Study of the spiramycin biosynthesis and its regulation in streptomyces ambofaciens

2.The biosynthesis of secondary metabolites by Streptomyces and related actinobacteria.. The isolation of secondary metabolites biosynthetic gene clusters in Streptomyces 29 2.4.. The

N°D’ORDRE

UNIVERSITE DE PARIS-SUD XI UFR SCIENTIFQUE D’ORSAY

THESE Présentée pour obtenir

LE GRADE DE DOCTEUR EN SCIENCE DE L’UNIVERSITE

DE PARIS-SUD XI

PAR

HOANG CHUONG NGUYEN

SUJET : Study of the spiramycin biosynthesis and its regulation in

N°D’ORDRE

UNIVERSITE DE PARIS-SUD XI UFR SCIENTIFQUE D’ORSAY

THESE Présentée pour obtenir

LE GRADE DE DOCTEUR EN SCIENCE DE L’UNIVERSITE

DE PARIS-SUD XI

PAR

HOANG CHUONG NGUYEN

SUJET : Study of the spiramycin biosynthesis and its regulation in

Remerciements

J’ecris ces remerciements en Français parce que, simplement, les gens auxquels je tiens à exprimer ma gratitude sont français ou bien qu’ils connaissent bien la langue française En plus, je trouve que le français est une langue qui convient bien pour exprimer des sentiments

La plupart de ce travail a été réalisé dans le Laboratoire de Microbiologie Moléculaire des Actinomycètes de l’Institut de Génétique et Microbiologie, UMR CNRS 8621 en collaboration avec le Laboratoire de Génétique de l’Université des Sciences Naturelles à Ho Chi Minh Ville dans le cadre d’une thèse en co-tutelle

Tout d’abord, je suis très reconnaissant à mes directeurs de thèse, Madame Thuy Duong Ho Huynh et Monsieur Jean-Luc Pernodet Madame Ho Huynh a initié cette collaboration, m’a donné la possibilité de venir travailler en France et m’a aidé pour toutes les démarches administratives Elle a toujours suivi mon travail et m’a conseillé Je tiens à remercier tout particulièrement Monsieur Pernodet pour la façon dont il a dirigé ma thèse : sa confiance en moi, sa grande disponibilité, sa générosité, son enthousiasme et ses connaissances qu’il a partagées avec moi Grâce à lui, j’ai pu également découvrir pleinement la vie ‘‘à la française’’ et passer des moments mémorables un peu partout en France, en Angleterre et

J’exprime mes profonds remerciements à tous les membres passés ou présents du Laboratoire de Microbiologie Moléculaire des Actinomycètes, du Laboratoire de Métabolisme Energétique des Streptomyces et de l’institut de Génétique et Microbiologie : Fatma Karray pour m’avoir guidé quand j’ai débuté mon travail de thèse, Sylvie Lautru pour sa grande aide aussi bien pour les expériences, en particulier celles de chimie analytique, que pour la rédaction, Muriel Decraene pour toutes les formalités françaises compliquées pour lesquelles

je ne sais pas me débrouiller sans elle, et aussi tous les autres pour leurs conseils, leur sympathie et leur soutien, Karine Tuphile, Annick Friedmann, Alain Raynal, Michel Cassan, Claude Gerbaud, Emmanuelle Darbon-Rongère, Marie-Hélène Blondelet-Rouault, Josette Gagnat, Catherine Esnault, Marie-Joëlle Virolle, Sylvain Pendino, Cécile Martel, Jean-Denis

Le Manach, Thierry Locatelli

Je tiens également à remercier les jeunes des labos, Sarka, Maud, Nicolas, Céline, Aleksandra, Noriyasu, Aleksei, Hanane, Amélie, Fabien, Emilie, Florence, Hasna, Delin qui ont suivi au fil de ma thèse mes espoirs aussi bien que mes découragements et m’ont fait bénéficier de leur aide, leur soutien constant et leur bonne humeur

Ces quatre dernières années ont été riches en moments de convivialité Les nombreux barbecues, les sorties (culturelles et/ou gastronomiques) et les soirées de fêtes resteront des souvenirs inoubliables

2.The biosynthesis of secondary metabolites by Streptomyces

and related actinobacteria 22

2.1 The importance of secondary metabolites for medicine

2.2 Biological activities of secondary metabolites

from Streptomycesand related actinobacteria 23

2.3 The genetics of secondary metabolism in Streptomyces 29

2.3.1 Initial genetic studies of secondary metabolism in S coelicolor 29

2.3.2 The isolation of secondary metabolites biosynthetic gene clusters in Streptomyces 29

2.4 The biosynthesis of polyketides 31

2.4.1 The assembly-line enzymology of polyketide synthases 31

2.5 The biosynthesis of nonribosomal peptides 38

2.6 The regulation of secondary metabolism in Streptomyces 40

2.6.1 Multiple factors influence the onset of secondary metabolism 40

2.6.4 Regulatory proteins controlling secondary metabolism in Streptomyces 45

3 Macrolide antibiotics and their biosynthesis in Actinobacteria 47 3.1 Definition and classification of macrolide antibiotics 47 3.2 The mode of action of macrolide antibiotics 50 3.3 Resistance to macrolide antibiotics 52

3.4 Biosynthesis of macrolides 55

3.4.4 Enzymes involved in glycosylation during macrolide biosynthesis 60

3.5 Combinatorial biosynthesis of new macrolide antibiotics 62

3.5.1 Combinatorial biosynthesis of the macrolactone ring 63 3.5.2 Modification of the glycosylation pattern through combinatorial biosynthesis 66

