Microbiology Comment provides a
platform for readers of Microbiology to
communicate their personal observations
and opinions in a more informal way than
through the submission of papers.
Most of us feel, from time to time, that
other authors have not acknowledged the
work of our own or other groups or have
omitted to interpret important aspects of
their own data. Perhaps we have
observations that, although not sufficient
to merit a full paper, add a further
dimension to one published by others. In
other instances we may have a useful piece
of methodology that we would like to
The Editors hope that readers will take full
advantage of this section and use it to raise
matters that hitherto have been confined to
a limited audience.
Christopher M. Thomas, Editor-in-chief
Cohesion of sister
This is a reply to an article published in a
previous issue of Microbiology (4). We first
demonstrated cohesion of replicated oriC
copies in E. coli cells (temperature sensitive
of chromosomal DNA replication [3; this
manuscript is not cited in the associated
same manuscript, we also demonstrated that
a single oriC focus, which was detected by
FISH (fluorescence in situ hybridization), was
located at the middle of the cell (more
correctly, at the middle of the nucleoid) just
before and after initiation of replication. A
single oriC focus appeared to separate into
two foci late in the chromosome replication
cycle. The result suggests that oriC is repli-
cated at the middle of the nucleoid and these
sister copies remain closely associated for a
long time. Replicated sister copies of other
chromosomal regions also remain associated
after replication until late in replication (11).
Sister oriC copies tend to locate as a single
focus near the middle of the nucleoid until the
point that two distinct foci can be seen (3).
However, sometimes it appears that a single
focus can move to a border of the nucleoid,
separate into two foci and these can migrate
asymmetrically toward both pole-proximal
borders of the nucleoid prior to the cell
division as described previously (6). Based on
the cohesion model, ‘duplication of an oriC
focus’ (6) does not mean ‘replication of the
oriC segment’, but ‘separation of cohesive
oriC copies’. Thus the distinction is based on
our experimental design in which replication
was initiated at a known time by a tem-
perature downshift in the dnaAts strain, and
then observation of separate foci was seen
only at a much later time.
The second subject is translocation of
replication apparatuses during chromosome
replication. Paired replication apparatuses
that are acting for bidirectional replication of
the chromosomal DNA are first closely lo-
cated with each other at the middle of the
and rapidly migrate to 1?4 and 3?4 cell
positions at the time that replication forks are
estimated to have progressed bidirectionally
approximately one third of the distance be-
tween oriC and terminus. This model, named
the ‘translocating replication apparatuses
model’ (3; Fig. 1) or ‘replication apparatuses
translocation model’, is based on analysis of
the subcellular localization of SeqA-bound
hemimethylated nascent DNA clusters (2, 3,
9, 11; for a review, see 1) and beta-clamp
clusters (DNA sliding clump or DnaN) of
DNA polymerase III holoenzyme (10). Ac-
cording to this model, copies of oriC and
the translocation of replication apparatuses
might remain localized as a single focus in the
middle of the nucleoid (Fig. 1B-b and c). On
the other hand, other genes that are replicated
after the translocation of replication appar-
atuses might be localized as a single focus
around the replication apparatus localized at
the 1?4 or 3?4 cell position until release from
cohesion (Fig. 1B-e and f). Our previous
results by FISH (7) are consistent with the
above speculation. After release from co-
hesion, sister chromosomes are rearranged
to form two separated nucleoids, in which
oriC copies are localized near pole-proxi-
mal borders of the nucleoid(s) and dif
copies are localized near mid-cell, that is, the
division site (Fig. 1B-f and g; 6, 7). The
migration of oriC copies toward pole-proxi-
mal borders of the nucleoid is presum-
ably due to an uncharacterized mechanism,
similar to the plasmid partition system
It is not so easy to determine the critical
time of initiation of chromosome replication
in exponentially growing cultures using the
of subcellular localization of fluorescent foci.
Furthermore, some complicated assumptions
are required for judging sister chromosome
cohesion in statistical results on the sub-
cellular localization of fluorescent foci in
exponentially growing cultures in the steady
state. We therefore analysed cells that were
synchronized for initiation of replication in
order to demonstrate the cohesion phenom-
enon (3, 11). More recently, we analysed
exponentially growing cells that carried a
cassette of multiple lacO sequences that was
inserted near oriC of the chromosome, using
the following two different methods; (i) FISH
in fixed cells to detect oriC foci and (ii)
fluorescence microscopy in living cells to
detect LacI–GFP foci. In both the methods,
approximately half of cells with a positive
signal cells had one fluorescent focus and the
Communications should be in the form
of letters and should be brief and to the
point. A single small Table or Figure may
be included, as may a limited number of
references (cited in the text by numbers,
and listed in alphabetical order at the end
of the letter). A short title (fewer than 50
characters) should be provided.
Approval for the publication rests with
the Editor-in-Chief, who reserves the
right to edit letters and?or to make a
brief reply. Other interested persons may
also be invited to reply. The Editors of
Microbiology do not necessarily agree
with the views expressed in
Contributions should be addressed to the
Editor-in-Chief via the Editorial Office.
Microbiology 148, December 20023755
Fig. 1. A model: translocation of replication apparatuses and cohesion of sister chromosomes in E.
coli. Fine lines and grey masses represent unreplicated or replicated fully methylated chromosomal
DNA. Bold lines and black masses represent nascent hemimethylated DNA strands. Black circles, oriC
sites; black triangles, terminus sites; open squares, replication apparatuses. Horizontal arrows in B-d
represent rapid bidirectional migration of SeqA clusters from the mid-cell position to the 1?4 and 3?4
other half had two foci, even though flow
cytometry indicated that the majority of
newborn cells had one replicating chromo-
some with two oriC copies and that cells had
four oriC copies prior to cell division (our
unpublished data). This discrepancy can be
explained by the sister chromosome cohesion
Mythird commentisaboutthe rolesof the
MukFEB complex (12, 13). MukB seems to
participate in sister chromosome cohesion
(11). Interestingly, MukB–GFP molecules are
recruited to clusters, although part of MukB–
GFP molecules diffuse throughout the nucle-
oid (8). MukB–GFP forms two clusters at 1?4
and 3?4 cell positions prior to migration of
SeqA–DNA clusters to the cell quarter pos-
itions in minimum glucose medium. After the
MukB–GFP cluster is separated into two,
resulting in four MukB–GFP clusters prior to
the cell division (8). The cluster formation
depends on the presence of both MukF and
MukE, suggesting that MukFEB complex
forms clusters. Our recent results on the
localization of nascent DNA that was pulse-
labelled with 5-bromo-2?-deoxyuridine indi-
cate that MukB participates in the correct
localization of replication forks at mid-cell or
the 1?4 and 3?4 cell positions (our unpub-
lished data). Thus, segregation of sister chro-
mosomes depends on functions of specific
Department of Radiation Genetics, Graduate
School of Medicine, Kyoto University,
Konoe-cho, Yoshida, Sakyo-ku, Kyoto
Author for correspondence: Sota Hiraga.
