- Email: [email protected]
Toxic and non-toxic strains of the cyanobacterium Microcystis aeruginosa contain sequences homologous to peptide synthetase genes
FEMS Microbiology Letters 135 (1996) 295-303
Toxic and non-toxic strains of the cyanobacterium Microcystis aeruginosa contain sequences homologous to peptide synthetase genes Kathrin MeiBner, Elke Dittmann, Thomas Biirner lnstitutfir Biologie, Humboldt-Uniuersitiit zu Berlin, halidenstr.
*
43, IO11.5 Berlin, Germany
Received 26 September 1995; accepted 17 November 1995
Abstract Toxic strains of Microcystis aeruginosa produce cyclic heptatoxins (microcystins) that are believed to be synthesized non-ribosomally by peptide synthetases. We analysed toxin-producing and non-toxic strains of M. aeruginosa with respect to the presence of DNA sequences potentially encoding peptide syntbetases. Hybridizations of genomic DNA of various M. aeruginosa strains with PCR-amplificated fragments possessing homologies to adenylate-forming domains of peptide synthetase genes provided first evidence for the existence of corresponding genes in cyanobacteria. Furthermore we isolated and sequenced from genomic libraries overlapping fragments of M. aeruginosa DNA with a total length of 2982 bp showing significant homology to genes encoding peptide synthetases and hybridizing exclusively with DNA from toxic strkins. Our results indicate that both toxic and non-toxic strains of M. aeruginosa possess genes coding for peptide synthetases and that hepatotoxin-producing and non-toxic strains differ in their content of genes for specific peptide synthetases. Keywords: Cyanobacteria; Microcystis aeruginosa; Peptide synthetase; Microcysti
1. Introduction Microcystis aeruginosa is one of the most common waterbloom-forming species of cyanobacteria. Various strains of M. aeruginosa were reported to produce toxic heptapeptides (microcystins) causing the death of animals and illnesses in humans. Moreover, the toxins are suspected to act as tumor promoters in humans. A large number of structural variations of microcystins have been described [1,2]. They share the common structure CydO(-D-Ala-L-X-
* Corresponding author. Tel.: f49 (30) 2897 2633; Fax: +49 (30) 2897 2799; E-mail: thomas = [email protected]. 037%1097/96/$12.00 SSDI 0378-l
oMeAsp-L-Z-Adda-D-Glu-Mdha) where X and Z represent variable L-amino acids. Small size, cyclic structure and in particular their content of unusual amino acids indicate that these cyanobacterial peptides are synthesized non-ribosomally rather than on ribosomes. Non-ribosomal peptide synthesis has been studied in antibiotic-producing organisms, like the Grampositive bacterium Bacillus subtilis and certain fungi [3,4]. The enzymes involved in non-ribosofnal peptide synthesis, peptide synthetases, activate the constituent amino acid or hydroxy acid as acyladenylates, a reaction corresponding to that catdysed by amino acyl tRNA synthetases, followed by a transfer of the intermediate to an acceptor thiol-group form-
0 1996 Federation of European Microbiological Societies. All rights reserved
097(95)00469-6
296
K. MeiJher et al. / FEMS Microhiolo~~
ing thioesters (‘thiotemplate multienzymic mechanism’) [3,5]. Comparison and analysis of DNA-derived amino acid sequences of several peptide synthetases of bacterial and fungal origin revealed a highly conserved structure [6,7]. The peptide synthetase genes possess a modular structure. Each module comprises the information for a peptide synthetase unit catalysing the activation and modification of a single component and the linking reaction with the nascent peptide. Comparison of amino acid sequences encoded by various modules has resulted in the definition of short sequence motifs that are highly conserved with respect to their position and amino acid sequence [6,7] providing a general approach for identifying new peptide synthetase genes [8,9]. They are in part located within the so-called adenylate-forming domain which is also found in other adenylate-forming enzymes like 4-coumarateCoA-synthetase and acetyl CoA synthetases [cf. 61. We used two conserved sequence motifs of the adenylate-forming domain of peptide synthetases (core I and II; [9]> to search for homologous sequences in toxic and non-toxic strains of M. aeruginosa. It is completely unknown why certain strains of M. aeruginosa (or other species of cyanobacteria) are able to produce toxins whereas others are not. Since it has been shown that the toxin production depends to certain extent on environmental factors and growth phase [ 10,111 it is unclear whether toxic and non-toxic strains differ with respect to their content of genes for toxin synthesis or whether the production of toxins is in principle feasible in all strains but down-regulated in the non-toxic ones. Peptide synthetases and their genes have not been analysed in cyanobacteria yet. Therefore, the aims of the present study were to investigate whether cyanobacteria contain genes encoding peptide synthetases of the type known from other organisms, and to find differences between toxic and non-toxic strains.
2; Materials and methods
2. I. Bacterial strains HUB-strains of M. aeruginosa were described in [12] with the exception of HUB X which was iso-
Letters 135 (1996) 295-303
lated from Lake Mliggelsee, Berlin, Germany, in 1984. All HUB strains were provided by Dr. J.-G. Kohl (Institut fur Biologie, Humboldt-University, Berlin, Germany), strains PCC 7813 and PCC 7820 by Dr. R. Rippka (Pasteur Culture Collection, Institut Pasteur, Paris) and strain PCC 7806 by Dr. J. Weckesser (Albert-Ludwigs-University, Freiburg, Germany). Strains HUB 5-2-4, PCC 78 13, PCC 7820, and PCC 7806 were found to produce toxins. HUB X, HUB 18 and HUB 5-3 are non-toxic strains. E. coli XL-l was used as a recipient strain for plasmid transformation. E. coli C 600 hfl served as host strain for recombinant lambda gt 10 phages. pBluescript KS( - > (Stratagene, Heidelberg, Germany) was used to subclone DNA for sequencing. 2.2. Preparation of DNA from M. aeruginosa, hybridization DNA from strains of M. aeruginosa was isolated following the CTAB method described for plant cells [ 131 with modifications concerning the incubation times after adding 2 X CTAB, 0.2 vol 5 X CTAB and the precipitation buffer which were performed for 30 min, 10 min and 20 min, respectively. Centrifugation steps were carried out for appropriate times necessary to separate two phases or to sediment a pellet. Digested cyanobacterial DNA was separated on 0.8% agarose gels and vacuum-transfered to Amersham Hybond-N membranes as described [14]. Oligonucleotides core I and core II [9] were end-labelled by standard methods [14]. DNA fragments were labelled using the Megaprime kit (Amersham, Braunschweig, Germany). Southern hybridizations were performed as described [14]. 2.3. DNA cloning, sequencing PCR products were subcloned in pGEM-T vector (Promega, Madison, USA). Plasmid-DNA was prepared using Tip 20 columns (Diagen, Dusseldorf, Germany). To construct genomic libraries, total DNA from M. aeruginosa HUB 5-2-4 was digested with restriction endonucleases RsaI or A/u1 (Amersham, Braunschweig, Germany) and ligated into lambda gtl0 vector using the RiboCloneeEcoRI Adaptor Ligation System and Packagene@ Lambda DNA Packaging System (Promega, Madison, USA). Phage
297
K. MeiJner et al. / FEMS Microbiology Letters 135 Cl9961 295-303
DNA was prepared as described [14]. Unidirectional deletions with exonuclease III were generated using the double stranded nested deletion kit (Pharmacia, Freiburg, Germany). DNA was sequenced using Sequenase Vers. 2.0 kit (USB, Cleveland, USA). Homology searches in EMBL, Genbank and SwissProt nucleic acid/protein databases were performed using FASTNSCAN/FASTPSCAN and sequence alignments were assembled using the programmes CLUSTAL and PALIGN from the PCGENE software package. 2.4. Polymerase
chain reaction (PCR)
The oligonucleotides BSCI (GGAATTCC TTA AAI GCI GGA(C) GG(c)A GCI TAT G(C)TG CCG C(G)TT GAT CC) derived from core 1 sequence and BSCII (GGAATTCC TTT I(C)GG I(~)T-I ICC ICC&T TGT(A) ICC IGA I(C)GT GTA A(C) derived from core 2 sequence were kindly provided by Dr. M.A. Marahiel (Philipps University Marburg, Germany). PCR experiments required 10 ng Microcystis DNA and 20 pmol of BSCI and BSCII and were carried out as follows: incubation for 7 min at 94°C cooling down to room temperature, addition of Taq-polymerase (Boehringer, Mannheim, Germany),
30 cycles (2.5 min at 72°C 1.O min at 94°C and 1.5 min at 6O”C), incubation for 7 min at 72°C.