3.6 The regulation of macrolide biosynthesis 68

3.6.2 Regulation of methymycin/pikromycin biosynthesis 69

3.7 Spiramycin biosynthesis in S ambofaciens 72

Chapter II: GLYCOSYLATION STEPS DURING SPIRAMYCIN BIOSYNTHESIS

IN STREPTOMYCES AMBOFACIENS: INVOLVEMENT OF THREE

GLYCOSYLTRANSFERASES AND TWO AUXILIARY PROTEINS 75

Chapter III: A POST-PKS PLATENOLIDE KETOREDUCTASE IS INVOLVED

IN SPIRAMYCIN BIOSYNTHESIS IN STREPTOMYCES AMBOFACIENS 103

Chapter IV: REGULATION OF SPIRAMYCIN BIOSYNTHESIS

IN STREPTOMYCES AMBOFACIENS 122

Chapter V: TRANSCRIPTIONAL ORGANIZATION OF THE SPIRAMYCIN CLUSTER

AND ACTIVITIES OF THE TRANSCRIPTIONAL ACTIVATORS SRMR AND SRMS 151

RESUME DE LA THESE EN FRANÇAIS 182

Chapter I

Introduction

1 The genus Streptomyces

1.1 Taxonomy

Streptomyces are Gram-positive bacteria They belong to the family of Streptomycetaceae that comprises also the genera Kitasatospora, Parastreptomyces, Streptacidiphilus, and Trichotomospora Streptomycetaceae is the only family of the suborder Streptomycineae that belongs to the order Actinomycetales (Figure 1) (Stackebrandt et al., 1997) This order belongs to the class of Actinobacteria, whose members are defined as

Gram-positive bacteria with a high G+C content in their DNA

Figure 1: Phylogenetic relatedness of the families of the class Actinobacteria Interclass relatedness of Actinobacteria, showing the orders and the 10 suborders of the order Actinomycetales, is based upon rDNA/rRNA sequence comparison The scale bar represents

5 nucleotide substitutions per 100 nucleotides From (Stackebrandt et al., 1997)

Actinobacteria, which constitute one of the largest bacterial phyla, include microorganisms exhibiting a wide spectrum of morphologies (coccoid: Micrococcus; rod- coccoid: Arthrobacter, fragmenting hyphal forms: Nocardia; highly differentiated branched mycelium: Streptomyces) and possessing highly variable physiological and metabolic

properties Various lifestyles are also encountered among this class, which includes

pathogens (e.g Mycobacterium spp., Corynebacterium spp., Tropheryma whipplei), soil inhabitants (e.g Streptomyces), plant commensals (Leifsonia spp.), nitrogen-fixing symbionts

of plants (Frankia) and gastrointestinal tract bacteria (Bifidobacterium spp.)

Several actinobacterial genomes have been sequenced and the diversity of

Actinobacteria is also visible in the characteristics of these genomes (Table 1) The sizes of these genomes vary from less than 1 Mb (Tropheryma whipplei) to more than 10 Mb (Streptomyces scabies) Most actinobacterial genomes are circular, but the genomes of Streptomyces are linear The G+C content is generally high, from 53% for Corynebacterium

to over 70% for Streptomyces, but it is of only 46% for the small genome of the obligate pathogen Tropheryma whipplei

Microrganism Genome

size (Mb)

Number of genes

% G+C Circular

(C) or linear (L)

Reference or Acc N°

Streptomyces scabies 10.14 Not yet

annotated

uk/Projects/S_scabies/

Table 1: Some characteristics of the genome from selected representatives of the

Actinobacteria

Among Actinobacteria, the genus Streptomyces groups mycelial spore-forming

bacteria and identified by both chemotaxonomic and phenotypic characters including 16S

rRNA homologies, cell wall analysis, fatty acid and lipid patterns (Wellington et al., 1992; Williams et al., 1989) Streptomycetes are believed to have originated about 440 millions

years ago (Embley & Stackebrandt, 1994)

1.2 The genome of Streptomyces

The genome of Streptomyces is remarkable for several reasons: its size, its high G+C

content, the linearity of the chromosome and the organization of the genes on the chromosome Moreover, some of type of accessory genetic elements has so far been found

only in Actinobacteria

1.2.1 Streptomyces chromosomes

The Streptomyces chromosomes have sizes in the range 8-10 Mb (see table 1) This

is about twice the size of the chromosome of bacteria such as Escherichia coli or Bacillus subtilis The number of genes carried by this chromosome is also quite large and S coelicolor (7 855 genes) was the first example of a bacteria possessing more genes than a unicellular eukaryote such as Saccharomyces cerevisiae (6 023 genes)

The average G+C content of Streptomyces chromosomal DNA is in the range 70-72

% (Table 1) In coding sequences, the codon usage is strongly biased towards codons with C

or G in the third position; this bias is used to predict probable coding sequences among all possible open reading frames (Wright & Bibb, 1992)

The linearity of the Streptomyces chromosome was first established for S lividans (Lin et al., 1993) and then demonstrated in other species including S ambofaciens (Leblond

& Decaris, 1994) and S coelicolor (Redenbach et al., 1996) It is now considered that all Streptomyces have a linear chromosome, but this is not true for all related mycelial actinobacteria with a large genome, as for instance Saccharopolyspora erythraea, which has

a circular chromsome (Oliynyk et al., 2007) Linearity has not been apparent from the circular genetic linkage map of S coelicolor (Hopwood, 1967) The Streptomyces linear chromosome possess terminal inverted repeats (TIR), the size of which varies from 174 bp for S avermitilis (Ikeda et al., 2003) to several hundred of kilobases, e.g 198 kb for S ambofaciens (Choulet et al., 2006) and about 550 kb for S rimosus (Pandza et al., 1997)

The chromosome is replicated bidirectionally from the oriC located in the centre of chromosome towards the telomeres at the two extremities (Jakimowicz et al., 2000) A

protein (terminal protein, TP) is covalently bound to both free 5’ ends This terminal protein, together with a telomere associated protein (TAP), is required for the synthesis of the last Okazaki fragment of the lagging strand when bidirectional DNA replication reaches the free ends

A remarkable feature of Streptomyces chromosomes is their highly

compartmentalized genetic organization This was already apparent with the sequence of the

S coelicolor chromosome, where all house-keeping genes were located in the central part (core region) of the chromosome (Figure 2) In S coelicolor no essential gene, with the single exception of argG, is located within 1.3 Mb of each extremity