Tel: ?81 75 753 4415. Fax: ?81 75 753
1. Hiraga, S. (2000). Dynamic localization of bacterial
and plasmid chromosomes. Annu Rev Genet 34, 21–59.
2. Hiraga, S., Ichinose, C., Niki, H. & Yamazoe, M.
(1998). Cell cycle-dependent duplication and
bidirectional migration of SeqA-associated DNA-protein
complexes in E. coli. Mol Cell 1, 381–387.
3. Hiraga, S., Ichinose, C., Onogi, T., Niki, H. &
Yamazoe, M. (2000). Bidirectional migration of SeqA-
bound hemimethylated DNA clusters and pairing of oriC
copies in Escherichia coli. Genes Cells 5, 327–341.
4. Nanninga, N., Roos, M. & Woldringh, C. L. (2002).
Models on stickiness of replicated Escherichia coli oriC.
Microbiology 148, 3327–3328.
5. Niki, H. & Hiraga, S. (1997). Subcellular distribution
of actively partitioning F plasmid during the cell division
cycle in E. coli. Cell 90, 951–957.
6. Niki, H. & Hiraga, S. (1998). Polar localization of the
replication origin and terminus in Escherichia coli
nucleoids during chromosome partitioning. Genes Dev
7. Niki, H., Yamaichi, Y. & Hiraga, S. (2000). Dynamic
organization of chromosomal DNA in Escherichia coli.
Genes Dev 14, 212–223.
8. Ohsumi, K., Yamazoe, M. & Hiraga, S. (2001).
Different localization of SeqA-bound nascent DNA
clusters and MukF-MukE-MukB complex in Escherichia
coli cells. Mol Microbiol 40, 835–845.
9. Onogi, T., Niki, H., Yamazoe, M. & Hiraga, S.
(1999). The assembly and migration of SeqA-Gfp fusion
in living cells of Escherichia coli. Mol Microbiol 31,
10. Onogi T., Ohsumi, K., Katayama, T. & Hiraga, S.
(2002). Replication-dependent recruitment of the β-
subunit of DNA polymerase III from cytosolic spaces to
replication forks in Escherichia coli. J Bacteriol 184,
11. Sunako, T., Onogi, T. & Hiraga, S. (2001). Sister
chromosome cohesion of Escherichia coli. Mol Microbiol
12. Yamanaka, K., Ogura, T., Niki, H. & Hiraga, S.
(1996). Identification of two new genes, mukE and
mukF, involved in chromosome partitioning in
Escherichia coli. Mol Gen Genet 250, 241–251.
13. Yamazoe, M., Onogi, T., Sunako, Y., Niki, H.,
Yamanaka, K., Ichimura, T. & Hiraga, S. (1999).
Complex formation of MukB, MukE and MukF proteins
involved in chromosome partitioning in Escherichia coli.
EMBO J 18, 5873–5884.
The minimal mobile
Horizontal transfer of genes is an integral
component of bacterial evolution and is
particularly associated with processes related
to environmental adaptation and virulence.
The association of many host adaptive and
virulence genes with mobile genetic elements
such as transposases and bacteriophages re-
flects this. Presumably the association of the
gene conferring some competitive advantage
for the recipient strain facilitates the dis-
semination of the mobile element. The pres-
ence of larger elements such as ‘pathogeni-
city islands’ or ‘islands of horizontal trans-
fer’ is also characteristic of some bacterial
bacterial genomes suggests a previously
utilizes natural transformation and homolo-
gous recombination, and that is indepen-
dent of transposases and other mobilization
When the genomes of related bacterial
species are compared, sites in which different
genes are present are identifiable, located
between highly conserved flanking genes. For
example, the neisserial pheS and pheT genes
Microbiology 148, December 20023756
Fig. 1. Nine different intergenic regions between pheS and pheT identified in the Neisseria spp. NMA
and NMB numbers refer to the respective annotations (7, 8). N. gonorrhoeae strain FA1090 (5) XNG
numbers refer to our own annotation: associated coding sequences available on request. N.
meningitidis strain FAM18: these sequence data were produced by the N. meningitidis serogroup C
ftp:??ftp.sanger.ac.uk?pub?pathogens?nm?. Described coding sequences are available on request. N.
meningitidis strain 01?241825 sequence is identical to GenBank AJ311178. Other sequences submitted
to GenBank: N. lactamica, AF542177; N. polysaccharea, AF542178; N. elongata, AF542173; and N.
flavescens, AF542175. Homologous coding sequences are indicated. ‘Dead’ indicates the presence of
multiple frame-shifts and?or in-frame termination codons, while ‘frame-shifted’ indicates a single
Institute andcan beobtained from
alpha and beta chains, respectively) flank two
genes of unknown function (NMA0935 and
NMA0936) encoding proteins of 354 and
172 aa in length in Neisseria meningitidis
strain Z2491 (7), and genes coding for two
different restriction?modification systems in
gonorrhoeae strain FA1090 (5) (Fig. 1). These
genes are adjacentin both Helicobacter pylori
strains (1, 9) and Haemophilus influenzae (3).
phenylalanyl-tRNA synthetase, In N. gonorrhoeae strain FA1090 glyQ and
glyS (encoding glycyl-tRNA synthetase, alpha
a short intergenic region of 116 nt. However,
in the meningococci glyQ and glyS flank a
putative phase-variable gene of unknown
and glyS flank genes encoding a nickel-
resistance type cation efflux system, and this
site in H. influenza strain Rd contains two
genes of unknown function.
