Received September 27, 1995
;
Revised and Accepted October 31, 1995
ABSTRACT
Our database on large ribosomal subunit RNA contained 334 sequences in July,
1995. All sequences in the database are aligned, taking into account secondary
structure. The aligned sequences are provided, together with incorporated
secondary structure information, in several computer-readable formats. These data can easily be obtained through the World Wide
Web. The files in the database are also available via anonymous ftp.
INTRODUCTION
Large ribosomal subunit RNA (further abbreviated as LSU rRNA) has proven an
interesting molecule to perform phylogenetic analysis (
1
). It is also very useful as a target for the detection of micro-organisms (e.g.
2
-
4
). For this type of studies, a database of aligned LSU rRNA sequences as
presented in this paper can be very useful. The database can also be used for
the design of PCR primers and the elucidation of the secondary structure of
newly determined sequences. It can also be an invaluable tool to find sequence
errors introduced during sequence analysis. These errors cause anomalies in the
sequence which can often be detected by alignment to a set of known sequences
and comparison of their possible secondary structure.
The database has been continuously updated by scanning the EMBL sequence
database (
5
) for updated or new rRNA sequences using the Current Sequence Awareness program
(a service of the Belgian EMBnet Node). These sequences are respectively used
to update older entries, or added to the database as new entries. They are then
aligned, and their secondary structure is investigated and incorporated into
the alignment using the program DCSE (
6
).
It is our goal to offer researchers easy on-line access to these LSU rRNA sequences and their alignments, together
with secondary structure information, literature references, accession numbers
and taxonomic information. All the data are obtainable in a number of formats
suitable for use in computer programs. Other databases on LSU rRNA are also
available, offering mutation data (
7
), predrawn secondary structure models (
8
) and also sequences and alignments (
9
).
CONTENTS OF THE DATABASE
Only complete or reasonably complete sequences are incorporated into the
database. Partial sequences are included only if the combined length of the
sequenced segments amounts to at least 70% of the estimated chain length of the
molecule. The chain length of a partially determined sequence is estimated by
comparing it to a complete sequence of a closely related species. The database
on LSU rRNA currently contains (July 1995) 334 sequences. This number comprises
46 eukaryotic, 17 archaeal, 102 bacterial, 38 plastidial and 131 mitochondrial
sequences.
Table
1
lists the number of representatives for each of the eukaryotic taxa in the
database. The taxonomic classification of the species is according to Brusca
and Brusca (
10
) for the Animalia, according to Cronquist (
11
) for the higher plants, according to Ainsworth
et al
. (
12
) for the zygomycetes and ascomycetes, according to Moore (
13
) for the basidiomycetes, and according to Margulis
et al
. (
14
) for the remaining eukaryotes, viz. the Protoctista.
.
Eukaryotic taxa represented in the database and number of their representatives
Kingdom Animalia
a
Phylum
Class
Number of sequences
b
N
M
Nematoda
Secernentea
1
2
Arthropoda
Insecta
2
10
Malacostraca
2
Mollusca
Bivalvia
1
Pulmonata
2
Polyplacophora
1
Echinodermata
Echinoidea
2
Chordata
Ascidiacea
1
Agnatha
1
Amphibia
3
2
Aves
18
Mammalia
3
35
Osteichtyes
12
Reptilia
4
Total:
10
92
Kingdom Fungi
Subphylum
Class
Number of sequences
b
N
M
Zygomycotina
Zygomycetes
1
Ascomycotina
Hemiascomycetes
5
3
Plectomycetes
3
Pyrenomycetes
2
Uncertain affiliation
1
Basidiomycotina
Heterobasidiomycetes
2
Total:
9
8
Kingdom Plantae
Phylum
Class
Number of sequences
b
N
M
P
Bryophyta
Marchantiopsida
1
1
Magnoliophyta
Liliopsida
1
2
3
Magnoliopsida
6
1
8
Total:
7
4
12
Kingdom Protoctista
c
Phylum
Class
Number of sequences
b
N
M
P
Apicomplexa
Coccidia
3
1
Hematozoa
3
Chlorophyta
Chlorophyceae
1
4
19
Ciliophora
2
5
Dictyostelida
1
1
Dinoflagellata
1
Euglenida
1
5
Granuloreticulosa
1
Oomycota
1
Phaeophyta
1
1
Plasmodial
Slime Molds
Myxomycota
2
Rhizopoda
Lobosea
1
1
Rhodophyta
1
1
Zoomastigina
Diplomonadida
3
Kinetoplastida
3
10
Total:
20
27
26
a
The Metazoan taxa are listed in the same order as they appear in (10).
b
The number of sequences listed in the database is larger than the number of
species, because for certain species multiple LSU rRNA sequences have been
determined, usually by different authors. The sequences are not necessarily
identical because they may have been determined for different varieties or
strains of a species, or for different genes of the same organism. The number
is listed for sequences of nuclear (N), mitochondrial (M) and plastid (P)
origin.
c
The Protoctist phyla and classes are ordered alphabetically.