Figure 2: Representation of the S coelicolor chromosome The outer scale is numbered

anticlockwise in megabases and indicates the core (dark blue) and arm (light blue) regions of the chromosome Circles 1 and 2 (from the outside in), all genes (reverse and forward strand, respectively) colour-coded by function (black, energy metabolism; red, information transfer and secondary metabolism; dark green, surface associated; cyan, degradation of large molecules; magenta, degradation of small molecules; yellow, central or intermediary metabolism; pale blue, regulators; orange, conserved hypothetical; brown, pseudogenes; pale green, unknown; grey, miscellaneous); circle 3, selected 'essential' genes (for cell division, DNA replication, transcription, translation and amino-acid biosynthesis, colour coding as for circles 1 and 2); circle 4, selected 'contingency' genes (red, secondary metabolism; pale blue, exoenzymes; dark blue, conservon; green, gas vesicle proteins); circle 5, mobile elements (brown, transposases; orange, putative laterally acquired genes); circle 6, G + C content; circle 7, GC bias ((G - C/G + C), khaki indicates values >1, purple

<1) The origin of replication (Ori) and terminal protein (blue circles) are also indicated From

(Bentley et al., 2002)

The comparative analysis of Streptomyces chromosome sequences revealed that the

core regions are highly syntenic and contain most of the genes conserved with other

Actinobacteria while the subtelomeric regions are species-specific (Bentley et al., 2002; Choulet et al., 2006; Ikeda et al., 2003) This is illustrated by the comparison of the S coelicolor and S avermitilis chromosomes presented in figure 3

Figure 3: Synteny between S avermitilis and S coelicolor A3(2) linear chromosomes Each

point in this figure is a reciprocal best hit These hits were obtained by pairwise BLASTP

searches of predicted S avermitilis proteins against those of S coelicolor A3(2) Each

protein pair is graphed according to the location of the corresponding gene on respective DNA molecules The bars above and at right of plot indicate the region conserved to circular Actinobacterium chromosomes (green), subtelomeric (blue), and backbone (red) regions in

S avermitilis and S coelicolor A3(2), respectively Arrows indicate the position of oriC A, B, and C indicate inverted regions between S avermitilis and S coelicolor A3(2) From (Ikeda

et al., 2003)

Comparison of the chromosome of S avermitilis with those of the two closely related species S coelicolor and S ambofaciens revealed that the size of the central region

conserved between species decreases as the phylogenetic distance between them

increases, whereas the specific terminal fraction reciprocally increases in size (Choulet et al.,

2006) Between highly syntenic central regions and species-specific subtelomeric regions, there is a notable degeneration of synteny due to frequent insertions/deletions

These terminal regions also suffer large deletions and amplifications which can occur

spontaneously with high frequency (Altenbuchner & Cullum, 1985; Chen et al., 2002; Schrempf et al., 1989) Regions of more than several hundreds of kilobases can be deleted

in the terminal parts without any deleterious effects on the viability of Streptomyces strains in

laboratory conditions (Volff & Altenbuchner, 1998) Interestingly, deletions affecting the

terminal regions can lead to chromosome circularization (Catakli et al., 2003; Lin et al., 1993; Redenbach et al., 1993)

1.2.2 Streptomyces extrachromosomal elements

Many Streptomyces strains contain plasmids, most of which are phenotypically

cryptic These plasmids can be circular or linear Linear plasmids have sizes in the range of

10 to 600 kb As the chromosome, they have a centrally located replication origin, terminal inverted repeats and depend for their replication on a covalently bound terminal protein and

on a telomere associated protein (Chater & Kinashi, 2007) Among circular plasmids, the larger ones (e.g SCP2, 31 kb) have a low copy number and replicate bidirectionally by a θ mode of replication The smaller plasmids (e.g pIJ101, 8.9 kb) tend to have higher copy numbers and replicate by a rolling circle mechanism via a circular, single-stranded replication

intermediate Their replication is initiated by a plasmid-encoded replication protein (Kieser et al., 2000a) Plasmid cloning vectors have been developed mostly from small circular

plasmids, but some vectors are derived from SCP2

A novel type of mobile genetic element was first discovered in Streptomyces and later

in some other Actinobacteria: the actinomycete integrative and conjugative elements (AICEs) (te Poele et al., 2008) These genetic elements belong to a broad class of integrative and conjugative elements (ICEs) (Burrus et al., 2002), which have both plasmid-and

bacteriophage-like features ICEs are normally integrated in the host chromosome, but they have the ability to excise, conjugate to a new host and integrate in the new host chromosome

by site-specific recombination AICEs represent a special class of ICEs, because unlike other ICEs, they have the ability to replicate autonomously like a plasmid For some of the AICEs, the integrated and replicative forms can even coexist Most of the AICEs integrate in a specific tRNA gene in the host chromosome and this gene is not inactivated after integration

because the AICE and the chromosome share a segment of identity (Mazodier et al., 1990) The most studied representatives of the AICEs are SLP1 from S coelicolor (Bibb et al., 1981), pSAM2 from S ambofaciens (Pernodet et al., 1984) and pMEA300 from Amycolatopsis methanolica (Vrijbloed et al., 1994) As they are able to replicate and to integrate site-specifically, AICEs have been used to develop cloning vectors (Cohen et al., 1985; Smokvina et al., 1990; Vrijbloed et al., 1995)

Most of the plasmids and all of the AICEs are conjugative mobile genetic elements

But the conjugation mechanism in Streptomyces is very different from the one in other

bacteria, the paradigm of which is the transfer mechanism of the F sexual factor from

Escherichia coli For the transfer of F, the products of about 30 transfer genes are required (Frost et al., 1994) These gene products first establish a stable mating pair via cell-to-cell

junctions DNA transfer is initiated by the relaxase that nicks DNA at the origin of transfer

(oriT) to generate a single-stranded molecule The resulting structure, the relaxosome, is

then coupled to the transferosome (related to type IV secretion systems) by the intermediate

of a coupling protein In Streptomyces, the conjugation machinery is actually much simpler

with a single membrane-associated ATPase (Tra) similar to DNA translocases, such as

