. 5w.
“5.2.“... a...
1H... h.

a». .r

4? .
3..
F...

r 1.. ".5

.14.)
a. .5. ‘

.5”

'1

m... was...

s
at,“ .

4%

. .
é. . :le
Iva. 4..

ix

 

.E.‘ .
{.uhhu:ﬂ. .
2:11P. v

\t T ‘

v.‘

 

 

 

 

«y..c.
1.1:. I

 

 

 

THE‘SBS

2002.

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

Transcriptional Regulation of Streptomyces
coelicolor Antibiotic-Specific Regulatory Genes

presented by
David J. Aceti

has been accepted towards fulﬁllment

of the requirements for

Ph.D Microbiology
degree in

 

 

Majorﬂofessor
Date //éiﬂ—7L (& 280 /

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRODﬂeDmpGS—p.”

TRANSCRIPTIONAL REGULATION OF STREPTOMYCES COELICOLOR
ANTIBIOTIC-SPECIFIC REGULATORY GENES

By

David J. Aceti

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

Doctor of Philosophy

Department of Microbiology

2001

ABSTRACT
TRANSCRIPTIONAL REGULATION OF STREPTOMYCES
COELICOLOR ANTIBIOTlC-SPECIFIC REGULATORY GENES
By
David J. Aceti

Antibiotics produced by streptomycetes are of great medical value
and may play an important role in soil ecology. The regulatory
mechanisms governing production of these antibiotics are not well
understood. In the model organism Streptomyces coelicolor, all four
antibiotics produced are affected by mutations in a putative two-component
signal transduction system encoded by absA1/absA2 and by mutations in a
putative RNAase encoded by absB. $1 nuclease protection assays were
used to assess whether these loci control synthesis of the antibiotics
actinorhodin and undecylprodigiosin by regulating transcript abundance
from biosynthetic and regulatory genes speciﬁc for each antibiotic. Strains
that were antibiotic-minus due to mutations in absA or absB were
examined. In the absA mutant strain, transcripts for the actinorhodin
biosynthetic genes ach-ORF1 and act], and for the pathway-speciﬁc
regulatory gene actlI-ORF4, were substantially lower in abundance than in
the parent strain. The level of transcript for the undecylprodigiosin
pathway-speciﬁc regulatory gene redD was similarly reduced in this
mutant. Additionally, a strain that exhibits precocious hyperproduction of

antibiotics (Pha phenotype) due to disruption of the absA locus contained

elevated levels of the ach-ORF1, actll-ORF4 and redD transcripts. In the
absB mutant strain, actVI-ORF1, actl, actlI-ORF4, and redD transcript
levels were also substantially lower than in the parent strain. The simplest

' explanation for these results is that the absA and the absB regulatory
pathways include transcriptional control of actll-ORF4 and redD. Studies
of transcriptional fusions between the actll-ORF4 promoter and the xyIE
reporter gene conﬁrmed the effects of absB on actlI-ORF4 transcription.
However, a mutation in absA had no apparent effect on the expression of
this fusion, suggesting that absA and absB have distinct regulatory
mechanisms. Of seven bld mutants tested, none appeared to signiﬁcantly
affect transcription from the actIl-ORF4zzxyIE fusion. In a separate study,
transcriptional reporter gene fusions between the actll-ORF4 promoter and
the M reporter gene were used in attempts to detect antibiotic synthesis in
natural soil. Expression of the fusion was detected during growth in sterile,
nutrient-amended soil microcosms; it was concluded, however, that the
technique used is unlikely to possess the sensitivity for detection under

more natural conditions.

ACKNOWLEDGEMENTS

I would like to thank my advisor, Professor Wendy Champness, and
my committee members, Professors Helmut Bertrand, Lee Kroos,
Richard Schwartz, and Michael Bagdasarian. I thank Dr. Paul Brian, Dr.
Brenda Price, Dr. Todd Anderson, Dr. Perry Riggle, and Dr. Trifon
Adamidis for helpful discussions and assistance. I would also like to
thank my parents, John and Carol, my sisters, Janet, Susan, and Sharon,
Dr. Klaus Nuesslein, and my friends at Michigan State University and the

University of Wisconsin-Madison for their support.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................ vii
LIST OF FIGURES ......................................................................................... viii
KEY TO ABBREVIATIONS ............................................................................... x
CHAPTER 1
INTRODUCTION .............................................................................................. 1
1.0 Introduction to Streptomycetes .............................................................. 1
1.1 Streptomycetes are antibiotic-producing, mycelial,
sporulating bacteria ................................................................................ 1
1.2 Streptomycetes are an important component of soil and
sediment microfauna ............................................................................... 5
1.3 Synthesis of pharmaceuticals by streptomycetes ............................... 9
2.0 Streptomyces coelicolor A3(2): A Model Organism ............................. 12
2.1 History .............................................................................................. 12
2.2 The S. coelicolor A3(2) genome ....................................................... 13
2.3 Genetic techniques ........................................................................... 15
2.4 Antibiotics produced by S. coelicolor A3(2) ...................................... 20
2.5 Antibiotic-speciﬁc regulatory genes .................................................. 25
3.0 Higher Order Regulation of Antibiotics ................................................. 27
3.1 Growth rate/nutrient limitation ........................................................... 27
3.2 bld loci .............................................................................................. 29
3.3 absA ................................................................................................. 32
3.4 absB ................................................................................................. 39
3. 5 afs loci and autoregulatory signaling actors ..................................... 42
3.6..(p)ppGpp . ...46
3. 7 RNA polymerase sigma factors ........................................................ 48
3. 8 Other potential regulators ................................................................. 50
3.9 Overview of antibiotic synthesis regulation in
S. coelicolor ............................................................................................ 52
4.0 This Work ............................................................................................. 53
CHAPTER 2

TRANSCRIPTIONAL REGULATION OF STREPTOMYCES COELICOLOR
PATHWAY-SPECIFIC ANTIBIOTIC REGULATORS

BY THE absA AND absB LOCI ...................................................................... 56
PART A: Transcriptional Regulation of Streptomyces coelicolor
Pathway-Speciﬁc Antibiotic Regulators by the absA and absB Loci .......... 57

Abstract .................................................................................................. 58
Introduction ............................................................................................. 58
Materials and Methods ........................................................................... 59

Results ........................................................................................................ 59
Regulation of the Act-speciﬁc regulator actll-ORF4 by the

absA and absB loci ............................................................................. 59
Regulation of Act biosynthetic genes by the absA and absB loci ....... 60
Regulation of expression of the Red-speciﬁc regulator redD
by the absA and absB loci .................................................................. 61
Effect of a disruption mutation of the absA locus on
expression of actII-ORF4, actVl-ORF1, and redD transcripts ............ 61
Discussion .................................................................................................. 62
References ................................................................................................. 62
PART B: Additional Results and Discussion .............................................. 65
Transcripts of an act biosynthetic gene are more severely
reduced than actlI-ORF4 transcripts in the abs mutants ....................... 65
Do the abs genes inﬂuence the timing of act and red gene
transcriptional onset? ............................................................................. 66
Identity of a secondary band in ach-ORF1 assays .............................. 67
Appendix: Complementation of the absB mutation with the cloned
absB gene partially restores actll-ORF4 and redD transcript levels ........... 69
CHAPTER 3

TRANSCRIPTIONAL REGULATION OF THE STREPTOMYCES
COELICOLOR actll-ORF4 PROMOTOR; REPORTER GENE FUSION

STUDIES ........................................................................................................ 73
Abstract ...................................................................................................... 73
Introduction ................................................................................................. 74
Materials and Methods ............................................................................... 75
Results ........................................................................................................ 76

Construction of low-copy and single-copy
actll-ORF4::xyIE transcriptional fusions ................................................. 76
The actIl-ORF4::xyIE fusions are subject to
regulation in S. coelicolor ....................................................................... 79
Expression from actll-ORF4::xyIE fusions in absA and
absB mutant backgrounds ...................................................................... 85
Expression of actlI-ORF4::xyIE fusions in bld mutant backgrounds ....... 89
Discussion .................................................................................................. 92
Appendix: Assessment of qu reporter gene fusions for monitoring
S. coelicolor antibiotic synthesis in soil ....................................................... 96

CHAPTER 4. CONCLUSIONS AND PROSPECTS .................................... 122
Appendix: Large-scale reproductions of Figures 1A, 2, 3A, 4A, and 5 ..... 133

REFERENCES ............................................................................................. 136

vi

LIST OF TABLES

Table 1. Comparison of reductions in actll-ORF4 and
actVl-ORFI transcript levels in abs mutants................... .66

Table 2. Light emission from liquid dilutions of qu-bean'ng
R. meliloti strain CV1 cells as detected by three instruments... ..107

Table 3. Light emission from soil dilutions of lax-bearing
R. meliloti strain CV1 cells as detected by three instruments... ..108

Table 4. Detection of light emission from S. coelicolor
strain M600/pDA7 in liquid and in liquid/soil mixture... ..1 12

vii

LIST OF FIGURES

Figure 1. Growth- -phase dependent expression of actll- ORF4
mRNA In plate-grown J1501.. 60

Figure 2. Expression of actll-ORF4 mRNA in J1501 (abs’),
C542 (absA), C120 (absB), and C430 (absA disruption)... ..60

Figure 3. Expression of ach-ORF1 mRNA in J1501 (abs*),
C542 (absA), C120 (absB), and C430 (absA disruption)... ..61

Figure 4. Expression of redD mRNA In J1501m(abs*)n C542m(absA),
0120(absB).. . . .....61

Figure 5. Expression of redD mRNA In J1501 m(abs*)m and C430
(absA disruption). ...........62

Figure 6. Expression of actll— ORF4 mRNA In J1501m(absA )- 0542
(absA‘), and C120 (absB).. .......... 68

Figure 7. Complementation of absB mutants by a cloned DNA
fragment 70

Figure 8. Construction of the actll- ORF4: ME... fusion vectors ”pDA4
and DA1.. . . ......77

Figure 9. Expression from actll- ORF4: nyIE and actl: xylE transcriptional
fusions' In S. coelicolor strain J1501.. . .82

Figure 10. Expression from actll-ORF4::xy/E transcriptional fusions in

absA*and absB‘ straIns 86
Figure 11. Expression of ”plasmid- borne actll- ORF4: :xyIE transcriptional
fusmns In bldmutants... .. ..91
Figure 12. Soil-grown S. coelicolor mycelia on a “buried slide" ........ 103
Figure 13. Construction of pDA7 109

Figure 14. Timecourse of luciferase expression from S. coelicolor strains
M600/pDA7 and J1501/le342 in liquid culture........................... 111

viii

Figure 15. Detection of qu fusion expression from
S. coelicolor M600lpDA7 growing in sterilized,
nutrient-amended soil microcosms” 114

Figure 16. Timecourse of light emission from S. coelicolor M600lpDA7
following addition of n-decanal....................................... 115

LIST OF ABBREVIATIONS

Abs"
Act

ASR
CDA
CDO
CF U
MCS
Mmy
ORF
PCR
Pha

Red

RLU
Sab

SARP

antibiotic synthesis deﬁcient
actinorhodin

antibiotic-speciﬁc regulator
calcium-dependent antibiotic

catechol 2,3-dioxygenase
colony-forming unit

multiple cloning site

methylenomycin

open reading frame

polymerase chain reaction

precocious hyperproduction of antibiotics
undecylprodigiosin

relative light unit

suppressor of abs

Streptomyces antibiotic regulatory proteins

CHAPTER 1

INTRODUCTION

1.0 Introduction to Streptomycetes

1.1 Streptomycetes are Antibiotic-Producing, Mycelial, Sporulating

Bacteria.

By one recent count, the bacterial genus Streptomyces includes more
than 450 recognized species (7). This number reﬂects the considerable
scientiﬁc and economic interest in streptomycetes that has been inspired
largely by their ability to produce a tremendous number and variety of
antibiotics and other useful pharmaceuticals. Less well known is the status
of these complex, mycelial, spore-forming bacteria as a signiﬁcant research
subject in the ﬁeld of cellular differentiation. A third, somewhat neglected,
aspect of streptomycete biology is their role in the environment, where they
are ubiquitous and often abundant in soil and sediments.

The genus Streptomyces, meaning “chain fungus”, was proposed by
Selman Waksman in 1943 to encompass that branch of the “aerobic,
saprophytic actinomycetes that form catenulate spores” (222). The mycelial
morphology of streptomycetes led researchers at that time to classify them
variously as fungi, as a form of life intermediate to bacteria and fungi, or as a
branch of bacteria separate from the true bacteria. Their classiﬁcation as

true bacteria was established in the 1950’s and was based on electron

microscopy showing the absence of a nuclear membrane (96) as well as
biochemical characteristics that included a gram positive-type cell wall.

Species classiﬁcations within the genus was long based primarily on
spore and mycelial morphology and on the synthesis of pigmented
compounds during growth on various media. The International Streptomyces
Project of the 1960’s standardized the growth media and the descriptive
methods used in characterization (80). Later, the computer-assisted
numerical taxonomy approach of Williams et al. (233) increased the number
of biochemical and morphological characteristics used in classiﬁcation.
Recent analyses based on rRNA and other molecular sequence information
confirmed earlier approaches in showing that the streptomycetes are a
taxonomically cohesive branch of the eubacteria (7,201). Currently, the
genus Streptomyces is placed within the Family Streptomycetaceae,
Suborder Streptomycineae, Order Actinomycetales, Class Actinobacteria (the
high G+C branch of the Gram positive bacteria), the Firmicutes (Gram
positive bacteria), and the Domain Bacteria (63,202).

Streptomycetes grow by apical hyphal extension and branching, forming
a dense interconnected mat of “vegetative” or “substrate” mycelia on plate
media. Chromosomes are distributed through the mycelia with cross walls
found between only approximately every tenth chromosome. A single growth
cycle on solid media, from spore germination to aerial hyphae bearing mature
spores, takes about four days under optimal conditions at 30°C. Spore

germination is followed by a period of rapid growth, then a short interval of

near-cessation of growth and macromolecular synthesis, thought to
correspond to a conversion from primary to secondary metabolism (164), and
another period of substantial mass accumulation as vertically-growing “aerial
hyphae” are formed (82). Vertical projection of'aerial hyphae is aided in
Streptomyces coelicolor by a small extracellular protein, SapB, that coats the
hyphae and is proposed to break surface tension and/or create a scaffold
(230). It is believed that energy and building blocks for the construction of
aerial hyphae are derived from the lysis and recycling of vegetative mycelia
(51). Little additional mass is accumulated during and after spore formation
and maturation (82). Three or more uninucleoid spores per hypha are
formed through growth of cell walls between chromosomes in the aerial
hyphae. Spore surfaces vary among species and may be smooth, hairy,
rugose, spiny, or warty. In liquid media, streptomycetes generally exist as
clumps or “pellets” that grow to macroscopic size, new pellets forming from
mycelial fragments shed by older pellets. Most species do not form aerial
hyphae or sporulate during growth in liquid media; a notable exception is
Streptomyces griseus.

Streptomycete chromosomes generally consist of 7-8 Mb of DNA
(118,136,226). Remarkably, at least some species possess linear
chromosomes, a discovery made in the early 1990’s that provided the ﬁrst
examples of linear prokaryotic chromosomes other than the small (1 kb)
chromosome of Borrelia burgdorfen‘ (136,226). Linear plasmids also appear

to be characteristic of the streptomycetes (122).

One of the more difﬁcult aspects of Streptomyces research is the
phenotypic and genetic variability of the organisms. This variability is evident
both in terms of phenotypic differences under superﬁcially identical growth
conditions and in the occurrence of spontaneous heritable mutations. It is
believed to be due in part to the linearity of (at least some) streptomycete
chromosomes; telomere-Iike ends contain repeats that may promote
recombination (226). Ampliﬁcation and deletion of large (sometimes greater
than 1 Mb) chromosomal segments are not uncommon and tend to occur at
“hotspots”; these are often found at chromosome ends (134,135,194) but are
not limited to them (181). Large-scale deletions and ampliﬁcations have also
been reported in connection with one of the more commonly encountered
variants in Streptomyces coelicolor cultures, the so-called “scarlet” variant,
which is characterized by changes in pigmentation and the loss of resistance
to chloramphenicol (68).

Antibiotic production may occur in both plate and liquid culture.
Microscopic examination of pigmented antibiotics or of antibiotic regulatory
protein/reporter protein fusions in Streptomyces colonies on solid media
show production only in older vegetative mycelia (46,131,209), and not in
aerial hyphae or spores. Antibiotics are almost always produced late in the
life cycle as growth rate is decreasing. A common hypothesis to explain this
timing proposes that antibiotics are produced to ﬂood the lysing vegetative
mycelia, protecting this vulnerable source of nutrients from competing

microorganisms until it can be recycled to form aerial hyphae and ultimately

spores (51). Presumably, antibiotics would be useful competitive weapons
during earlier growth stages as well; however, synthesis may be too costly in
terms of resources during early growth and the antibiotics may be less
effective at lower cell density.