3. Results and discussion 3.1. Hybridization qf cyanobacterial DNA with oligonucleotides derived from two conserved sequence motifs of peptide synthetase genes We used oligonucleotides (core I and core II, [9]) corresponding to conserved amino acid sequences of adenylate-forming domains as probes for hybridization of Southern blots with the DNAs of toxic and non-toxic strains of M. aeruginosa. The sequence ‘SGTTG’ (corresponding to core II> is conserved in all polypeptides belonging to a group of enzymes called either the ‘family of adenylate-forming enzymes’ or according to the SWISSProt database definition the ‘AMP-binding enzymes’. The sequence ‘LKAGGA’ (core I) is highly conserved in peptide synthetases only, whereas other types of adenlyate-forming enzymes show a low degree of homology in this region [9]. Hybridization of the radioactively labelled probes with digested DNA from six different strains re-
PCRl PCR2 PCR3 PCR4
TGACTATCCCCCTGAACGTCTTCAGTTTATGTTAGACTTC TGACTATCCCATTGAACGTCTTCAATTTATGTTAGAAGATGTTT GGATTATCCTACCGAGCGCTTGGGGGATATCCTCTCAGATTCGGGGGTTT TAATTATCCCCCTGAGCGCTTGGATTATATGATATCCGATTCAGCTATTT ..* ***** ** ** * * ** . . . . . . *** . .l*
PCRl PCR2 PCR3 PCR4
CTTTTTTGATTACCCAACGTTCTTTATTAGCAAAATTGCC CATTTTTGATTACCCAACGTTCTTTATTGGCCGTATTACCTCCTTCTCAA CTTTGGTGTTAACTCAGGACTTTAGGGGATTTTCTTCCCCA?mYGGG CTCTGTTGTTAACTCAGCAATCCTTAGTCCAATTTCTGCCAGAA?iATCAA ._** *** . . * *..a*.*.** ** . . . *.** . . *
100 100 100 100
PCRl PCR2 PCR3 PCR4
GCGACTCTTATTTGTTTAGATCACATCCAAGAGCAGATTTCTCAATATTC GCTAATGTTATTTGTTTAGATGAGATTGAAGAGCAGGTTTCTCAATATCC GCCGAGTTACTGTGTTTAGATAGGGATTGGGAAAAGATAGCTACCTATAG GCCGAAATACTGTGTTTAGATACAGATTGGTCAAGGATAGCTAATTATAG . . * .********* l ** ** . *. . . . . . *.* ..**
150 150 150 150
PCRl PCR2 PCR3 PCR4
TCAAGATAATCTTCAAAGTGAGTTAACTCCTCCTTCCAATTTAGCTAACGTT TCAAGATAATCTTCAAAATGGATTAACGATTGCTAGTTTAGCGAATGTG TCCAGAAAATCCCTTCAATCTAACGACTCCTGAGAATTTAGCCTATGTT TCAAGAGAATTTGACCTCTCCAGTTACCCCAGAAAATTTAGCTTATGTT ** *** *** * * . **tt** .* ** . . . ** . .
50 50 50 50
199 199 199 199
Fig. 1. Nucleotide sequence alignment of four PCR products representing the regions between motifs core 1 and core 2 in putative peptide synthetase modules of M. aeruginosa. PCRl and 2 were amplified with DNA from the toxic strain HUB 5-2-4 and PCR3 and 4 were amplified with DNA from the toxic strain PCC 7820. Identical bases are indicated by asterisks, transitions by dots.
K. Megner et al. / FEMS Microbiology Letters 135 (19%) 295-303
298 PCRl ACUAEMRNI
- DYPPERLQFMLEDSQVPFL---ITQ----RSLLA-KLPASQATLICLDHI - ITTDEGKRGGKVIGTKRIVDEGLKQCPDVSTVLVYKRTGAEVPWTEGRDI * . .* .* *
PCRl ACUAEMRNI
- --QEQISQYSQDNLQSELTPSN-LANV - WWHEEVEKYPAYIAPDSVNSEDPLFLL * *. *
*
-42 -50
-66 -17
PCRl GRSBI
- DYPPERLQFMLEDSQVPFLITQRSLLAKLPASQATLICLDHIQEQISQYS - AYPQERIQYLLEDSGAALLLTQSHVLNKLPVD---IEWLDLTDEQNYVED ** ** ** ** *..**** *.** .* ***
PCRl GRSB4
- QDNLQSELTPSNLANV - GTNLPFMNQSTDLAYI ** . l*
-50 -47
-66 -63
.
Fig.2. Comparison of the amino acid sequence deduced from PCRl with sequences synthetase 2 (CiRSB4, [4]) and another adenylate-forming enzyme (acetyl-co&yme indicates identical residues, a dot indicates residue similarity.
vealed patterns of signals indicating that sequences similar to core 1 and core 2 occur in all strains of M. aeruginosa investigated (not shown) and suggesting that all strains (toxic and non-toxic) of M. aeruginosa carry sequences homologous to motifs of the adenylate-forming domain of peptide synthetase modules.
of the leucine-activating domain of gramicidin S [16]). An asterisk A synthetase, ACUAEMENI,
3.2. Amplification
of sequences from cyarwbacten’al DNA using primers derived from conserved sequences of peptide synthetases
Since sequences with homology to the conserved motifs core 1 and core 2 were obviously present in toxic and in non-toxic strains, we amplified by PCR
A
B
C
D
E
123456
123456
123456
123456
123456
11.5 kb-,
Fig. 3. Southern blot hybridization with radioactively labelled PCR products. A: Electrophoresis of XbaI digested DNA from M. aeruginosa: lane 1: HUB 18, lane 2: HUB X, lane 3: HUB 5-3, lane 4: PCC 7813, lane 5: PCC 7820, lane 6: HUB 5-2-4. B: Hybridization with PCRl. C: Hybridization with PCR2. D: Hybridization with PCR3. E: Hybridization with PCR4. Filters were washed with 0.1 SSC/O. 1 SDS at 60°C.
299
K. Meijher et al. / FEM.7 Microbiology Letters 135 (19%) 295-303
DNA regions located between these conserved motifs in the toxic strains HUB 5-2-4 and PCC 7820. These regions were described to be of less homology compared with the core motifs themselves. The PCR fragments (PCRl and 2 amplified from HUB 5-2-4, PCR3 and 4 from PCC 7820) were cloned and sequenced (Fig. 1). Four different PCR products
were obtained and compared with sequences of other peptide synthetases and adenylate-forming enzymes in the data bases. The homology of the four PCR products from M. aeruginosa DNA was dstinctly higher to corresponding regions between code 1 and core 2 of genes encoding peptide synthetases than to those sequences in genes for other adenylate-forming
M?+PEPl YDTSRFTQDTIERMAAHLQTILTGIVTDTRQRVTQLPILTTQDQHQLLVE-50 GRSBl
N--------------PHV------LVQD-------VPLLTKQEKQHLL~-23 *.** * * * *
****
BOX A MAPEPl
WNNTEADYPLDKSLHQLFEEQVAQNPQGIAVIFEGHKLTYQQ-L-100
GRSBl
LHDSITEYP-DKTIHQLFTEQVEKTPEWAWFEDEKVTlL-72 * l * ** *.** ,.** ** ,.**** ***
*
l
***
Box B MAPEPl
AHCLRDKGVGPESLVGIFMERSLEMVIGLLGI
LAAMUTVPLDPDYPTER-150
GRSBl
ARFLREKGVKKESIIGIMMERSVEMIVGILGI -IDPEYPKER-122 l ** *** ** ..** ****.**.,*,*********,**.**,**
**
MAPEP
LGDILSDSGVSLVLTQESLGDFLPQTGAELLCLDRDWEKIATYSP~PFN-200
GRSBl
IGYML-DS-VRLVLTQRHLKDKFAFTKETIVIEDPS---ISHELTEEIDY-l67 * ,* ** * ***** * * * * * . *
MAPEPl
LTTPENLAYVIYTSOBTGKPKGVMNIHRGICNTLTYTIGHYNITSEDRIL-25O
GRSBl
1NESEDLFY11YTSGTTGXPKGVMLEHKN1VNLLHFTFEKTN1NFSDKVL-217 * * * ***** ******** *, * * * * ** * .l
MAPEPl
QIISLSFDGSVWEIFSSLISGASLWAKPDGYKDIDYLIDLIVQEQ~YF-3OO
GRSBl
QYTNAVLTCVTKKFFSTLLSGGQLYLIRKETQRDVEQLFD-267 * **.* ** * . . . .* . . * **
Box C
*..