SpoIIIE of Bacillus subtilis (Wu & Errington, 1997), sufficient for plasmid transfer (Reuther et al., 2006) A plasmid-borne cis-acting sequence of about 50 bp, called clt, is recognized by Tra (Reuther et al., 2006) Another original feature is the transfer of double-stranded DNA (Possoz et al., 2001) It was confirmed recently that no DNA cleavage occurs at the clt locus (Ducote & Pettis, 2006) The protein Tra is located at the tip of the growing hyphae (Reuther

et al., 2006) Conjugative mobile genetic elements are also able to mobilize chromosomal gene markers, but in contrast to the recent knowledge on plasmid transfer in Streptomyces,

chromosomal DNA mobilization is very poorly documented, except that it was proven that conjugative elements stimulate chromosomal DNA transfer (Pettis & Cohen, 1994)

1.2.3 Streptomyces phages

Most of Streptomyces phages have been isolated from soil because nearly all soil samples readily yield phages that give plaques on S lividans 66, a generally permissive host for Streptomyces phages rare (Kieser et al., 2000a) Besides, Streptomyces themselves

provide a second source of phages because some phages have been discovered after release by natural lysogenic strains Phages isolated from soil have generally a wide host–range In contrast, phages isolated from lysogenic strains have generally a narrow host-

range Nearly all the Streptomyces phages examined have polyhedral heads and long tails

As with other bacteriophages of similar morphology, they contain double-stranded DNA

(Lomovskaya et al., 1980) Streptomyces phages are both lytic and temperate and it is infrequent that Streptomyces phages infect other genera of actinomycetes Most phages are

active on germ tubes and young mycelium but some can be active on old mycelium

1.3 Morphological development of Streptomyces

Streptomyces are mycelial spore-forming bacteria, remarkable for their morphological

development and considered to be among the most complex of bacteria (Chater & Chandra, 2006; Flardh & Buttner, 2009) A spore germinates to give a vegetative mycelium Then aerial hyphae, which are at least partially parasitic on the vegetative mycelium, emerge

These aerial hyphae finally differentiate to form chains of spores These steps are presented

in Figure 4 and the resulting developmental cycle is represented in Figure 5

Figure 4: Mycelial forms during the development of Streptomyces From (Flardh & Buttner,

been identified Much of these genes were identified by the isolation and characterization of

S coelicolor mutants that failed complete normal development and were affected either in aerial growth (“bald” or bld mutants) or in the formation and maturation of spores (“white” or whi mutants) Most of the genes thus identified are regulatory genes and their interactions

have been partly characterized The different steps of the developmental cycle and the major genes involved are detailed below

When a spore encounters favourable conditions, it germinates and one or two germ tubes emerge These germ tubes grow as thread-like hyphae by tip extension and branching

to form a vegetative or substrate mycelium This mycelium often goes deep into the surrounding substrate New cell wall material is synthesized only at the hyphal tip This process involves the coiled-coil protein DivIVA DivIVA acts as a landmark protein that recruits directly or indirectly the cell wall biosynthetic machinery The hyphal tip is probably also a hot spot where peptidoglycan, teichoic acid, cell-surface proteins, and membrane lipids are secreted and assembled Apart from DivIVA, there are other proteins residing at the hyphal tip regions and participating in the apical growth such as the cellulose synthase-like protein CslA which interact with DivIVA to add a so far uncharacterised β-linked glucan in the cell envelope, or, in a later stage, the surfactant protein SapB (spore-associated protein B) which has a function in the aerial mycelium development

In response to nutrient depletion and other signals, the morphological differentiation is initiated and the formation of aerial mycelium starts This is often accompanied by modifications at the metabolic level and the commitment to secondary metabolism The aerial mycelium breaks the surface tension, escaping the aqueous environment of substrate

mycelium and grows into the air To do that, Streptomyces has to coat its aerial hyphae in a

hydrophobic sheath that is absent from the substrate mycelium The SapB protein is produced to allow efficient formation of aerial mycelium This protein is only produced on rich media On minimal media, other proteins replace SapB to be the major components of the

hydrophobic sheath They are called the chaplins and the rodlins Streptomyces strains that

cannot make both SapB and chaplins are bald, i.e they cannot form aerial mycelium, under

all growth conditions When studied in vitro, both SapB and chaplins have powerful

surfactant activities at air-water interface They lower the surface tension from 72 mJ per m2

to 27 mJ per m2 The rodlin proteins have less impact on the formation of aerial mycelium Mutants that lack rodlin proteins still form a hydrophobic sheath and aerial mycelium Instead

of having normal surface ultrastructure, the rodlin mutants have a disordered network of fine

filaments The S coelicolor bld mutants are blocked in aerial mycelium formation, but when

different mutants are grown near to each other on certain media, most of them exhibit directional interactions that lead to aerial mycelium formation These interactions have been

interpreted as resulting from the existence of an extracellular signalling cascade The signalling molecules are unknown This cascade is represented in Figure 6

Figure 6: Extracellular signalling cascade dependent on bld genes in S coelicolor From

(Chater & Chandra, 2006)

In other Streptomyces species, different extracellular interactions controlling aerial mycelium formation have been described In particular, in S griseus, A-factor, a compound

of the gamma-butyrolactone family, is a extracellular signal required for aerial mycelium formation but also for the switch to the synthesis of the antibiotic streptomycin

The last step of the cycle is the differentiation of aerial hyphae into spores (Figure 7)

It starts with the formation of a long, non-septated, apical compartment, called sporogenic cell, each of which contains 50 or more copies of chromosome Then, the sporogenic cell stops its extension and begins synchronous, multiple cell division, under the effect of the

whiA and whiB gene products The next step is the formation of septum, directed by the

bacterial tubulin homologue FtsZ In virtually all bacteria, FtsZ assembles into a cytokinetic ring, the Z ring that defines the site of division and recruits other cell-division proteins During the stage of septation, DNA is segregated into unigenomic spores, but complete partitioning into individual nucleoids does not occur until the final stage of septation The DNA segregation involves at least two systems: ParAB and FtsK Sporulation septa constrict over unsegregated nucleoids Finally, a thick, lysozyme-resistant, spore wall is formed This wall

is laid down after sporulation septation is complete and is associated with the rounding up of the cylindrical prespore into an ovoid shape This step involves MreB Then, one cycle is finished and the spore can be disseminated and give rise to a new colony