Many strains of Streptomyces produce more than one antibiotic, and a
particular antibiotic may be produced by more than one species. Genes for
antibiotic production and resistance appear to have been exchanged
frequently in evolutionary history, not only between streptomycetes but
among different genera of bacteria and even between bacteria and fungi

(88,224).

1.2 Streptomycetes are an Important Component of Soil and Sediment
Microfauna.

Actinomycetes are ubiquitous in soils around the world and are
sometimes present at concentrations as high as 108 per gram of soil
(152,221). The genus Streptomyces is often the most numerous subclass of
actinomycete isolated, typically making up 1-20% of the total bacterial viable
count from growth of soil inocula (130). Streptomycetes are easily isolated
from lake, river, and ocean sediments as well. It has been argued that their
presence in sediments results from the survival of dormant spores washed
from terrestrial soil (221 ); however, recent analysis of 16S rRNA isolated from
coastal marsh sediments detected metabolically active streptomycetes and

found that they comprised 2-5% of the active microbial community (160).

Despite their abundance, the role of streptomycetes in nature has been little
studied.

Growth of streptomycetes in most soil environments is believed to be
intermittent; long periods of dormancy alternate with short growth periods
induced by the occasional inﬂux of water and nutrients (unlike some soil
bacteria, streptomycetes appear to make only limited use of root exudates)
(232). An estimated mean generation time in woodland soils of 1.7 days,
compared to 1-2 hours in laboratory liquid culture, may reﬂect intermittent
periods of rapid growth instead of slow, constant growth (232). In
experiments using soil microcosms seeded with spores of Streptomyces
coelicolor or Streptomyces Iividans, a 5—6 day cycle of spore germination,
growth, and sporulation was observed, apparently correlating with
dehydration of the soil (225).

Their non-fastidious feeding habits and ability to sporulate gives
streptomycetes a competitive advantage during extended periods of
starvation and dessication, as shown by their increasing numbers as a
percentage of total soil bacteria in dry or carbon-poor soils (232).
Streptomycete spores are hydrophobic and may be adapted for dispersal by
soil arthropods (186).

Growing mycelia prefer moderate temperatures and pH; no true
‘extremophiles’ have been found among the streptomycetes (152). They are
strict aerobes but can grow in microaerophilic environments. Obligate

heterotrophs, streptomycetes produce a range of degradative enzymes

including cellulases, xylanases, and ligninases that enable them to degrade
and use a variety of organic compounds as carbon and energy sources
(127). Concentrations of “biodegradative” actinomycetes have been
estimated at 105 - 108 per gram of soil and may play signiﬁcant roles in the
degradation of recalcitrant plant materials, including cellulose and lignin
(152). Streptomycetes are thought to have potential as biopesticides (72,
236) and bioremediation agents (196, 237).

To date, no convincing demonstration of antibiotic production by
streptomycetes in nature has been published. However, it is generally
accepted today that antibiotic production occurs in nature and is not simply a
laboratory phenomenon (57). The most compelling evidence is the discovery
in the last several decades that streptomycetes devote substantial genetic
resources to the synthesis and regulation of antibiotics, often greater than
1% of the total genome. However, this indirect evidence is unsatisfying and
tells little about the role of antibiotics in nature. Efforts to obtain more direct
evidence have succeeded for certain fungi and bacteria (27, 215) but have
failed for streptomycetes. The reason for this disparity is unclear; perhaps
streptomycetes produce lower quantities of antibiotics over the course of a
life cycle.

What criteria should be set for a rigorous proof of antibiotic synthesis in
nature? First, it can be agreed that the use of sterilized soil must be
disallowed; although it is relatively easy to recover antibiotics from sterile soil

microcosms seeded with streptomycetes (27, 79, 234), such environments

resemble laboratory cultures more than natural soil. Second, concentrations
of antibiotic-producers seeded into soil must not greatly exceed
concentrations found in natural soil. Third, if biological assays (i.e., inhibition
of a sensitive organism by an antibiotic-producing organism) are used,
isogenic antibiotic-producing and non-producing strains must be compared
so that inhibition by causes other than antibiotics can be ruled out. Finally, is
nutrient amendment to soils allowable? Many soil microbes, including
streptomycetes, are thought to be essentially dormant except when induced
by the occasional inﬂux of nutrients (including water). Therefore, the addition
of nutrients that might be normally encountered in natural soil should not be
disallowed. Plant matter or compounds such as starch and chitin that could
be derived from plant debris, fungal hyphae, or insect carcasses are
acceptable.

Several experiments that approach these standards were performed
soon after the discovery of antibiotics in the 1950’s, when interest in their
natural role was at an early peak. In 1952, Gottleib and Simonoff (79)
extracted chloramphenicol from soil inoculated with Streptomyces
venezuelae; this soil had been pre-sterilized but did contain other microbial
life in the form of co-inoculated Bacillus subtilis. More convincingly, Gregory
et al. (83) and Krasilnikov (27) recovered traces of substances inhibitory to
fungi from non-sterilized, nutrient-amended soil inoculated with
streptomycetes. However, the inhibitory substances were not further

characterized nor were streptomycete inoculum concentrations reported. In

contrast, Wright et al. (27) were able to recover and identify antibiotics
produced by various fungi seeded into natural soil but were unsuccessful
with Streptomyces aerofaciens, S. griseus, and S. venezuelae.

Important questions in soil microbiology cannot be approached until a
method for monitoring streptomycete antibiotic synthesis in soil is found.
Given their abundance, it is reasonable to speculate that streptomycetes are
a major factor in soil ecology. They may also be involved in suppression of
soil-bome plant diseases and biodegradation of xenobiotics. What role do
antibiotics play in these phenomena? It is generally believed that, with few
exceptions, streptomycete antibiotics are used as weapons against
competing microbes. Are they used to protect lysing vegetative mycelia as a
nutrient source for the construction of aerial hyphae as proposed by Chater
and Merrick (51)? This theory depends on the temporal correllation of
antibiotic synthesis and sporulation apparent in laboratory culture but
unproven in nature. By what criteria do streptomycetes that produce multiple
antibiotics “decide” which to synthesize under various conditions? ls
antibiotic synthesis coordinated among closely located colonies and, if so, by

what mechanism?

1.3 Synthesis of Pharmaceuticals by Streptomycetes
Streptomycin was ﬁrst isolated from Streptomyces griseus in the
laboratory of Selman Waksman; one of the ﬁrst medically important

antibiotics, it was used against tuberculosis beginning in 1944. Over the next

several decades, streptomycetes were intensively isolated from soil and
screened for the production of useful antibiotics and other pharmaceuticals.
Nearly 7,000 bioactive compounds resulted, including the majority of useful
antibiotics (206); these include tetracycline, chloramphenicol, cycloserine,
kanamycin, Iincomycin, neomycin, nystatin, cycloheximide, chlortetracycline,
and vancomycin. Other pharmaceuticals derived from streptomycetes
include such immunosuppressants as tacrolimus (FK—506), cyclosporin, and
rapamycin, the antitumor drug adriamycin, the antiparasitic ivermectin, and
the fungicidal agent polyoxin. Screening strains for new compounds
continues to be a signiﬁcant activity for pharmaceutical companies (206).
Over the last decade, both academic and commercial laboratories have
worked towards the development of novel antibiotics through rational
recombination of antibiotic biosynthetic genes and in vivo synthesis.
Antibiotics of the polyketide type are particularly amenable to “mixing and
matching” of genes due to the modular nature of gene clusters encoding
polyketide synthase (PKS) enzymes. Polyketides are constructed through
the action of one or more ketosynthase/acetyltransferase units and acyl
carrier proteins, often with numerous auxiliary modifying enzymes. PKS’s
have been shown to be generally tolerant to modiﬁcation and substrate
alteration (110, 128, 154). Aspects of polyketide synthesis that can be
altered by selective deletion or replacement of antibiotic biosynthetic genes
include degrees of reduction, chain length, stereochemistry, and post-PKS

processing, (39, 110). In addition, feeding of non-natural substrates can alter

1O

the carboxylic acid units used in priming and extension of the chain.
Hundreds of variations on bacterially—synthesized polyketide antibiotics have
already been produced by engineering of antibiotic biosynthetic pathways;
the potential exists for innumerable new bioactive compounds, many of which
would be difﬁcult or impossible to synthesize chemically.

In addition to providing antibiotic biosynthetic genes for the creation of
recombinant systems, streptomycetes are a source of regulatable promoters
and can serve as hosts for the expression of recombinant systems. In a
recent example, the Taxol-like polyketide anti-cancer agent epothilone was
synthesized using Streptomyces coelicolor as a host to express biosynthetic
genes cloned from a Myxobacten’um species (214). Promotors from the
actinorhodin gene cluster of Streptomyces coelicolor have been commonly
used to express recombinant biosynthetic clusters (153, 183). The
genetically well-characterized S. coelicolor and its close relative
Streptomyces Iividans have often been used as expression hosts (15, 128,
153), taking advantage of well-established large-scale fermentation methods
for streptomycetes.

Recently, several vectors that may simplify the construction and
expression of recombinant antibiotic biosynthetic clusters have been
reported. Actinomycete bacterial artiﬁcial chromosomes capable of carrying
100 kb or more of DNA, that replicate in E. coli, and that integrate site-
speciﬁcally into the chromosomes of S. Iividans and S. coelicolor were

constructed by Sosio et al. (200). Another integrative vector for the

11

controllable high-level expression of recombinant antibiotic biosynthetic
clusters in actinomycetes (using the actl promotor of S. coelicolor) was

reported by Rowe et al. (183).

2.0 Streptomyces coelicolor A3(2): A Model Organism

2.1 History

As a graduate student at Cambridge University in the early 1950’s, David
Hopwood chose a Streptomyces strain named “A3(2)” as a research subject
(96). Streptomycetes appeared to be an interesting intermediate between
bacteria and fungi, and strain A3(2) produced a striking blue pigment that
promised to be a useful marker for genetic crosses. Hopwood gave this
strain the species name coelicolor, meaning “sky color”. It was later
discovered that this name had been previously applied to a different
streptomycete and that strain A3(2) properly belonged within the species
“violaceoruber’; however, the name “Streptomyces coelicolor A3(2)” had
become established and is still used by most researchers. Taxonomists
place S. coelicolor strain A3(2) within a group that includes Streptomyces
Iividans 66 and the spiramycin-producer Streptomyces ambofaciens (201 ).

Although research efforts have been scattered over a large number of
Streptomyces species due to interest in the antibiotics they produce, S.

coelicolor A3(2) has become a model organism, particularly for genetic

12

studies. Several properties make it attractive for the study of antibiotic
regulation: (1) it produces four antibiotics, each from a distinct chemical
class, thus facilitating the search for higher order/global regulators of
antibiotics, (2) two of the antibiotics are pigmented and therefore easily
assayable, (3) none of the four antibiotics are commercially useful, promoting
the free exchange of information, and (4) it is thought to be more stable
phenotypically and genotypically than many other streptomycetes.

It is believed that the study of antibiotic regulation in S. coelicolor is likely
to be widely applicable to other streptomycetes since hybridization studies
have shown homologs of many S. coelicolor antibiotic regulatory genes in
other species (8, 149, 178). In addition, S. coelicolor genes cloned into other
streptomycetes tend to be temporally regulated, suggesting conservation of
regulatory mechanisms (145, 157).

Of several S. coelicolor strains commonly used in genetic studies, a
strain known as J1501 and its derivatives are probably the most used. J1501
has useful genetic markers and is a good host for plasmids and phage (49).
The pedigree of strain J1501 is somewhat convoluted; its derivation from the
ancestral A3(2) strain is comprised of a minimum of ﬁve steps, resulting in a

number of both known and uncharacterized alterations to its genome (118).

2.2 The S. coelicolor A3(2) Genome
The S. coelicolor chromosome consists of nearly 8 Mb of DNA (118) and

has a typical streptomycete G+C composition of approximately 74% (77). Its

13

size (approximately twice that of the E. coli genome) and the tendency of
related genes to be located on opposite sides of the genome has led to the
suggestion that the S. coelicolor genome may have resulted from an
ancestral duplication (96). A detailed genetic linkage map of the
chromosome has been constructed (180). A physical map, constructed using
pulsed ﬁeld gel electrophoretic analysis of fragments created by Asel and
Dral digestion, was reported in 1993 (118). The physical and genetic maps
are well correlated. The linearity of the S. coelicolor chromosome was
revealed in 1993 (136, 226). Chromosome ends contain repeated identical
sequences and are covalently bound by proteins (226).

A cosmid library containing nearly the complete genetic complement of
S. coelicolor has been constructed (180) and used, since 1997, in an effort to
sequence the complete S. coelicolor genome by 2001. This project, jointly
organized at The John Innes Center (Norwich, UK) and the Sanger Center
(Cambridge, UK), had completed approximately 90% of the genome as of
December 2000. The Sanger Centre Streptomyces coelicolor Genome

Sequencing Project provides sequence information and annotations to the

a;

public at http:/fwm'v'.sangerac.ukxFroiects; coelicolor. Based on data
collected thus far, it has been predicted that the S. coelicolor genome will
contain approximately 7000 genes (96).

One peculiarity of streptomycete genetics is a tendency towards

leaderless transcripts (10, 113, 205). A high degree of promotor

heterogeneity is also characteristic of streptomycete genes (205), and many

14

genes are transcribed from more than one promotor (111, 113, 205); this may
reﬂect the need for subtle transcriptional regulation in an organism that
undergoes a complex developmental program.

A number of plasmids and transposable elements have been discovered
in S. coelicolor A3(2). SCP1, a giant (350 kb) linear plasmid, exists in self-
replicating form at approximately four copies per chromosome (123) or
‘ integrated into the S. coelicolor chromosome at 9 o’clock on the genetic map
(99, 124). SCP1 is self-transmissable and is responsible for much of the
fertility used in genetic linkage mapping in this organism. SCP2 and its more
fertile variant SCP2* are 30 kb, low copy-number, self-transmissable
plasmids accounting for much of the remaining fertility of S. coelicolor strains
(21, 195). The genetic element SLP1 is found integrated site-speciﬁcally into
the chromosome (25). I811? (formerly known as the 2.6 kb mini-circle) is an
insertion sequence integrated at two sites in the chromosome; it can also

exist independently as a very low copy number plasmid (143).

2.3 Genetic Techniques

2.3.1 Linkage Mapping. A fairly extensive suite of techniques has been
developed for the genetic manipulation of S. coelicolor. Conjugal
recombination was the basis for early genetic studies and is still useful today.
Genetic linkage mapping may be approached using crosses mediated by the
self-transmissable plasmids SCP1 and SCP2 (95). Alternatively, crosses

accomplished by protoplast fusion result in probable complete diploidy and

15

produce recombination at very high frequency, but give results that may be
more difﬁcult to interpret (14).

2.3.2 Vectors. A number of vectors have been developed for introducing
and maintaining DNA in S. coelicolor. Many plasmid vectors are based on
either the low copy-number (2-3/chromosome) SCP2 (21) or the S. Iividans-
derived moderate copy number (40—300 copies/chromosome) le101 (120).
E. coli/Streptomyces shuttle vectors have been developed (59). Plasmids
with E. coli origins of replication (which do not replicate autonomously in
streptomycetes) have been adapted for use as suicide vectors (129). A
polyethylene glycol (PEG)—assisted transformation protocol for the
introduction of plasmids and phage DNA into S. coelicolor protoplasts has
been in use since 1978 (22). Chromosomal DNA can be introduced by
transformation when entrapped in liposomes (144). Plasmids can also be
conjugated from E. coli to Streptomyces (151). The temperate Streptomyces
bacteriophage ¢C31, with a native genome of 41 kb, has been widely used to
integrate DNA into streptomycete chromosomes (31, 129). A number of
useful variants of this phage have been developed that carry selective
markers, convenient cloning sites, and the ability to integrate either at the
attB site or (in attP strains) by homologous recombination between cloned S.
coelicolor DNA and the identical chromosomal sequence (31, 182).

2.3.3 Transposon mutagenesis. Transposon mutagenesis of cloned S.
coelicolor DNA has been accomplished in vitro with manner or with a

derivative of Tn5 (75). Mutagenesis in viva of the closely related S. Iividans

16

has been demonstrated using another Tn5 derivative (220). In each case,
insertion seems to have been fairly random.