To extend this observation, the regions
between the pheS and pheT genes were
amplified from a small collection of unrelated
neisserial strains and species (Table 1). PCR
products were digested with DpnII and XbaI,
and product sizes and digest patterns were
compared on agarose gels. The amplified and
genome sequence-derived sequences were
sorted into a number of similar groups by
these tests (Table 1). An example of each type
that differed in size or digest pattern was
sequenced (previously unavailable sequences
GenBank accession nos: Neisseria cinerea,
AF542174; Neisseria elongata, AF542173;
Neisseria flavescens, AF542175; Neisseria
AF542176), and after accounting for altered
digest patterns due to single base poly-
morphisms, a total of 9 major variants were
found between these two phe genes (Fig. 1).
Previous work on this region in N. menin-
gitidis identified a region (AF238948) similar
to that in sequence strain FAM18 (2) (Fig. 1).
One insert, sequenced from N. meningitidis
strain 01?241825, was identical to a pre-
viously submitted sequence that did not in-
clude the pheS?pheT flanking sequences
(AJ311178). Two short intergenic sequences
were identified: one of 123 bp was present in
N. elongata and N. subflava, while the other,
sequenced from N. flavescens and N. cinerea
was 103 bp in length. These two short inter-
genic regions are not homologous and do not
include any probable coding regions. The
restriction?modification system genes, which
are believed to confer self-selectable pheno-
types (4, 6), which is consistent with the
Additional examples of functionally re-
lated, metabolic, gene pairs flanking insertion
sites containing different genes were sought
genome sequences. This revealed five more
such sites: the alpha and beta subunits of
ribonucleoside-diphosphate reductase, nrdA
and nrdB; the small and large subunits of
carbamoyl-phosphate synthase, carA and
carB; two components of an Fe–S cluster
associated cysteine metabolism system, nifU
and nifS; LPS biosynthetic genes, rfaD and
rfaE; and the L and M subunits of NADH
dehydrogenase, nuoL and nuoM. This list
does not include all genes likely to be mobile
by this mechanism within this and related
conservation being differentially variable be-
tween different pairs of strains and species.
In the case of pheS and pheT sequence
comparisons of the ends of these genes in the
with recombination, with divergent bases
present in different combinations indicating a
mosaic structure associated with recombi-
nation, that clustered differently between the
two genes (Fig. 2). This is consistent with a
Microbiology 148, December 20023757
Table 1. Neisseria spp. grouped by pheS–pheT regions based on PCR amplification and restriction analysis
N. polysaccharean?aD. Stephens§
Microbiology 148, December 2002 3758
Table 1. (cont.)
Species Strain SerogroupSource
N. meningitidis 01?24182529E MRUPHL*
*Meningococcal Reference Unit, Public Health Laboratory, UK
†WHO Reference Laboratory
‡University of Oxford, UK
§Emory University, USA
¶Drexel University, USA
?UK National Reference Laboratory
model in which recombination leading to the
exchange of sections of these genes is being
driven by the phenotypes of the intervening
gene(s). While one might predict that the
changes in the flanking genes are normally
neutral, it is also possible that minor changes
in the functions of these genes may sometimes
The essential feature of each case is that
the flanking genes are highly conserved meta-
a transcriptional unit. In the pheS–pheT
example, the inserted genes are in the same
orientation as the conserved flanking genes,
which may indicate their inclusion in the
transcriptional unit. Insertions that generate
polar effects in such a way as to disrupt the
be selected against. If a gene becomes inserted
into such a site it has acquired highly con-
served flanking sequences that can act as
substrates for homologous recombination be-
tween other strains and, depending upon the
there may be no specific feature targeting
or ‘selfish’ gene has inserted into such a site
it will then gain the capacity to be more
readily transferred. This probably represents
the minimal mobile element and should be
considered in analyses of horizontal gene
N.J.S. is supported by a Wellcome Trust
Advanced Research Fellowship.
Nigel J. Saunders and Lori A. S. Snyder
Bacterial Pathogenesis and Functional
Genomics Group, Sir William Dunn School of
Pathology, University of Oxford, South Parks
Road, Oxford. OX1 3RE.
Author for correspondence: Nigel J.
Saunders. Tel: (?44) 1865 275521.
Fax: (?44) 1865 275515.
1. Alm, R. A., Ling, L. S., Moir, D. T. & 20 other
authors. (1999). Genomic-sequence comparison of two
unrelated isolates of the human gastric pathogen
Helicobacter pylori. Nature 397, 176–180.
2. Claus, H., Friedrich, A., Frosch, M. & Vogel, U.
(2000). Differential distribution of novel restriction-
modification systems in clonal lineages of Neisseria
meningitidis. J Bacteriol 182, 1296–1303.
3. Fleischmann, R. D., Adams, M. D., White, O. & 37
other authors (1995). Whole-genome random sequencing
and assembly of Haemophilus influenzae Rd. Science
4. Kobayashi, I. (2001). Behavior of restriction-
modification systems as selfish mobile elements and their
impact on genome evolution. Nucleic Acids Res 29,
5. Lewis, L., Gillaspy, A. F., McLaughlin, R. E. & 22
other authors. (1997).
GenBank No AE004969.
6. Nakayama, Y. & Kobayashi, L. (1998). Restrictions-
modification gene complexes as selfish gene entities: roles
of a regulatory system in their establishment,
maintenance, and apoptotic mutual exclusion. Proc Natl
Acad Sci USA 95, 6442–6447.