Figure 7: Aerial hyphae differentiation into spore chains in S coelicolor From (Chater &

Chandra, 2006)

Aerial growth is partly parasitic on the primary colony, which is digested and reused for aerial growth This late phase of development is coordinated with the production of antibiotics, which may protect the colony against invading bacteria during aerial growth

Among Actinobacteria, the Streptomycetes are the only ones for which development

was studied in detail and many of the genes involved in aerial mycelium development and spore formation have been identified A comparative genomic analysis suggests that the

evolution of Streptomyces development has probably involved the stepwise acquisition of

laterally transferred DNA Each of these acquisitions gave rise to regulatory changes or to

changes in cellular structure and morphology (Figure 8) (Chater & Chandra, 2006; Ventura

et al., 2007)

Figure 8 Proposed pathway for the evolution of Streptomyces development by sequential

acquisition of developmental genes The scheme was drawn to minimize the number of gene

acquisition and gene loss events From (Ventura et al., 2007)

1.4 Ecology

Streptomyces are mostly living in soil, where they represent a large proportion of the

cultivable bacteria In soil, they can exist in the form of spores or vegetative mycelium In

soils, Streptomyces are present mostly as dormant spores, providing a large proportion of the Streptomyces colony forming units from soils The spore has extraordinary capabilities of survival in nature Viable Streptomyces spores were recovered from 70-year-old soil samples

(Morita, 1985)

In the soil environment, Streptomyces produce many extracellular enzymes These

enzymes hydrolyse the biomaterials that are present in soil to supply nutriments for the

development of Streptomyces This biodegradation plays an important role in the carbon, nitrogen and phosphorus cycles in soils (Hernandez-Perez et al., 1997; Raweesri et al.,

2008) Apart from natural biomaterials, some toxic compounds are also degraded by

Streptomyces There are for instance studies on the possibility to use Streptomyces to treat soils contaminated with herbicides (Duraes Sette et al., 2004; Shelton et al., 1996)

Streptomyces ssp are very seldomly pathogens, as very few diseases are caused by Streptomyces In human, Streptomyces somaliensis and Streptomyces sudanensis are

known to cause the actinomycetoma form of mycetoma Mycetoma is a chronic progressive inflammatory disease affecting the foot It is endemic in Africa, India, and Central and South

America (Quintana et al., 2008; Trujillo & Goodfellow, 2003) In plants, some Streptomyces strains cause diseases, which have economical impacts in agriculture Streptomyces scabies {Lambert, 1989 #73}, Streptomyces acidiscabies {Lambert, 1989 #74} and Streptomyces turgidiscabies (Miyajima et al., 1998) cause potato scab disease that damages the surface of the tubers and also render the flesh inedible (Figure 9) Streptomyces ipomoea causes sweet

potato soil pox, which causes yield lost and disfiguring lesions to the tubers (Kennedy &

Alcorn, 1980) Some other Streptomyces (e.g Streptomyces parvulus, Streptomyces sparsogenes) which are found in xylem vessels of maples cause early decay and dieback of tree branches (Sutherland et al., 1979)

Figure 9: Potato disease by Streptomyces scabies

(http://www.gardenguides.com/pests/pestinfo/potatoscab.asp) and

(http://www.apsnet.org/Education/IllustratedGlossary/PhotosA-D/actinomycete.htm)

Antibiotic produced by Streptomyces might antagonise the other microorganisms cohabiting in the same environment When S bikiniensis was co-inoculated with Samonella dusseldorf, it decreased strongly the viable count of this pathogen (Turpin et al., 1992)

Antibiotics have not been detected in natural untreated soils, but production has been detected in nonsterile amended, sterile amended and unamended soil inoculated with

antibiotic producing strains of Streptomyces (Marsh & Wellington, 1994; Wellington et al.,

1993; Williams, 1982) Experiments by Weiner (Wiener, 1996) showed that antibiotic

production by Streptomyces can prevent invasion by competing Bacillus subtilis but does not improve the ability to invade an established population of sensitive B subtilis

As antibiotic producers, Streptomyces possess resistance mechanisms to protect them against their own antibiotics As the resistance mechanisms found in Streptomyces are

similar to those encountered in antibiotic resistant pathogens, and as resistance genes

homologous to those found in Streptomyces have been identified on mobile genetic elements

reservoir of antibiotic resistance genes (Davies, 1994) However, a recent and direct transfer

seems quite improbable, as soil-dwelling Streptomyces and pathogenic bacteria do not

generally live in the same environment Indeed, some studies, such as the one performed on

genes conferring macrolide resistance by target modification (Lau et al., 2004), found no

evidence of recent horizontal transfer between producers and pathogens Therefore, if antibiotic resistance genes originated from antibiotic producers, the transfer is most probably not recent and it might have involved many transfer steps and intermediate hosts between the antibiotic-producing streptomycetes and the pathogens Recent studies showed that among Streptomycetes isolated from soil, a very large proportion was highly resistant to several antibacterial agents, which were generally not produced by the resistant strain, and which even included antibacterial agents of synthetic origin (Figure 10) (Canton, 2009;

D'Costa et al., 2006; Martinez, 2008) This suggests that the presence of antibiotics in the

environment might have promoted the acquisition or independent evolution of specificresistance genes in the absence of innate antibiotic production and that various telluric bacteria might constitute an important reservoir of resistance genes

Figure 10 Antibiotic resistance profiling of 480 soil-derived streptomycetes bacterial isolates

(A) Schematic diagram illustrating the phenotypic density and diversity of resistance profiles

The central circle of 191 black dots represents different resistance profiles, where a line

connecting the profile to the antibiotic indicates resistance (B) Resistance spectrum of soil

isolates Strains were individually screened from spores on solid Streptomyces isolation

media (SIM) against 21 antibiotics at 20 µg of antibiotic per ml of medium Resistance was

defined as reproducible growth in the presence of antibiotic (C) Resistance levels against

each antibiotic of interest From (D'Costa et al., 2006)