2.3.4 Generalized transduction. The isolation of generalized
transducing phages for S. coelicolor and other streptomycetes were reported
only very recently (32) though they have been sought for many years. These
phages are capable of transducing chromosomal and plasmid DNA between
strains of S. coelicolor at frequencies comparable to that of the wild—type P22
phage in Salmonella. lmportantly, they appear to possess broad host ranges
among streptomycetes, implying that they may facilitate the analysis and
improvement of strains that produce commercially important bioactive
compounds but that have resisted genetic manipulation.

2.3.5 Reporter Genes. A number of foreign genes have been used
with some success as reporters of transcription in streptomycetes. The
Pseudomonas-derived xylE , encoding catechol 2,3-dioxygenase (CDO), is
most widely used (31, 108, 238); others include bacterial lux (191, 199),
ampC (71), neo (223), cat (19), melC (171), aph (140), and gfp (209). The
common reporter gene IacZ from E. coli is not useful in streptomycetes due to
multiple endogenous B-galactosidase activities and poor translation of the E.
coﬁgene.

The xylE reporter gene bears further discussion due to its popularity and
its relevance to the work presented in this dissertation. Its use in
streptomycetes was ﬁrst reported in 1989 by Ingram et al. (108) who

constructed a plasmid-borne fusion between xyIE and an S. Iividans

17

galactose-utilizatien operon promotor, galP1. A good correlation between
dot blot-quantiﬁcation of fusion transcripts and xylE-encoded CDO activity
was demonstrated (however, note that this test does not determine whether
transcription of the fusion reﬂects that from the native galP1 promoter). In
1994, Paget et al. reported a lack of reproducibility from xylE fusions
(although data was not shown) and touted the use of the S. glaucescens
meIC gene as an alternative (171). Lindley et al., in 1995, reported a
discrepancy between results obtained from strB1::xylE and strB1::aph fusions
in Streptomyces griseus, although the fusion at fault was not determined
(140). Mogk et al. (158), in 1996, fused the reporters xyIE, cat and bgaB
(beta-galactosidase of Bacillus stearothennophilus) to B. subtilis heat shock
protein promoters dnaK and groE. Although increased activity was measured
from all fusions in response to heat shock, wide variations were observed;
again, which if any of these fusions accurately reported transcription rates
was not determined. In 1999, Hu et al. reported a positive (though rough)
correlation between trpC::xylE fusion expression in S. coelicolor and
quantitation of the trpC transcript itself by St nuclease protection (106).
Craster et al. in 1999 (55) reported sometimes wide variations in CDO activity
in replicate S. coelicolor cultures expressing xyIE from amino acid
biosynthetic gene promoters; however, the authors were inclined to attribute
these variations not to the reporter gene but to small differences in culture

inoculum that became ampliﬁed during growth. Craster et al. also observed a

18

correlation of fusion-encoded CDO activity from these promoters with the
activity of the biosynthetic enzymes themselves (55, 176).

Several cautionary notes regarding xylE have been published. Hassett
et al. (89) showed that CDO activity is very sensitive to hydrogen peroxide
and therefore questioned its reliability, particularly under conditions of
oxidative stress. Gonzalez-Ceron et al. (78) note that xylE in its native state
is cotranscribed with xle, a gene encoding a ferredoxin that functions to
recycle CDO after each round of catalysis. Several fusion constructs
expressing xleE exhibited up to ﬁve-fold greater CDO activity as a result of
this recycling activity compared to fusions expressing xyIE alone. The use of
xyITE promises to increase the sensitivity of this reporter gene assay and
may also improve the accuracy of reporting under conditions in which
recycling of CDO by endogenous electron-transport proteins is limiting.

In summary, the reputation of xylE as a reporter gene is mixed; however,
the data at this time does not seem to support its abandonment. All reporter
genes (and, indeed, all methods of transcript assay) have limitations as well.
The luxAB reporter, for example, was shown to affect the activity of a
promoter to which it was fused, such that different results were obtained with
3 quAB fusion compared to fusions with lacZ, cat, or galK (70). Warnings
have also been published recently about the reliability of green ﬂuorescent
protein (GFP) at higher temperatures and higher concentrations of the

protein (163, 193). However, recent successes in the use of GFP in

19

streptomycetes (86, 131) indicate that it may become the preferred reporter
gene for future studies.

2.3.6 Sequence Analysis. Analysis of streptomycete DNA is aided by
a computer program called “FRAME" that was devised in the early 1980’s by
Bibb and co-workers (20). The high G+C content of streptomycete DNA
results in a bias towards G’s and C’s in the third (wobble) position of codons
in ORF’s, resulting in a typical G+C content of 50%, 70%, and 90% at the
ﬁrst, second, and third positions of the codon, respectively. FRAME

identiﬁes ORF’s by searching for this pattern.

2.4 Antibiotics Produced by S. coelicolor A3(2).

S. coelicolor produces at least four compounds with antibiotic activity;
actinorhodin, undecylprodigiosin, methylenomycin A, and calcium-dependent
antibiotic. Each is chemically distinct and is produced through the action of a
unique set of tightly clustered and contiguous genes. Each cluster includes
genes encoding biosynthetic enzymes, antibiotic resistance, and at least one
regulatory gene speciﬁc to that antibiotic. Antibiotics are produced late in
growth under most laboratory conditions, beginning in transition or stationary
phases (81, 93, 210). S. Iividans, which is closely related to S. coelicolor,
also has the biosynthetic ability to make Act and Red but does not do so
under most conditions; induction of the pigmented antibiotics in S. Iividans
through introduction of cloned S. coelicolor DNA has proven to be a useful

method for identifying genes that may regulate antibiotics.

20

2.4.1 Actinorhodin or “Act” is an isochromanequinone antibiotic that is
blue at alkaline pH and red at neutral or acidic pH. It has weak antibiotic
activity against a range of Gram positive bacteria (235). The blue pigment
visible in cultures of S. coelicolor A3(2) and often in the surrounding media is
actually a mixture of actinorhodin and its lactone derivative gamma-
actinorhodin; the former (mycelium-bound) form appears to be derivatized to
the latter concurrently with expert (37).

Act is a member of a class of natural compounds known as polyketides.
Varied and widely distributed among organisms, pelyketides share a common
biosynthetic mechanism in which simple carboxylic acid metabolites are
iteratively condensed to form chains of various lengths. Decarboxylative
condensation of these metabolites is frequently followed by reduction, giving
rise to the keto groups that give the class its name. Many other
modiﬁcations, including dehydrations, cyclizatiens, isomerizations,
aromatizations, and additional reductions are possible. These potential
modiﬁcations, as well as the choice of staIter and extender carboxylic acid
units and ﬁnal chain length, make possible a great variety of polyketides.
The “Polyketide Synthases” (PKS’s) that catalyze polyketide construction
occur in two classes: Type I, consisting of multiple modules of a core PKS
with auxiliary enzymes, each of which is used once in assembly line fashion,
and Type II, multienzyme complexes in which individual active sites are used

repeatedly.

21

Act is formed by the condensation of an acetyl-CoA starter unit and
seven malenyl-CoA extender units to form an aromatic sixteen-carbon
octaketide; two octaketides are dimerized to make Act (28). The Type II
multienzyme complex that catalyzes Act synthesis is encoded by a cluster of
genes located at 6 o’clock on the S. coelicolor genetic map (235) and
covering a maximum of 26 kb (146). Early cosynthesis studies grouped the
act genes into seven categories (actl-VII) (184). Sequencing has identiﬁed
23 ORF's in the act cluster that are transcribed as at least 7 mRNA’s (98).
The Act biosynthetic pathway and the functions of most of the biosynthetic
enzymes have been elucidated through analyses of mutants, chemical
intermediates, and shunt products. The core PKS consists of a ketosynthase
or “condensing enzyme” which carries the polyketide chain as it is
constructed; it is encoded by the actl-ORF1 and actl-ORF2 genes (67). An
acyl-carrier protein, encoded by actl-ORF3, carries new building units (67).
Other act genes encode modifying enzymes including reductases, cyclases,
and dehydratases, etc.

Mutants representing the actll class did not cosynthesize Act when
grown in pairwise combinations with other mutants, suggesting a regulatory
function (184). A clone restoring Act synthesis to the actll mutant was
isolated from a library (146). Introduction into an Act+ strain in multi-copy
resulted in precocious hyperproduction of Act. Actll-ORF4 has been

characterized as a positive regulatory protein (65).

22

Self-resistance to Act is conferred by proteins encoded by actll-ORFZ
and actll-ORF3 that are believed to export Act from the mycelia (37, 65). The
product of actll-ORF1 is predicted to negatively regulate transcription of
these resistance genes (65).

2.4.2 Undecyprodigiosin or “Red” is the major component of a mixture
of prodigionine compounds (216) that are red at acidic pH and yellow at
alkaline pH. Red possesses antibiotic activity against certain gram-positive
bacteria (185). A highly non-polar tripyrrole compound, Red does not visibly
diffuse from the mycelia (185) as Act does. Proline is probably a biosynthetic
precursor of Red (76); in fact, it has been suggested that Red may be a sink
for excess proline, possibly reﬂecting proline’s role as an osmoregulant in
some bacteria (121). The Red biosynthetic pathway is well characterized
through the analysis of blocked mutants and cosynthesis tests. The products
of at least 18 genes (54, 64, 185) clustered within a 35.7 kb segment of DNA
(147) and located at 5 o’clock on the genetic map (185) are involved. Much
of the cluster has been sequenced and possible functions assigned to a
number of ORFs (54, 64). The red cluster includes a positive regulatory
gene named redD that shares extensive sequence homology with actll-ORF4
(161).

2.4.3 Calcium-dependent antibiotic or “CDA” is a colorless
lipopeptide consisting of an eleven-residue non-ribosomally-synthesized
peptide with a hydroxylated fatty acid at the N-terminus (53). In the presence

of calcium ions, it apparently acts as a monovalent cation ionophore and is

23

active against a wide range of Gram-positive bacteria (132). CDA is assayed
by placing plugs of growing colonies onto low-calcium nutrient agar with or
without added calcium and seeded with a CDA-sensitive bacterial strain
(commonly Staphylococcus aureus); zones of inhibition are caused by the
presence of the antibiotic (132). The cda gene cluster, located at 10 o’clock
on the genetic map (100), was not cloned until 1998 (53). Identiﬁed within
the cluster are genes encoding three peptide synthetases with modular
structures consistent with the synthesis of an eleven amino acid peptide (53,
187); disruption of one of these genes inactivated CDA synthesis (187). Also
identiﬁed are a putative regulatory locus (cdaR) with homology to actll-ORF4
and redD, a putative efﬂux protein, a potential ABC transporter that may be
involved in resistance, and predicted amino acid and lipid biosynthetic
enzymes (187). In an unexpected development, it was recently discovered
that the genes encoding the AbsA1/AbsA2 two-component signal
transduction system, a major focus of the research described in this work, are
located within the cda cluster (187).

2.4.4 Methylenomycin A or “Mmy” is a small (182 Da), colorless,
cyclopentanone antibiotic active against both gram-positive and gram-
negative bacteria (93). In 1976, it was discovered that the genes coding for
Mmy biosynthesis are clustered on the large linear plasmid SCP1 (126, 235)
leading to speculation that antibiotic genes might generally be found in
clusters borne on plasmids. Clustering has proven to be a general paradigm

for antibiotic biosynthetic genes, but to this day the mmy genes are the only

24

conﬁrmed plasmid-borne antibiotic gene cluster. The mmy genes cover a
maximum of 28 kb (98). Under the growth conditions used by Hobbs et al.,
Mmy synthesis begins during the transition from exponential to linear growth
and peaks as stationary phase is reached (93). Self-resistance to Mmy is
believed to occur through the action of a membrane-bound export protein
encoded by the mmr gene (162). Also located within the cluster is mmyR
(48, 98), one of the few known negatively acting antibiotic-speciﬁc regulators

among Streptomyces species.

2.5 Antibiotic-Speciﬁc Regulatory Genes

The best studied of the S. coelicolor antibiotic-speciﬁc regulatory genes
are actll-ORF4 and redD. Under most or all conditions, the proteins encoded
by these genes appear to be necessary and sufﬁcient to induce transcription
of their cognate biosynthetic genes. Strains carrying null mutations of these
genes fail to transcribe biosynthetic genes, while the introduction of multiple
copies of actll-ORF4 and redD into S. coelicolor results in the production of
Act and Red even in exponential phase (65, 210). actll-ORF4 and redD
transcript levels increase dramatically as liquid cultures enter stationary
phase, although low levels of transcripts can be detected in the exponential
phase (81, 98, 210).

The ActIl-ORF4 and RedD proteins share 33-37% amino acid identity
(47, 65) with each other and with Dan, the activator of the duanorubicin

biosynthetic cluster in Streptomyces peucetius (207). Actll-ORF4 and Dnrl

25

can substitute for each other in activation of antibiotic production (207). It
has long been postulated that ActlI-ORF4, RedD, and Dnrl activate
transcription by binding to the promoters of most or all of the biosynthetic and
resistance genes. However, none of these proteins contains a traditional
helix-turn-helix DNA-binding motif (65, 159, 161, 207). Nevertheless, in 1996
Dnrl was shown to bind to dnr biosynthetic gene promoters (213). Imperfect
inverted repeats with the consensus sequence 5’-TCGAG-3’ found in several
dnr biosynthetic genes promoters were proposed as possible binding targets.
In 1997, Wietzorrek and Bibb reported that ActIl-ORF4, RedD, Dnrl, and a
number of other streptomycete antibiotic regulators share a high degree of
amino acid sequence identity with OmpR-type regulatory proteins, which
contain DNA-binding domains that apparently consist of an alpha-helix and
two beta strands (229). The Streptomyces antibiotic regulators are predicted
to contain this secondary structure as well. Therefore, the Streptomyces
proteins were proposed to comprise a new class of proteins named SARP’s,
for §treptomyces Antibiotic Regulatory Eroteins (229). Potential binding sites
consisting of two or three tandem heptameric repeats with the consensus
sequence 5’-TCGAGCG/C-3’ were also identiﬁed in the promoter regions of a
number of act and dnr biosynthetic genes. These sequences are repeated at
either 11 or 22 bp intervals and are, therefore, on the same face of the
promoter DNA. In some cases, the sequences overlap -35 regions on the
face opposite from that of RNA polymerase binding, suggesting that they

could promote binding. Recently, Actll-ORF4 was shown to bind to the

26

sequence 5’-TCGAG-3’ in the -35 region of several act biosynthetic gene
promoters (12).

Additional levels of antibiotic-speciﬁc regulation have been identiﬁed in
the red and dnr clusters. Each contains a gene encoding a protein with
similarity to the response/regulator proteins of two-component signal
transduction systems but lacking the phosphorylation site typical of that
class. These genes are named redZ and dnrN. Studies of disruption
mutants of redZ and dnrN indicate that their products activate transcription of
redD and dnrl, respectively (74, 85, 228). Neither the red nor dnr clusters
appear to contain a sensor/kinase-encoding gene, which are often found
closely linked to their cognate response-regulators. Recently, yet another
apparent regulator of the dnr cluster, dnrO, has been discovered; this gene
encodes a DNA-binding protein that appears to be necessary for dnrN
transcription (170). In contrast with red and dnr, there is no evidence for

antibiotic-speciﬁc regulation in the act cluster other than actll-ORF4.

3.0 Higher Order Regulation of Antibiotics

3.1 Growth Rate/Nutrient Limitation
Given its practical value, our knowledge of conditions that trigger
antibiotic synthesis in streptomycetes is surprisingly limited. One reason for

this is the complexity of the relationships involved; it is clear that many

27

factors can inﬂuence the synthesis of a single antibiotic and that individual
antibiotics produced by a single strain tend to respond to different stimuli. It
is generally believed that a decrease or cessation in growth rate and/or the
depletion of speciﬁc nutrients triggers synthesis of most antibiotics. Addition 1
of glucose, phosphate, or ammonium has been shown to repress synthesis of
many antibiotics (47). Other factors proposed to inﬂuence production of
speciﬁc antibiotics are mycelial density, change in pH, and build-up of excess
metabolites or waste products.

Synthesis of Act by S. coelicolor can be elicited by nitrogen or phosphate
depletion and addition of the same nutrients suppresses Act production (60,
92). However, Act production in continuous cultures is inversely proportional
to growth rate regardless whether glucose, ammonium, or phosphate is
limiting (116), suggesting that growth rate and not depletion of a speciﬁc
nutrient is the triggering condition.