MAPEPl
TCVPSILRVFLQHPK--SKDCHCLKRVIVGGE--ALSYELNQRFFQQLNC-346
GRSBl
SFPVAFLKFIFNEREFINRFPTCVKHIITAGEQLVVNNEFK-RYLHEHNV-316 **. .* ** * * . . *. . .
BOX
*
D
MAPEPl
ELYNAYGPTEVAVETTIWCCQPNSQIS----IGTPIANAQVYILDSYLQP-392
GRSBl
HLHNHYGPSETHWTTY-TINPEAEIPELPPIGKPISNTWIYILDQEQQL-365 * * *** * * ** ** **,*. .**** .* .*
*
Fig. 4. Alignment of MAPEPl with the amino acid sequence of the proline-activating synthetase unit of gramicidin S synthetase 2 (GRSB 1, [4]). Sequences corresponding to the conserved motifs of peptide synthetases (cf. [20]) are printed in bold type. The nucleotide sequence encoding MAPEPl has accession number 228338 in the EMBL data library (Heidelberg). Asterisks indicate identical residues, dots indicate similar residues.
K. Mei&er et al. / FEMS Microbiology Letters 135 11996)295-303
300
enzymes like 4-coumarate-CoA-synthetase [15] and acetyl-CoA-synthetases [16]. As an example, the comparison of the deduced amino acid sequence of PCRl from M. aeruginosu HUB 5-2-4 with corresponding sequences of the leucine-activating unit of gramicidin S-synthetase 2 from B. breuis [4] and acetyl-coenzyme A synthetase from Aspergillus niduZuns [16] is shown in Fig. 2. Accordingly, we conclude that we cloned sequences that form part of four different adenylate-forming domains of peptide synthetase genes of toxic Microcystis strains.
3.3. Hybridization of cyanobacterial DIVA with the amplified fragments
In order to investigate the strain-specific distribution of these putative adenylate-forming domains, the PCR products were radioactively labelled and used as hybridization probes. Interestingly, the four fragments generated different patterns with DNAs from the strains investigated. All four fragments hybridized with specific fragments of digested total DNA of strain HUB 5-2-4. In addition, PCR frag-
Box E
Box F
MAPEPl
VPIGVAGI#LHIGGMQLARGYLWRL~LTQEKFISNPF-AQ-441
GRSBl
QPQGIVt3ELYIS GMVGRQYLMU~~LTAI!XFFADPFRPNERMYRTGDLAR-415 . ***** *** *** ** * * . *** * * . . .* .******
MAPEPl
YLPEGNIEnoRIDN~VKLRGLRIIU&EIQTVLETHPNVEQTWIMREDS-49l
GRSBl
WLPEGNIEFLQFtADliQVKIRG~I~LQEIEAQLLNCKGVKEAWIDKADD-465 * .**.****,*** * ***.** ***t*** . * *** . *
MAPEPl
LYNQRLVAYVIRKNSLLTPQDLRRFLQQQLPAYMVPSAFVLLSASPLNNlU-541
GRSBl
KGGKYLCAYVVMEVE * *** .
MAPEPl
GKIDFtKKLPIPDETSIIESAYIAPRNEKESLLAQIWEDVLQVSKIGVSDN-59l
GRSBl
QKIDRKSLPNLEGIVNTNAKYWPTNELEEKLAKIWEEVLGISQIGIQDN-564 ****** ** . . * . * ** * ** *** .** .* **
Box G
Box H
Box
V-NDSELREYLGKALPDYMIPSFFVPLDHVRLIiLN-514 ** ** ** ** * * * . .** .*
I
**
Box J MAPEPl
FFELGI3%5LKiUSLVSKIQEKLGQSLPIKQVFAHPTIAEQAALLSTVTPL-64l
GRSBl
FFSLQGHSLK?AITLISRMNKECNVDIPLRLLFEAPTIQEISNYINGAKKE-~~~ ** ***********.*_,_ .* . . .* *** * . .
MAPEPl
TVATIPLVSAQETYETSBAQ
GRSBl
SYVAIQPVPEQEYYPVSSVQKRMFILNEFDRSGTAYNLPGVMFLDGKLNY-664 l *** * **. . .* * ** , . .* . . *
MAPEPl
DVFEKAIQLLISRHEBLRTSFILINGEPQQKILQNPSFWIGNFKDLDKIN-74l
GRSBl
RQLEZAVKKLVXREZALRTSFHSINGEPVQRVHQNVELQIAYSE-----S-709 * * . * . ***.***** l **** *.. ** *
Box K
Box
RRFYVLQQMDLNNVAYHIVSTLKIAGDFSP-691
L
Fig.4.(continued).
301
K. MeiJner et al. / FEMS Microbiology Letters 135 (19961295-303
ments 3 and 4 were found to generate signals with fragments of all toxic strains (HUB 5-2-4, PCC 7820, PCC 78131, but not with DNA from non-toxic strains. PCR fragment 1 hybridized with DNA fragments of the toxic strain (HUB 5-2-4) and three non-toxic strains (HUB 18, HUB 5-3, HUB X) (Fig. 3) supporting our conclusion that both types of strains possess genes encoding peptide synthetases. While peptide synthetases in strains producing hepatotoxins could be involved in toxin formation, peptide synthetases of non-toxic strains have to serve other purposes, i.e. are probably involved in the synthesis of non-toxic peptides. It is known from previous studies that M. aeruginosa indeed produces also non-toxic peptides of unknown function [17,18]. Therefore, we suggest that Microcystis strains carry specific sets of genes for peptide synthetases.
3.4. Identification of sequences potentially an entire peptide synthetase unit
PCR product 3 which hybridized selectively to DNA from toxic strains was used to identify corresponding sequences in genomic libraries of M. aeruginosa HUB 5-2-4. Recombinant phages carrying a 1157 bp RsaI fragment, a 2.8 kb and a 2.0 kb A/u1 fragment (overlapping with the 5’- and the 3’-end of the RsaI fragment), respectively, were identified and sequenced. The continous sequence of 2982 bp determined so far has been deposited in the EMBL data library with accession number ‘228338. The amino acid sequence of the potentially iencoded polypeptide (called MAPEPl) was compared with sequences in the EMBL and SwissProt data bases. The comparison revealed significant homology (34
Box M MAPEPl
PDEEILETIAKERKPFDLEKSPLVRSKIYKLSPNEYILEL~I~IIC~-79l
GRSBl
TEDQVERIIAEFMQPFALEVRPLLRVGLVKLEAERHLFIMDMIiliIIS~-l59 ** ** ** ** .* ..,**** ** . ** .. .
MAPEPl
SMSLLAKECLQYYSDLAKGLQPSIEPLPIQYKDYAGWPNNLLRS---ENN-838
GRSBl
SMQIMIQE----IADLYKEKE--LPTLGIQYKDFTVWHNRLLQSDVIEKQ-803 ** * * ***** . * * ** * **.. *
*
MAPEPl
PPNLDYWRPKLDNGQLTRVHLPTDFKRPQIKTFKGSHLSWTFDRETISKL-888
GRSBl
EASLAERICRRDSSIESTDRLP---KYQPFKALMVKDLH-SVQESSLWM-849 * * ** * * .* ..
.
Box N MAPEPl
RKICQEGRNTLFMAL-VAPVQILLYRYSGRZZDITIGT~IA'MkSHPQLQSL-937
GRSBl
YTRWQQKQEQHYIWFYLLRIMFFLSKYSGQDDIWGTPIAGRSHADVENM-899 * .*** * ** ** l * *** .. . .
BOX
.
