|
| |
Journal of Bacteriology, January 2004, p . 267-269, Vol . 186,
No . 2
Digging
with Experimental Pick and Computational Shovel: a New Addition to the Histidine
Kinase Superfamily
Igor B . Zhulin*
School of Biology, Georgia Institute of Technology, Atlanta, Georgia
30332-0230
Experimental microbiology in the postgenomic era has survived the
first wave of uncertainty: a perspective of being replaced by in
silico microbiology . Although it is clear now that experimental
science is irreplaceable, we are facing yet another wave: the
frustration of having a half million microbial genes in databases but
lacking simple ways of getting biologically relevant information .
Therefore, it is reassuring to see some clear water between the
waves: examples of experimental research that takes advantage of a
comparative genomic approach one step at a time . The paper by Karniol
and Vierstra in this issue (10) describing a new
family of histidine kinases shows how the use of simple bioinformatics
tools by microbiologists can (i) identify new targets for experimental
work and (ii) provide important feedback for improving these
tools .
|
|
HISTIDINE KINASES: MAJOR ENVIRONMENTAL SENSORS IN
BACTERIA |
Environmental sensing by histidine kinases is a fundamental property
of a microbial cell . Extensive research on this topic during the last
fifteen years has been recently summarized in two books (7,
8), a major review (15), and dozens of
more-specialized reviews . Histidine kinases appear to be the major
class of environmental sensors in bacteria . A histidine kinase is a
perfect sensor because of its modular architecture . The sensing
capabilities lie within an input module (14),
which contains one or more sensory domains that detect various
physicochemical parameters . The input module communicates the
information to a transmitter module within the same protein (14),
which in turn sends the signal in the form of a phosphoryl group to
another protein, a cognate response regulator . The activated response
regulator triggers a cellular response, usually on the level of
transcription . Figure 1 shows a domain
representation of FixL from Bradyrhizobium japonicum (5),
a typical histidine kinase where the input module contains two
sensory PAS domains and the transmitter module contains dimerization
(HisKA) and ATP-binding (HATPase_c) domains .
|
FIG . 1 . Domain architecture of classical (FixL_BRAJA) and HWE (Sma2063)
histidine kinases derived by SMART (11) . The FixL
protein contains two sensory PAS domains (each consists of PAS and PAC
subdomains) in its N-terminal half and HisKA and HATPase_c domains in
the C terminus . Therefore, FixL is clearly predicted to be a sensor
histidine kinase, even if its function would not have been known . In
contrast, no known domains or motifs are predicted in the SMa2063
protein, except for a short low-complexity region (a purple rectangle)
in the central portion of the protein . The study by Karniol and Vierstra
(10) not only predicts that the protein of unknown
function Sma2063 is a histidine kinase but also proves it
experimentally.
|
|
|
|
HISTIDINE KINASES: EASY TARGETS IN MICROBIAL GENOME
ANNOTATION |
Histidine kinases are easy targets for genome annotators because of
the significant sequence conservation within the dimerization and
ATP-binding domains . Profile hidden Markov models (HMMs) were
designed for these domains, enabling their rapid detection and
visualization in protein sequences in two primary domain databases,
Pfam (2) and SMART (11) . Recent
implementation of Pfam and SMART domains into the conserved domain
database (12) and InterPro (13)
tools results in identification of histidine kinases in any routine
similarity search of primary protein sequence databases, nr-NCBI and
SWISS-PROT . Using SMART and Pfam domain models and conventional BLAST
(1) searches, hundreds of histidine kinases have
been identified in completely sequenced prokaryotic genomes (6) .