7. Parkhill, J., Achtman, M., James, K. D. & 25 other
authors (2000). Complete DNA sequence of a serogroup
A strain of Neisseria meningitidis Z2491. Nature 404,
8. Tettelin, H., Saunders, N. J., Heidelberg, J. & 39
Microbiology 148, December 20023759
Fig. 2. Regions from 11 strains with different genes between pheS and pheT. Divergence within the
coding region of the flanking genes of the minimal mobile element is indicative of the recombination-
generated mosaic structure of these regions. Note that the sequences from these strains have clustered
differently at the two ends of this region. A. Region containing the termination codon of pheS (white
letters). B. Region containing the initiation codon of pheT (white letters). FAM18: N. meningitidis
strain FAM18, sequence from the Sanger Centre; N.l., N. lactamica, sequence submitted to GenBank
(AF542177); N.p., N. polysaccharea, sequence submitted to GenBank (AF542178); Z2491, N.
meningitidis strain Z2491 (7); MC58, N. meningitidis strain MC58 (8); 29E, N. meningitidis strain
01?241825, identical pheS?pheT region to AJ311178; N.e., N. elongata, sequence submitted to
GenBank (AF542173); N.s., N. subflava, sequence submitted to GenBank (AF542176); N.f., N.
flavescens, sequence submitted to GenBank (AF542175); N.c., N. cinerea, sequence submitted to
GenBank (AF542174); FA1090, N. gonorrhoeae strain FA1090 (5).
other authors (2000). Complete genome sequence of
Neisseria meningitidis serogroup B strain MC58. Science
9. Tomb, J. F., White, O., Kerlavage, A. R. & 39 other
authors (1997). The complete genome sequence of the
gastric pathogen Helicobacter pylori. Nature 388,
between the channel-
forming domains of
voltage-gated ion channel
proteins and the C-
terminal domains of
secondary carriers of the
Integral membrane transport proteins are
believed to have evolved as a distinct class of
proteins independently of other protein types
such as enzymes, structural proteins and
regulatory proteins (14). The ancestral ele-
mentswere small channel-forming oligomeric
proteins or peptides analogous in structure
and function to the present-day potassium
channel of Streptomyces lividans KscA (4)
and the toxic bacteriocins (5, 12). Genes
encoding these one or two transmembrane
segment (TMS) peptides duplicated or were
fused to other genetic elements to give larger
and functionally more complex transport
systems (9). Superimposition of catalytic pro-
teins onto channel-forming proteins or car-
riers gave rise to primary active translocators
(14, 17). The specific pathways taken for the
evolution of several families of transporters
have been elucidated (16). The proposed
pathway taken for the evolutionary appear-
ance of different types of transporters (14) is:
channel peptides?channel proteins
primary active transporters
Of the over 300 families of transport
proteins currently recognized, three of these
predominate (15). These families are the
voltage-gated ion channel (VIC) superfamily
(TC ?1.A.1; 9), the major facilitator super-
family (MFS) of secondary carriers (TC
?2.A.1; 10, 18) and the ATP-binding cassette
(ABC) superfamily of primary active trans-
porters (1). No previous publication has
provided evidence for a structural, functional
or evolutionary connection between these
three superfamilies. In this article we provide
VIC superfamily and the MFS.
MFS proteins arose by an internal gene
duplication event in which a primordial 6
encoding element (see 10 and references cited
Microbiology 148, December 20023760
Fig. 1. Model of the 6 TMS channel unit of
many VIC family members including the Na+
channel protein of Bacillus halodurans, NaChBac.
S1 (white) is the sequence involved in insertion of
the protein into the lipid bilayer membrane. S2,
S3 and S4 (black) jointly form the voltage sensor;
S5, the P-loop and S6 (grey) are components of
the pore-forming structure. The transmembrane
helices and the short pore (P) helix are displayed
Fig. 2. Alignment of the channel-forming domain of NaChBac with the last three TMSs (TMSs 10–12)
in the E. coli SetC protein. Positions of TMSs 5 and 6 and the P-loop in NaChBac are as presented in
ref. 11. Positions of the last three TMSs in the 12 TMS SetC protein are based on hydropathy profiles
but correlate with the positions of established TMSs in other MFS permeases (see 10, 18 and cited
references). TMSs are boxed and labelled by number while the P-loop is so labelled. Identities are
indicated by vertical lines between identical residues; close and distant similarities are indicated by
colons and dots, respectively. The alignment was generated using the GAP program (3).
Fig. 3. Proposed pathway for the appearance of a monomeric 12 TMS carrier of the MFS from a
primordial tetrameric 2 TMS channel-forming protein of the voltage-gated ion channel (VIC)
superfamily involving two intermediate structures, I1 and I2. Note that the two proposed intragenic
duplication events shown in steps 2 and 3, give rise to a pseudotetrameric structure, superficially
resembling the established tetrameric VIC family structure (4). The evidence for step 1 is presented in
this report. While no evidence currently supports the duplication event depicted in step 2, that shown in
step 3 is well established (see 10 and references cited therein).
TMSs although some have 14 due to insertion
of the central loop (between TMSs 6 and 7 in
the 12 TMS proteins) into the membrane.
Further, a 6 TMS MFS homologue is encoded
within the genome of Bacillus subtilis (YitZ;
19). One family within the MFS, the sugar
?2.A.1.20), which like other MFS families,
includes members with two homologous
halves, seems to be primarily concerned with
sugar efflux (7, 8). By contrast, the prevalent 6
TMS K+ channels of the VIC superfamily
arose by fusion of a 4 TMS-encoding genetic
element (the voltage sensor domain) to a 2
TMS-encoding element (the channel domain)
(seeFig. 1; 9). Thelatter 2 TMSs separated by
the P-loop comprise the common element of
all members of the VIC superfamily. Gene
fusion and gene duplication events have given
rise to members of the VIC superfamily with
2, 4, 6, 8, 12, and 24 TMSs per polypeptide
chain (16). However, in all cases the basic
structure is equivalent to a simple tetramer
Recently, the first voltage-gated Na+chan-
nel in a prokaryote was identified and charac-
terized (11). This 6 TMS protein, NaChBac
from Bacillus halodurans, most resembles in
sequence animal Ca?+ channels and is in-
hibited by classical calcium blockers (11).
However, it has the 6 TMS topology that is
characteristic of K+ channels of the VIC
superfamily. While investigating NaChBac,
we were surprised to find that the channel-
forming domain of this bacterial protein
exhibits striking sequence similarity to the C-
terminal regions of well-characterized pmf-
driven sugar efflux carriers of E. coli, SetA,
SetB and SetC (7, 8). The degree of sequence
similarity between NaChBac of the VIC
superfamily and the Set paralogues of the
MFS is sufficient to strongly suggest a com-
mon evolutionary origin.
Fig. 2 shows a sequence alignment of the
C-terminal parts of NaChBac and SetCEco.
