Sequences and evolutionary analyses of eukaryotic-type protein kinases from Streptomyces coelicolor A3(2)

Hiroshi Ogawara1, Narumi Aoyagi1, Mami Watanabe1 and Hiroaki Urabe1

Department of Biochemistry, Meiji Pharmaceutical University, Noshio-2, Kiyose, Tokyo 204-8588, Japan1

Author for correspondence: Hiroshi Ogawara. Tel: +81 424 95 8474. Fax: +81 424 95 8474. e-mail: hogawara{at}my-pharm.ac.jp


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Four eukaryotic-type protein serine/threonine kinases from Streptomyces coelicolor A3(2) were cloned and sequenced. To explore evolutionary relationships between these and other protein kinases, the distribution of protein serine/threonine kinase genes in prokaryotes was examined with the TFASTA program. Genes of this type were detected in only a few species of prokaryotes and their distribution was uneven; Streptomyces, Mycobacterium, Synechocystis and Myxococcus each contained more than three such genes. Homology analyses by GAP and Rdf2 programs suggested that some kinases from one species were closely related, whilst others were only remotely related. This was confirmed by examining phylogenetic trees constructed by the neighbour-joining and other methods. For each species, analysis of the coding regions indicated that the G+C content of protein kinase genes was similar to that of other genes. Considered with the fact that in phylogenetic trees the amino acid sequences of STPK from Aquifex aeolicus and some other eukaryotic-type protein kinases in prokaryotes form a cluster with protein kinases from eukaryotes, this suggests that the eukaryotic-type protein kinases were present originally in both prokaryotes and eukaryotes, but that most of these genes have been lost during the evolutionary process in prokaryotes because they are not needed. This conclusion is supported by the observation that the prokaryotes retaining several of these kinases undergo complicated morphological and/or biochemical differentiation.

Keywords: eukaryotic-type protein kinase, Streptomyces, protein serine/threonine kinase, evolution, phylogenetic tree

Abbreviations: cAMPPK, cAMP-dependent protein kinase; CaMPK, calmodulin-dependent protein kinase; S6-2PK, S6-2 protein kinase

The GenBank accession numbers for the sequences determined in this work are AB016932, AB018799, AB019394 and AB021679.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Protein phosphorylation is one of the most important regulatory mechanisms in the signal transduction systems of eukaryotes as well as prokaryotes. However, the main phosphorylation sites in prokaryotes are on histidine and aspartic acid residues (Parkinson & Kofoid, 1992 ), whereas in eukaryotes they are on serine, threonine and/or tyrosine (Hunter & Cooper, 1985 ; Hanks et al., 1988 ). Numerous eukaryotic protein kinase genes have been cloned. Although these kinases are classified into several groups on the basis of their amino acid sequences, a series of 11 subdomains (I–XI) have been conserved (http://www.sdsc.edu/kinases/). Recently, eukaryotic-type protein kinases have been detected in a small number of prokaryotes, including Streptomyces (Matsumoto et al., 1994 ; Urabe & Ogawara, 1995 ; Hirakata et al., 1998 ), Myxococcus (Munoz-Dorado et al., 1991 ; Zhang et al., 1992 ) and Mycobacterium (Fsihi et al., 1996 ; Cole et al., 1998 ).

Since complete nucleotide sequences of the chromosomal DNA of over 10 bacterial species have now been determined (Doolittle, 1998 ), it would be interesting to know whether these contain eukaryotic-type protein kinases, and if so, how such kinases are distributed. This paper describes the distribution of eukaryotic-type protein kinases in prokaryotes and analyses evolutionary relationships among these kinases in prokaryotes and eukaryotes. It also reports the nucleotide sequences of four newly isolated protein kinase genes from Streptomyces coelicolor A3(2).


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and growth conditions.
Escherichia coli K-12 JM109 was used as a host for pBluescript II SK(+) and its derivatives. E. coli and pUC119 were purchased from Takara Shuzo. E. coli was cultivated in 2xYT medium (Sambrook et al., 1989 ). S. coelicolor A3(2) M145 was used as a DNA donor. It was cultured in liquid YEME medium (Hopwood et al., 1985 ).

DNA manipulations.
Preparation of plasmid DNA and in vitro DNA manipulations were carried out as described by Sambrook et al. (1989) . Enzymes for DNA manipulations were obtained from Takara Shuzo.