I (as well as many other experimental and computational scientists)
was under impression that if you have a protein sequence, you will
know in a few seconds whether or not it is a histidine kinase .
|
|
THE HWE HISTIDINE KINASE FAMILY: A DIFFICULT CASE
|
The paper by Karniol and Vierstra (10) describes a new family
of histidine kinases exemplified by a well-studied protein, the
BphP2 light sensor histidine kinase from Agrobacterium tumefaciens
(9) . The family is named HWE after uniquely conserved
histidine, tryptophan, and glutamate residues . The family consists of
dozens of homologs, and many of them have no obvious characteristics
of histidine kinases . Figure 1 shows that scanning a
protein SMa2063 from Sinorhizobium meliloti, a member of the
newly identified HWE family (10), against the
SMART database results in a prediction of "protein of unknown
function" because no known domain can be detected . SMART is a
professionally curated domain database specialized in signal
transduction; therefore, its opinion regarding this protein can be
viewed as an expert one . For example, SMART easily detects all
domains, not only those that are well conserved, such as HATPase_c,
but also those that are poorly conserved, such as HisKA and PAS, in
an experimentally characterized histidine kinase FixL (Fig.
1) . Searching the Pfam database with the SMa2063
sequence results in ambiguous results, where the HATPase_c domain is
detected at the borderline of statistical significance and is
overlapped with another unrelated domain predicted with a similar
unreliable statistical score . Both SMART and Pfam utilize the HMMer
program for domain detection (3) . Changing the program
to reverse-position-specific BLAST (1), a domain search
tool implemented in the conserved domain database (13),
does not improve the situation: no conserved SMART or Pfam domains
can be detected in SMa2063 . Thus, it comes as no surprise that
although 37 histidine kinases were annotated in the completely
sequenced genome of S . meliloti (4), the
SMa2063 protein was not among them; it received the familiar label of
"hypothetical protein."
Nevertheless, Karniol and Vierstra (10) were able to
convincingly predict that SMa2063 is a histidine kinase . This has
been done by detecting the SMa2063 protein in BLASTP searches (1)
initiated with the ATP-binding domain of the BphP2 light sensor
histidine kinase from A . tumefaciens (9)
followed by a thorough analysis of its alignment with homologous
domains . Special attention was paid to conserved motifs that have a
functional role in the histidine kinase activity . The reader is
referred to the paper for details of this analysis and experimental
results demonstrating that SMa2063 and other proteins similar to
BphP2 are indeed histidine kinases .
The importance of the results obtained by Karniol and Vierstra for
microbial signal transduction is obvious . The new family of histidine
kinases, which includes many previously unrecognizable members, is a
big step forward . These sensor molecules initiate important
regulatory cascades, and their discovery will facilitate experimental
research aimed at understanding cellular properties that are
controlled by two-component systems in a given microbial species . New
findings also prompt new questions for experimental research . How
significant is the deviation of the kinase module in terms of
structure and function? Is there any distinct feature in cognate
response regulators, etc.? The impact of this finding on
bioinformatics is less obvious, but it is as important . It shows that
current domain detecting tools need significant improvement when it
comes to signal transduction proteins . The failure of SMART and Pfam
to recognize a version of a conserved domain, such as HATPase_c,
clearly calls for adjustments to current HMMs . New (for the HWE
family) and improved (for the entire superfamily) models will ensure
better automated detection of histidine kinases in all available and
newly sequenced genomes .
|
|
HOW TO STORE KNOWLEDGE IN THE SILICON AGE
|
The main difference between biological research in the pre- and
postgenomic eras is in the numbers . Traditionally, an exciting
finding of a novel function for a given protein meant that it was for
this protein only . Nowadays, this finding can be extrapolated to
numerous homologs that have never been (and most of them will never
be) in the hands of experimentalists and exist mainly in the virtual
realm of databases . It is vitally important for the future of biology
to make sure that such extrapolation is (i) applied and (ii) applied
correctly . The second point is a subject of serious discussions and
debates; however, even the first one is not very clear . Postgenomic
biology is experiencing a normal disease of growth, where
experimental and genomic information exist in parallel, rarely
interacting worlds (Fig . 2) . The scientific
community has a traditional way of learning from reports published in
peer-reviewed journals . Searching genomic databases with a BLAST
program (1) became the second way of learning for many
biologists (the BLAST paper has been cited
 10,000
times, and many more publications refer to BLAST without citation) .