This alignment of 109 positions exhibited
Microbiology 148, December 20023761
34% identity, 44% similarity and a com-
parison score using the GAP program (500
random shuffles) of 12?8 (3). The prob-
ability that this degree of sequence similarity
10?? (2, 13). As shown in Fig. 2, putative TMS
12 in SetC (corresponding to the established
12th TMS in many MFS permeases) corre-
sponds in position to TMS 6 in NaChBac.
The hydrophilic C-terminal tails of NaChBac
and SetC are cytoplasmic, and both TMSs
thus have the same orientation in the mem-
brane (out to in). TMS 11 in the aligned
sequence of SetC overlaps the residues in
NaChBac that comprise the P-loop (Fig. 2).
While the fairly hydrophobic P-loop dips into
the membrane and re-emerges on the same
side of the membrane in all VIC family
members, this region is erroneously predicted
tobeaTMSusing severaltopology predicting
computer programs. TMS 11 in SetC, on the
the established topologies of many MFS
permeases (see 10, 18 for reviews). TMS 5 in
NaChBac (11) corresponds in position pre-
cisely to the predicted TMS 10 in SetC. TMS
5 in NaChBac is of opposite orientation in the
membrane compared to TMS 10 of the MFS
permeases. It is interesting to note that the
regions exhibiting sequence similarity be-
tween NaChBac and SetCEco are among the
families. Intragenic duplication of primordial
genes encoding membrane proteins with odd
numbers of TMSs is known to give rise to
homologous integral membrane domains ex-
(6, 13, 20).
Multiple alignments of several SET family
proteins of the MFS with several prokaryotic
the residues conserved between SetCEco and
NaChBac (Fig. 2) are many that are well-
conserved between both families (data not
shown). These observations strengthen the
significance of the sequence similarity shown
in Fig. 2. They suggest that the evolutionary
pathway for the appearance of MFS carriers
from a primordial 2 TMS VIC family channel
occurred via the pathway shown in Fig. 3. In
the proposed pathway, the 2 TMS channel-
a primordial 3 TMS element that first under-
went an intragenic duplication event to give a
6 TMS protein with the two homologous
halves of opposite orientation in the mem-
brane, followed by a later intragenic dupli-
cation event to give a 12 TMS protein with
these two homologous halves of the same
orientation inthe membrane.Thus,the origin
of the 6 TMS topology for VIC family
members, shown in Fig. 1, which resulted
from the fusion of two dissimilar elements,
and the origin of the 6 TMS repeat unit in the
12 TMS MFS permeases is different. Only S5,
the P-loop and S6 are proposed to have been
used for construction of the MFS carriers.
The simple postulate presented in Fig. 3
should be subject to confirmation or refu-
tation by a combination of experimental and
This work was supported by NIH grant
assistance in the preparation of this manu-
Rikki N. Hvorup and Milton H. Saier, Jr
Division of Biology, University of California
at San Diego, La Jolla, CA 92093–0116, USA
Author for correspondence: Milton H. Saier,
Jr. Tel: ?1 858 534 4084. Fax: ?1 858 534
1. Dassa, E. & Bouige, P. (2001). The ABC of ABCs: a
phylogenetic and functional classification of ABC
systems in living organisms. Res Microbiol 152, 211–229.
2. Dayhoff, M. O., Barker, W. C. & Hunt, L. T. (1983).
Establishing homologies in protein sequences. Methods
Enzymol 91, 524–545.
3. Devereux, J., Haeberli, P. & Smithies, O. (1984). A
comprehensive set of sequence analysis programs for the
VAX. Nucleic Acids Res 12, 387–395.
4. Doyle, D. A., Morais Cabral, J., Pfuetzner, R. A.,
Kuo, A., Gulbis, J. M., Cohen, S. L., Chait, B. T. &
MacKinnon, R. (1998). The structure of the potassium
channel: molecular basis of K+conduction and
selectivity. Science 280, 69–77.
5. Ennahar, S., Sashihara, T., Sonomoto, K. & Ishizaki,
A. (2000). Class IIa bacteriocins: biosynthesis, structure
and activity. FEMS Microbiol Rev 24, 85–106.
6. Jack, D. L., Yang, N. M. & Saier, M. H., Jr. (2001).
The drug?metabolite transporter superfamily. Eur J
Biochem 268, 3620–3639.
7. Liu, J. Y., Miller, P. F., Gosink, M. & Olson, E. R.
(1999). The identification of a new family of sugar efflux
pumps in Escherichia coli. Mol Microbiol 31, 1845–1851.
8. Liu, J. Y., Miller, P. F., Willard, J. & Olson, E. R.
(1999). Functional and biochemical characterization of
Escherichia coli sugar efflux transporters. J Biol Chem
9. Nelson, R. D., Kuan, G., Saier, M. H., Jr & Montal,
M. (1999). Modular assembly of voltage-gated channel
proteins: a sequence analysis and phylogenetic study.
J Mol Microbiol Biotechnol 1, 281–287.
10. Pao, S. S., Paulsen, I. T. & Saier, M. H., Jr (1998).
The major facilitator superfamily. Microbiol Mol Biol
Rev 62, 1–32.
11. Ren, D., Navarro, B., Xu, H., Yue, L., Shi, Q. &
Clapham, D. E. (2001). A prokaryotic voltage-gated
sodium channel. Science 294, 2372–2375.
12. Riley, M. A. (1998). Molecular mechanisms of
bacteriocin evolution. Annu Rev Genet 32, 255–278.
13. Saier, M. H., Jr (1994). Computer-aided analyses of
transport protein sequences: gleaning evidence
concerning function, structure, biogenesis, and evolution.
Microbiol Rev 58, 71–93.
14. Saier, M. H., Jr (2000a). Vectorial metabolism and
the evolution of transport systems. J Bacteriol 182,
15. Saier, M. H., Jr (2000b). A functional-phylogenetic
classification system for transmembrane solute
transporters. Microbiol Mol Biol Rev 64, 354–411.
16. Saier, M. H., Jr (2001). Evolution of transport
proteins. In Genetic Engineering: Principles and
Methods, vol. 23, pp. 1–10. Edited by J. K. Setlow. New
York: Kluwer Academic?Plenum.
17. Saier, M. H., Jr & Tseng, T.-T. (1999). Evolutionary
origins of transmembrane transport systems. In
Transport of Molecules Across Microbial Membranes
(Society for General Microbiology Symposium no. 58),
pp. 252–274. Edited by J. K. Broome-Smith, S.