2 The biosynthesis of secondary metabolites by Streptomyces and related

Actinobacteria

2.1 The importance of secondary metabolites for medicine and agriculture

The definition of secondary metabolites has been discussed among biologists and there is no clear-cut definition (Challis & Hopwood, 2003) However, the term secondary metabolite is used to design compounds that are produced by living organisms, mainly plants and microorganisms, and that are characteristic of narrow taxonomic groups of organisms, and have diverse, unusual, and often complex chemical structures Secondary metabolites are nonessential for the growth of the producing organism, at least under the conditions studied, and are often made after the phase of active vegetative growth They present a wide range of biological activities, including the inhibition or killing of other microorganisms (the narrow definition of an antibiotic), but also toxic effects against multicellular organisms The fact that some of these compounds are for instance involved in microbial differentiation or in metal transport blurs the distinction between primary and secondary metabolites Secondary metabolites are also called natural products

For thousands of years, natural products have constituted the core of human remedies and medicines During the last century, secondary metabolites, especially antibiotics, have exerted a major impact on the control of infectious diseases and other medical conditions In the USA, their use has contributed to an incease in the average life expectancy from 46 years in 1900 to 74 years (for men) and 80 years (for women) in 2000 (Lederberg, 2000) The development of organic chemistry, however, has challenged their place in our pharmacopoeia Moreover, the recent development of techniques such as combinatorial chemistry and high throughput screening has contributed to lessen, over the last 15 to 20 years, the interest of the pharmaceutical industry towards natural products

(Newman et al., 2003) Yet today, natural products continue to occupy a preponderant place

among biologically active molecules: almost 50% of new drugs introduced into the market between 1985 and 2006 were natural products or their derivatives and these coumpounds

represent over 40 billion US dollars in sales (Newman & Cragg, 2007; Newman et al., 2003).

Furthermore, studies have shown that natural products populate, in the chemical diversity space, a larger volume than molecules originating from combinatorial synthesis (Feher & Schmidt, 2003) It is now widely recognised that natural products constitute privileged structures in the search for biological activity, due to their intrinsic ability to interact with biological targets Taken together, this is contributing a renewed interest in the study of natural products and their biosynthesis

The total number of natural products produced by plants has been estimated to be between 500,000 and 600,000 (Berdy, 2005; Mendelson & Balick, 1995) The major groups

of compounds found in plants are alkaloids; flavones, flavonoids, and flavonols; essential oils and terpennoids; lectins and polypeptides; polyphenols and phenolics; tannins Plants or plant extracts play an important role in traditional medicine Presently, in India, 85% of people use higher plants as effective antimicrobial agents for the treatment of various diseases (Kamboj, 2000) Recently programs to explore the wealth of natural products from plants have been initiated in developing countries such as China and Vietnam and they gave promising results in the fields of antibacterial and antitumor agents

About 100 000 secondary metabolites of molecular weight less than 2 500 have been characterised and some 50 000 are produced by microorganisms (Berdy, 1995; Roessner & Scott, 1996) Concerning antibiotics, defined as low molecular weight natural products made

by microorganisms which are active against other microorganisms, about 12 000 of them

were known in 1995 Of these, 55% were produced by Actinobacteria of the genus Streptomyces, 11% by other Actinobacteria, 12% by non-filamentous bacteria and 22% by

filamentous fungi (Berdy, 1995)

Besides the terrestrial microbes, marine microbes contribute to the wealth of natural products: 129 bioactive compounds were isolated from marine microbes from 2000 to 2003

(Liu et al., 2004)

2.2 Biological activities of secondary metabolites from Streptomyces and related

Actinobacteria

The chemical diversity of the secondary metabolites synthesized by Streptomyces is

remarkable and these compounds have a wide range of biological activities that are exploited for various purposes in human and animal medicine but also for agricultural applications 2.2.1 Antibacterial and antifungal agents

As seen above, Actinobacteria synthesize about two thirds of the known antibiotics, the genus Streptomyces being at the origin of more than half of the known antibiotics The

proportions are roughly the same for the compounds that are use in human and animal

medicine Streptomyces and closely related genera produced compounds belonging to all

major families of antibiotics: beta-lactams, cyclines, macrolides, aminoglycosides, glycopeptides, polyenes… These antibiotics are used as antibacterial and antifungal agents

in medicine and agriculture The chemical diversity of these compounds is illustrated in figure

11

(phenicol, Streptomyces venezuelae)

Figure 11 Structure of some antibacterial and antifungal agents produced by Streptomyces

or related Actinobacteria The family to which the compound belongs and the producer

organism are indicated

2.2.2 Antitumor agents

Many compounds from Streptomyces are used in the treatment of cancer:

actinomycin D, anthracyclines (e.g daunorubicin, doxorubicin, epirubicin, pirirubicin and valrubicin) bleomycin, mitomycin C These compounds bind to DNA Among them, actinomycin D is the oldest metabolite used in cancer therapy It is produced by

Streptomyces antibioticus, isolated by Waksman and Woodruff (Waksman & Woodruff,

1941) Actinomycin D binds DNA at the transcription initiation complex and thus preventing elongation by RNA polymerase (Sobell, 1985) The structure of some of these compounds is presented in figure 12

Bleomycin (Streptomyces verticillus)

Mitomycin C (Streptomyces lavendulae)

Daunorubicin (Streptomyces peucetius)

Actinomycin D (Streptomyces lavendulae)

Figure 12 Structure of some antitumor agents produced by Streptomyces

2.2.3 Enzymes inhibitors

Acarbose produced by Streptomyces glaucescens is an inhibitor of

alpha-glucosidase It is used in the treatment of diabetes patients, enabling them to better utilize starch- or sucrose-containing diets by slowing down the intestinal release of glucose

(Yamagishi et al., 2009) Clavulanic acid produced by Streptomyces clavuligerus is a

beta-lactamase inhibitor that is often associated with beta-lactam antibiotics to prevent their

inactivation by pathogenic bacteria secreting beta-lactamases (Saudagar et al., 2008).