However, the real problem is that much (if not most) biological
information in the databases is not only poorly peer reviewed but
also annotated by people who cannot possibly be experts in all areas
of biological research or are not biologists at all . Experimental
scientists have little control over this process . For example, how
will the important finding reported by Karniol and Vierstra (10)
make it into the primary databases? One way is that curators at
the National Center for Biotechnology Information, Swiss-Prot, Pfam,
and SMART will find the paper and take time to connect each relevant
record in the database to their publication (current automated tools
usually fail in doing so) or convert printed alignments into models
and make appropriate descriptions for domain databases . What if they
miss the paper or are not willing to deal with the alignments from
scratch (currently, most curators will ask for a ready-to-go
alignment file)? Well, then SMa2063 and dozens of other proteins in
these databases will remain hypothetical . There is a way out of this
situation . For example, authors who wish to publish a new protein
family can be asked to submit their alignments and descriptions that
they feel are appropriate to primary domain databases if the paper is
accepted for publication . This will ensure that the new finding,
which has been peer reviewed, finds its place in the database and
that its description will be provided by experimentalists themselves
and not by database curators .
|
FIG . 2 . Relationship between experimental and genomic information in the
area of signal transduction . The current status of the information flow
is shown by solid lines . The scientific community gets information by
reading the results of experimental work and searching genome sequence
databases for similarities . Future developments are shown by dashed
lines . Computational biologists can retrieve relevant information from
primary genomic databases and carry out more-specific annotation than
that given in the primary database . This can be stored in a separate
database (knowledge environment) accessible to all scientists interested
in signal transduction . They can carry out online annotation of proteins
they study, link them to the appropriate peer-reviewed publications, and
communicate their suggestions on database improvements directly to the
database developers . Thus, information will be presented in a form which
suits the scientific community the best.
|
|
Another attempt to bridge experimental and genomic information is the
development of specialized knowledge databases rather than sequence
databases . At this time, most of such databases are organism
oriented . For example, EcoCyc (http://ecocyc.org/)
and CYORF (http://cyano.genome.ad.jp/)
serve the Escherichia coli and cyanobacterial research
communities, respectively . There is an urgent need, however, for the
creation of databases of functions that span many different
organisms . Figure 2 illustrates the idea of such an
interactive knowledge environment for those interested in signal
transduction . Having such databases would ensure that all relevant
biological discoveries and their extrapolation on the genomic data
are stored, managed, and available to the scientific community in a
peer-reviewed, user-friendly form . Wouldn't it be nice to do less
BLASTing and more reading?
* Mailing address: School of Biology, Georgia Institute of
Technology, 310 Ferst Dr., Atlanta, GA 30332-0230 . Phone: (404) 385-2224 . Fax:
(404) 894-0519 . E-mail:
igor.zhulin@biology.gatech.edu .
The views expressed in this Commentary do not necessarily
reflect the views of the journal or of ASM.
- Altschul, S . F., T . L . Madden, A . A . Schäffer, J . Zhang, Z .
Zhang, W . Miller, and D . J . Lipman. 1997 . Gapped BLAST and PSI-BLAST: a
new generation of protein database search programs . Nucleic Acids Res . 25:3389-3402 .
- Bateman, A., E . Birney, L . Cerruti, R . Durbin, L . Etwiller,
S . R . Eddy, S . Griffiths-Jones, K . L . Howe, M . Marshall, and E . L . Sonnhammer.
2002 . The Pfam protein families database . Nucleic Acids Res . 30:276-280 .
- Eddy, S . R. 1998 . Profile hidden Markov models .
Bioinformatics 14:755-763.
- Galibert, F., T . M . Finan, S . R . Long, A . Puhler, P . Abola,
F . Ampe, F . Barloy-Hubler, M . J . Barnett, A . Becker, P . Boistard, G . Bothe, M .
Boutry, L . Bowser, J . Buhrmester, E . Cadieu, D . Capela, P . Chain, A . Cowie, R .
W . Davis, S . Dreano, N . A . Federspiel, R . F . Fisher, S . Gloux, T . Godrie, A .