Nucleotide sequence determination.
Nucleotide sequences were determined by the method of Sanger et al. (1977) with the Thermosequenase II kit (Amersham Pharmacia Biotech).

Search for protein kinases.
Eukaryotic-type protein kinases were investigated by two methods. First, the word ‘kinase’ was searched for in the GenBank data available for the genomes of 16 species (Table 1) for which nucleotide sequences have already been determined (http://www.genome.ad.jp/kegg/java/org-list.html). The amino acid sequences of the kinases detected were compared to determine whether they really contained eukaryotic-type protein kinase domains. Second, the 16 nucleotide sequences and the amino acid sequences of subdomains I–IX of PkaA, PkaB, Pkn2 and cAMP-dependent protein kinase (cAMPPK) from Saccharomyces cerevisiae, calmodulin-dependent protein kinase (CaMPK) from Mus musculus and S6-2 protein kinase (S6-2PK) from Homo sapiens were inspected thoroughly by the TFASTA program for the presence of protein kinases (Pearson & Lipman, 1988 ). The protein kinases investigated in this paper are listed in Table 2, which includes some protein kinases other than those detected by the above procedure.


View this table:
[in this window]
[in a new window]
 
Table 1. Distribution of eukaryotic-type protein kinases in bacteria

 

View this table:
[in this window]
[in a new window]
 
Table 2. Protein kinases used in this work

 
Homology analyses.
GAP (Needleman & Wunch, 1970 ; Lipman & Pearson, 1985 ) and Rdf2 (Pearson & Lipman, 1988 ) were used to compare pairs of proteins. FASTA (Pearson & Lipman, 1988 ) was used for comparisons of more than three proteins.

Phylogenetic trees.
Phylogenetic trees were constructed mainly by a neighbour-joining method (CLUSTAL W; Thompson et al., 1994 ). MOLPHY (a maximum-likelihood method; Adachi & Hasegawa, 1992 ), PHYLIP (a maximum-parsimony method; Felsenstein, 1995 ) and ODEN (a neighbour-joining method; Ina, 1991 ) were also used.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Nucleotide sequences
In a previous paper, we reported the cloning of seven DNA fragments containing putative protein serine/threonine kinase genes from S. coelicolor A3(2) (Hirakata et al., 1998 ). The nucleotide sequences of four of them, pkaD, pkaE, pkaF and pkaG, were determined in this work and are available in DDBJ/GenBank/EMBL, accession numbers AB016932, AB018799, AB019394 and AB021679, respectively. Preceding the translation start sites of these genes, Shine–Dalgarno-like sequences were detected (GCGGGG, AGGGCG, ACGCGG and GGGAGG, respectively). Each of the kinases contained the 11 protein serine/threonine kinase subdomains I–XI predicted for eukaryotic-type enzymes. Kinase PkaD, corresponding to Sch5 in the previous paper, consisted of 598 amino acids; PkaE, corresponding to Sch1, consisted of 487 amino acids; PkaF, corresponding to Sch3, consisted of 667 amino acids; and PkaG, corresponding to Sch4, consisted of 592 amino acids. At least two of these enzymes, PkaE and PkaG, were shown to have protein serine/threonine and/or tyrosine kinase activity (unpublished data).

Distribution of eukaryotic-type protein kinases in bacteria
Investigation of the protein kinase sequences in the complete nucleotide and/or amino acid sequences of 16 bacterial genomes determined at the present time detected the eukaryotic-type enzyme in only a few species; however, in some species, there were more than three protein kinases present (Tables 1 and 2). Among the proteobacteria, YjbH (accession no. U00096) in E. coli is reported to have a protein kinase ATP-binding signature. However, none of the subdomains of protein kinases could be detected, so we conclude that there is no eukaryotic-type protein kinase gene in E. coli. From similar inspections, Haemophilus influenzae does not possess any protein kinase, and the one (HPO432) in Helicobacter pylori is of the eukaryotic-type.

Myxococcus xanthus, a Gram-negative soil bacterium, has more than five protein kinases (Zhang et al., 1996 ), and among Gram-positive bacteria, Bacillus subtilis has two (YloP and YbdM), whilst Mycoplasma genitalium and Mycoplasma pneumoniae each have one (MG109 and ORF389, respectively). However, within the high G+C group of Gram-positive bacteria, Mycobacterium tuberculosis and S. coelicolor A3(2) each have more than five eukaryotic-type protein kinases. Mycobacterium leprae is also reported to have more than three kinases (Fsihi et al., 1996 ); Chlamydia trachomatis has two (Pkn5 and PknD), whereas spirochaetes (Borrelia burgdorferi and Treponema pallidum) do not have any kinase gene. In the O2-reducing bacterium, Aquifex aeolicus, one kinase (STPK) was detected.

