, we have examined the compositional characteristics of AT rich genomes viz. Methanococcus jannaschii,
Sulfolobus solfataricus, Sulfolobus tokodaii and Nanoarcheum equitans by their comparison with four mesophiles having
similar genomic GC content. The analysis revealed a significant increase in purine content of ORFs due to increase in
guanine content. Moreover, the influence of dinucleotide composition on protein thermostability was found even larger.
Accordingly, increased usage of codons that are constituted of dinucleotides RR was observed. Arginine, proline, valine
and tyrosine were most abundant amino acids in hyperthermophilic proteomes, and similar bias was seen when dipeptidic
composition of proteins was compared. Further amino acid composition analysis of alpha helices indicates an increased
usage of E, K, R and decreased usage of N and Q. Summing up, the study suggested that elevated growth temperature im-
pose selective constraints at all the three molecular levels- nucleotide composition, codon usage and amino acid content.
Keywords: Hyperthermophiles; nucleotide bias; codon usage; amino acid composition.
INTRODUCTION
Hyperthermophiles constantly face the challenge of
maintaining the stability of their genome. Increasing the
melting point of their DNA by keeping relatively higher GC
[1] is one of the methods they have constituted to address the
issue. However, the GC content of the genomes does not
correlate with optimal growth temperature (OGT) [2, 3].
Various additional attributes have been suggested that con-
tribute in maintaining the stability of genomic DNA of hy-
perthermophiles [2, 3]. Infact a number of hyperthermo-
philes have GC content of their DNA lesser than 40% [1].
On the other hand GC content of rRNA and tRNA show
strong correlation with optimal growth temperature [4, 5].
Various studies have established that these living organisms
are subject to a variety of selection pressures that act not
only at the level of global phenotype but at each level of the
cell’s organization i.e. DNA, RNA and proteins [6]. For ex-
ample, there is evidence that the proteins of thermophiles are
characterized by a distinct pattern of amino acids [7-10].
Moreover a difference in the pattern of synonymous codon
usage between thermophiles and mesophiles has been ob-
served [7, 10].
Although considerable studies have focused on under-
standing the mechanisms that makes life possible under these
conditions, it still remains unclear that whether it is due to
external conditions or natural selection [4, 7, 11-14]. In order
to infer the molecular mechanistic adjustments to the thermal
stress, it is desired to compare the genomic characteristics of
hyperthermophiles with mesophilic genera. Singer and
Hickey [14] made such an attempt considering the genera
that show optimal growth temperature (OGT) near or above
*Address correspondence to this author at the Bioinformatics Center, School
of Information Technology, Jawaharlal Nehru University, New Delhi
110067, Ind ia; E-mail: smagarwal@yahoo.co.in
50°C, while the AT-rich hyperthermophilic genomes were
ignored in their analysis. Das et al. [5] looked into some of
the hyperthermophilic genomes that had their GC content
lower than 50%. A shortcoming of this study was the broad
range of OGT (>13°C) over which the genera under study
varied. Thus, in order to minimize the ascertainment bias in
terms of codon usage and nucleotide composition between
different species we have picked up various mesophiles and
hyperthermophiles in an even narrower OGT range of 7.8°C
for comparative analysis among mesophiles and hyper-
thermophiles.
The hyperthermophilic archaebacteria, Nanoarchaeum
equitans is one of the interesting examples qualifying for this
kind of analysis. The archaebacteria is known to host small-
est non viral genome to date, which spans 490 Kb and is
constituted of 537 protein coding genes [15]. The genome
displays short intergenic regions, large number of split
genes, few pseudogenes, and lacks many of the vital meta-
bolic genes [15]. Further phylogenetic analysis suggested
that it diverged early in archaeal lineage even before the
emergence of Euryarchaeota and Crenarchaeota, represent-
ing basal archaeal lineage [16]. Considering Nanoarcheum
equitans to be one of the simplest genome of cellular organ-
isms and of course simplest among the genera under study,
genomic features of N. equitans have been dealt as a special
case within the hyperthermophilic group.
Thus, the present paper outlines comparisons of nucleo-
tide bias, codon usage patterns and amino acid bias drawn
between mesophiles and hyperthermophiles having average
GC content close to 31%.
MATERIALS AND METHODS
The coding sequences (CDS) and the corresponding
amino acid sequences for all of the eight genomes (Table 1)
were downloaded from ftp site of GenBank. Following CDS