Goffeau, B . Golding, J . Gouzy, M . Gurjal, I . Hernandez-Lucas, A . Hong, L .
Huizar, R . W . Hyman, T . Jones, D . Kahn, M . L . Kahn, S . Kalman, D . H . Keating,
E . Kiss, C . Komp, V . Lelaure, D . Masuy, C . Palm, M . C . Peck, T . M . Pohl, D .
Portetelle, B . Purnelle, U . Ramsperger, R . Surzycki, P . Thebault, M .
Vandenbol, F . J . Vorholter, S . Weidner, D . H . Wells, K . Wong, K . C . Yeh, and
J . Batut. The composite genome of the legume symbiont Sinorhizobium
meliloti. Science 293:668-672.
- Gilles-Gonzalez, M . A. 2001 . Oxygen signal transduction .
IUBMB Life 51:165-173.
- Grebe, T . W., and J . B . Stock. 1999 . The histidine
protein kinase superfamily . Adv . Microb . Physiol . 41:139-227.
- Hoch, J . A., and T . J . Silhavy (ed.). 1995 . Two-component
signal transduction . ASM Press, Washington, D.C.
- Inouye, M., and R . Dutta (ed.). 2003 . Histidine kinases
in signal transduction . Academic Press, New York, N.Y.
- Karniol, B., and R . D . Vierstra. 2003 . The pair of
bacteriophytochromes from Agrobacterium tumefaciens are histidine
kinases with opposing photobiological properties . Proc . Natl . Acad . Sci . USA
100:2807-2812 .
- Karniol, B., and R . D . Vierstra. 2004 . The HWE histidine
kinases, a new family of two-component sensor kinases with potentially diverse
roles in environmental signaling . J . Bacteriol., 445-452.
- Letunic, I., L . Goodstadt, N . J . Dickens, T . Doerks, J .
Schultz, R . Mott, F . Ciccarelli, R . R . Copley, C . P . Ponting, and P . Bork.
2002 . Recent improvements to the SMART domain-based sequence annotation
resource . Nucleic Acids Res . 30:242-244 .
- Marchler-Bauer, A., J . B . Anderson, C . DeWeese-Scott, N . D .
Fedorova, L . Y . Geer, S . He, D . I . Hurwitz, J . D . Jackson, A . R . Jacobs, C . J .
Lanczycki, C . A . Liebert, C . Liu, T . Madej, G . H . Marchler, R . Mazumder, A . N .
Nikolskaya, A . R . Panchenko, B . S . Rao, B . A . Shoemaker, V . Simonyan, J . S .
Song, P . A . Thiessen, S . Vasudevan, Y . Wang, R . A . Yamashita, J . J . Yin, and
S . H . Bryant. 2003 . CDD: a curated Entrez database of conserved domain
alignments . Nucleic Acids Res . 31:383-387 .
- Mulder, N . J., R . Apweiler, T . K . Attwood, A . Bairoch, D .
Barrell, A . Bateman, D . Binns, M . Biswas, P . Bradley, P . Bork, P . Bucher, R .
R . Copley, E . Courcelle, U . Das, R . Durbin, L . Falquet, W . Fleischmann, S .
Griffiths-Jones, D . Haft, N . Harte, N . Hulo, D . Kahn, A . Kanapin, M .
Krestyaninova, R . Lopez, I . Letunic, D . Lonsdale, V . Silventoinen, S . E .
Orchard, M . Pagni, D . Peyruc, C . P . Ponting, J . D . Selengut, F . Servant, C . J .
Sigrist, R . Vaughan, and E . M . Zdobnov. 2003 . The InterPro database, 2003
brings increased coverage and new features . Nucleic Acids Res . 31:315-318 .
- Parkinson, J . S., and E . C . Kofoid. 1992 . Communication
modules in bacterial signaling proteins . Annu . Rev . Genet . 26:71-112.
- Stock, A . M., V . L . Robinson, and P . N . Goudreau. 2000 .
Two-component signal transduction . Annu . Rev . Biochem . 69:183-215.
Free Online Full-text Article
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
|