0
MAPEPl
IGLFLNTLVIRDQIEPEKGYKNLLAKVRQTVTEALEHSDYPFDILVEKLP-987
GRSBl
LmRSRLNNEDTFKDFLANVKQTALHAYENPDYPFDTLVEKLG-949 * * ** *.** * * ***** .* .* .*** **
MAPEPl
VSRAAAT-994
GRSBl
IQRDLSR-956 *
encoding
Fig.4.(continued)
*****
.
302
K. MeiJner et al. / FEMS Microbiology Letters I35 (1996) 295-303
to 44% identity) to peptide synthetases from other organisms. Interestingly, the highest homology was found not to an adenylate-forming part of the only other peptide synthetase hitherto known from a Gram-negative bacterium (enterobactin synthetase of E. co& [19]) but to the proline-activating module of gramicidin S synthetase 2 from B. breuis [4] (Fig. 4). We could identify the conserved motifs typical of adenylate-forming domains observed in peptide synthetases and other adenylate-forming enzymes (motifs A-I, Fig. 4). Moreover, all motifs specific for peptide synthetases (motifs J-O; Fig. 4) were recognized in the sequence from M. aeruginosa, too, classifying its putative product as a unit of a peptide synthetase. The consensus motifs J to 0 differ between epimerizing and non-epimerizing peptide synthetase units [20]. Motifs P and Q have hitherto been observed only in enzymes with an epimerizing function [20]. A comparison of the deduced amino acid sequence from M. aeruginosu with the amino acid sequences of epimerizing and non-epimerizing peptide synthetases revealed a significantly higher homology to non-epimerizing peptide synthetase units (see conserved motifs J-O in Fig. 4, [20]) grouping the sequence from &l. ueruginosu clearly with the peptide synthetases units without epimerizing function. Our results suggest that non-ribosomal peptide synthesis using multienzyme complexes of the conserved type occurs in cyanobacteria. Furthermore, strains of M. ueruginosu were found to differ in their content of sequences belonging to adenylateforming domains of peptide synthetases. The DNA fragments isolated from HUB 5-2-4 potentially coding for a complete unit of a peptide synthetase were observed to hybridize with DNA from all toxic strains (HUB 5-2-4, PCC 7806, 7813, 7820) but not with DNA from the non-toxic strains (HUB X, 18, 5-3) applying stringent conditions (not shown). Obviously, the putative gene product exists exclusively in the investigated microcystin-producing strains of M. ueruginosu. Therefore, the ability of a strain to produce microcystins may depend first of all on its possession of the gene(s) required. Further studies have to determine whether their expression is controlled by environmental conditions. Although the number of investigated strains was not large enough
to rule out the possiblity that there is no causal relationship between the presence of this sequence in certain strains and their ability to produce toxins, one should emphasize that we dealt mostly with entirely unrelated strains, a fact increasing the significance of our observation. Most of the investigated toxic and non-toxic strains were originally collected at completely different locations [12]. Thus, the analysed nucleotide sequence may encode a peptide synthetase unit of a multienzyme complex involved in the production of microcystins.
Acknowledgements We are indebted to Dr. M. Marahiel for providing the oligonucleotides core I, core II, BSCI, and BSCII and to Dr. H. von Dijhren for stimulating discussions. We are grateful to Drs. H.-J. Kohl, R. Rippka, and J. Weckesser for providing cyanobacterial strains. This work was supported by a NaFijG fellowship of the Berlin Senate to K.M. and by a grant of the Deutsche Forschungsgemeinschaft to T.B.
References [I] Carmichael, [2]
[3]
[4]
[5]
[6]
[7]
W.W. (1994) The toxins of cyanobacteria. Scient. Am. 270, 78-86. Sivonen, K., Namikoshi, M., Evans, W.R., Fardig, M., Carmichael, W.W. and Rinehart, K.L. (1992) Three new microcystins, cyclic heptapeptide hepatotoxins, from Nosfoc sp. strain 152. Chem. Res. Toxicol. 5, 464-469. Kleinkauf, H. and von DGhren, H. (1990) Nonribosomal biosynthesis of peptide antibiotics. Eur. J. Biochem. 192, I-15. Turgay, K., Krause, M. and Marahiel, M.A. (1992) Four homologous domains in the primary structure of GrsB are related to domains in a superfamily of adenylate-forming enzymes. Mol. Microbial. 6, 529-546. Babbitt, P.C., Kenyon, G.L., Martin, B.M., Charest, H., Sylvestre, M., Scholten, J.D., Chang, K.-H., Liang, P.-H. and Dunaway-Mariano, D. (1992) Ancestry of the 4-chlorobenzoate-dehalogenase: analysis of amino acid sequence identities among families of acyl:adenyl ligases, enoyl-CoA hydratases/isomerases, and acyl-CoA thioesterases. Biochemistry 3 I, 5594-5604. Dieckmann, R., Lee, Y.-O., van Liempt, H., von Diihren, H. and Kleinkauf, H. (1995) Expression of an active adenylateforming domain of peptide synthetases corresponding to acyl-CoA-synthetases. FEBS Lett. 357, 212-216. Stachelhaus, T. and Marahiel, M.A. (1995) Modular structure
K. MeiJner
et al. / FEMS Microbiology
of genes encoding multifunctional peptide synthetases required for non-ribosomal peptide synthesis. FEMS Microbiol. Lett. 125, 3-14. [8] Tutgay, K. and Marahiel, M.A. (1994) A general approach for identifying and cloning peptide synthetase genes. Peptide Res. I, 238-241. [9] Borchert. S., Patil, S.S. and Marahiel, M.A. (1992) Identification of putative multifunctional peptide synthetase genes using highly conserved oligonucleotide sequences derived from known synthetases. FEMS Microbial. Lett. 92, 175180. [lo] Van der Westhuizen, A.J.J.N. and Eloff, J.N. (1985) Effect of temperature and light on the toxicity and growth of the blue-green alga Miciocystis aeruginosa (UV-006). Planta 163. 55-59. [III Watanabe, M.F. and Oishi, S. (1985) Effects of environmental factors on toxicity of a cyanobacterium (Microcystis aeruginosa) under culture conditions. Appl. Environ. Microbiol. 49, 1342- 1344. [121Schwabe, W., Weihe, A., Biimer, T., Henning, M. and Kohl, J.-G. (1988) Plasmids in toxic and nontoxic strains of the cyanobacterium Microc~.stis aeruginosa. Cm. Microbial. 17. 133-137 1131Rogers. S.O. and Bendich, A.J. (1985) Extraction of DNA from milligram amounts of fresh, herbarium and mummified tissues. Plant Mol. Biol. 5, 69-76. 1141 Sambrook, J., Fritsch. E.F. and Maniatis, T. (1989) Molecular cloning: A laboratory manual. 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
Letters
135 (1096) 295-303
303
[15] Lozoya, E., Hoffmann, H., Douglas, C., Schulz, W., Scheel, D. and Hahlbrock, K. (1988) Primary structure and catalytic properties of isoenzymes encoded by the two 4coumarate:CoA ligase genes in parsley. Eur. J. Biochem. 176, 661-667. [ 161 Connerton, I.F., Fincham, J.S., Sandeman, R.A. and Hynes, M.J. (1990) Comparison and cross-species expression of the acetyl-CoA synthetase genes of the ascomycete fungi, Aspergillus niduluns and Neurospom crassa. Mol. Microbial. 4, 451-460. [171 Birk, 1.M.. Dierstein, R., Kaiser, I., Matem, U.. Konig, W.A.. Krebber, R. and Weckesser, J. (1989) Noutoxic and toxic oligopeptides with Bamino acids and unusual residues in Microcystis aeruginosa PCC 7806. Arch. Microbial. 15 I, 41 l-415. El81Tsukamoto. S., Painuly, P.. Young, K.A.. Yang, X. and Shimizu, Y. (1993) Microcystilide A: a novel cell-differentiation-promoting depsipeptide from Microcystis aeruginosa NO-15-1840. J. Am. Chem. Sot. 115, 11046-llod7. 1191Rusnak, F., Sakaitani, M., Drueckhammer. D., Reichert, J. and Walsh, C.T. (1991) Biosynthesis of the Eschericlzia coli siderophore enterobactin: sequence of the entF gene, expression and purification of EntF, and analysis of covalent phosphopantethein. Biochemistry 30. 2916-2927. M., von Dohren, H. and DO1Pfeifer, E., Pavela-Vrancic, Kleinkauf. H. (1995) Characterization of tyrocidine synthetase I (TY I ): requirement of post-translational modification for peptide biosynthesis. Biochemistry 34. 7450- 7459.