Protease inhibitors are potentially powerful tools for inactivating target proteases in some diseases such as emphysema, arthritis, AIDS, cancer Some protease inhibitors from

Streptomyces are antipain of Streptomyces yokosukaensis, leupeptin of Streptomyces roseochromogenes, chymostatin of Streptomyces hygroscopius (Imada, 2004). Some examples of the structure of these products are given in figure 13

Chymostatin (Streptomyces hygroscopicus)

Clavulanic acid (Streptomyces clavuligerus) Acarbose (Streptomyces glaucescens )

Figure 13 Structure of some enzyme inhibitors produced by Streptomyces

as an immunosupressant in liver transplantation (Demain & Sanchez, 2009) Its structure is presented in figure 14

Figure 14 Structure of tacrolimus

2.2.5 Insecticides and antiparasitic drugs

These compounds are used in agriculture, medicine, industry, and households The use of insecticides is believed to be one of the major factors behind the increase in agricultural productivity in 20th century (Demain & Sanchez, 2009) Natural insecticides are

often less toxic than synthetic insecticides Spinosyns from Saccharopolyspora spinosa are

used in agriculture as a potent insect control agent with exceptional safety to non-target

organisms (Waldron et al., 2001) Some others insecticides produced by Streptomyces are piericidins from Streptomyces pactum, which shows a strong inhibitory effect on mitochondrial respiration (Yoshida et al., 1977); prasinon A and prasinon B of S prasinus which are highly active against sheep blowfly larvae (Box et al., 1973) Avermectins, produced by Streptomyces avermitilis possess activity against a broad range of nematodes

and arthropods, are non toxic to mammals and lack significant antibacterial or antifungal

activity (Burg et al., 1979) Avermectins are used in human medicine for the treatment of

onchocerciasis (river blindness) and in veterinary medicine for the prevention and treatment

of various parasitic infections Avermectins and spinosyns both belong to the macrolide family and their structure are presented in figure 15

Avermectins (Streptomyces avermitilis) Spinosyn A (Saccharopolyspora spinosa)

Figure 15 Structures of avermectins and spinosyn A

2.2.6 Herbicides

Herbicides inhibit normal plant growth and development and are widely used in agriculture for weed management While chemical herbicides have many severe effects for the environment, bioherbicides are as effective and much more environmentally friendly One

of the best-known herbicide produced by Streptomyces is bialaphos, also called phosphinothricin tripeptide It is produced by Streptomyces hygroscopius and by Streptomyces viridochromogenes Tü494 and is a commercially important herbicide (Anzai et al., 1987) It consists of a glutamic acid analogue moiety, called phosphinothricin and two

alanine residues (Figure 16) It is an inhibitor of glutamine synthetase and has not only

herbicidal activity but also antibacterial and antifungal properties (Schwartz et al., 2004) A

gene encoding a bialaphos acetyl-transferase conferring bialaphos resistance has been isolated from the producer and introduced into some plants to obtain transgenic plants

resistant to bialaphos (Block et al., 1987) Some other herbicides have been isolated from Streptomyces as herbimycin of Streptomyces hygroscopius which has herbicidal activity against mono- and dicotyledonous plants (Omura et al., 1979); geldanamycin and nigericin from Streptomyces hygroscopius which cause a 50% reduction in garden grass radicle at 1-2

ppm and nearly complete inhibition at 3-4 ppm (Heisey & Putnam, 1986)

Bialaphos (Streptomyces viridochromogenes and Streptomyces hygroscopicus)

Figure 16 Structure of bialaphos

2.3 The genetics of secondary metabolism in Streptomyces

2.3.1 Initial genetic studies of secondary metabolism in S coelicolor

Streptomyces coelicolor, the genetically best characterized strain, played a very important role in the development of the genetics of secondary metabolism in Streptomyces

This strain was shown to produce different antibiotics: methylenomycin, actinorhodin, undecylprodigiosin, and a calcium dependent antibiotic called CDA The fact that two of these secondary metabolites, actinorhodin and undecylprodigiosin, are coloured compounds was exploited to easily isolate non-producing mutants, to do genetic experiments with them and later for the cloning of biosynthetic genes The first studies on the genetics of antibiotic

production were performed in S coelicolor in the 1970s They revealed that the genes

directing the biosynthesis of the antibiotic methylenomycin were located on the plasmid

SCP1 (Kirby et al., 1975; Vivian & Hopwood, 1970; Wright & Hopwood, 1976b) The

actinorhodin biosynthetic genes were shown to be clustered on the chromosome (Rudd & Hopwood, 1979; Wright & Hopwood, 1976a) This clustering of biosynthetic regulatory and resistance genes then turned out to be the paradigm for secondary metabolite biosynthetic

genes in Streptomyces and in other bacteria

The first gene cloning experiments in Streptomyces were performed in the 1980s and the genes cloned were antibiotic resistance genes from antibiotic producing Streptomyces cloned in S coelicolor or the closely related species Streptomyces lividans (e.g methylenomycin resistance gene (Bibb et al., 1980) thiostrepton resistance gene

(Thompson & Cundliffe, 1980) This provided marker genes for the development of cloning vectors and facilitated further work The first antibiotic-biosynthetic genes were then cloned: they were involved in the biosynthesis of methylenomycin (Chater & Bruton, 1983) and undecylprodigiosin (Feitelson & Hopwood, 1983) Soon, genes for a wholepathway, the act

genes for actinorhodin biosynthesis, were isolatedand expressed in a different Streptomyces

host (Malpartida & Hopwood, 1984) The first hybrid antibiotic was then produced by genetic

engineering; when specific segments of the act genecluster were introduced on a plasmid vector into the producer of the related antibiotic medermycin The transformed strain was found to synthesize mederrhodin, a compound withstructural features of both medermycin

and actinorhodin (Hopwood et al., 1985) These initial studies opened the way to the use of

molecular genetics for the detailed characterization of secondary metabolite biosynthetic

gene clusters in S coelicolor but also in many other Streptomyces

2.3.2 The isolation of secondary metabolites biosynthetic gene clusters in Streptomyces

Different strategies have been used to isolate secondary metabolites biosynthetic genes As the genes involved in biosynthesis of a given secondary metabolite are clustered,

the cloning of one of the genes will give access to all others In addition, if the producer organism must protect itself against the metabolite produced, one or more resistance genes might be linked to the biosynthetic genes