Baumberg, C. J. Stirling and F. B. Ward. Cambridge:
Cambridge University Press.
18. Saier, M. H., Jr, Beatty, J. T., Goffeau, A. & 11
other authors (1999). The major facilitator superfamily.
J Mol Microbiol Biotechnol 1, 257–279.
19. Saier, M. H., Jr, Goldman, S. R., Maile, R. R.,
Moreno, M. S., Weyler, W., Yang, N. & Paulsen, I. T.
(2002). Transport capabilities encoded within the
Bacillus subtilis genome. J Mol Microbiol Biotechnol 4,
20. von Heijne, G. (1999). Recent advances in the
understanding of membrane protein assembly and
structure. Q Rev Biophys 32, 285–307.
Quorum sensing in
context: out of molecular
biology and into microbial
has undergone something of a renaissance
over the past decade. In part, this can be
attributed to the appeal of the ‘quorum
sensing’ hypothesis, which suggests that the
selective advantage driving the establishment
and conservation of intercellular signalling
mechanisms (ISMs) rests in the ability of
intercellular signals to reflect cell density (5).
Largely, however, the explosion in published
the expression of virulence factors in a hand-
ful of agriculturally or medically significant
synthesis of, and response to, diffusible ‘sig-
nalling’ metabolites. At the centre of all this
research attention is the gene expression
lactone (AHL), which generically involves the
action of an AHL synthase protein (LuxI
homologue) and an AHL-responsive tran-
scriptional activator protein (LuxR homol-
ogue). The accumulation of AHL produced
by LuxI homologues causes LuxR homo-
logues to direct the transcription of specific
‘structural’ genes. Examples of structural
genes regulated in this way include those
encoding elastase production in the oppor-
tunistic human pathogen Pseudomonas aeru-
ginosa and cellulase production in the plant
pathogen Erwinia carotovora (5).
expression has largely arisen from the po-
tential opportunities it offers to control un-
desirable bacterial activities, many questions
relating to the molecular biology and bio-
chemistry of the mechanism have been
answered. For example, we now know the
substrates required by LuxI homologues to
synthesize AHLs and that AHLs interact
directly with the N-terminal domain of LuxR
homologues, thereby inducing multimeric
changes that culminate in DNA binding and
recruitment of RNA polymerase to specific
promoters with dyad symmetry (5). In con-
trast, fundamental questions pertaining to the
ecology and evolution of AHL-mediated gene
Microbiology 148, December 20023762
Table 1. Phylogenetic association of known AHL systems
ClassOrderNo. families with
No. genera with
No. species with
All orders (6)
All orders (6)
All orders (14)
All orders (7)
All orders (1)
All orders (34)
*Note that lists of bacterial species with AHL systems are notoriously difficult to keep up-to-date.
For example, only one hypothesis (quorum
sensing) describing the selective advantage of
AHL-mediated gene expression is widely
known (3). Surprisingly, the quorum sensing
hypothesis has never been tested in the en-
vironment. Indeed, the role of the AHL ISM
in mediating interactions between bacteria
and their biotic and abiotic environment
globally is poorly understood, if ever con-
As a case in point, we are still unaware of
the taxonomic and environmental distribu-
tion of AHL-mediated gene expression sys-
tems amongst bacteria. Statements suggesting
that AHL-mediated gene expression is a
widespread phenomenon in Gram-negative
bacteria often appear in reviews (4) or intro-
ductions on the subject (and certainly in grant
applications) and yet a simple detailed survey
of its distribution amongst bacterial taxa has
never appeared in the literature.
A cursory glance at the distribution of
known AHL-producing bacterial taxa (Table
1) listed in Bergey’s Manual of Systematic
Bacteriology (1) reveals that they are con-
strained to three out of five classes (alpha,
bacterial phyla (Proteobacteria). This is gen-
erally appreciated. What you don’t often hear
is that within the alpha Proteobacteria only
7 out of 133 (5?3%) genera harbour known
AHL-producing species, within the beta Pro-
teobacteria only 4 out of 56 (7?1%) genera
harbour known AHL-producing species and
within the gamma Proteobacteria only 10 out
of 180 (5?5%) genera harbour known AHL-
(2?2 % of the total number of bacterial
genera) contain species known to produce
AHLs. At the species level the percentage
of AHL producers drops to a fraction of a
We believe this reflects two things. The
first is that a comprehensive survey of bac-
terial species for AHL production has never
been conducted. The beta Proteobacteria
seem to be suffering especial neglect in this
regard. Even so, because no AHL-producing
species have ever been found outside of the
Proteobacteria (leading us to assume they will
not be found), the true percentage of AHL-
producing genera is not likely to deviate far
fromthatbased onour currentyetincomplete
dataset.Note thatthis assumption ignores the
possibility that a large pool of anaerobic or
uncultured AHL producers or AHL-respon-
sive non-producers (such as Escherichia coli)
is lurking undetected by conventional bio-
assays. Secondly, the above inspection reflects
Microbiology 148, December 20023763
the fact that AHL production is restricted to a
limited number of Proteobacteria, let alone
Gram-negative bacteria, and should therefore
not be considered widespread. Phylogenetic
analyses of LuxR and LuxI homologues
suggest that these genes were present in the
last common ancestor of the Proteobacteria
(2) and have therefore been discarded by the
vast majority of descendants over evolution-
ary time. These assertions foster critical
appraisal of the ecological significance of
AHL-mediated gene expression.
Appraising the taxonomic distribution of
AHL-mediated gene expression is only the
first step in assessing the significance of this
mechanism both as a survival strategy and in
maintaining ecosystem function. We must
also measure the abundance and distribution
of the relevant taxonomic groups in the
environment and decipher through in situ
experiments what ecological interactions are
mediated by the AHL ISM. Finally, there is a
need to establish whether or not all the other
proposed bacterial intercellular signalling
molecules (AI-2, PAME, peptides, butyro-
lactones, quinolones, cyclic dipeptides) serve
an ecological function analogous to that of
AHLs. The widespread use of the term
quorum sensing suggests that many re-
searchers assume all ISMs do serve the same
function. We believe that hypotheses de-
scribing the ecological function of these sig-
nalling systems rarely receive critical ap-
praisal and even less frequently are tested
experimentally. This serves as a final illus-
tration of the need for the field to appraise
AHL-mediated gene expression, and indeed
intercellular signalling generally from an eco-
logical and evolutionary perspective.