Toxic and non-toxic strains of the cyanobacterium Microcystis aeruginosa contain sequences homologous to peptide synthetase genes Kathrin MeiBner, Elke Dittmann, Thomas Biirner lnstitutfir Biologie, Humboldt-Uniuersitiit zu Berlin, halidenstr.
*
43, IO11.5 Berlin, Germany
Received 26 September 1995; accepted 17 November 1995
Abstract Toxic strains of Microcystis aeruginosa produce cyclic heptatoxins (microcystins) that are believed to be synthesized non-ribosomally by peptide synthetases. We analysed toxin-producing and non-toxic strains of M. aeruginosa with respect to the presence of DNA sequences potentially encoding peptide syntbetases. Hybridizations of genomic DNA of various M. aeruginosa strains with PCR-amplificated fragments possessing homologies to adenylate-forming domains of peptide synthetase genes provided first evidence for the existence of corresponding genes in cyanobacteria. Furthermore we isolated and sequenced from genomic libraries overlapping fragments of M. aeruginosa DNA with a total length of 2982 bp showing significant homology to genes encoding peptide synthetases and hybridizing exclusively with DNA from toxic strkins. Our results indicate that both toxic and non-toxic strains of M. aeruginosa possess genes coding for peptide synthetases and that hepatotoxin-producing and non-toxic strains differ in their content of genes for specific peptide synthetases. Keywords: Cyanobacteria; Microcystis aeruginosa; Peptide synthetase; Microcysti
1. Introduction Microcystis aeruginosa is one of the most common waterbloom-forming species of cyanobacteria. Various strains of M. aeruginosa were reported to produce toxic heptapeptides (microcystins) causing the death of animals and illnesses in humans. Moreover, the toxins are suspected to act as tumor promoters in humans. A large number of structural variations of microcystins have been described [1,2]. They share the common structure CydO(-D-Ala-L-X-
* Corresponding author. Tel.: f49 (30) 2897 2633; Fax: +49 (30) 2897 2799; E-mail: thomas = [email protected]. 037%1097/96/$12.00 SSDI 0378-l
oMeAsp-L-Z-Adda-D-Glu-Mdha) where X and Z represent variable L-amino acids. Small size, cyclic structure and in particular their content of unusual amino acids indicate that these cyanobacterial peptides are synthesized non-ribosomally rather than on ribosomes. Non-ribosomal peptide synthesis has been studied in antibiotic-producing organisms, like the Grampositive bacterium Bacillus subtilis and certain fungi [3,4]. The enzymes involved in non-ribosofnal peptide synthesis, peptide synthetases, activate the constituent amino acid or hydroxy acid as acyladenylates, a reaction corresponding to that catdysed by amino acyl tRNA synthetases, followed by a transfer of the intermediate to an acceptor thiol-group form-
0 1996 Federation of European Microbiological Societies. All rights reserved
097(95)00469-6
296
K. MeiJher et al. / FEMS Microhiolo~~
ing thioesters (‘thiotemplate multienzymic mechanism’) [3,5]. Comparison and analysis of DNA-derived amino acid sequences of several peptide synthetases of bacterial and fungal origin revealed a highly conserved structure [6,7]. The peptide synthetase genes possess a modular structure. Each module comprises the information for a peptide synthetase unit catalysing the activation and modification of a single component and the linking reaction with the nascent peptide. Comparison of amino acid sequences encoded by various modules has resulted in the definition of short sequence motifs that are highly conserved with respect to their position and amino acid sequence [6,7] providing a general approach for identifying new peptide synthetase genes [8,9]. They are in part located within the so-called adenylate-forming domain which is also found in other adenylate-forming enzymes like 4-coumarateCoA-synthetase and acetyl CoA synthetases [cf. 61. We used two conserved sequence motifs of the adenylate-forming domain of peptide synthetases (core I and II; [9]> to search for homologous sequences in toxic and non-toxic strains of M. aeruginosa. It is completely unknown why certain strains of M. aeruginosa (or other species of cyanobacteria) are able to produce toxins whereas others are not. Since it has been shown that the toxin production depends to certain extent on environmental factors and growth phase [ 10,111 it is unclear whether toxic and non-toxic strains differ with respect to their content of genes for toxin synthesis or whether the production of toxins is in principle feasible in all strains but down-regulated in the non-toxic ones. Peptide synthetases and their genes have not been analysed in cyanobacteria yet. Therefore, the aims of the present study were to investigate whether cyanobacteria contain genes encoding peptide synthetases of the type known from other organisms, and to find differences between toxic and non-toxic strains.
2; Materials and methods
2. I. Bacterial strains HUB-strains of M. aeruginosa were described in [12] with the exception of HUB X which was iso-
Letters 135 (1996) 295-303
lated from Lake Mliggelsee, Berlin, Germany, in 1984. All HUB strains were provided by Dr. J.-G. Kohl (Institut fur Biologie, Humboldt-University, Berlin, Germany), strains PCC 7813 and PCC 7820 by Dr. R. Rippka (Pasteur Culture Collection, Institut Pasteur, Paris) and strain PCC 7806 by Dr. J. Weckesser (Albert-Ludwigs-University, Freiburg, Germany). Strains HUB 5-2-4, PCC 78 13, PCC 7820, and PCC 7806 were found to produce toxins. HUB X, HUB 18 and HUB 5-3 are non-toxic strains. E. coli XL-l was used as a recipient strain for plasmid transformation. E. coli C 600 hfl served as host strain for recombinant lambda gt 10 phages. pBluescript KS( - > (Stratagene, Heidelberg, Germany) was used to subclone DNA for sequencing. 2.2. Preparation of DNA from M. aeruginosa, hybridization DNA from strains of M. aeruginosa was isolated following the CTAB method described for plant cells [ 131 with modifications concerning the incubation times after adding 2 X CTAB, 0.2 vol 5 X CTAB and the precipitation buffer which were performed for 30 min, 10 min and 20 min, respectively. Centrifugation steps were carried out for appropriate times necessary to separate two phases or to sediment a pellet. Digested cyanobacterial DNA was separated on 0.8% agarose gels and vacuum-transfered to Amersham Hybond-N membranes as described [14]. Oligonucleotides core I and core II [9] were end-labelled by standard methods [14]. DNA fragments were labelled using the Megaprime kit (Amersham, Braunschweig, Germany). Southern hybridizations were performed as described [14]. 2.3. DNA cloning, sequencing PCR products were subcloned in pGEM-T vector (Promega, Madison, USA). Plasmid-DNA was prepared using Tip 20 columns (Diagen, Dusseldorf, Germany). To construct genomic libraries, total DNA from M. aeruginosa HUB 5-2-4 was digested with restriction endonucleases RsaI or A/u1 (Amersham, Braunschweig, Germany) and ligated into lambda gtl0 vector using the RiboCloneeEcoRI Adaptor Ligation System and Packagene@ Lambda DNA Packaging System (Promega, Madison, USA). Phage
297
K. MeiJner et al. / FEMS Microbiology Letters 135 Cl9961 295-303
DNA was prepared as described [14]. Unidirectional deletions with exonuclease III were generated using the double stranded nested deletion kit (Pharmacia, Freiburg, Germany). DNA was sequenced using Sequenase Vers. 2.0 kit (USB, Cleveland, USA). Homology searches in EMBL, Genbank and SwissProt nucleic acid/protein databases were performed using FASTNSCAN/FASTPSCAN and sequence alignments were assembled using the programmes CLUSTAL and PALIGN from the PCGENE software package. 2.4. Polymerase
chain reaction (PCR)
The oligonucleotides BSCI (GGAATTCC TTA AAI GCI GGA(C) GG(c)A GCI TAT G(C)TG CCG C(G)TT GAT CC) derived from core 1 sequence and BSCII (GGAATTCC TTT I(C)GG I(~)T-I ICC ICC&T TGT(A) ICC IGA I(C)GT GTA A(C) derived from core 2 sequence were kindly provided by Dr. M.A. Marahiel (Philipps University Marburg, Germany). PCR experiments required 10 ng Microcystis DNA and 20 pmol of BSCI and BSCII and were carried out as follows: incubation for 7 min at 94°C cooling down to room temperature, addition of Taq-polymerase (Boehringer, Mannheim, Germany),
30 cycles (2.5 min at 72°C 1.O min at 94°C and 1.5 min at 6O”C), incubation for 7 min at 72°C.