Red synthesis appears to be directed by slowing or cessation of growth
rate and not the availability of particular nutrients, at least under certain
conditions. Takano et al. (210) reported that the stationary phase onset of
Red synthesis in batch cultures proceeded without signiﬁcant delay whether
carbon, nitrogen, or phosphate was in apparent excess, suggesting that none
of these repress and/or inhibit synthesis of Red. In continuous culture, Red
production peaks at a doubling time of about 13.9 hours, decreasing at both

higher and lower growth rates (116).

28

Mmy synthesis begins during the transition from exponential to linear
growth and concurrently with a rapid drop in medium pH caused by the efﬂux
of pyruvate and alpha-ketoglutarate (93). An artiﬁcially induced drop in pH
alone can cause transient Mmy synthesis (90); based on this phenomenon, it
has been proposed that acid shock reﬂecting an imbalance in carbon

metabolism may trigger synthesis (47).

3.2 bld Loci

The existence of genetic regulatory elements common to antibiotic
synthesis and sporulation is demonstrated by the existence of S. coelicolor
mutants affected in both processes. Such mutants are named bld (for bald)
because they are fail to form the fuzzy aerial mycelia that precede normal
sporulation. Each is blocked for the synthesis of at least one antibiotic and
generally all four. _bld mutants are currently divided into 12 distinct types; A,
B, C, D, G, H, I, J, K, L, M, and N. Given the existence of a large number of
bld mutants that are not yet classiﬁed, it is likely that this number will
increase with time.

Many of the bld mutants can be induced to form aerial mycelia when
grown on poor carbon sources (e.g., mannitol) instead of standard complex
or glucose-containing media (43, 156). In some cases, the formation of
aerial hyphae can also be induced by growing bld colonies next to wild type
or certain other bld mutants. In fact, most of the bld mutants can be placed in

a cross-feeding hierarchy, suggesting the presence of a biochemical or

29

signaling pathway (165, 231). Many bld mutants are affected in the
regulation of carbon/energy metabolism; for example, bld A, B, C, D, G, and
H are deregulated in expression of the galactose-dependent, glucose-
sensitive promotor galP1. It has been proposed, therefore, that the inability
of at least some bld mutants to sporulate and produce antibiotics may be due
to an inability to sense starvation (175). The BldJ phenotype is apparently
due to acidiﬁcation of growth medium due to an inability to metabolize
organic acids (218). Some bld mutants (B, C, and H) also differ from parental
strains by their inability to modify certain proteins that normally are ADP-
ribosylated at approximately the time of differentiation (197).

Only a few conditions are known that allow the recovery of antibiotic
synthesis by any bld mutants, indicating a division in regulatory and/or
biochemical pathways. Growth on low-phosphate media allows synthesis of
Red by bldA mutants (137), mutants in bIdH are completely unblocked on a
number of non-glucose carbon sources (43), and bldD mutants produce
signiﬁcant quantities of Act on complex media (62).

The best-characterized bld mutant is bldA. The bldA gene encodes a
tRNA translating the leucine codon UUA (133). This codon occurs only
rarely in the S. coelicolor genome and so far the great majority have been
found in genes involved in secondary metabolism and sporulation (137).
While primary growth is unaffected, bldA mutants produce no antibiotics
under most conditions and morphological differentiation is severely limited;

aerial hyphae are constructed but are not projected vertically (51, 156).

30

These observations led to the proposal of a novel mechanism in which
differentiation would be translationally regulated through the availability of
the active tRNAUUA (133, 137, 138). Act and Red synthesis would be subject
to this postulated regulatory mechanism, since actll-ORF4 and redZ
transcripts each contain a single UUA codon (65, 85). The hypothetical bldA-
controlled regulon has been estimated at approximately 100 genes (96).
This mechanism is as yet unproven, and apparently conﬂicting evidence has
been published regarding the essential point of temporal availability of active
tRNAUU" (81, 139). Alternatively, the presence of UUA codons primarily in
transcripts of secondary metabolic genes may simply result from evolutionary
drift of more essential vegetative genes from the use of a coden that may be
inefﬁciently translated. A similar mechanism of translational control has been
proposed in Clostridium acetobutylicum, where a rare coden may be involved
in the regulation of the secondary metabolites acetone and butanel (189).
Four other bld genes have been cloned. bIdB (174) and bldD (61)
encode small, highly charged proteins that are suspected to have regulatory
functions. Each possesses a potential DNA binding helix-turn-helix motif and
a functionally essential tyrosine as a potential phosphorylation site. Each
appears to negatively regulate its own transcription. As cultures near
stationary phase, transcription of bIdB increases. BIdD has been shown to
bind, and negatively regulate expression from, two developmental sigma
factor genes (62). The sequence of bldG has a high degree of similarity to

Bacillus subtilis anti-sigma/anti-anti-sigma factors (23). The bIdK mutation

31

occurs in one of 5 ORFs that appear to encode subunits of an ABC-type
transporter that is speculated to be an oligopeptide pennease (166);
therefore, it has been proposed that an oligopeptide signal may begin a
cascade leading to sporulation in S. coelicolor. A candidate factor has been

partially puriﬁed from medium conditioned by growth of S. coelicolor (165).

3.3 absA

Postulating the existence of higher order genetic loci globally regulating
the synthesis of antibiotics without affecting sporulation, Champness and co-
workers began a search for such loci in the late 1980’s. At that time, only the
afsB locus (discussed later) was known to affect multiple antibiotics (Act and
Red, but not CDA or Mmy). Champness and then-graduate student Trifen
Adamidis subjected spores of S. coelicolor strain J1501 to random
mutagenesis by ultraviolet light and N-methyl-N’nitro—N-nitrosoguanidine
(NTG). To improve the chances of identifying loci regulating multiple
antibiotics, a screen for the simultaneous loss of the two pigmented
antibiotics was employed. Of approximately 800,000 survivors of
mutagenesis screened in plate culture, eight exhibited the simultaneous loss
of Act and Red while retaining normal sporulation (5). Four of the isolates
had suffered simultaneous mutations in act and red biosynthetic genes and
were discarded (5). The four remaining isolates (named C505, C542, 0554,
and 0577) were genetically mapped in close proximity to each other at

approximately 10 o’clock on the circular map. This location distinguished

32

them from all known antibiotic-related sites except for the cda cluster.
Although resolution from the cda locus was unsuccessful, sufﬁcient
differences in phenotype with any known mutations in the cda region
prompted the classiﬁcation of these mutations as a new locus to be named
absA, for antibiotic synthesis deﬁcient.

Each absA mutant strain was tightly Act-minus and Red-minus on all
media tested and deﬁcient in CDA and Mmy synthesis as well. Introduction
of actll-ORF4 or redD in multi-copy restored Act and Red synthesis,
respectively (4, 42), demonstrating that the mutants remained metabolically
capable of producing the antibiotics. Strain 0505 was chosen for a
comparison of growth rate and mycelial accumulation and found comparable
to the parental strain (5).

To clone the wild type absA gene, a library of parental strain J1501
chromosomal DNA was constructed using the low copy-number plasmid
vector le922 (26). Initial attempts at mutant complementation through
transformation of the library into absA mutant strain C542 were complicated
by two interesting characteristics of the mutant strain: ﬁrst, protoplasts of
0542 were found to be less transforrnable (by more than an order of
magnitude) than the parental type; second, apparent reversion of the absA
strains to a pigment-producing phenotype occurred at high frequency
(approximately 0.1%), making identiﬁcation of complemented mutants
difﬁcult. The apparent revertants also recovered the transformation

efﬁciency of the parental type, making them more numerous on

33

transformation plates. Therefore, the DNA library was conjugated into strain
0542 using the mobilizing ability of le922. Eight clones from the library
restored production of Act, Red, and CDA to all four mutant strains (26).
Mmy was also restored in the two representative transconjugants tested. A
hybridization experiment showed that six of the clones shared some DNA
sequence. A 1.95 kb subcloned fragment complemented all four absA
mutant strains when a single copy was introduced by integration of a
recombinant phage containing the fragment. Phage released from lysogens
and used to infect the parental strain created an antibiotic-minus phenotype
at a high frequency. When used to probe the S. coelicolor chromosomal
DNA cosmid library, the 1.95 kb fragment hybridized to a clone
corresponding to the 10 o’clock location of the genetic map previously
determined for the absA locus. These experiments demonstrated the identity
of the cloned fragment with the absA locus.

Sequencing of the 1.95-kb fragment revealed two incomplete ORFs,
divergently oriented with approximately 130 intervening nucleotides (26).
The larger ORF included the region previously shown by marker exchange to
contain the absA mutations; this ORF was named absA1. Sequencing of a
downstream fragment revealed the remainder of the absA1 sequence, for a
total length of 1715 nt. The deduced amino acid sequence of AbsA1 showed
signiﬁcant similarity to the histidine protein kinases of bacterial two-
component signal transduction systems, especially the Deg/Uhp subgroups.

Typical motifs of this class were present, including the conserved sequence

34

surrounding and including a histidine likely to be a site of
autophosphorylation, and N-terminal hydrophobic regions proposed to span
the cell membrane.

Just downstream of absA1 and oriented in the same direction was
another ORF that would be cotranscribed with absA1. This ORF consisted of
668 nt that showed 39% end-to-end identity with the response regulators
DegU and NarP. Named absA2, the ORF contained typical conserved
motifs, notably a putatively phosphorylated aspartyl residue and a C-terminus
with a helix-turn-helix DNA-binding motif.

Paul Brian, a post-doctoral researcher in the Champness laboratory,
constructed several deletion mutants in the absA1/absA2 system (26). In the
ﬁrst, all conserved regions of absA1 were deleted, probably resulting in a
polar effect on absA2 transcription. This strain, named C420, exhibited a
phenotype essentially opposite to the original absA mutants; i.e., precocious
hyperproduction of Act and Red. Both Act and Red were produced 612
hours earlier than in the parental type and quantities of the antibiotics were
ﬁve-fold and eight-fold greater, respectively, after 75 hours of growth. This
phenotype was given the name “Pha” for Precocious hyperproduction of
gntibiotics. Colony morphology was also affected in C420; the colony
surface was crenulated, sporulation sparse, and transformation efﬁciency
signiﬁcantly decreased. Another disruption mutant of absA1, strain C430,
probably did not have a polar effect on absA2 but resulted in the same

phenotype. Experiments monitoring transcription from the acﬂ biosynthetic

35

gene promoter using a transcriptional fusion to the xylE reporter gene
demonstrated detectable promoter activity 6-12 hours earlier and a
signiﬁcantly higher peak of activity in C420 compared to the parental strain
(26). Mutants deleted in the proposed transmitter domain (including both
kinase and phosphatase activities) of absA1 exhibited the Pha phenotype, as
did mutants deleted in proposed essential regions of absA2 (10). From these
results, it was concluded that absA negatively regulates antibiotics, directly
or indirectly.

Phosphorylated AbsA2 is the negatively regulating form of the response
regulator protein. This was demonstrated through the creation of strains with
mutations in the proposed phosphorylation sites of AbsA2 (054) and AbsA1
(H202), both of which exhibited the Pha phenotype. In addition to their
effects on the Act and Red antibiotics, precocious hyperproduction of CDA
was noted in these mutants as well (10).

Analysis of absA transcription revealed that absA1 and absA2 are
cotranscribed (10). This transcript is leaderless (transcription and translation
beginning at the same nucleotide), a not uncommon phenomenon in
streptomycetes. Transcription is positively autoregulated (apparently by the
phosphorylated form of AbsA2) and growth phase regulated, with transcript
abundance rising by approximately ﬁve-fold between 18 hours and 30 hours
in liquid cultures (10).

Based on these results, a model for AbsA function has been proposed.

In this model, low concentrations of AbsA1 and AbsA2 are present (as

36

indicated by modest transcript levels) during early growth, when the organism
is probably not competent to synthesize antibiotics. Responding to an
increase in the concentration of an unknown signal during the transition
phase, AbsA1 is converted to kinase mode and phosphorylates AbsA2,
resulting in repression of antibiotic synthesis. Simultaneously, AbsA2~P
positively regulates transcription from its own promotor, resulting in an
approximately 5—fold increase in absA1/absA2 transcripts (10). Later, signal
depletion or degradation results in derepression of antibiotic synthesis
through conversion of AbsA1 to the phosphatase form and the resulting
dephospherylation of AbsA2.

Marker rescue experiments showed that the four original absA*
(antibiotic deﬁcient) isolates contained mutations in absA1 (26). Two of the
mutants were characterized at the sequence level (9). Strain C577 carries a
substitution, L253R, within a region proposed to be involved in
aspartylphosphatase activity. Strain 0542 carries two closely placed
substitutions, I360L and R3650, in a region involved in nucleotide binding; it
is not known whether both mutations are required for the antibiotic minus
affect. It is now thought that these mutations lock AbsA1 into a kinase
dominant form; in the case of C577 this would occur through inactivation of
phosphatase activity, in 0542 through the creation of a signal-independent
kinase.

The AbsA system apparently functions to delay/decrease antibiotic

synthesis. Why is this important to the cell? One speculation is based on

37

the sparse sporulation of knockout (Pha phenotype) mutants; early
expression or overexpression of antibiotics may inhibit sporulation (8).

As noted previously, absA mutants are characterized by the frequent
occurrence of spontaneous apparent revertants. Although antibiotic-
producing apparent revertants appear occasionally in other non-producing
mutants, they are much less common than in absA mutants and most are of a
known type resulting from a spontaneous chromosomal deletion (68). Those
arising from absA strains can be isolated and stably cultured (5, 26). A
number of such mutant strains have been characterized at the sequence
level for the information they might reveal about the mechanism of absA1/A2
action. In all cases, they were found to be not true revertants but second-site
suppressive mutations; these were named sab mutants (for suppressors of
_ags). sab mutants occur in two forms: Sabl, resembling the parental strain,
and Sabll, which overproduce Act and Red for a Pha-like phenotype. The
absA mutation in strain 0542 (located in the nucleotide binding region of
absA1) was suppressed to a Sabl phenotype by the substitution G252V in
the absA1 aspartyl phosphatase region. Of two Pha-like sabll strains in the
0542 background, one was associated with a S171W substitution occurring
in the DNA-binding region of absA2, the second with a nonsense coden in
the middle of absA1 with a possible polar effect on absA2. The L253R
substitution in the proposed aspartyl phosphatase region of absA1 strain

0577 was suppressed to a Sabl phenotype by the substitution V29A in an

38

alpha-helical region of absA2, and to a Sabll phenotype by a deletion
including much of absA1, all of absA2, and additional downstream sequence.
The absA1/absA2 locus has recently been localized to within the cda
cluster (187), near one end between the cdaR regulatory gene and the bulk
of the biosynthetic genes. This discovery raises interesting evolutionary
questions; for example, did the absA system evolve with the cda cluster and
only later become a global regulator of antibiotics? If so, does absA control
CDA synthesis to a greater degree or in a different manner than the other

antibiotics?

3.4 absB

The mutant screen that led to the discovery of the absA mutants also
produced a “leaky” mutant that synthesized small amounts of pigment on the
complex “R5” medium used in the screen (3). A second mutant hunt was
undertaken to isolate more representatives of this type, resulting in ﬁve
additional isolates from a screen of 120,000 survivors of UV light— and NTG-
treated spores. This frequency is consistent with loss-of—function mutations.
The six isolates (named 0120, 0170, C175, 0246, 0252, and 0576) were
genetically mapped to an apparent single locus at 5 o’clock, which was
named “absB". Strain 0120 was tested for Act and Red synthesis on seven
types of minimal or complex media and produced none on media other than
R5. Multiple copies of actll-ORF4 or redD bypass the mutation, showing that

the strain is metabolically able to make Act and Red (3, 4). absB mutants

39

also produce much less CDA and Mmy than the parent. Strain 0120 was not
impaired in growth rate, although sporulation was somewhat less than normal
on thin R5 plates. Strains 0175 and 0252 formed smaller than normal
colonies on some media and their antibiotic-minus phenotype was more
severe than 0120’s (178). In contrast with absA* strains, the occurrence of
spontaneous suppressing mutants in absB strains are rare.

Triton Adamidis isolated a clone from the parental J1501 chromosomal
DNA library that restored synthesis of all four antibiotics to the absB mutants
(178). Brenda Price, also a graduate student at that time, continued
characterization of the cloned fragment. The original fragment was reduced
to a 1.4 kb sequence that was still capable of complementing each of the
absB mutants. Sequencing revealed an ORF encoding a putative 276 amino
acid protein with 41% amino acid sequence identity to E. coli RNAase Ill
(encoded by the mo gene) as well as homology to RNAase lll’s from other
organisms. The E. coli RNAase III is a double-stranded endo-RNAase that
processes rRNAs and many mRNAs. Although it is not essential for growth,
RNAase Ill-minus E. coli strains grow slower than normal. Processing of
transcripts by RNAase lll results in either upregulation or downregulation of
as many as 10% of E. coli proteins. The E. coli RNAase lll contains a dsRNA
binding domain and other conserved residues that are also predicted for the
AbsB protein. The target sequence in E. coli is difﬁcult to deﬁne, but often

includes a stem-loop with an ~20 bp double-helical region. The S. coelicolor

40

RNAase III was shown to process 308 rRNA, as does the E. coli enzyme
(178).