Mike Manefield and Sarah L. Turner
CEH Oxford, Mansfield Road, Oxford
OX1 3SR, UK
Author for correspondence: Mike
Manefield. Tel: ?44 1865 281 630. Fax: ?44
1865 281 696.
1. Garrity, G. M., Winters, M. & Searles. D. B. (2001).
Bergey’s Manual of Systematic Bacteriology, 2nd edn.
New York: Springer.
2. Gray, K. M. & Garey J. R. (2001). The evolution of
bacterial LuxI and LuxR quorum sensing regulators.
Microbiology 147, 2379–2387.
3. Redfield, R. (2002). Is quorum sensing a side effect of
diffusion sensing? Trends Microbiol 10, 365.
4. Swift, S., Williams, P. & Stewart, G. S. A. B. (1997).
N-acylhomoserine lactones and quorum sensing in
Proteobacteria. In Cell-Cell Signalling in Bacteria,
pp. 291–313. Edited by G. M. Dunny & S. C. Winans.
Washington, DC: American Society for Microbiology.
5. Whitehead, N. A., Barnard, A. M. L., Slater, H.,
Simpson, N. J. L. & Salmond, G. P. C. (2001). Quorum-
sensing in Gram-negative bacteria. FEMS Microbiol Rev
A putative transcription
factor inducing mobility in
Mycoplasma genitalium and Mycoplasma
pneumoniae are among the smallest free-
living cells with only a few hundred genes.
Their genomes were among the first ones to
be completely sequenced (8, 6) and assumed
to comprise a ‘minimal gene set’ coding for a
‘minimal metabolism’ (10, 13). By compara-
tive genome analysis (9, 7, 5), for M. pneumo-
niae, for 458 [??67%; 349 (??51%)
before the recent re-annotation] of the 688
(667) ORFs, functional features have been
predicted. M. genitalium has a subset of the
M. pneumoniae genes and a conserved gene
order. M. pneumoniae and M. genitalium
show many abnormalities with respect to
cellular biology, biochemistry and gene ex-
pression. For example, no Rho factor is
known and only one σ factor was reported.
On one hand this is puzzling since, for
example, Bacillus subtilis has at least 14 σ
factors (2), on the other hand symbionts and
parasites seem to require a smaller number of
their own transcription factors (12).
Sigma factors induce the binding of core
RNA polymerase (CPB) to specific initiation
sites at the ?35 and ?10 regions and then
detach. Two groups, σ?? and σ??, are known.
σ?? factors group in essential primary, non-
essential primary and alternative σ?? factors
σ?? factors, regulate transcription of specific
subsets of genes. Sigma factors vary strongly
in length from around 150 to 450 aa. Variants
significant divergence and various domain
combinations occur (14, 17, 12). Promoter
sequences in M. pneumoniae are very het-
erogeneous and the ?35 region is poorly
conserved (16). This suggests that there are
more groups and thus more σ factors than
previously assumed. The lack of knowledge
it appears that the pathogenicity of M. pne-
umoniae is due to adaptation of its bio-
chemistry and gene expression in response to
external stimuli (1). Consequently, it is
conceivable that additional σ factors may
exist but have been overlooked by compu-
tational genome analysis so far. Therefore,
we decided to re-investigate the genome with
bioinformatics tools to look for possible
further σ factors. We describe here just the
most important results while methods, de-
tailed computations and further information
about identities and accession numbers of
individual proteins are described in the sup-
A search for a helix-turn-helix (HTH)
motif against the set of translated potential
ORFs from M. pneumoniae returned 22
significant hits. (Proteins shorter than 50
residues were not considered further.) The
resulting set contained the known σ factor A
(P78022). The remaining set was used for
Smith-Waterman and searches (1).
Protein P75169 [172 residues, see (1)] from
M. pneumoniae has a strong sequence simi-
larity to P47667 (171 residues) from M.
genitalium and to the σ factor D, P10726 (254
residues) from B. subtilis. The
alignment is shown with secondary structure
prediction for P75169 in Fig. 1. P75169 and
P10726 are identical at 46 residues (27% or
18% respectively) and have similarity at
another 68 residues (40%, 27%). Especially
in the putative RNA core polymerase binding
region (residues 56–76; box 1 in Fig. 1) and a
stretch of 50 residues comprising the DNA
binding HTH motif (aa 226–246; box 3) the
similarities are around 70%. By contrast, the
‘random similarity’ to an arbitrarily chosen
HTH motif (from P62678), is around 30%.
shows high homology and so does a stretch of
15 aa which is approximately in the other
region (aa 101–125; box 2) that is relevant for
core binding (17). Since the alignment pro-
duced virtually no insertions into the B.
subtilis σ factor, P75169 and P47667 may well
P10726 by loss of domains and thus may be
orthologues. Secondary structure prediction
complies well with the helical architecture of
σ?? factors (see Fig. 1). Only one large gap
occurs within a helix and none in either of the
two functional association domains (CPB and
To understand why such strong homo-
logies have not been recognized earlier we
carried out pairwise and profile-based data-
base searches. Results are summarized in the
supplement (1). In essence, it was possible in
direct and reverse searches to pull out all σ
factors correctly but no subgroups could be
specified despite using different seed align-
ments or levels of sensitivity. However, this
was true for several other σ?? factors as well.
This agrees with numerous claims about the
strong divergence in σ factors (11, 17, 12).
Therefore,it is likely thatso far the homology
has simply been overlooked.