3. Results and discussion 3.1. Hybridization qf cyanobacterial DNA with oligonucleotides derived from two conserved sequence motifs of peptide synthetase genes We used oligonucleotides (core I and core II, [9]) corresponding to conserved amino acid sequences of adenylate-forming domains as probes for hybridization of Southern blots with the DNAs of toxic and non-toxic strains of M. aeruginosa. The sequence ‘SGTTG’ (corresponding to core II> is conserved in all polypeptides belonging to a group of enzymes called either the ‘family of adenylate-forming enzymes’ or according to the SWISSProt database definition the ‘AMP-binding enzymes’. The sequence ‘LKAGGA’ (core I) is highly conserved in peptide synthetases only, whereas other types of adenlyate-forming enzymes show a low degree of homology in this region [9]. Hybridization of the radioactively labelled probes with digested DNA from six different strains re-
PCRl PCR2 PCR3 PCR4
TGACTATCCCCCTGAACGTCTTCAGTTTATGTTAGACTTC TGACTATCCCATTGAACGTCTTCAATTTATGTTAGAAGATGTTT GGATTATCCTACCGAGCGCTTGGGGGATATCCTCTCAGATTCGGGGGTTT TAATTATCCCCCTGAGCGCTTGGATTATATGATATCCGATTCAGCTATTT ..* ***** ** ** * * ** . . . . . . *** . .l*
PCRl PCR2 PCR3 PCR4
CTTTTTTGATTACCCAACGTTCTTTATTAGCAAAATTGCC CATTTTTGATTACCCAACGTTCTTTATTGGCCGTATTACCTCCTTCTCAA CTTTGGTGTTAACTCAGGACTTTAGGGGATTTTCTTCCCCA?mYGGG CTCTGTTGTTAACTCAGCAATCCTTAGTCCAATTTCTGCCAGAA?iATCAA ._** *** . . * *..a*.*.** ** . . . *.** . . *
100 100 100 100
PCRl PCR2 PCR3 PCR4
GCGACTCTTATTTGTTTAGATCACATCCAAGAGCAGATTTCTCAATATTC GCTAATGTTATTTGTTTAGATGAGATTGAAGAGCAGGTTTCTCAATATCC GCCGAGTTACTGTGTTTAGATAGGGATTGGGAAAAGATAGCTACCTATAG GCCGAAATACTGTGTTTAGATACAGATTGGTCAAGGATAGCTAATTATAG . . * .********* l ** ** . *. . . . . . *.* ..**
150 150 150 150
PCRl PCR2 PCR3 PCR4
TCAAGATAATCTTCAAAGTGAGTTAACTCCTCCTTCCAATTTAGCTAACGTT TCAAGATAATCTTCAAAATGGATTAACGATTGCTAGTTTAGCGAATGTG TCCAGAAAATCCCTTCAATCTAACGACTCCTGAGAATTTAGCCTATGTT TCAAGAGAATTTGACCTCTCCAGTTACCCCAGAAAATTTAGCTTATGTT ** *** *** * * . **tt** .* ** . . . ** . .
50 50 50 50
199 199 199 199
Fig. 1. Nucleotide sequence alignment of four PCR products representing the regions between motifs core 1 and core 2 in putative peptide synthetase modules of M. aeruginosa. PCRl and 2 were amplified with DNA from the toxic strain HUB 5-2-4 and PCR3 and 4 were amplified with DNA from the toxic strain PCC 7820. Identical bases are indicated by asterisks, transitions by dots.
K. Megner et al. / FEMS Microbiology Letters 135 (19%) 295-303
298 PCRl ACUAEMRNI
- DYPPERLQFMLEDSQVPFL---ITQ----RSLLA-KLPASQATLICLDHI - ITTDEGKRGGKVIGTKRIVDEGLKQCPDVSTVLVYKRTGAEVPWTEGRDI * . .* .* *
PCRl ACUAEMRNI
- --QEQISQYSQDNLQSELTPSN-LANV - WWHEEVEKYPAYIAPDSVNSEDPLFLL * *. *
*
-42 -50
-66 -17
PCRl GRSBI
- DYPPERLQFMLEDSQVPFLITQRSLLAKLPASQATLICLDHIQEQISQYS - AYPQERIQYLLEDSGAALLLTQSHVLNKLPVD---IEWLDLTDEQNYVED ** ** ** ** *..**** *.** .* ***
PCRl GRSB4
- QDNLQSELTPSNLANV - GTNLPFMNQSTDLAYI ** . l*
-50 -47
-66 -63
.
Fig.2. Comparison of the amino acid sequence deduced from PCRl with sequences synthetase 2 (CiRSB4, [4]) and another adenylate-forming enzyme (acetyl-co&yme indicates identical residues, a dot indicates residue similarity.
vealed patterns of signals indicating that sequences similar to core 1 and core 2 occur in all strains of M. aeruginosa investigated (not shown) and suggesting that all strains (toxic and non-toxic) of M. aeruginosa carry sequences homologous to motifs of the adenylate-forming domain of peptide synthetase modules.
of the leucine-activating domain of gramicidin S [16]). An asterisk A synthetase, ACUAEMENI,
3.2. Amplification
of sequences from cyarwbacten’al DNA using primers derived from conserved sequences of peptide synthetases
Since sequences with homology to the conserved motifs core 1 and core 2 were obviously present in toxic and in non-toxic strains, we amplified by PCR
A
B
C
D
E
123456
123456
123456
123456
123456
11.5 kb-,
Fig. 3. Southern blot hybridization with radioactively labelled PCR products. A: Electrophoresis of XbaI digested DNA from M. aeruginosa: lane 1: HUB 18, lane 2: HUB X, lane 3: HUB 5-3, lane 4: PCC 7813, lane 5: PCC 7820, lane 6: HUB 5-2-4. B: Hybridization with PCRl. C: Hybridization with PCR2. D: Hybridization with PCR3. E: Hybridization with PCR4. Filters were washed with 0.1 SSC/O. 1 SDS at 60°C.
299
K. Meijher et al. / FEM.7 Microbiology Letters 135 (19%) 295-303
DNA regions located between these conserved motifs in the toxic strains HUB 5-2-4 and PCC 7820. These regions were described to be of less homology compared with the core motifs themselves. The PCR fragments (PCRl and 2 amplified from HUB 5-2-4, PCR3 and 4 from PCC 7820) were cloned and sequenced (Fig. 1). Four different PCR products
were obtained and compared with sequences of other peptide synthetases and adenylate-forming enzymes in the data bases. The homology of the four PCR products from M. aeruginosa DNA was dstinctly higher to corresponding regions between code 1 and core 2 of genes encoding peptide synthetases than to those sequences in genes for other adenylate-forming
M?+PEPl YDTSRFTQDTIERMAAHLQTILTGIVTDTRQRVTQLPILTTQDQHQLLVE-50 GRSBl
N--------------PHV------LVQD-------VPLLTKQEKQHLL~-23 *.** * * * *
****
BOX A MAPEPl
WNNTEADYPLDKSLHQLFEEQVAQNPQGIAVIFEGHKLTYQQ-L-100
GRSBl
LHDSITEYP-DKTIHQLFTEQVEKTPEWAWFEDEKVTlL-72 * l * ** *.** ,.** ** ,.**** ***
*
l
***
Box B MAPEPl
AHCLRDKGVGPESLVGIFMERSLEMVIGLLGI
LAAMUTVPLDPDYPTER-150
GRSBl
ARFLREKGVKKESIIGIMMERSVEMIVGILGI -IDPEYPKER-122 l ** *** ** ..** ****.**.,*,*********,**.**,**
**
MAPEP
LGDILSDSGVSLVLTQESLGDFLPQTGAELLCLDRDWEKIATYSP~PFN-200
GRSBl
IGYML-DS-VRLVLTQRHLKDKFAFTKETIVIEDPS---ISHELTEEIDY-l67 * ,* ** * ***** * * * * * . *
MAPEPl
LTTPENLAYVIYTSOBTGKPKGVMNIHRGICNTLTYTIGHYNITSEDRIL-25O
GRSBl
1NESEDLFY11YTSGTTGXPKGVMLEHKN1VNLLHFTFEKTN1NFSDKVL-217 * * * ***** ******** *, * * * * ** * .l
MAPEPl
QIISLSFDGSVWEIFSSLISGASLWAKPDGYKDIDYLIDLIVQEQ~YF-3OO
GRSBl
QYTNAVLTCVTKKFFSTLLSGGQLYLIRKETQRDVEQLFD-267 * **.* ** * . . . .* . . * **
Box C
*..