The native S. coelicolor absB gene was ampliﬁed from J1501
chromosomal DNA by PCR (178). The native gene was able to complement
absB mutants as either a high- or low copy-number clone. absB was mapped
by hybridization to a position on the physical map that corresponded to the 5
o’clock position previously determined by genetic mapping. The genetic
structure of the absB region suggests that it may be cotranscribed with two
upstream ORF’s, one predicted to encode a 508 ribosomal protein, the other
of unknown function. Homologs of absB have been detected in other
streptomycetes by hybridization.

The absB gene was sequenced in three mutant strains: 0120, 0175,
and 0252 (178). In strain 0120, the gene contained a mutation that would
replace a Leu residue with a Pro, probably disrupting an alpha-helix. In
strain 0175, a mutation results in a nonsense codon at residue 8. In strain
0252, a nucleotide duplication of nucleotides 505 to 511 would result in a
frameshifl mutation.

Disruption of the absB gene resulted in essentially the same phenotype
as did the original point mutations, i.e., creation of a fairly tight Abs'
phenotype but with the more severe antibiotic deﬁciency typical of strains
0175 and 0252 (178). If the putative AbsB RNAase lll functions as a
regulator of antibiotics, it would appear to be a positive regulator.

Alternatively, AbsB may not be a true regulator of antibiotics, but more simply

41

a processing enzyme that confers stability upon transcripts from the

antibiotic-speciﬁc regulatory genes or higher order regulators.

3.5 afs Loci and Autoregulator Signaling Factors

The best studied of a group of small, diffusible, signaling molecules
produced by streptomycetes is “A-factor”, a y—butyrolactone autoregulatory
molecule produced by Streptomyces griseus that is necessary for both
streptomycin synthesis and sporulation in that organism. A-factor is similar in
structure and proposed function to the homoserine lactones (HSL’s) found in
various gram-negative bacteria. HSL’s are bound with high afﬁnity and
speciﬁcity by receptor proteins and are believed to function as quorum
sensors in some bacteria (101, 102). Based on the observation that A-factor
is produced mainly during transition and stationary phases, Bibb (17) has
proposed that it acts not as a quorum sensor but in response to physiological
or environmental signal(s). In contrast, Neumann et al. (164) found evidence
for a programmed developmental cycle in which the presence of low levels of
A-factor during exponential growth were necessary to commit a colony to
later differentiation; this program appeared to be independent of nutritional
state.

S. coelicolor A3(2) produces at least four compounds structurally very
similar to A-factor (17, 211). These compounds are capable of cross-
stimulating S. griseus but, under conditions studied thus far, are not required

for antibiotic synthesis or sporulation in S. coelicolor (87). However, one of

42

these (named 8081) causes precocious production of Act and Red when
added exogenously (47, 117).

Several S. coelicolor genetic loci affect synthesis of both antibiotics and
A-factor-like compounds; these are named afs (for A—jactor synthesis). A
long-standing mystery regarding the identity of one of these loci, afsB, was
recently solved. First reported in 1983, afsB mutants are defective in Act,
Red, and $081 (6, 87, 103); CDA and Mmy are unaffected (3), as are all
other observed characteristics. The effects on Act and Red are apparently
mediated through signiﬁcant decreases in transcription of actll-ORF4 and
redD; interestingly, redZ transcription is unaffected (6, 105). Exogenous
addition of 8081 does not restore Act and Red (6). The afsB mutation is
now known to be located in an RNA polymerase sigma factor, 0”“ (6) (see
later discussion of sigma factors).

Another afs locus was discovered when a cloned fragment of S.
coelicolor DNA, in multi-copy, restored Act, Red, and A-factor-like
compounds to an afsB mutant strain. It was later discovered that the
fragment did not correspond to the afsB locus (103, 203). Two ORFs were
identiﬁed on this fragment and were named ast (69, 103, 203) and afsS
(150, 219). ast alone, in multi-copy, can suppress the AfsB' phenotype and
stimulate precocious hyperproduction of Act and Red in AfsB+ S. coelicolor
with corresponding increases in actll-ORF4 and redD transcription. afsS
alone, in multi-copy, can also stimulate Act and Red; its effect on Act

synthesis is moderated through actll-ORF4 transcription. ast encodes a

43

993-residue protein with a potential DNA-binding helix-turn-helix motif, ATP-
binding consensus sequences, and an N-terminus with signiﬁcant sequence
similarity to both actll-ORF4 and redD (104). AfsS encodes a 63-residue
protein of unknown function (150, 219). The Ast protein is phosphorylated
by a membrane-bound kinase named Asz (94, 149) which belongs to a class
of serine/threonine kinases generally associated with eukaryotic two-
component signal transduction systems. Deletion of ast results in the
absence or reduction of Act, Red, and CDA synthesis under certain
nutritional conditions, while Mmy is unaffected (69). Phosphate
concentration is a key nutritional factor in this conditional phenotype, leading
to speculation that Ast may play a role in a cascade that regulates
antibiotics in response to phosphate depletion (69). Interestingly, actll-ORF4
and redD transcript abundance were not affected by disruption of ast, but
transcripts of the act biosynthetic gene actlll were severely reduced (69).
However, despite this and the sequence similarity between ast and the
antibiotic-speciﬁc regulators, ast cannot substitute for actll-ORF4 and redD
(69). When introduced into absA and absB mutant strains in multi-cepy, a
DNA fragment containing both ast and afsS restores Act and Red synthesis
in a medium-dependent manner but does not restore CDA or Mmy (44).

Two recently discovered S. coelicolor genes encode proteins with
homology (35% identity) to S. gn'seus A-factor receptor proteins (169).
These proteins, CprA and CprB, are predicted to contain DNA-binding helix-

turn-helix motifs. Although they share 91% amino acid sequence identity,

44

CprA and CprB have quite different effects on antibiotic production. CprA
appears to be a positive regulator of Act and Red, as disruption of cprA
severely reduced and signiﬁcantly delayed synthesis of those antibiotics,
while introduction of several plasmid-borne copies into the parental strain
resulted in overproduction. CprB appears to be a negative regulator of Act,
as disruption of the encoding gene resulted in Act overproduction while
additional copies reduced Act synthesis (Red was unaffected) (169).

A fragment of S. coelicolor DNA that stimulated Act, Red, and A-factor
synthesis when introduced into S. lividans led to the discovery of the gene
afsQ1 (109). This gene is homologous to the response/regulator
components of two-component signal transduction systems. The
corresponding histidine protein kinase was subsequently discovered
downstream and named afsQ2. Disruption of both genes in S. coelicolor had
no noticeable affect on antibiotic synthesis, therefore this system may be
non-essential for antibiotic production or may operate under conditions that
have not yet been deﬁned. The ability of an afsQ1-carrying low copy-number
plasmid to suppress an S. coelicolor absA* mutation (109) tends to support
the latter conclusion.

An S. coelicolor gene that may be involved in an autoregulator signaling
pathway distinct from those discussed above has been discovered. This
gene, known as spaA, is homologous to the RspA protein of E. coli (192).
RspA has been implicated in a starvation-sensing pathway; it is proposed to

affect induction of the stationary-phase sigma factor as by degrading or

45

synthesizing homoserine lactones (107). Colonies of spaA—disrupted S.
coelicolor strains exhibited delayed and decreased Act and Red production
when growing on a poor carbon source and at low colony density. At high
colony density, Act was overproduced. This phenotype was not rescued by
the presence of spaA+ colonies next to spaA' mutants, suggesting an
intracellular signaling pathway. Its effect on antibiotic production and its
proposed intracellular role points toward involvement of spaA in a pathway

distinct from that of A-factor.

3-6 (PIPPGPP

Guanosine tetraphosphate [(p)ppGpp] has been proposed as a possible
signal in intracellular nutrient-limitation detection pathways. In E. coli,
(p)ppGpp is an important signal molecule for the regulation of growth rate-
related gene expression. Depletion of amino acids in E. coli cultures triggers
the “stringent response”, which is initiated as uncharged coden-speciﬁc
tRNAs are bound by a ribosomal protein (168) encoded by relC. This leads
to activation of a (p)ppGpp synthetase encoded by the reIA gene.
Carbon/energy depletion can also result in (p)ppGpp synthesis through the
action of another synthetase encoded by spoT. Ultimately, a reduction in
RNA synthesis and an increase in the expression of amino acid biosynthetic
genes occurs in response to the (p)ppGpp signal. A number of antibiotic-
producing streptomycetes (67) exhibit a spike in (p)ppGpp concentration

near the end of exponential growth in amino acid-limited media (204). The

46

temporal correlation between (p)ppGpp synthesis and the onset of production
of many antibiotics has led to suggestions that (p)ppGpp may serve as an
intracellular signal for antibiotic synthesis, at least under some nutritional
conditions (204, 210).

S. coelicolor homologs of the E. coli relA and relC genes have been
cloned (41 , 148, 167, 208). A reIA null mutant of S. coelicolor strain M600
produced no detectable (p)ppGpp and made neither Act nor Red when grown
under nitrogen-limiting conditions; CDA synthesis was unaffected (40).
Failure to produce Act and Red was reﬂected in severe reductions in actll-
ORF4 and redD transcripts (40, 91). In the case of actll-ORF4, this effect
was linked directly to (p)ppGpp levels through the use of a cloned, inducible
copy of relA introduced into a ArelA mutant (91, 208). Induction of the reIA
construct during late exponential growth led to a spike in (p)ppGpp, followed
by increased transcription from actll-ORF4 and an act biosynthetic gene, and
Act production (however, this response was not obtained in mid-exponential
phase, indicating the necessity of other factors). Although low levels of redD
and red biosynthetic gene transcription and Red antibiotic were obtained in
response to modest concentrations of (p)ppGpp, higher (p)ppGpp
concentrations failed to elicit signiﬁcant responses, suggesting the necessity
of another factor or factors. Growth in phosphate-limited conditions allowed
Act and Red synthesis by a reIA mutant, a result consistent with the theory
that reIA is involved in sensing nitrogen limitation instead of starvation

conditions in general (41). In continuous culture, relA proved to be important

47

for optimal production of Act and Red under glucose-limited conditions as
well (1 16).

A second gene with similarity to relA and spoT was recently discovered
in the S. coelicolor genome (208); named rshA, its role is uncertain at this
time since deletion of the gene has no apparent effect. A reIC deletion
mutant did not produce (p)ppGpp, Act, or Red (168). These results strongly
imply that (p)ppGpp plays a role in triggering the synthesis of Act and Red, at

least under certain nutritional conditions.

3.7 RNA Polymerase Sigma Factors

One method by which organisms can regulate gene expression is by
altering the speciﬁcity of RNA polymerase through its exchangeable sigma
subunit. S. coelicolor possesses at least nine unique sigma factors. This
number is likely to increase as additional described transcribing activities and
sigma-homologous DNA sequences are claSsiﬁed; it was recently predicted
that the ﬁnal count will be greater than twenty (96). Of the current nine, two
have been linked to general vegetative gene expression (30, 114, 227), two
with sporulation (50, 177), one with extracytoplasmic functions (142, 173),
and one with response to oxidative stress (115, 172). Of these, four have
also been linked in some way to antibiotic synthesis.

The S. coelicolor sigma factor genes hrdA, hrdB, hrdC, and hrdD were
identiﬁed through sequence similarity to a conserved sequence in the

principal sigma factor RpoD of E. coli and B. subtilis (212) and M. xanthus

48

(34) (hrd stands for homolog of meg). The sigma factor product of the gene
hrdD, a”), is the major in vitre transcribing activity of the antibiotic-speciﬁc
regulatory genes actll-ORF4 and redD (73). hrdD is transcribed in liquid-
grown cultures (34) and translated into a protein of Mr =46 kDa (73). o”"‘° is
controlled by 05 (142), which itself has recently been found to be controlled
by a putative two-component signal transduction system which may respond
to signals from the cell envelope (173). However, disruption of hrdD results
in no obvious defects in antibiotic synthesis or any other function, placing in
doubt a role for 0”“0 for in vivo antibiotic synthesis (35).

em8 (Mr=66 kDa) is believed to be essential based on the failure of
attempts to isolate hrdB-disrupted mutants and may be a “housekeeping”
sigma factor equivalent to E. coli 07° (30). Somewhat higher 0"” -
dependent transcriptional activity is detected in exponential phase compared
to stationary phase (114). In vitro, 0”"8 recognizes a ribosomal RNA
promotor (114), several other promoters that are likely to be expressed in
exponential phase (30, 36), and supports the synthesis of minor quantities of
actll-ORF4 and redD transcripts (73). It is now quite certain that am” is the
major (and perhaps only) sigma factor supporting transcription of the
antibiotic-speciﬁc regulators actll-ORF4 and redD in vivo. As discussed
earlier, a mutation in cm"8 was recently found to be the cause of the Act',
Red", “AfsB” phenotype. Transcription of actll-ORF4 and redD are strongly
reduced in the presence of this mutation, a G243D substitution located in a

highly conserved region of unassigned function. How could a mutation in

49

elm have such narrow consequences? Aigle et al. (6) have hypothesized,
based on the location of the mutation, that it may result in ineffective binding
to either a regulatory factor or to RNA polymerase. It is easily imagined how
the ferrner case could lead to highly speciﬁc effects. Aigle et al. have
proposed that, in the latter case, the mutant 0”“‘3 might bind to RNA
polymerase well enough to be sufﬁciently effective during exponential growth
when present in high concentrations, but at low concentrations during
transition phase it is ineffective. It is not known whether 0””3 has any role in
regulating antibiotic synthesis.

Another sigma factor known as 052 (Mr = 52 kDa) (36, 114) and a
putative sigma factor with Mr = 31 kDa (114) each recognizes one of two
identiﬁed promoters of the actlll biosynthetic gene in vitro. Each has higher

activity in stationary than in exponential phase (114).

3.8 Other Potential Regulators

The cutR/cutS two-component signal transduction system was
discovered fortuitously within a cloned fragment of S. Iividans DNA (217) that
negatively regulates Act and Red production in S. coelicolor when in high
copy-number (45). Disruption results in antibiotic overproduction. Little else
is known of this system.

abaA was isolated as a fragment of S. coelicolor DNA causing Act
synthesis when introduced into S. lividans in a high copy-number plasmid

(66). This fragment maps to a different location from the abs and afs genes,

50

at approximately 2 o'clock. Two ORFs (orfA and orfB) on the fragment were
shown to be responsible for the phenotype; orfA shows similarity to the
transmitter region of two-component systems, while orfB did not show
similarity to any known genes. Disruption of orfB gene in S. coelicolor results
in the complete loss of Act and signiﬁcant reductions in Red and CDA
synthesis without affecting Mmy or sporulation.

The mia (for multi-copy inhibition of sntibiotics) sequence was
discovered fortuitously by Adamidis and Champness on an S. coelicolor DNA
fragment that inhibited production of Act, Red, and CDA when introduced in
copy numbers greater than 2-3 (42). As little as 90 bp of sequence from the
fragment (which includes a portion consistent with the formation of a large
stem loop) is capable of causing this effect (9). The 90 bp sequence includes
the likely promoter region and beginning of an ORF; the antibiotic-inhibitory
effect of mia is therefore likely to result from titration of a DNA-binding
protein. The mia promotor appears to transcribe an operon, tentatively
named absC1 (188). The N-terminus of AbsC1 aligns with two-component
histidine kinases, the C-terminus with protein phesphatases.

The recently discovered ClpP ATP-dependent proteases (0IpP1 and
0IpP2) and their putative regulators 0le and ClpC (56) may also regulate
antibiotic production. Overproduction of 0IpX accelerates Act production in
S. coelicolor and induces Act production in S. lividans. Proteases of this
family affect stationary phase or differentiation phenomena in E. coli and B.

subtilis through degradation of speciﬁc proteins. Degradation of competing

51

primary phase regulatory proteins may be an important function of these
proteases since slew rates of cell growth and division at the time of

differentiation prevents dilution of such regulators.

3.9 Overview of Antibiotic Synthesis Regulation in S. coelicolor.

It is becoming clear that streptomycetes possess complex and ﬁnely
tuned genetic systems that control antibiotic production in response to many
environmental and internal signals. Not only are the onset and shutdown of
antibiotic synthesis controlled, it appears that quantities are modulated as
well and that each of the four antibiotics produced by Streptomyces coelicolor
is (to some degree) individually controlled.