Among cyanobacteria, Synechocystis sp. and Anabaena sp. have more than five protein kinases (Zhang & Libs, 1998 ). Among archaea, only Pyrococcus horikoshii has one kinase-like protein (PHCJ009), although this shows very low homology with other protein kinases. Recently, the complete nucleotide sequence of chromosomal DNA in Rickettsia prowazekii has been determined (Andersson et al., 1998 ), but no eukaryotic-type protein kinase gene was detected within it, although four histidine kinase genes were found.

Homology analyses
Table 1 shows the result of analysis by the GAP program using the amino acid sequences of subdomains I–XI. Protein kinases HPO432 from Helicobacter pylori and SLL0776 from Synechocystis sp. exhibited high similarity values with only one of the six protein kinases in the comparison, including those from eukaryotes; and kinase PknM from Mycobacterium tuberculosis showed no homology at all (Table 1). These proteins, except for SLL0776, lacked kinase subdomains VIa, VII and VIII. The Rdf2 analysis (Table 3) indicates that PknM from Mycobacterium tuberculosis scored very low for homology with 40 other protein kinases. Therefore, it is uncertain whether PknM really has kinase activity. Indeed, the corrected version of the PknM sequence excluded this protein from the list of eukaryotic-type protein kinases in Mycobacterium tuberculosis (Cole et al., 1998 ). On the other hand, HPO432, SLL0776 and PHCJ009 gave high scores for homology with several kinases: HPO432 with PknK, CTPknD and PH009; SLL0776 with PkaD, AfsK, PknE, CTPknD, SLR1225, SLR1443, SLR1697 and SLR0152; and PHCJ009 with PkaE, YloP, PknH, PknL, CTPkn5 and CTPknD (Table 3). Interestingly, in FASTA analyses, STPK from Aquifex aeolicus showed higher homology scores with protein kinases from eukaryotes such as a serine/threonine kinase of Caenorhabditis elegans (par-1, accession no. U40858), human protein p78 (accession no. M80359) and a serine/threonine kinase of Rattus norvegicus (accession no. Z83868) than with protein kinases from prokaryotes. par-1, human protein p78 and the serine/threonine kinase of Rattus norvegicus showed 31·51% identity in a 238 amino acid overlap (Smith–Waterman score of 412), 30·83% identity in a 240 amino acid overlap (Smith–Waterman score of 392) and 29·83% identity in a 238 amino acid overlap (Smith–Waterman score of 388) with STPK, respectively. The protein with the highest score in prokaryotes was SLR0152 of Synechocystis sp., but it showed 30·83% identity in a 240 amino acid overlap (Smith–Waterman score of 363) and ranked only 35th in similarity scores. Consistent with this, STPK forms a cluster with eukaryotic protein kinases such as cAMPPK, S6-2PK and CaMPK in the phylogenetic tree. This cluster includes some other eukaryotic-type protein kinases from various species of prokaryotes, as described below.


View this table:
[in this window]
[in a new window]
 
Table 3. Relationship between pairs of eukaryotic-type protein kinase sequences analysed by the Rdf2 program

 
Phylogenetic trees
Fig. 1 shows a phylogenetic tree constructed on the basis of 41 amino acid sequences of the conserved subdomains I–XI (excluding PknM) by using a neighbour-joining method (the CLUSTAL W program). Generally, protein kinases from the same species cluster in the phylogenetic tree, but this is not true of all them. For example, PkaA, PkaB, PkaE, PkaG and AfsK from S. coelicolor form a branch, but PkaD and Pks3 from the same species belong to different branches. Similarly, YbdM and YloP from Bacillus and CTPknD and CTPkn5 from Chlamydia trachomatis are located in different branches. Comparable results were obtained in phylogenetic trees constructed with other programs, including PROTML (a maximum-likelihood method) in the MOLPHY package, TREENJ (a neighbour-joining method) in the ODEN package and PROTPAR (a maximum-parsimony method) in the PHYLIP package (data not shown). As described above, STPK from Aquifex aeolicus, PknG from Mycobacterium tuberculosis and HPO432 from Helicobacter pylori were closely related to CaMPK, which is of eukaryotic origin. MG109 and ORF389 from Mycoplasma form a single branch, although they come from different species. The close relationship between MG109 and ORF389 is confirmed by the high value obtained from an Rdf2 analysis (Table 3).