One obvious strategy is to clone resistance genes, which can be easily selected in a host susceptible to the antibiotic This strategy was used to gain access to erythromycin

biosynthetic genes from Saccharopolyspora erythraea (Stanzak et al., 1986), spiramycin biosynthetic genes from Streptomyces ambofaciens (Richardson et al., 1987; Richardson et al., 1990) or streptomycin biosynthetic genes from Streptomyces griseus (Distler et al.,

1987) Another strategy is the complementation of mutants blocked in biosynthesis of the

antibiotic It was used for example in the cloning of streptomycin biosynthetic genes (Ohnuki

et al., 1985)

In some cases, an enzyme of the secondary metabolite biosynthetic pathway had been purified and some protein sequences have been obtained, thus allowing a reverse genetic approach This strategy was used to clone some of the genes for the biosynthesis of

tylosin in Streptomyces fradiae: degenerate oligonucleotides deduced from the protein

sequence of the N-terminal region of TylF, the enzyme catalyzing the final step in the

biosynthesis of tylosin, were used as probes (Fishman et al., 1987) The cluster of genes directing the biosynthesis of albonoursin in Streptomyces noursei was isolated by a reverse

genetic strategy, using a PCR approach with primers designed from internal amino acid

sequences of the enzyme catalyzing the final step of the biosynthesis (Lautru et al., 2002)

As more and more clusters were isolated and sequenced, it became apparent that homologous genes were found not only in clusters directing the synthesis of closely related secondary metabolites, but also in clusters directing the synthesis of secondary metabolites sharing only a limited structural similarity or even in clusters directing the synthesis of very different secondary metabolites This allowed the use of some genes as probes or the design

of primers for PCR-based strategies Genes involved in the biosynthesis of the macrolide antibiotic erythromycin were for instance used as probes to clone genes involved in the

synthesis of other macrolides such avermectin (MacNeil et al., 1992), oleandomycin (Swan

et al., 1994) and rapamycin ((Schwecke et al., 1995) Many secondary metabolites are

glycoslated compounds comprising deoxyhexose moieties The enzyme dNDP-glucose dehydratase is involved in the early stages of deoxyhexose biosynthesis Oligonucleotide primers were designed and successfully applied to amplify DNA fragments of dNDP-glucose dehydratase genes from actinomycete species producing natural compounds which contain

deoxysugar moieties (Decker et al., 1996) This approach gave for instance access to the genes for the biosynthesis of methymycin/picromycin in Streptomyces venezuelae (Xue et al., 1998)

The study of these clusters and of the associated biosynthetic pathways revealed that secondary metabolites with very different chemical structures could be synthesized by similar biosynthetic pathways encoded by homologous genes Many secondary metabolites are derived from polyketides, assembled by polyketide synthases, or from non ribosomal peptides, assembled by non ribosomal peptide synthetases More details on these enzymes will be given below

2.4 The biosynthesis of polyketides

2.4.1 The assembly-line enzymology of polyketide synthases

Polyketides are a very large family of natural products possessing various biological activities such as antibacterial (erythromycin), antifungal (nystatin), anti-cancer (daunorubicin), insecticide (avermectin) More than a third of natural products or natural product-derived compounds approved as drug in the 2005-2007 period were polyketides (Butler, 2008) Despite their structural heterogeneity, all polyketides are assembled by the same biochemical process, which resembles that of fatty acid biosynthesis The chemical logic for the assembly of polyketides is that a set of monomer units (carboxylic acid derivatives) are incorporated into a linear oligomer by successive condensation steps, followed in some case by reductive modification of the resulting keto groups These reactions are catalyzed by multienzymatic complexes, the polyketide synthetase (PKS) PKS are often compared to assembly lines in which the identity and order of each protein domain in the line specifies the sequence of monomer units activated and incorporated, the chemical modifications that occur on the elongating chain, and the length of the final product (Fischbach & Walsh, 2006; Hopwood, 2004; Weissman, 2009)

The different steps of the biosynthesis of a polyketide chain are shown in Figure 17

In the example shown in figure 17, the carboxylic acids are acetic acid and malonic acid They are converted to their coenzyme A (CoA) esters and then attached, by specific acyl transferases, to components of the polyketide synthase: acetyl-CoA is attached to the active site of the ketosynthase, and malonyl-CoA to a structural component of the PKS called the acyl carrier protein (ACP) Condensation of the two units is catalyzed by the ketosynthase One carbon is lost from malonyl-CoA as carbon dioxide This produces a four-carbon chain attached to the ACP Before the next round of condensation, this chain is transferred back to the ketosynthase Further rounds of condensation with malonyl-CoA (as shown in the figure)

or other chain extender units produce a polyketide chain If the acyl-CoA used contains more than 3 carbons, the main linear polyketide is always extended by two carbons, the additional carbons forming side chains The acyl-transferase is responsible for the choice of the acyl-

CoA incorporated The fact that PKSs are able to incorporate different acyl-CoA units explains in part the chemical diversity of the resulting polyketides

Figure 17 The different steps of the biosynthesis of a polyketide chain From (Hopwood, 2004)

Another source of variability is due to the reactions which are performed on the keto group present on a carbon of each acyl incorporated This function may remain unchanged, but it can also be subjected to reductive modifications (Figure 18A) In the first step, catalyzed by a ketoreductase, the keto group is converted into a hydroxyl group In the second step, catalyzed by a dehydratase, dehydration eliminates the hydroxyl group and there is formation of a double bond between two carbons The third and last stage of hydrogenation, catalyzed by an enoylreductase, leads to a fully saturated carbon Each of the keto groups may remain intact or can be reduced, the reduction stops after one, two or three steps, giving a carbon chains in which there are different types of side groups (Figure 18B) The stereochemistry of hydroxyl groups, methyl groups and other side chains introduces additional structural diversity