The genetic circuitry governing metabolic and morphological
differentiation in S. coelicolor is often pictured as a regulatory tree with bld
genes at the trunk; mutations at these loci generally result in a global
blockage of antibiotics and arrest sporulation at its earliest steps. The bld
loci that have been cloned and sequenced are predicted to encode proteins
with a wide variety of functions including extracellular signaling, signal
transduction, and both transcriptional and postranscriptional regulation.
Environmental and internal signals feeding into this regulatory level are likely
to include carbon and/or energy starvation and, through an extracellular
oligopeptide signal, colony density, growth rate, or coordination of a group

response. It is likely that some of the bld loci will be found to encode

52

“housekeeping” functions that are necessary for differentiation but do not
truly regulate it.

At the next level, the regulatory tree forms branches consisting of master
regulators dedicated to sporulation (whi genes) or antibiotics. Genetic
elements at this level may be controlled to some degree by bld genes.
Master regulators of antibiotics include some that globally inﬂuence the four
antibiotics (absA, absB, and possibly the mia/absC1 locus) and others that
affect two or more antibiotics (asz/ast, afsS, afsQ2/Q1, cutR/cutS, cprA,
cprB, spaA. relC, and relA). Some signals that are likely to feed into
regulation at this level are Scbl and other gamma-butyrolactones (possibly
communicating growth rate or colony density) and nitrogen, phosphate, and
carbon starvation. Signal transduction is accomplished by phosphorelay in
many of these systems. Other conditions that may be monitored (as
suggested by some evidence or precedent) are levels of trace elements, pH,
temperature, and waste products; some of these are likely to be detected
directly, others monitored through a general mechanism such as growth rate.

At the ﬁnal level of regulation are antibiotic-speciﬁc regulators. It is very
likely that many of the higher order regulatory pathways target these
antibiotic speciﬁc regulators, while others inﬂuence antibiotic biosynthetic

genes directly.

53

4.0 This Work

A primary goal of researchers studying streptomycetes is to understand,
at the molecular genetic level, the regulation of secondary metabolism and
cellular differentiation in these organisms. In addition to valuable basic
knowledge, an enormous practical value is associated with the potential for
efﬁcient, knowledge-based manipulation of the expression of antibiotics and
other pharmaceuticals produced by streptomycetes.

A signiﬁcant body of knowledge has been amassed and continues to
accumulate regarding antibiotic biosynthetic genes and their protein
products. A number of conﬁrmed or putative “higher order” regulators of
antibiotics have been discovered; in most cases, the mechanism by which
they affect antibiotic production is poorly understood. The majority of the
work described in this dissertation seeks to deﬁne unknown aspects of
several of these regulatory pathways; speciﬁcally, potential transcriptional
regulation of several S. coelicolor antibiotic-speciﬁc regulatory and
biosynthetic genes by the absA1/absA2, absB, and various bld loci were
investigated. Transcription was monitored in two ways: by St nuclease
protection transcript assays and by transcriptional fusions between promoters
of interest and the reporter gene xyIE. The use of two methods has allowed
comparison of the techniques as well as exploitation of the strengths of each.

A second facet of this work is the investigation of techniques to monitor

antibiotic synthesis by streptomycetes growing in soil, their natural

54

environment. A number of researchers have tried and failed to detect
antibiotic synthesis by streptomycetes in soil. A new approach was made
possible by the relatively recent development of reporter genes encoding
enzymes that catalyze light-producing reactions. Several approaches were
investigated towards the goal of detecting transcription of lux reporter gene-

fused S. coelicolor antibiotic biosynthesis-related genes in soil microcosms.

55

CHAPTER 2

TRANSCRIPTIONAL REGULATION OF Streptomyces coelicolor

PATHWAY-SPECIFIC ANTIBIOTIC REGULATORS BY THE absA AND

absB LOCI

56

PART A

Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of

Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100-3106.

Reprinted from Journal of Bacteriology, with permission.

Note: Higher quality reproductions of the ﬁgures on pages 60, 61, and 62

can be found on pages 133-135.

57

p.3100
Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100-3106.

58

p.3101
Aceti, D. J. and W. 0. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100-3106.

59

p. 3102
Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100-3106.

60

p.3103

Aceti, D. J. and W. 0. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100—3106.

61

p. 3104

Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100-3106.

62

p. 3105
Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of

Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100—3106.

63

p. 3106

Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of
Streptomyces coelicolor Pathway-Speciﬁc Antibiotic Regulators by the absA

and absB Loci. 180:3100—3106.

PART B

ADDITIONAL RESULTS AND DISCUSSION

Transcripts of an act biosynthetic gene are more severely reduced
than actll-ORF4 transcripts in the abs mutants. Figures 2 and 4 (above)
show that actll-ORF4 and redD transcripts are present at higher levels in the
abs mutant strains than might be expected based on the absence of
detectable antibiotic production in these strains. Substantial levels of redD
transcript were again detected in the absA-542 mutant in recent experiments
of Ryding and Champness as well (187). Therefore, it was of interest to
determine whether antibiotic biosynthetic gene transcript levels were similarly
disproportionate to antibiotic synthesis. The comparison presented in Table
1, below (which incorporates experiments presented in Part A of this work as
well as additional, unpublished data), indicates that transcripts of the ach-
ORF1 biosynthetic gene are reduced substantially more than are actll-ORF4
transcripts in both abs mutants. The degree of reduction in biosynthetic gene
transcripts may, in some cases at least, explain the actinorhodin-deﬁcient
phenotypes of these mutants (a similar comparison of the effects of abs
mutations on red regulatory and biosynthetic genes could not be made due to
the lack of a well-characterized red biosynthetic gene). However, it should
be noted that Act production is not observed even in cases where substantial

quantities of actVl-ORFI transcripts are present (i.e., Figure 3 and Table 1

65

Timecourse 3). Possible implications of this phenomenon are discussed in

Chapter 4 of this work, “Conclusions and Prospects”.

Table 1. Comparison of reductions in actll-ORF4 and actVl-ORFI
transcript levels in abs mutants.

 

 

 

 

 

RNA Fold transcript reduction in Fold transcript reduction in
Time- 0542 (absA*) relative to J1501, 0120 (absB) relative to J1501,
course3 for each assay erfonned for each assay performed
actll-ORF4 acNI-ORF1 actll-ORF4 ach-ORFI
#1 6, 5" >50 2, 2 12, 25
#2 4, 4 11 4, 4 18
#3 3° 4, 66 12° 13", 48

 

 

 

 

 

 

 

3Total RNA was isolated as described in Materials and Methods from 48, 54,
and 60 hour-old minimal mannitol medium plate cultures (Timecourse #1),
from 39, 48, 54, and 60 hour-old PGA medium plate cultures (Timecourse #2),
and from 39, 48, 54, and 65 hour-old PGA plate cultures (Timecourse #3).
”Transcripts were quantitated as described in Material and Methods. Data
deemed unreliable due to low signal-to-noise ratio is not presented.

°Derived from experiment shown in Figure 2, this work.

clDerived from experiment shown in Figure 3, this work.

Do the abs genes influence the timing of onset of transcription from
act and red genes? Clearly, the abs genes inﬂuence actll-ORF4, redD, and
actVI-ORFI transcript abundance. Less clear is whether the timing of
antibiotic regulatory and biosynthetic gene transcriptional onset is affected by
the abs genes. The experiment shown in Figure 4 (above) suggests that the

timing of redD transcriptional onset is not (or is not greatly) altered in the abs

66

mutants; the ﬁrst appearance of substantial quantities of redD transcripts
occurs at the 48-hour timepoint in both of the mutants and in J1501.
Similarly, an increase in actll-ORF4 transcript levels was detected at the 48-
hour timepoint in strains J1501, 0542, and‘0120 in Figure 6 (below). In each
of these experiments, however, temporal shifts of less than nine hours could
have occurred without detection.

The above data suggest that the abs genes do not (or do not greatly)
affect the timing of transcriptional onset of actll-ORF4 and redD. Additional
data bearing on this question, derived from reporter gene assays presented
in the following chapter, as well as the work of Brian at al. (26), Anderson et

al. (10), and Ryding et al. (187), are discussed in Chapter 4 of this work.

Identity of a secondary band in actVl-ORF1 assays. In assays of
actVl-ORF1 transcripts performed in this work (e.g., Figure 3, above), minor
protected fragments corresponding to a length of approximately 170 bp were
observed in addition to those resulting from the predicted 191 nucleotide
transcript. These minor bands were unexplained at the time of publication of
that ﬁgure. Data published since that time by Arias et al. (12) suggests a
possible explanation. Footprinting studies identiﬁed a secondary ActIl-ORF4
binding site within the actVI-ORF1 gene that, if transcription were initiated at
this position, would result in a protected fragment of approximately the size
observed. Furthermore, a potential GTG start codon is located in the same

area,16 bp downstream from the primary transcription initiation site.

67

J1501 C542 C120

 

 

 

 

Time (h) 39 48 54 60 39 48 54 60 39 48 54 60 SM (nt)
527
FLP actll- >
ORF4

404
actll-ORF4 >

FLP glk > 307

242
238

glk (3) > 217

 

201

Figure 6. Expression of actll-ORF4 mRNA in J1501 (abs+), 0542 (absA*)
and 0120 (absB-). RNA was isolated from PGA plate-grown cultures at the
times indicated and used in S1 nuclease protection assays. The size markers
(SM) and probes for actll-ORF4 and glk were as described in Figure 1.

"FLP actll-ORF4" indicates the position of full-length probe; transcriptional
readthrough from the upstream promoter for acﬂl-ORF3 likely contributes to
the signal as well.

68

APPENDIX

Complementation of the absB mutation with the cloned absB gene
partially restores actll-ORF4 and redD transcript levels. As part of an
effort to isolate the genetic locus represented by the absB mutation, I
assayed actll-ORF4 and redD transcripts in a putative complemented strain
of the absB mutant 0120. Brenda Price, then a graduate student, had
isolated a candidate 2.6 kb sequence from an S. coelicolor genomic DNA
library (178). After cloning this fragment into the Streptomyces suicide vector
le963, Brenda transformed the resulting construct (pBK310) into the absB
strain 0120. Transformants exhibiting the le963-encoded hygromycin
resistance contained pBK310 integrated into the chromosome by
homologous recombination through the cloned fragment. Several Hng
colonies were obtained; these produced Act and Red at levels similar to the
parental strain J1501 (178).

To determine whether restoration of actll-ORF4 and redD transcription
had occurred coincident with antibiotic synthesis, l assayed total RNA
isolated from cultures of strains J1501, 0120, and 0120/pBK310. The
0120/pBK310 plate cultures from which RNA was obtained began producing
visible quantities of Act at approximately 39 hours compared to 48 hours in
J1501 cultures. Both J1501 and 0120/pBK310 cultures had produced
copious quantities of Red by 39 hours. S1 nuclease protection assays were

used to determine actll-ORF4 and redD transcript abundance (Figure 7).

69

Figure 7. Complementation of absB mutants by a cloned DNA
fragment. RNA was isolated from PGA plate cultures at the times
indicated and used in S1 nuclease protection assays. The size
markers (SM) and probes for actll-ORF4, redD, and glk were as
described in Figures 1 and 4. “F LP” indicates full-length probe.

70

C120/

 

 

 

(A) J1501 c120 PBK31°
Time(h) 39 48 54 39 48 54 48 54 SM (nt)
527
FLP actll->
ORF4
actll-ORF4 > 404
307
242
238
glk(3)> 2‘7
201
(B) c1201

J1501 C120 pBK310 E
SM (nt) {39 48 54 65i39 48 54 65 ‘39 48 54 65(Time(h)

404

< redD

< FLP glk
307

< glk (2+3)
242
238

217

 

< glk (3)

Figure 7

71

The quantity of actll-ORF4 transcript in strain 0120/pBK310 was nearly
that of the parental J1501 (Figure 7A). Restoration of redD transcript levels,
however, was less complete (Figure 78); following adjustment for the glucose
kinase internal control transcript, redD was found to be 2.5-fold higher in
0120/pBK310 than in 0120 but still 5-feld lower than in J1501. This
apparently incomplete restoration of redD transcription may be an artifact
caused by assaying at timepoints when transcription was dropping rapidly,
resulting in magniﬁcation of small differences in growth phase. As described
by Price et al. (178), an 828 nucleotide open reading frame located within the

2.6 kb fragment of pBK310 was determined to be the absB gene.

72

CHAPTER 3

TRANSCRIPTIONAL REGULATION OF THE actll-ORF4 PROMOTOR:
REPORTER GENE FUSION STUDIES

ABSTRACT

Transcriptional fusions between the promoter of the actll-ORF4 antibiotic
regulatory gene of Streptomyces coelicolor and the xylE reporter gene were
created in a low-copy plasmid and in an integrative phage. These fusions
were used to investigate the possible effects on actll-ORF4 transcription of
mutations in the proposed master regulators absA, absB, and several bld
genes. When introduced into the parental strain J1501, expression patterns
reﬂected those observed in previous direct assays of actll-ORF4 transcripts,
suggesting normal regulation of the promoter. Transcript levels in absB
mutant cultures were about 50% less than in the parental strain when
averaged over a timecourse, a result compatible with prior direct transcript
assays. In contrast, transcription was not altered signiﬁcantly in the absA
mutant; this result contrasts with prior S1 nuclease protection experiments
showing a 3 to 6-fold reduction in actll-ORF4 transcription in the absA
mutant. Several hypotheses are proposed to explain this apparent conﬂict.
In addition, transcription was not altered signiﬁcantly in bldA, bIdB, bldD,

bldG, bIdH, bldl, or bldA95 mutant strains.

73

INTRODUCTION

Prior to investigating the possible roles of absA and absB in regulating
antibiotic-speciﬁc regulatory gene transcription as described in the previous
section (through S1 nuclease protection assays), an effort was made to
address these questions using reporter gene fusions. The reporter gene
approach possessed several potential advantages: (1) A number of reporter
genes have previously been used with success in streptomycetes, (2) Once
constructed, fusions are often easily and quickly introduced into a large
number of backgrounds, and (3) The relative ease of reporter gene assays
may allow the survey of many timepoints with many assay repetitions,
potentially resulting in more precise data.

The catechol 2,3—dioxygenase (0DO)-encoding xyIE gene derived from
Pseudomonas putida was chosen for these studies due to its superior
characterization as the most commonly used reporter gene in Streptomyces
tresearch. Despite its popularity, certain potential pitfalls exist in its use (as is
the case with any reporter gene, or indeed with any method of transcript
quantiﬁcation); these were discussed in more detail in the Introduction to this
work.

Transcriptional fusions between the actll-ORF4 promoter and xylE were
constructed in a low-copy replicative plasmid and in an integrative phage.
These were introduced into the J1501 parental type and derivative absA,

absB, and bld mutant strains. Preliminary results, in which CDO activity was

74

assayed visually (uncorrected for changes in cell mass), suggested abnormal
expression of the fusions; in particular, an apparent lag of greater than 20
hours between transcription from actll-ORF4 and from an act biosynthetic
gene suggested constitutive expression of the former. Therefore, the
reporter gene approach was initially abandoned in favor of S1 nuclease
protection assays. However, later reassessment of the data led to additional
experiments that are presented in this chapter. These experiments were
improved over those performed earlier by spectrophotometric determination
of CD0 activity with normalization for changes in cell mass. They were
designed to extend the study of transcriptional regulation of actll-ORF4 to a
number of bld strains, and to test the viability of the xylE reporter gene

method through comparison with S1 nuclease protection assay data.

MATERIALS AND METHODS
Bacterial strains. S. coelicolor strains used were: J1501 (hisA1, uraA1,
strA1, pg! )(49), 0542 (hisA1, uraA1, strA1, pgr, absA1-542) (5), 0120
(hisA1, uraA1, strA1 , pgf, absB-120) (3), J1501/K0900 (31 ), and bid strains
(1) 0103 (bldA301), 0186 (bldB186), 0112 (bldB112), 0181 (bldH181),
0536 (bldG536), C109 (bldH109), 0249 (bld/249), J774 (bldD53) (156), and
A95 (2),
Growth conditions. Cultures used for reporter gene studies were grown by
spreading spore suspensions (1x105 cfu/plate) onto cellophane discs placed

on SpMR agar media containing 6.25 pg/mL thiostrepton.