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 1. Phylogenetic tree constructed on the basis of amino acid sequences of the conserved subdomains I–XI in 41 eukaryotic-type protein kinases using a neighbour-joining method (CLUSTAL W). The numbers at the branchpoints are bootstrap probabilities. The scale-bar indicates 0·1 substitutions per site.

 
Fig. 2 shows a phylogenetic tree of protein kinases from S. coelicolor A3(2). Although these kinases come from the same species, they are only distantly related. They show long distances after branching, indicating that they are now related only remotely having separated early and evolved independently of each other. This situation was confirmed by Rdf2 analysis (Table 3) and examination with the PROTDIST program using the Dayhoff PAM model (Table 4). Similar relationships were observed in the kinases from Mycobacterium and those from cyanobacteria. Phylogenetic trees constructed by other programs supported the conclusion reached above (data not shown).



View larger version (8K):
[in this window]
[in a new window]
 
Fig. 2. Phylogenetic tree constructed on the basis of amino acid sequences of the conserved subdomains I–XI in eight eukaryotic-type protein kinases from S. coelicolor A3(2) using a neighbour-joining method (CLUSTAL W). The numbers at the branchpoints are bootstrap probabilities. The scale-bar indicates 0·1 substitutions per site.

 

View this table:
[in this window]
[in a new window]
 
Table 4. Distance matrix for amino acid sequences of eukaryotic-type protein kinases from Streptomyces coelicolor A3(2) computed by the PROTDIST program using the Dayhoff PAM model

 
Codon usage
Table 5 shows the mean G+C content of the genes analysed to date in various species (obtained from http://www.dna.affrc.go.jp/) and those of the coding region of the protein kinases. It indicates that almost all the protein kinase genes showed G+C contents similar to those of other genes in the same species. In addition, the G+C contents in the first, second and third bases of codons of STPK are 47·15, 28·23 and 55·31%, respectively. These values are closer to the mean values of all the genes analysed so far in Aquifex aeolicus than to those in the genes in eukaryotes such as Saccharomyces cerevisiae and Homo sapiens. The mean values in Aquifex aeolicus are 50·44, 32·37 and 47·93%, respectively. Similar relationships were observed for other protein kinases. Furthermore, the G+C content in the third letter of codons in the protein kinase genes from S. coelicolor was over 90%, whilst the mean value of the genes analysed in the present work in Streptomyces is 92·82%. These results indicate that codon usage for these kinases follows that of other proteins in each species.


View this table:
[in this window]
[in a new window]
 
Table 5. G+C contents of eukaryotic-type protein kinases in various species

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The nucleotide sequences of four newly isolated putative eukaryotic-type protein serine/threonine kinase genes from S. coelicolor A3(2) were determined. All of them contained 11 conserved subdomains, and at least two of them showed protein serine/threonine and/or tyrosine kinase activity (unpublished data). Although the number of amino acids in these and the other kinases from S. coelicolor differed appreciably (from 417 to 799), the number of amino acids in the subdomains was limited to 233 (PkaG and AfsK) to 249 (PkaA). On this basis, at least eight eukaryotic-type protein kinases were detected in S. coelicolor A3(2).