Table
2
covers the bacterial and archaebacterial LSU rRNA sequences. Their
classification is based on the construction of evolutionary trees. In short,
evolutionary trees are constructed by the neighbor-joining method (
15
) for all new sequences retrieved from the EMBL (
5
) nucleotide sequence library. According to the phylogenetic position observed,
the species are assigned to one of the taxa described by Woese and co-workers (
16
,
17
) and our research group (
18
,
19
). For the Archaea, a distinction is made between the divisions Crenarchaeota
and Euryarchaeota (
20
). The latter division is further subdivided into seven subdivisions.
.
Prokaryotic taxa represented in the database and number of their
representatives
Bacteria
Division
Number of sequences
a
Cyanobacteria
1
Flavobacteria and relatives
2
Gram Positives and relatives, Low G+C
45
Gram Positives and relatives, High G+C
12
Green Sulfur
1
Planctomyces and relatives
1
Proteobacteria [alpha]
11
Proteobacteria [beta]
8
Proteobacteria [gamma]
13
Proteobacteria [epsilon]
2
Radioresistant micrococci and relatives
1
Spirochetes
4
Thermotogales
1
Total:
102
Archaea
Division
Subdivision
Number of sequences
a
Euryarchaeota
Archaeoglobales
1
Halobacteria
5
Methanobacteriales
1
Methanococcales
1
Methanomicrobium group
1
Thermococcales
1
Thermoplasma
1
Crenarchaeota:
6
Total:
17
a
The number of sequences listed in the database is larger than the number of
species, because for certain species multiple LSU rRNA sequences have been
determined, usually by different authors. The sequences are not necessarily
identical because they may have been determined for different strains, or for
different genes of the same strain.
HETEROGENEITY IN SEQUENCE AND CHAIN LENGTH
An intriguing feature of the LSU rRNA is the large variation in length between
different species as illustrated in Figure
1
. Whereas bacterial, archaeal and plastidial LSU sequences have a relatively
constant length of ~2900 nt, eukaryotic sequences show a great diversity in length, ranging
from sizes comparable with those of the bacteria to over 5000 bases in the
Homo sapiens
sequence. The presence of extra nucleotides seems to be restricted mainly to
several extremely variable insertion regions, which occupy a constant position
relative to the more conserved parts of the sequences (
21
,
22
). The variation in sequence length is even larger in mitochondria. The
ribosomal RNAs found in animal and kinetoplastid mitochondria even miss large
parts of the sequence conserved in other LSU rRNAs, and can be <1000 nt in size. Plant and fungal mitochondrial LSU rRNAs have chain lengths
comparable with or larger than those found in bacteria.
SECONDARY STRUCTURE MODEL
The secondary structure model followed in the database conforms largely to the
model developed in earlier studies (
8
,
23
-
25
). It is illustrated in Figure
2
for the LSU rRNA of the archaebacterium
Sulfolobus acidocaldarius
. The secondary structure of the molecule is treelike, with the helices forming
branches which end either in a hairpin or in a multibranched loop. The stem of
the tree joins the 5' and 3' ends of bacterial LSU rRNAs. From this stem emanates a central
multibranched loop. In Archaea this stem is generally shorter than in Bacteria,
and not always present. In Eucarya, the stem helix has completely disappeared.
AVAILABILITY AND FORMAT OF THE DATABASE
Each LSU rRNA sequence, together with gaps, secondary structure information and
reference information is stored in a separate file in a special distribution
format. The easiest way to obtain the data is through the World Wide Web (WWW).
The LSU rRNA home page can be reached at http://
rrna.uia.ac.be/rrna/lsuform.html. Using forms, a file containing any selection
of sequences can be obtained in a number of formats. The sequences can be
selected either one by one using list boxes, or by whole groups using check
buttons, or a combination of both (see Fig.
3
). The desired format should be indicated in the appropriate selection box.
Clicking on the button labelled `Get sequences' will create and transfer the
resulting file. Currently supported formats are DCSE (
6
) alignment and reference files, EMBL, NBRF/PIR, the distribution format, and a
printable form in which the alignment has been cut into blocks which fit onto a
page. The latter format is limited to a selection of 100 sequences.
ACKNOWLEDGEMENTS
Our research is supported by the BIOTECH programme of the Commission of European
Communities (contract BIO2-CT94-3098), by the Programme on Interuniversity Poles of Attraction of
the Office for Scientific, Cultural, and Technical Matters of the Belgian State
(contract 23), and by the National Fund for Scientific Research. We thank
Sabine Chapelle for the computer drawings of the secondary structure models.
Peter De Rijk and Yves Van de Peer are Research Assistants of the National Fund
for Scientific Research.
REFERENCES
1 De Rijk,P., Van de Peer,Y., Van den Broeck,I. and De Wachter,R. (1995) J. Mol. Evol., 41, 366-375.MEDLINE Abstract
2 Bastyns,K., Chapelle,S., Vandamme,P., Goossens,H. and De Wachter,R. (1994) System. Appl. Microbiol., 17, 563-568.