75

Enzyme assays. Cell extracts were prepared and catechol dioxygenase
activity was determined spectrophotometrically essentially as described by
Ingram et al. (108), except that cell material scraped from plates and
suspended in 1 mL sample buffer was substituted for liquid cultures. Where
activity was visually determined, plates were sprayed with 0.5 M catechol
solution and incubated for 1 hour at 30°C before color development was
assigned. Protein concentration was determined by the method of Bradford
(24) using reagent purchased from Sigma with bovine serum albumin as the
standard.

Note: lmagg iﬂhis dissertation are presenteg in color.

RESULTS

Construction of low-copy and single-copy actll-ORF4::xyIE
transcriptional fusions. Two types of actll-ORF4::xylE transcriptional
fusion were created. A low-copy (1-3/chromosome) replicating plasmid
construct was made by ligating a fragment of S. coelicolor DNA containing
the actll-ORF4 promotor (Figure 8A) into the plasmid le2840 (18), which
carries a promoterless xyIE gene (Figure 88). This plasmid construct, named
pDA4, had the advantage of ease of construction and introduction into

Streptomyces hosts. Next, a construct that would be maintained at

76

Figure 8. Construction of the actll-ORF4zzxylE fusion vectors pDA4
and DA1. A 1.6 kb S. coelicolor DNA fragment that included the actll-
ORF4 promoter region (A) was ligated into the vectors depicted (B and 0)
to produce transcriptional reporter gene fusions. The promoter fragment
was prepared by subcloning from pAT107, an actll-ORF4-containing clone
isolated by T. Adamidis (3), into the pU018 derivative le2925 (112) and
re-isolating it with Bglll ends to obtain the fragment depicted (“le2925
MCS” is the multiple cloning site sequence of that vector). The promoter
fragment was ligated into the BamHI site of le2840 (B) to produce the
xyIE fusion construct pDA4. The phage construct DA1 was produced by
ligating the promoter fragment into phage K0859 (C) that had been
digested with BamHI and Bglll. Both vectors contain the tsrthiostrepton
resistance gene as a selective marker. “id" is the major transcriptional
terminator of coliphage fd. “to” is the terminator of coliphage lambda.

77

 

 

 

 

 

 

 

 

 

 

. . . . actl l-OR F3 (truncated) k actll-OR F4
89L" ﬂ J) adj..."
£1 I 3’ If I 191bmruncatedonr= '

a"? ‘0; «98
«5f
(B)
le2840
17.4 kb
(0)
I W I 2: —
/\
Bglll BamHl
Figure 8 K0859 (39 kb)

78

single-copy in the S. coelicolor chromosome was made by ligating the actll-
ORF4 promotor fragment into K0859 (31), a derivative of the 41031
actinophage with a promoterless xyIE gene (Figure 80). The resulting
phage, DA1, integrated at single copy into the chromosomal attBisite when
transfected into S. coelicolor. Several attempts to create a third vector by
which the promoterless xyIE could be integrated into the native actll-ORF4
gene were unsuccessful.

The actll-ORF4::xyIE fusions are subject to regulation in S.
coelicolor. Before using the actll-ORF4::xylE fusions to investigate the
effects of mutant backgrounds on transcriptional activity, it was necessary to
establish expression patterns in the parental strain. Therefore, pDA4 and
DA1 were introduced into S. coelicolor strain J1501 by transformation and
transfection/integration, respectively. It was desired to compare the timing of
actll-ORF4 expression with that from an act biosynthetic gene (i.e., a gene
known to be regulated by actll-ORF4). The actinorhodin biosynthetic gene
chosen for comparison in this study was acﬂ, which consists of three ORF’s
encoding the central B-ketoacyI-synthase/acyl carrier/acyl transferase
activities of the actinorhodin polyketide synthase (67). Therefore, a
transcriptional actl::xyIE fusion was created using K0900 (31), a
bacteriophage 41031 derivative supplied by Keith Chater (John Innes
Institute, UK). In this attP phage, an internal fragment of the actl gene

sequence cloned upstream of a promoterless xyIE allows homologous

79

recombination into the S. coelicolor chromosome upon transfection, fusing
xylE to the native actl promotor. Such a fusion was created in strain J1501.

Expression patterns from the actll-ORF4 and actl fusions were expected
to correspond generally to those observed by S1 nuclease protection assays
for actll-ORF4 and ach-ORF1, respectively, in Chapter 2. Additional
precedent was provided by the studies of Gramajo et al. (81), who monitored
transcripts of actll-ORF4 and the act biosynthetic genes actlll and ach-
ORF1 in liquid cultures by St nuclease protection assays. Gramajo et al.
captured early timepoints in the growth cycle, showing that actll-ORF4 is
transcribed at low rates from the earliest assayable timepoints in exponential
phase but increases dramatically as a culture nears and passes through
transition phase, then decreases in stationary phase. Transcription from
actlll and ach-ORF1 was detectable 1 to 2 hours after the onset of major
actll-ORF4 transcription and was followed almost immediately by visible Act
production. In summary, it was expected that as cultures aged, rising levels
of regulatory gene transcripts would be followed closely by biosynthetic gene
transcripts, itself followed by Act production.

Differences in the growth media used in these studies should be noted
here due to their effects on the timing of transition to secondary metabolism.
Gramajo et al. reported visible Act production after approximately 14 hours of
incubation in stirred casamino acid-supplemented liquid minimal medium
cultures. In Chapter 2 of this work, RNA was isolated from cultures grown on

Peptone Glucose Agar plate medium; Act was visibly produced by J1501

80

(Abs+) cultures beginning at approximately 42 hours. In these two studies,
act genes were strongly expressed shortly before the appearance of Act. A
relatively rich solid medium, SpMR, was used for the reporter gene studies
reported in this chapter since attempts to use PGA medium were
unsuccessful due to poor growth (possibly stemming from variation in
medium components). Act was visibly produced on SpMR after
approximately 73 hours of growth, and it was expected that actll-ORF4
expression in the current experiments would occur shortly before visible Act
formation, as it did in the previous experiments. It was unclear whether the
longer lag times between events (i.e. transcription from actll-ORF4, act
biosynthetic genes, and Act production) would also result from a lengthened
growth cycle.

Fusion strains were plated at equal densities on selective agar medium
and CD0 speciﬁc activity monitored overtime (Figure 9). Transcription
patterns from the actll-ORF4zzxyIE fusions generally correlate with those
expected based on the experiments presented in Chapter 2 of this work
(Figures 1 and 2) and on the data of Gramajo et al. (81). The time span
between major actll-ORF4 expression and the onset of actl expression
appears to be between 0 and 24 hours or 0 and 12 hours based on the data
shown in Figures 9B and 9D, respectively. Therefore, this span may not be
signiﬁcantly greater than the 1-2 hours noted by Gramajo et al. or it may be

considerably longer.

81

Figure 9. Expression from actll-ORF4zzxylE and actl::xyIE
transcriptional fusions in S. coelicolor strain J1501. Cultures were
grown and assayed as describe in Materials and Methods. Strains
J1501/K0859 and J1501/le2840 carry the parental (promoterless xyIE)
vectors as negative controls. Graphs A and B represent replicate
experiments, as do graphs 0 and D, but with an additional timepoint (at 36
hours) in graphs B and D. A single experiment using concurrently-grown
cultures resulted in the data shown in graphs A and 0; data was divided
into two graphs for readability, with the J1501::K0900 timecourse used in
both. Graphs B and D also result from a single experiment with similar
treatment of data. With the exception of 36 hour timepoints, each data
point represents the mean of three determinations of catechol
dioxygenase speciﬁc activity in a single cell extract; the former

represent triplicate assays of each of two extracts. Error bars indicate
standard deviation. A unit of catechol dioxygenase activity was deﬁned as
AOD375x106/sec. The onset of actinorhodin synthesis was determined
visually as the ﬁrst detectable appearance of red pigment that became
blue upon addition of sodium hydroxide solution.

82

(A)

Cateehol dioxygenase
Speciﬁc Activity (Unlts/ug protein)

(B)

Catechol dioxygenase
Speciﬁc Activity (Unltslug protein)

Figure 9A.

 

 

  

J1501/K0900
(actl::xylE)

    
 
 

  

J1501/pDA4
(actllOR F4: :xy/E)

 
 
   

J1501 pufeao

(vector control)

 

r a! 7.7- ”I:
V l

 

0 i—‘IV‘TF‘T I
30

Culture Age (hours)

40

80 90 100 110 120 130 140 150

  
  

r I ' I

Onset of actinorhodin production

J1501/pDA4
(actll-ORF4::xy/E)
v i

J1501/K0900
(actl::xylE)

J1501 leZ’c‘sAO
(vector Control)

 

 

80 90 100 110 120 130 140 150

Culture Age (hours) T

Onset of actinorhodin production

83

    
  
 

J1501/K0900
(actl::xylE)

 
  
   
 

J1501/DA1
(actll-ORF4::xy/E)

Catechol dioxygenase
Speciﬁc Activity (Unitslug protein)
0|

J1501/K0859

3O 40 50 60 70 80 90 100110120130140150
Culture Age (hours) T

Onset of actinorhodin production

(D)

    
    

J1501/DA1
(actll-ORF4ztxylE)

Catechol dioxygenase
Speciﬁc Activity (Unlts/ug protein)

J1501/K0900

30 4e 50 so 70 so 901001101201301401
Culture Age (hours)
Onset of actinorhodin production
Figure QB.

84

 

Expression from actll-ORF4::xylE fusions in absA and absB mutant
backgrounds. The fusion constructs pDA4 and DA1 were introduced into
the absA mutant strain 0542 and the absB mutant strain 0120. In Figure 10,
transcription from actll-ORF4 in these strains and in strain J1501 are
compared. A general reduction in expression is noticeable in the absB
strain; in experiments A, B, C, and D, respectively, expression in 0120 was
56%, 35%, 31% and 62% of that in J1501 for an overall mean value of 46%
(percentages were determined from the mean of all data points for C120 and
for J1501 in each timecourse). In the absA1-542 strain, however, differences
are substantially less; values in 0542 were 68%, 86%, 78%, and 155% of
those in J1501 for experiments A, B, 0, and D, respectively, for a combined

mean of 97%.

85

Figure 10. Expression from actll-ORF4 fusions in absA' and absB'
strains. Experiments were done essentially as described in Figure 9.
Graphs A and B represent replicate experiments with the exception of the
36 hour timepoint, as do graphs 0 and D. Actinorhodin was visible at the
time indicated in the J1501 cultures; no pigment was visible in the 0542
and C120 cultures over the timecourses.

86

E

 

101

  
  

J1501/pDA4

    

0542/pDA4

Speciﬁc Activity (Units/ug protein)
OI

Catechol dioxygenase

 

120/pDA4

 

 

 

0 4 1 - I . I . I . r . . . . . . t .
30 4o 50 so 70 so so 100 110 120 130 140 150
Culture Age (hours)

Onset of actinorhodin production

@

 

  
    
      

 

 

’5‘

2

g 10

g. C542/pDA4
3 .3:
2 5
g V
E E
s 2 5 .
g 0 T. J1501/pDA4
at
f l
0 (I)

l C120/pDA4
0

 

so 40 50 so 70 so 90100110120130140150
Culture Age (hours)

Onset of actinorhodin production

Figure 10A.

87

(C)

Cetechel dioxygenase
Speciﬁc Activity (Units/ug protein)

 

30 40 50 60 70 80 90 100 110 120 130 140 150
Culture Age (hours) T
Onset of actinorhodin producilen

(D)

 

25-

2° ‘. ' J1501/DA1

Speciﬁc Activity (Units/ug protein)

Catechol dioxygenase

C120/DA1

 

 

 

 

30 40 50 60 70 30 90 100 110 120 130 140 150
Culture Age (hours) T

Onset of actinorhodin production

Figure 108

88

Expression of actll-ORF4zzxylE fusions in bld mutant backgrounds.
The actll-ORF4::xyIE fusion constructs were introduced into a number of bld
mutant strains; these are characterized by blocks in both sporulation and
(generally) in antibiotic synthesis. It is important to note that two of the bld
strains tested are exceptions to this antibiotic-minus rule; strain J774
(bldD53) made substantial quantities of Act and strain 0181 (bldH181) made
slight quantities (none was detected in the bldH109 strain) as is common for
these strains (4, 43). Therefore, it was expected that these strains would
express actll-ORF4 to some degree. In addition, it was predictable that the
bldA mutation would not affect expression from the fusion since that mutation
is known to affect actll-ORF4 expression at the level of translation (65).

The phage construct DA1 was introduced into strains 0301 (bldA301),
0112 (bld8112), J774 (bld053), 0536 (bldG536), 0109 (bldH109), 0181
(bldH181), 0249 (bld/249), and A95 (no bld locus assigned). Each strain was
also lysogenized with the parent phage K0859 to determine background
levels of expression. Plate cultures were assayed visually for catechol
dioxygenase activity at 44 hours and at 60 hours post-inoculation. All bld
DA1 lysogens, J1501/DA1, and J1501/K0900 were strongly positive at both
timepoints, while all K0859 lysogens were negative (data not shown). These
results suggest that the lack of Act synthesis in the Act-minus bld mutants is
not due to a lack of actll-ORF4 transcription. The pDA4 construct was

transformed into bld strains C103 (bldA301), C112 (bldB112), C186

89

(bldB186), 0536 (bldG536), 0109 (bldH109), 0181 (bldH181), and 0249
(bld/249). The experiment shown in Figure 11 compares expression from the
actll-ORF4 promoter in these bld strains, in J1501, and from the actl
promoter in strain J1501/K0900. It is important to note that, in contrast with
the experiments shown in Figures 9 and 10, catechol dioxygenase activity
was assayed visually in this experiment and was not corrected for changes in
cell mass. The expression pattern in each of the bld mutant backgrounds
tested generally resembled that of the parental strain; therefore, these results
again suggest that actll-ORF4 transcription is not affected by these bld
mutations. This conclusion must be considered with the caveat that the actll-
ORF4 promoter in this fusion may not respond normally to all regulatory

mechanisms (see Discussion)

90

 

 

 

   

 

 

 

 

 

 

121
.é‘
s to
o '1
<
2
at 8 1
3
.9
D
e 6 ‘
.C
.8.
8 —e— C301(bldA301)/pDA4
e 4 ‘ ~0- 0112(bld8112)/pDA4
.2 —e— 0536(bld6536)lpDA4
iii -v- 0181(bldH181)/pDA4
it -a— 0249(bldl249)/pDA4
e: 2 ‘ —o— J1501/pDA4

—n— J1501/le2840
1 + J1501/KC”
o It”; >3I.,G. .3434 1:}.
20 30 40 {it 50 60 70 80

Incubation TIme (hours) l

Aerial hyphae visible

Figure 11. Expression of plasmid-home actll-ORF4zzxylE transcriptional
fusions in bld mutants. Spores of each pDA4-carrying strain, J1501/K0900
(actl::xylE), and J1501/le2840 (promoterless xylE) were streaked onto R5
medium containing thiostrepton. At each timepoint, fresh plates incubated at
30°C for the time indicated were sprayed with catechol and assayed visually
for catechol dioxygenase activity, which was scored on a relative scale.
Values represent the mean of two experiments. Strains carrying alternate
alleles of bIdB and bldH not shown in the ﬁgure (bldB186 and bldH109) gave
similar results to those shown and are not included in the ﬁgure.

91

DISCUSSION

Summary:
-Transcriptional patterns from the. actll-ORF4zzxylE fusions generally
conformed to what is known about the actll—ORF4 promotor, indicating that
the fusions are responsive to at least some regulatory functions.
-Transcription from the actll-ORF4zzxyIE fusion was signiﬁcantly reduced in
the absB mutant strain.
-Transcription from the actll-ORF4zzxylE fusion was not signiﬁcantly
affected by the absA1-542 mutation or the bld mutations tested.

-Several hypotheses are proposed to explain these results.

Based on the following observations, the actll-ORF4 promotor appears
to be subject to regulation in both the plasmid-borne and the chromosomally-
integrated reporter gene constructs; (1) Expression from the promoter varies
signiﬁcantly over the cultures’ life cycle instead of exhibiting an essentially
constant rate as would be expected of an unregulated promoter, (2) These
variations in expression generally correspond to expected patterns relative to
culture age, actl transcription, and actinorhodin production, (3) The time lag
between actll-ORF4 and actl expression is not inconsistent with independent
data.