We were interested in determining whether these protein kinases are distributed ubiquitously in bacteria and archaea, and if so, what evolutionary relationship exists between those in prokaryotes and eukaryotes. Our search showed that the eukaryotic-type protein kinases have a restricted distribution in prokaryotes. On the other hand, almost all of the prokaryotes analysed here have histidine kinase genes. There are 30, 4, 4, 36, 14, 4, 3, 4, 32, 15, 24, 1 and 4 histidine kinase genes in E. coli, Haemophilus influenzae, Helicobacter pylori, Bacillus subtilis, Mycobacterium tuberculosis, Borrelia burgdorferi, T. pallidum, Aquifex aeolicus, Synechocystis, Methanobacterium thermoautotrophicum, Archaeoglobus fulgidus, P. horikoshii and Rickettsia prowazekii, respectively. The results suggest, therefore, that the eukaryotic-type protein kinases are not essential for primary metabolism in prokaryotes. In addition, the presence of eukaryotic-type protein kinases is not directly related to the classification of prokaryotes. In fact, it is hard to correlate the distribution of these kinase genes in prokaryotes with the classification of prokaryotes on the basis of 16S rRNA. For example, in archaea, which are believed to be more closely related to eukaryotes, only one eukaryotic-type protein kinase could be detected (PHCJ009 in P. horikoshii). Both Myxococcus xanthus and E. coli are Gram-negative bacteria, but while Myxococcus xanthus has more than five eukaryotic-type protein kinases, no such enzyme could be detected in E. coli. The common feature among the bacteria that contain more than three kinase genes is that they exhibit various morphological forms and/or carry out biochemical reactions in response to changes in their environment. For example, Myxococcus xanthus forms a spore-filled fruiting body and migrates on semi-solid surfaces by gliding. Streptomyces undergoes complicated morphological differentiation into aerial mycelium and chain-like spores and produces secondary metabolites such as antibiotics. Mycobacterium tuberculosis and Mycobacterium leprae grow very slowly and form rudimentary mycelium under some conditions. Synechocystis has thylakoids thought to be the origin of chloroplasts in eukaryotes. Furthermore, among the proteins listed in Table 2, three from Myxococcus xanthus and five from S. coelicolor have protein serine/threonine and/or tyrosine kinase activity and engage in morphological and/or biochemical differentiation (Matsumoto et al., 1994 ; Zhang et al., 1996 ; Udo et al., 1997 ; H. Ogawara & H. Urabe, unpublished data). Thus, we speculate that the eukaryotic-type protein kinases have been preserved only in these specific species during the course of evolution.

The phylogenetic tree (Fig. 1) constructed by comparing the amino acid sequences of consensus regions shows that the protein kinases from one species do form clusters as suggested by Zhang (1996) , but not always. CTPknD and CTPkn5 from Chlamydia trachomatis belong to different branches and the protein kinases from Mycobacterium tuberculosis scatter divergently in the tree. Similar situations were observed for protein kinases from other bacteria, such as YbdM and YloP from Bacillus subtilis and those from Synechocystis. Furthermore, the substitution frequencies are very high after branching. This is more clearly shown in the tree constructed on the basis of amino acid sequences of protein kinases from a single species, S. coelicolor (Fig. 2); Rdf2 and PROTDIST analyses confirmed the high substitution frequencies within species (Tables 3 and 4). We suggest, therefore, that these eukaryotic-type protein kinases from prokaryotes are only remotely related to each other, even if they come from a single species. However, critical amino acid residues involved in the kinase reaction are almost always conserved.

The G+C contents of the genes of the protein kinases resemble the mean values of the genes from the same species, indicating that codon usage is retained. The only exceptions are Pks3 from S. coelicolor and SLL0776 from Synechocystis. The reason for the discrepancy is not yet clear. These genes might come from other species by horizontal transfer. In both cases, it is the third letter that differs most: the G+C contents of the first, second and third letters in Pks3 and SLL0776 are 68·67, 50·62 and 76·46; and 51·29, 40·41 and 27·13, respectively. The mean G+C contents are 71·75, 50·45 and 92·82 in Streptomyces and 56·08, 39·92 and 49·98 in Synochocystis, respectively. These results strongly suggest that the eukaryotic-type protein kinases were already present before the branching of eukaryotes from prokaryotes and that these kinases disappeared in some prokaryotes because the enzymes were unnecessary. The alternative explanation, that the kinases were acquired in prokaryotes after the divergence of prokaryotes from eukaryotes, is less plausible, although a few kinases might be the result of horizontal transfer.


   ACKNOWLEDGEMENTS
 
We thank M. Bibb of the John Innes Institute for providing us with S. coelicolor A3(2), M. Hasegawa and Y. Cao of the Institute of Statistical Mathematics for their valuable advice on MOLPHY analysis, and the National Institute of Genetics and the Institute of Medical Sciences, University of Tokyo, for permission to use their computers for genetic analyses. This work was supported in part by the special fund of the Meiji Pharmaceutical University.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Adachi, J. & Hasegawa, M. (1992). molphy: programs for molecular phylogenetics. I. PROTML: maximum likelihood inference of protein phylogeny (Computer Science Monographs no. 27), Institute of Statistical Mathematics, Tokyo.