Phylogenetic trees for the full sequences
were indecisive. Again, the groupings for the
functional domains (CPB and HTH) were
clearer. In particular, the HTH trees grouped
P75169 with flagellar σ?? factors. Since HTH
further supports the assumption that P75169
and P47667 are involved in the induction of
We tried to find further evidence for our
claim. Location on the genome also suggests
involvement in regulation. Other genes such
as the known σ?? factor P78022 and genes
[see (1) for more details]. Furthermore, the
gene appears to be essential since transposon
mutagenesis suggests that viable cells must
Microbiology 148, December 20023764
Microbiology Comment Download full-text
Fig. 1. Alignment of the suspected σ factors from M. pneumoniae (M.p.; (P75169) and M. genitalium (M.g.; P47667) with Sig D from
B. subtilis (B.s.; P10726). The first line (sec) shows the secondary structure prediction (h for helix, c for coiled region), the last one the
degree of conservation (*, completely conserved; :, almost conserved; ., amino acids of similar properties). The three boxes from N to C
terminus indicate the conserved regions. Box 1, potential core binding region; box 2, region involved in core binding; box 3, helix-turn-
helix region. See text for further details.
contain P75169 (10). The protein could not be
identified in a 2-D gel protein expression
study. This is not too surprising since tran-
scription factors often occur only in a small
number of copies in the cell and have a
relatively short retention time. Furthermore,
the protein may be expressed only under very
peculiar conditions which are to date largely
unknown. P75169 is poorly enhanced on
transcription level under standard growth
conditions (18). Putative high-scoring pro-
moter sequences can be found with the
method described in (16) at ?35 (TGCAAA)
and ?10 (TAAATT).
The B. subtilis homologue P10726 belongs
to the σ?? family group 3 (17), known to
influence expression of genes involved in
autolysis and flagellar-based motility. It bears
strong resemblance to the fliA gene from
Escherichia coli. M. pneumoniae has been
reported to show some mobility and ability to
detach (3, 4) but the mechanism of induction
andthe corresponding control elements are to
date unreported. Thus P75169 is a reasonable
candidate for these functions.
Annotation by computations is useful but
should be confirmed or disproved in the
laboratory (15). Further experiments should,
therefore, focus on the precise biochemical
role as suggested by our analysis, i.e. finding
experimental conditions under which this
gene is more strongly expressed, and cloning
and functional analysis, for example two-
hybrid analysis. We hope that experimental-
ists will find our suggestions useful and
conduct research to verify or falsify our
J.W. is supported through the Graduierten-
kolleg ‘Pathogene Mikroorganismen’, a DFG
grant (He 780?10-1) and the ‘Fonds der
Chemischen Industrie’. We thank R. Herr-
mann and D. Pollack for useful discussions,
and J. Ross and R. Gauges for computational
help. E.B.-B. is supported by an MRC in-
nowledges support by the KTF during the
early stages of the project. We thank C.
Zimmermann and H. Goehl for providing
Erich Bornberg-Bauer1and January Weiner
1School of Biological Sciences,
The University of Manchester,
2.205 Stopford Building, Oxford Road,
Manchester M13 9PT, UK
2ZMBH, University of Heidelberg, D69120
Author for correspondence: Erich Bornberg-
Bauer. Tel: ?44 161 275 7396. Fax: ?44 161
275 5082. e-mail: ebb?bioinf.man.ac.uk
1. Supplementary data. See http:??mic.sgmjournals.org
2. Bairoch, A. & Appweiler, R. (2000). The SWISS-
PROT protein sequence database and its supplement
TrEMBL in 2000. Nucleic Acids Res 28, 45–48.
3. Bredt, W. (1968). Motility and multiplication in
Mycoplasma pneumoniae: a phase contrast study. Pathol
Microbiol 32, 321–326.
4. Bredt, W. (1973). Motility of mycoplasmas. Ann N Y
Acad Sci 225, 246–259.
5. Dandekar, T., Huynen, M., Regula, J. T. & 10 other
authors (2000). Re-annotating the Mycoplasma
pneumoniae genome sequence: adding value, function
and reading frames. Nucleic Acids Res 28, 3278–3288.
6. Fraser, C., Gocayne, J. D., White, O & 22 other
authors (1995). The minimal genome complement of
Mycoplasma genitalium. Science 270, 397–403.
7. Herrmann, R. & Reiner, B. (1998). Mycoplasma
pneumoniae and Mycoplasma genitalium. Curr Opin
Microbiol 1, 572–578.
8. Himmelreich, R., Hilbert, H., Plagens, H., Pirlkl, E.,
Li, B. C. & Herrmann, R. (1996). Complete sequence
analysis of the genome of the bacterium Mycoplasma
pneumoniae. Nucleic Acids Res 24, 4420–4449.
9. Himmelreich, R., Plagens, H., Hilbert, H., Reiner, B.
& Herrmann, R. (1997). Comparative analysis of the
genomes of the bacteria Mycoplasma pneumoniae and
Mycoplasma genitalium. Nucleic Acids Res 25, 701–712.
10. Hutchison, C. A., Peterson, S. N., Gill, S. R., Cline,
R. T. White, O., Fraser, C. M., Smith, H. O. & Venter,
J. C. (1999). Global transposon mutagenesis and a
minimal Mycoplasma genome. Science 286, 2165–2169.
11. Lonetto, M., Gribskov, M. & Gross, C. A. (1992).
The sigma 70 family: sequence conservation and
evolutionary relationships. J Bacteriol 174, 3843–3849.
12. Mittenhuber, G. (2002), An inventory of genes
encoding RNA polymerase sigma factors in 31
completely sequenced eubacterial genomes. J Mol
Microbiol Biotechnol 4, 77–91.
13. Mushegian, A. R. & Koonin, E. V. (1996). A minimal
gene set for cellular life derived by comparison of
complete bacterial genomes. Proc Natl Acad Sci USA 93,
14. Pao, G. M. & Saier, M. H. (1995). Response
regulators of bacterial signal transduction systems:
selective domain shuffling during evolution. J Mol Evol
15. Pollack, J. P. (1997). Mycoplasma genes: a case for
reflective annotations. Trends Microbiol 5, 413–419.
16. Weiner, J., 3rd, Herrmann, R. & Browning, G. F.
(2000). Transcription in Mycoplasma pneumoniae.
Nucleic Acids Res 28, 4488–4496.
17. Wo ? sten, M. M. (1998). Eubacterial sigma-factors.
FEMS Microbiol Rev 22, 127–150.
18. Zimmermann, C. U., Go ? hlmann, H. W. H., Weiner,
J., 3rd & Herrmann, R. (in preparation). Transcription
profile of the bacterium Mycoplasma pneumoniae grown
at different temperatures.
Microbiology 148, December 20023765