MAPEPl
TCVPSILRVFLQHPK--SKDCHCLKRVIVGGE--ALSYELNQRFFQQLNC-346
GRSBl
SFPVAFLKFIFNEREFINRFPTCVKHIITAGEQLVVNNEFK-RYLHEHNV-316 **. .* ** * * . . *. . .
BOX
*
D
MAPEPl
ELYNAYGPTEVAVETTIWCCQPNSQIS----IGTPIANAQVYILDSYLQP-392
GRSBl
HLHNHYGPSETHWTTY-TINPEAEIPELPPIGKPISNTWIYILDQEQQL-365 * * *** * * ** ** **,*. .**** .* .*
*
Fig. 4. Alignment of MAPEPl with the amino acid sequence of the proline-activating synthetase unit of gramicidin S synthetase 2 (GRSB 1, [4]). Sequences corresponding to the conserved motifs of peptide synthetases (cf. [20]) are printed in bold type. The nucleotide sequence encoding MAPEPl has accession number 228338 in the EMBL data library (Heidelberg). Asterisks indicate identical residues, dots indicate similar residues.
K. Mei&er et al. / FEMS Microbiology Letters 135 11996)295-303
300
enzymes like 4-coumarate-CoA-synthetase [15] and acetyl-CoA-synthetases [16]. As an example, the comparison of the deduced amino acid sequence of PCRl from M. aeruginosu HUB 5-2-4 with corresponding sequences of the leucine-activating unit of gramicidin S-synthetase 2 from B. breuis [4] and acetyl-coenzyme A synthetase from Aspergillus niduZuns [16] is shown in Fig. 2. Accordingly, we conclude that we cloned sequences that form part of four different adenylate-forming domains of peptide synthetase genes of toxic Microcystis strains.
3.3. Hybridization of cyanobacterial DIVA with the amplified fragments
In order to investigate the strain-specific distribution of these putative adenylate-forming domains, the PCR products were radioactively labelled and used as hybridization probes. Interestingly, the four fragments generated different patterns with DNAs from the strains investigated. All four fragments hybridized with specific fragments of digested total DNA of strain HUB 5-2-4. In addition, PCR frag-
Box E
Box F
MAPEPl
VPIGVAGI#LHIGGMQLARGYLWRL~LTQEKFISNPF-AQ-441
GRSBl
QPQGIVt3ELYIS GMVGRQYLMU~~LTAI!XFFADPFRPNERMYRTGDLAR-415 . ***** *** *** ** * * . *** * * . . .* .******
MAPEPl
YLPEGNIEnoRIDN~VKLRGLRIIU&EIQTVLETHPNVEQTWIMREDS-49l
GRSBl
WLPEGNIEFLQFtADliQVKIRG~I~LQEIEAQLLNCKGVKEAWIDKADD-465 * .**.****,*** * ***.** ***t*** . * *** . *
MAPEPl
LYNQRLVAYVIRKNSLLTPQDLRRFLQQQLPAYMVPSAFVLLSASPLNNlU-541
GRSBl
KGGKYLCAYVVMEVE * *** .
MAPEPl
GKIDFtKKLPIPDETSIIESAYIAPRNEKESLLAQIWEDVLQVSKIGVSDN-59l
GRSBl
QKIDRKSLPNLEGIVNTNAKYWPTNELEEKLAKIWEEVLGISQIGIQDN-564 ****** ** . . * . * ** * ** *** .** .* **
Box G
Box H
Box
V-NDSELREYLGKALPDYMIPSFFVPLDHVRLIiLN-514 ** ** ** ** * * * . .** .*
I
**
Box J MAPEPl
FFELGI3%5LKiUSLVSKIQEKLGQSLPIKQVFAHPTIAEQAALLSTVTPL-64l
GRSBl
FFSLQGHSLK?AITLISRMNKECNVDIPLRLLFEAPTIQEISNYINGAKKE-~~~ ** ***********.*_,_ .* . . .* *** * . .
MAPEPl
TVATIPLVSAQETYETSBAQ
GRSBl
SYVAIQPVPEQEYYPVSSVQKRMFILNEFDRSGTAYNLPGVMFLDGKLNY-664 l *** * **. . .* * ** , . .* . . *
MAPEPl
DVFEKAIQLLISRHEBLRTSFILINGEPQQKILQNPSFWIGNFKDLDKIN-74l
GRSBl
RQLEZAVKKLVXREZALRTSFHSINGEPVQRVHQNVELQIAYSE-----S-709 * * . * . ***.***** l **** *.. ** *
Box K
Box
RRFYVLQQMDLNNVAYHIVSTLKIAGDFSP-691
L
Fig.4.(continued).
301
K. MeiJner et al. / FEMS Microbiology Letters 135 (19961295-303
ments 3 and 4 were found to generate signals with fragments of all toxic strains (HUB 5-2-4, PCC 7820, PCC 78131, but not with DNA from non-toxic strains. PCR fragment 1 hybridized with DNA fragments of the toxic strain (HUB 5-2-4) and three non-toxic strains (HUB 18, HUB 5-3, HUB X) (Fig. 3) supporting our conclusion that both types of strains possess genes encoding peptide synthetases. While peptide synthetases in strains producing hepatotoxins could be involved in toxin formation, peptide synthetases of non-toxic strains have to serve other purposes, i.e. are probably involved in the synthesis of non-toxic peptides. It is known from previous studies that M. aeruginosa indeed produces also non-toxic peptides of unknown function [17,18]. Therefore, we suggest that Microcystis strains carry specific sets of genes for peptide synthetases.
3.4. Identification of sequences potentially an entire peptide synthetase unit
PCR product 3 which hybridized selectively to DNA from toxic strains was used to identify corresponding sequences in genomic libraries of M. aeruginosa HUB 5-2-4. Recombinant phages carrying a 1157 bp RsaI fragment, a 2.8 kb and a 2.0 kb A/u1 fragment (overlapping with the 5’- and the 3’-end of the RsaI fragment), respectively, were identified and sequenced. The continous sequence of 2982 bp determined so far has been deposited in the EMBL data library with accession number ‘228338. The amino acid sequence of the potentially iencoded polypeptide (called MAPEPl) was compared with sequences in the EMBL and SwissProt data bases. The comparison revealed significant homology (34
Box M MAPEPl
PDEEILETIAKERKPFDLEKSPLVRSKIYKLSPNEYILEL~I~IIC~-79l
GRSBl
TEDQVERIIAEFMQPFALEVRPLLRVGLVKLEAERHLFIMDMIiliIIS~-l59 ** ** ** ** .* ..,**** ** . ** .. .
MAPEPl
SMSLLAKECLQYYSDLAKGLQPSIEPLPIQYKDYAGWPNNLLRS---ENN-838
GRSBl
SMQIMIQE----IADLYKEKE--LPTLGIQYKDFTVWHNRLLQSDVIEKQ-803 ** * * ***** . * * ** * **.. *
*
MAPEPl
PPNLDYWRPKLDNGQLTRVHLPTDFKRPQIKTFKGSHLSWTFDRETISKL-888
GRSBl
EASLAERICRRDSSIESTDRLP---KYQPFKALMVKDLH-SVQESSLWM-849 * * ** * * .* ..
.
Box N MAPEPl
RKICQEGRNTLFMAL-VAPVQILLYRYSGRZZDITIGT~IA'MkSHPQLQSL-937
GRSBl
YTRWQQKQEQHYIWFYLLRIMFFLSKYSGQDDIWGTPIAGRSHADVENM-899 * .*** * ** ** l * *** .. . .
BOX
.
0
MAPEPl
IGLFLNTLVIRDQIEPEKGYKNLLAKVRQTVTEALEHSDYPFDILVEKLP-987
GRSBl
LmRSRLNNEDTFKDFLANVKQTALHAYENPDYPFDTLVEKLG-949 * * ** *.** * * ***** .* .* .*** **
MAPEPl
VSRAAAT-994
GRSBl
IQRDLSR-956 *
encoding
Fig.4.(continued)
*****
.