Andersson, S. G. E., Zomorodipour, A., Andersson, J. O., Sicheritz-Ponten, T., Alsmark, U. C. M., Podowski, R. M., Naslund, A. K., Erissson, A. S., Winkler, H. H. & Kurland, C. G. (1998). The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396, 133-140.[Medline]

Cole, S. T., Brosch, R., Parkhill, J. & 39 other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 396, 537–544.

Doolittle, R. F. (1998). Microbial genomes opened up. Nature 392, 339-342.[Medline]

Felsenstein, J. (1995). PHYLIP, version 3.5c, University of Washington, USA.

Fsihi, H., De Rossi, E., Salazar, L. & 7 other authors (1996). Gene arrangement and organization in an ~76 kb fragment encompassing the oriC region of the chromosome of Mycobacterium leprae. Microbiology 142, 3147–3161.[Abstract]

Hanks, S. K., Quinn, M. & Hunter, T. (1988). The kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 241, 42-52.[Medline]

Hirakata, T., Kieser, H., Hopwood, D., Urabe, H. & Ogawara, H. (1998). Putative protein serine/threonine kinase genes are located in several positions on the chromosome of Streptomyces coelicolor A3(2). FEMS Microbiol Lett 159, 1-5.[Medline]

Hopwood, D. A., Bibb, M. J., Chater, K. F. & 7 other authors (1985). Genetic Manipulation of Streptomyces: a Laboratory Manual. Norwich: John Innes Foundation.

Hunter, T. & Cooper, J. A. (1985). Protein-tyrosine kinases. Annu Rev Biochem 54, 897-930.[Medline]

Ina, Y. (1991). ODEN, molecular evolutionary analysis system for DNA and amino acid sequences. DDBJ, National Institute of Genetics, Japan.

Lipman, D. J. & Pearson, W. R. (1985). Rapid and sensitive protein similarity searches. Science 227, 1435-1441.[Medline]

Matsumoto, A., Hong, S. K., Ishizuka, H., Horinouchi, S. & Beppu, T. (1994). Phosphorylation of the AfsR protein involved in secondary metabolism in Streptomyces species by a eukaryotic-type protein kinase. Gene 146, 47-56.[Medline]

Munoz-Dorado, J., Inouye, S. & Inouye, M. (1991). A gene encoding a protein serine/threonine kinase is required for normal development of M. xanthus, a gram-negative bacterium. Cell 67, 995-1006.[Medline]

Needleman, S. B. & Wunch, C. D. (1970). A general method applicable to the search for similarities in the amino acid sequence of two proteins. J Mol Biol 48, 443-453.[Medline]

Parkinson, J. S. & Kofoid, E. (1992). Communication modules in bacterial signalling proteins. Annu Rev Genet 26, 71-112.[Medline]

Pearson, W. R. & Lipman, D. J. (1988). Improved tools for biological sequence comparison. Proc Natl Acad Sci USA 85, 2444-2448.[Abstract]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74, 5463-5467.[Abstract]

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignments through sequence weighting, position-specific GAP penalties and weight matrix choice. Nucleic Acids Res 22, 4673-4680.[Abstract]

Udo, H., Inouye, M. & Inouye, S. (1997). Biochemical characterization of Pkn2, a protein Ser/Thr kinase from Myxococcus xanthus, a gram-negative developmental bacterium. FEBS Lett 400, 188-192.[Medline]

Urabe, H. & Ogawara, H. (1995). Cloning, sequencing and expression of serine/threonine kinase-encoding genes from Streptomyces coelicolor A3(2). Gene 153, 99-104.[Medline]

Zhang, C.-C. (1996). Bacterial signalling eukaryotic-type protein kinases. Mol Microbiol 20, 9-15.[Medline]

Zhang, C. C. & Libs, L. (1998). Cloning and characterization of the pknD gene encoding an eukaryotic-type protein kinase in the cyanobacterium Anabaena sp. PCC7120. Mol Gen Genet 258, 26-33.[Medline]

Zhang, W., Munoz-Dorado, J., Inouye, M. & Inouye, S. (1992). Identification of a putative eukaryotic-like protein kinase family in the developmental bacterium Myxycoccus xanthus. J Bacteriol 174, 5450-5453.[Abstract]

Zhang, W., Inouye, M. & Inouye, S. (1996). Reciprocal regulation of the differentiation of Myxococcus xanthus by Pkn5 and Pkn6, eukaryotic-like Ser/Thr protein kinases. Mol Microbiol 20, 435-447.[Medline]

Received 17 March 1999; revised 6 July 1999; accepted 9 August 1999.