302
K. MeiJner et al. / FEMS Microbiology Letters I35 (1996) 295-303
to 44% identity) to peptide synthetases from other organisms. Interestingly, the highest homology was found not to an adenylate-forming part of the only other peptide synthetase hitherto known from a Gram-negative bacterium (enterobactin synthetase of E. co& [19]) but to the proline-activating module of gramicidin S synthetase 2 from B. breuis [4] (Fig. 4). We could identify the conserved motifs typical of adenylate-forming domains observed in peptide synthetases and other adenylate-forming enzymes (motifs A-I, Fig. 4). Moreover, all motifs specific for peptide synthetases (motifs J-O; Fig. 4) were recognized in the sequence from M. aeruginosa, too, classifying its putative product as a unit of a peptide synthetase. The consensus motifs J to 0 differ between epimerizing and non-epimerizing peptide synthetase units [20]. Motifs P and Q have hitherto been observed only in enzymes with an epimerizing function [20]. A comparison of the deduced amino acid sequence from M. aeruginosu with the amino acid sequences of epimerizing and non-epimerizing peptide synthetases revealed a significantly higher homology to non-epimerizing peptide synthetase units (see conserved motifs J-O in Fig. 4, [20]) grouping the sequence from &l. ueruginosu clearly with the peptide synthetases units without epimerizing function. Our results suggest that non-ribosomal peptide synthesis using multienzyme complexes of the conserved type occurs in cyanobacteria. Furthermore, strains of M. ueruginosu were found to differ in their content of sequences belonging to adenylateforming domains of peptide synthetases. The DNA fragments isolated from HUB 5-2-4 potentially coding for a complete unit of a peptide synthetase were observed to hybridize with DNA from all toxic strains (HUB 5-2-4, PCC 7806, 7813, 7820) but not with DNA from the non-toxic strains (HUB X, 18, 5-3) applying stringent conditions (not shown). Obviously, the putative gene product exists exclusively in the investigated microcystin-producing strains of M. ueruginosu. Therefore, the ability of a strain to produce microcystins may depend first of all on its possession of the gene(s) required. Further studies have to determine whether their expression is controlled by environmental conditions. Although the number of investigated strains was not large enough
to rule out the possiblity that there is no causal relationship between the presence of this sequence in certain strains and their ability to produce toxins, one should emphasize that we dealt mostly with entirely unrelated strains, a fact increasing the significance of our observation. Most of the investigated toxic and non-toxic strains were originally collected at completely different locations [12]. Thus, the analysed nucleotide sequence may encode a peptide synthetase unit of a multienzyme complex involved in the production of microcystins.
Acknowledgements We are indebted to Dr. M. Marahiel for providing the oligonucleotides core I, core II, BSCI, and BSCII and to Dr. H. von Dijhren for stimulating discussions. We are grateful to Drs. H.-J. Kohl, R. Rippka, and J. Weckesser for providing cyanobacterial strains. This work was supported by a NaFijG fellowship of the Berlin Senate to K.M. and by a grant of the Deutsche Forschungsgemeinschaft to T.B.
References [I] Carmichael, [2]
[3]
[4]
[5]
[6]
[7]
W.W. (1994) The toxins of cyanobacteria. Scient. Am. 270, 78-86. Sivonen, K., Namikoshi, M., Evans, W.R., Fardig, M., Carmichael, W.W. and Rinehart, K.L. (1992) Three new microcystins, cyclic heptapeptide hepatotoxins, from Nosfoc sp. strain 152. Chem. Res. Toxicol. 5, 464-469. Kleinkauf, H. and von DGhren, H. (1990) Nonribosomal biosynthesis of peptide antibiotics. Eur. J. Biochem. 192, I-15. Turgay, K., Krause, M. and Marahiel, M.A. (1992) Four homologous domains in the primary structure of GrsB are related to domains in a superfamily of adenylate-forming enzymes. Mol. Microbial. 6, 529-546. Babbitt, P.C., Kenyon, G.L., Martin, B.M., Charest, H., Sylvestre, M., Scholten, J.D., Chang, K.-H., Liang, P.-H. and Dunaway-Mariano, D. (1992) Ancestry of the 4-chlorobenzoate-dehalogenase: analysis of amino acid sequence identities among families of acyl:adenyl ligases, enoyl-CoA hydratases/isomerases, and acyl-CoA thioesterases. Biochemistry 3 I, 5594-5604. Dieckmann, R., Lee, Y.-O., van Liempt, H., von Diihren, H. and Kleinkauf, H. (1995) Expression of an active adenylateforming domain of peptide synthetases corresponding to acyl-CoA-synthetases. FEBS Lett. 357, 212-216. Stachelhaus, T. and Marahiel, M.A. (1995) Modular structure
K. MeiJner
et al. / FEMS Microbiology
of genes encoding multifunctional peptide synthetases required for non-ribosomal peptide synthesis. FEMS Microbiol. Lett. 125, 3-14. [8] Tutgay, K. and Marahiel, M.A. (1994) A general approach for identifying and cloning peptide synthetase genes. Peptide Res. I, 238-241. [9] Borchert. S., Patil, S.S. and Marahiel, M.A. (1992) Identification of putative multifunctional peptide synthetase genes using highly conserved oligonucleotide sequences derived from known synthetases. FEMS Microbial. Lett. 92, 175180. [lo] Van der Westhuizen, A.J.J.N. and Eloff, J.N. (1985) Effect of temperature and light on the toxicity and growth of the blue-green alga Miciocystis aeruginosa (UV-006). Planta 163. 55-59. [III Watanabe, M.F. and Oishi, S. (1985) Effects of environmental factors on toxicity of a cyanobacterium (Microcystis aeruginosa) under culture conditions. Appl. Environ. Microbiol. 49, 1342- 1344. [121Schwabe, W., Weihe, A., Biimer, T., Henning, M. and Kohl, J.-G. (1988) Plasmids in toxic and nontoxic strains of the cyanobacterium Microc~.stis aeruginosa. Cm. Microbial. 17. 133-137 1131Rogers. S.O. and Bendich, A.J. (1985) Extraction of DNA from milligram amounts of fresh, herbarium and mummified tissues. Plant Mol. Biol. 5, 69-76. 1141 Sambrook, J., Fritsch. E.F. and Maniatis, T. (1989) Molecular cloning: A laboratory manual. 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
Letters
135 (1096) 295-303
303
[15] Lozoya, E., Hoffmann, H., Douglas, C., Schulz, W., Scheel, D. and Hahlbrock, K. (1988) Primary structure and catalytic properties of isoenzymes encoded by the two 4coumarate:CoA ligase genes in parsley. Eur. J. Biochem. 176, 661-667. [ 161 Connerton, I.F., Fincham, J.S., Sandeman, R.A. and Hynes, M.J. (1990) Comparison and cross-species expression of the acetyl-CoA synthetase genes of the ascomycete fungi, Aspergillus niduluns and Neurospom crassa. Mol. Microbial. 4, 451-460. [171 Birk, 1.M.. Dierstein, R., Kaiser, I., Matem, U.. Konig, W.A.. Krebber, R. and Weckesser, J. (1989) Noutoxic and toxic oligopeptides with Bamino acids and unusual residues in Microcystis aeruginosa PCC 7806. Arch. Microbial. 15 I, 41 l-415. El81Tsukamoto. S., Painuly, P.. Young, K.A.. Yang, X. and Shimizu, Y. (1993) Microcystilide A: a novel cell-differentiation-promoting depsipeptide from Microcystis aeruginosa NO-15-1840. J. Am. Chem. Sot. 115, 11046-llod7. 1191Rusnak, F., Sakaitani, M., Drueckhammer. D., Reichert, J. and Walsh, C.T. (1991) Biosynthesis of the Eschericlzia coli siderophore enterobactin: sequence of the entF gene, expression and purification of EntF, and analysis of covalent phosphopantethein. Biochemistry 30. 2916-2927. M., von Dohren, H. and DO1Pfeifer, E., Pavela-Vrancic, Kleinkauf. H. (1995) Characterization of tyrocidine synthetase I (TY I ): requirement of post-translational modification for peptide biosynthesis. Biochemistry 34. 7450- 7459.