Transcription from actll-ORF4 is reduced approximately 2-feld in the

absB background, a result that conforms with the 2- to 12-fold reductions

92

observed by S1 nuclease protection assays (Chapter 2). The absB-encoded
RNAase is predicted to function by cleaving a transcript or transcripts,
thereby affecting message stability. The target transcript(s) are unknown,
those from antibiotic-speciﬁc regulatory genes such as actll-ORF4 being one
possibility (increased transcript stability would be predicted in this case,
since mutational inactivation of absB results in lower levels of detectable
actll-ORF4 transcript). However, the following observations suggest that
actll-ORF4 is not a target. First, approximately two thirds of the 3-prime end
of actll-ORF4 is deleted in these fusions and replaced with xyIE , yet absB
continues to affect transcript abundance; this implies that the 3-prime region
of the actll-ORF4 transcript is not the target. Second, if the transcript were
cleaved at the 5-prime end, one would expect to see evidence of a longer
(albeit less stable) transcript in the absB background; there is no evidence of
this in the S1 nuclease protection assays presented in Chapter 2, Figure 2.

Expression was affected very little, if at all, in the absA1-542 mutant; this
contrasts with the three to six-fold decreased expression observed by S1
nuclease protection assays in Chapter 2. Although no satisfying explanation
for this phenomenon has been found, several possibilities are considered
below.

Given the variability in the data, it must be considered that this result
could be an artifact. Additional repetitions of the experiments presented here
would be helpful in settling this issue, as would investigation of fusion

expression in other absA mutants such as the Pha type. However, this

93

explanation seems unlikely, since in none of the four timecourses presented
was there a reduction in expression approaching that seen in S1 nuclease
protection assays.

A second possibility is that the regulatory locus targeted by absA (and
possibly bld genes) could be missing from the fusion construct, while those
for absB (and possibly other regulators) are present. Since greater than 200
bp of sequence downstream and about 1400 bp upstream of the
transcriptional start site are included in the construct, the absA target would
have to be a novel cis-acting locus distant from the actll-ORF4 promoter
region. This seems unlikely, partly due to the distance required. While
regulatory mechanisms operating at such a distance are not unknown (58),
they are presumably rare. In addition, an actll-ORF4 promotor unresponsive
to regulation by absA would be expected to exhibit not normal levels of
expression but hyperexpression resembling that in an absA-knockout Pha-
type mutant. This was not observed.

Thirdly, it is possible that the reporter gene constructs are not faithfully
reporting transcription. This explanation also seems unlikely, since the actll-
ORF4::xy|E construct appears to be regulated by absB and exhibits an
expression pattern in J1501 consistent with regulation.

The variability of signals within and between experiments in this study is
a concern. One of the likely causes for this variability is that cell material
was not consistently harvested from within lawns of conﬂuent growth on

plates; this is desirable since mycelia at the middle and at the edges of

94

growth often develop at different rates. Nevertheless, overall mean activity
was fairly consistent between replicate experiments for each strain, with the
exception of strain J1501/K0900 (actl::xylE). This strain exhibited
approximately 50% less activity in the replicate experiment (Figure 9, graphs
B and D) than in the ﬁrst trial (Figure 9, graphs A and 0). The reason for this
difference is not obvious. On the possibility that a fraction of the mycelia in
the culture used in the replicate experiment had lost the ability to express the
fusion, inocula used in that experiment were plated for isolated colonies and
assayed visually for CDO activity; no signiﬁcant differences between colonies
were seen. A second possibility (which was not tested) is that the spore
suspension used as inoculum was not accurately titered; although it was
ensured that equivalent amounts of cell material were used in CDO assays
(through normalization of signals for cell extract protein content), differences
in growth density could affect antibiotic expression.

Relatively few inferences can be drawn from results involving bld
mutants since there is no independent data regarding the affect of the bld
mutations on actll-ORF4 transcription. Therefore, the simplest explanation
for lack of response of actll-ORF4 expression to the presence of bld
mutations is that Act production does not occur through regulation of actll-
ORF4 transcription, instead targeting a downstream function such as

biosynthetic gene expression.

95

APPENDIX

ASSESSMENT OF LUX REPORTER GENE FUSIONS FOR
MONITORING S. COELICOLOR ANTIBIOTIC SYNTHESIS IN SOIL

ABSTRACT

Antibiotic production by streptomycetes in nature has never been
conclusively demonstrated. A new approach to this question, in which
transcriptional activity of streptomycete antibiotic regulatory or biosynthetic
genes would be monitored through fusions to qu reporter genes, was
investigated. A fusion between the S. coelicolor actll-ORF4 promoter and the
V. harveyi quAB genes was created in a plasmid vector and introduced into
S. coelicolor. Expression from this strain was detected during growth in
sterile, nutrient-amended soil microcosms. Also, several light-detecting
instruments were compared for their abilities to detect expression of a lux-
expressing bacterium mixed into soil; an ATP Photometer proved to be the
most sensitive of the three instruments tested. Finally, a technique for the
rapid extraction and microscopic localization of mycelia from soil was tested
for potential use in monitoring qu expression by microscopic imaging of
individual soil-grown mycelia or colonies. Based on this work, it was

concluded that further advances in reporter gene technology or detection

96

methods may be required before this approach can be used successfully to

monitor antibiotic synthesis in natural soil.

INTRODUCTION

Streptomycetes produce the majority of medically useful antibiotics.
However, no convincing demonstration of antibiotic synthesis by
streptomycetes in nature has been reported. Researchers attempting to
investigate this matter have taken two major approaches. The more direct of
these is extraction and detection (13) of the antibiotic itself from a soil
microcosm seeded with the producing strain and incubated under
appropriate conditions. The failure of this approach is most likely due to the
small quantities of antibiotics produced, degradation by other microbes, and
sequestering of antibiotic compounds by soil colloids. The more indirect
“bioassay’ approach monitors variation in the population of a co-inoculated
organism sensitive to the antibiotic produced by the streptomycete. In this
type of study, researchers have often failed to exclude possible interactions
between producer and indicator strains that are unrelated to antibiotic
synthesis.

Developments in reporter gene technology have suggested a new
approach to this problem. A reporter gene fused to an antibiotic biosynthetic

or regulatory gene promoter could allow detection of expression, by

97

established and sensitive methods, of gene activity reﬂecting antibiotic
synthesis. Reporter genes encoding enzymes that catalyze light-producing
reactions are particularly attractive for soil in situ assays; light can be
detected with great sensitivity and speed, with minimal disturbance of
cultures, and is not bound by soil as either antibiotics themselves or
chemical reporter gene products may be. The luxAB genes derived from
Vibn'o harveyi are in common use as reporters. They encode a
heterodimeric luciferase that catalyzes the following reaction, where RCHO
represents a long-chain aldehyde substrate that can be supplied
exogenously (commonly as n-decanal) and cellular FMNHz provides

reducing potential;

FMNH; + RCHO + Oz -) FMN + RCOOH + H20 + light

Fusions involving lux genes have been used successfully for soil in situ
studies; for example, King et al. (125) reported detection of qu activity from a
pseudomonad in soil slurries. The lux reporter has also been used to study
streptomycetes in plate culture (191).

For this work, it was proposed to construct fusions driven by S. coelicolor
actinorhodin regulatory and biosynthetic gene promoters. Detection of
transcription from these promoters in mycelia growing under conditions as
“natural” as possible in soil microcosms could provide convincing evidence

for the production of antibiotics as well as a tool for further soil ecology

98

studies. Two somewhat different approaches to detection of light emission
were considered. In one, “bulk” light emission from a sample would be
collected and reported as a single cumulative numerical result. Several
instruments capable of this type of assay were available, including an ATP
Photometer, a Photonics System or “Photon Camera”, and a Luminometer.
In addition, a “Fiber Optic System” was speciﬁcally designed and constructed
for light detection in situ by Prof. Mark Worden of the MSU Dept. of Chemical
Engineering; this system ampliﬁes and quantiﬁes light transmitted through a
ﬁber optic cable, the tip of which can be inserted into a soil or other sample.
The second approach, undoubtedly less sensitive but potentially more
informative, would use microscopic imaging to discriminate between light
emission from different parts of a mycelium or colony. Localization of lux-
based light emission to individual bacterial cells has been demonstrated by
(among others) Cai and Welk (38), who showed heterocyst-speciﬁc
expression of genes related to nitrogen uptake in Anabaena. This was
achieved using a Hamamatsu Photonics System or “Photon Camera”, an
instrument capable of detecting the very small number of photons generated
by single lax-expressing cells in the ﬁeld of a light microscope. Patterns of
qu-based light emission collected by the instrument can be superimposed
onto visual microscope images of the organism to localize expression. Using
a similar instrument, Schauer et al. (191) have demonstrated spatial and
temporal patterns of qu fusion expression within streptomycete colonies. An

additional challenge was faced in the current study; that of extracting mycelia

99

from soil and completing the imaging process within a short period of time so
that any signals detected could be attributed to expression during soil growth

and not to disturbance of the culture.

MATERIALS AND METHODS

Bacterial Strains, Vectors, and Growth Conditions. S. coelicolor
strains J1501 (49) and M600 (119) used in this work were grown in liquid
YEME medium (97) supplemented with thiostrepton (10 pg/mL, ﬁnal
concentration) for selection where appropriate. R. meliloti strains 1021 (155)
and CV1 (179) were grown in TY medium (16) containing 200 pg/mL
kanamycin. The qu reporter gene constructs used were le342 (190, 198)
and pDA7 (this work).

Luciferase Assays. All assays were performed with instruments set for
maximum sensitivity (i.e., maximum digital gain and, where applicable, widest
possible aperture) and represent 60 second cumulative signal collections
unless otherwise stated. Signal-to-noise ratio was calculated by dividing
peak signal by the root mean square of noise. The luciferase substrate n-
decanal was used as an emulsion prepared the day of use by 5 minutes
sonication of a mixture of n—decanal (Sigma Chemical Co., St. Louis, MO) at
a concentration of 0.1% vlv in deionized water containing 20 mg/mL fatty-
acid free Bovine Serum Albumin (Sigma). For comparison of luciferase

activity from S. coelicolor M600/pDA7 and R. meliloti strain 0V1, 0.2 mL

100

samples of cultures diluted 1:5 with fresh medium were mixed with 10 pl
decanal solution and, after a 5 second delay, a 10 second cumulative assay
was initiated. Three replicate assays of each culture were done. Dry weights
were determined (four replications each culture) by pelleting‘ cells from 1 mL
culture by centrifugation, washing pellets with distilled water, transferring to
pre-baked and pre-weighed alumininum weigh beats, and drying for 40 hours
at 42°C before re-weighing.

Soil and soil microcosm preparation. Agricultural CAPAC soil from
Kellogg Biological Station was collected and stored at 4°C in sealed plastic
bags until use. Prior to inoculation, soil was spread in a thin layer and air-
dried for several days followed by sieving to eliminate clumps. Soil amended
with starch and chitin was prepared by mixing powdered soluble starch
(Sigma) and powdered crab shell chitin (Sigma) each to 1% w/w and
sterilizing by autoclaving twice for 30 minutes each. Glucose-amended soil
was prepared by mixing 200 pl 1M glucose solution into 19 soil. Soil
microcosms consisted of 0.1 g of sterile, nutrient-amended soil placed in 1.5
mL (8x50 mm) polypropylene luminemeter assay tubes (Turner Designs) that
had been disinfected by soaking for 24 hours in 70% ethanol. S. coelicolor
spore inocula were prepared as previously described (97) and stored at -
20°C until use; spores were pelleted by centrifugation and resuspended in
sterile deionized water before using as inocula.

Note: Images in this (ﬁssertatjon arﬂresentedr in color.

101

RESULTS

Test of the “buried slide” method for extraction of soil-grown
Streptomyces for microscopic imaging. One approach considered for this
study was to attempt observation of qu fusion expression from individual
Streptomyces mycelia or colonies extracted from soil using a light
microscope linked to a Hamamatsu Photonics System (Hamamatsu
Photonics K. K., Hamamatsu City, Japan). This approach would require
rapid extraction of mycelia from soil, location by microscope, and imaging for
light emission; a time limit of several minutes for this process would ensure
that any luciferase activity observed resulted from transcription during soil
growth and not as a result of disturbance. Therefore, the “buried slide” or
“Chelodny” method (52) for extraction of soil-grown bacteria for microscopy
was attempted. Sterile microscope cover slips were partially buried in
samples of sterilized soil that had been inoculated with spores of the
prototrophic S. coelicolor strain M600. After 4 days incubation, cover slips
were withdrawn from the soil cultures. A light microscope was used to search
for mycelia clinging to the cover slip. Mycelia could be located within several
minutes of extraction, and both undifferentiated vegetative mycelial and
differentiated spore chains were observed (Figure 12). Therefore, this

technique appears to be suitable for the desired purpose.

102

 

“-H h .
m
*7 .a—r‘

Figure 12. Soil-grown S. coelicolor mycelia on a “buried slide”. A 50
gram sample of sterilized agricultural soil was inoculated with approximately
1x10° spores of S. coelicolor strain M600 suspended in 5 mL sterile
deionized water. After partially inserting sterile microscope cover slips into
the soil, it was incubated loosely covered at 30°C for 4 days. Cover slips
were then withdrawn, cleaned on one side, then suspended (using a ring of
petroleum jelly around the edge of the cover slip) "dirty" side down over a
microscope slide. Both differentiated spore chains (photographs A and 0)
and vegetative mycelia (B) are visible. Photographs are at 1000x
magniﬁcation. Photography by Prof. Frank Dazzo.

103

150

natal;
5 into
ingot

BIB
dC)

 

Figure 12.

Detection limits of three light-detecting instruments for lax-based
light emission from bacteria in liquid or mixed into soil (in collaboration
with Daniel Ragatz, Leann Matta, Mark Worden, and Frans Debruijn). To
determine whether available light-detecting instruments were capable of
detecting “bulk” (cumulative from all sources in the sample) lux expression
from bacterial concentrations likely to be found in soil, the detection limits of
three instruments for a lux-expressing bacterium were tested. The
instruments tested were; i) An “ATP Photometer” (Lab-Line Instruments,
Melrose Park, IL), ii) A Hamamatsu Photonics System or “Photon Camera”
and, iii) A “Fiber Optic System” designed and constructed by Prof. Mark
Worden (MSU Dept of Chemical Engineering). Rhizobium meliloti strain CV1,
which carries an uncharacterized tux-bearing Tn5 insertion (179), was used
as a test strain. Strain 0V1 was grown to ODsoo = 1.0 in rich liquid medium;
the culture was then serially diluted with fresh medium and the dilutions
mixed into soil. Bacterial concentration was determined by plate counts of
the liquid culture.

The detection limit for liquid dilutions of bacterial culture and for
soil/liquid culture mixtures were tested for each instrument. Samples
consisting of 10 mL diluted liquid culture or 4 9 soil mixed with 0.4 mL diluted
liquid culture were placed into glass “scintillation” vials (2.5 cm diameter).
Addition of 0.2 mL n-decanal solution, mixing, and 5 minutes room

temperature incubation were followed by immediate assay in one of the

105

instruments. For ATP Photometer assays, vials were placed into the
instrument’s light-tight chamber and light emission from the sample assayed
through a window located directly below the vial. For Photon Camera
assays, vials were placed uncovered into a light-tight box and a video
camera located above the vial activated to collect light emission; the signal
distribution displayed on a video monitor was converted to a numerical value.
F or assays using the Fiber Optic System, the ﬁber optic probe was inserted
14-16 mm into the sample and vials were placed in a light-tight box; light
absorbed by the tip of the ﬁber optic cable was transmitted to the
photomultiplier/detector and reported as a numerical value. For each
instrument, cumulative signal from 60 second assays were reported.

First, the ability of the instruments to detect light emission from cells in
liquid was tested. Concentrations as low as 1x10‘ cfu (colony-forming units)
per mL were detectable with the ATP Photometer (Table 2; signals
considered signiﬁcant are in bold type). In contrast, the lowest
concentrations reliably detected by the Photon Camera or Fiber Optic System
were 1x107 cfu/mL. Thus, under these conditions, the ATP Photometer was
approximately 1000-fold more sensitive than either the Photon Camera or

Fiber Optic System.

106

Table 2. Light emission from liquid dilutions of lax-bearing
R. meliloti strain CV1 cells as detected by three instruments.

 

 

 

 

 

 

 

 

 

 

CONC. LIGHT DETECTED, FOR EACH INSTRUMENT
(CFU/ML) of

R, MEMO” ATP PHOTON FIBER OPTIC
STRAIN CV1 PHOTOMETER CAMERA SYSTEM

08 1b 90 831

1 X 103 0 ND ND

1 x 10‘ 162 160 475

1 x 105 2320 163 644

IXW 30010 577 981

1 X107 ND° 40013 9188

1 x 103 ND ND 21512

 

 

 

 

 

°R. meliloti strain 1021, the lux-minus parent of 0V1, was used
here as a control at 2x109 cfu/mL.
°Each value is in Relative Light Units speciﬁc to the instrument
and is the mean two independent experiments, each consisting
of three replicate assays.
cNot Determined.
In soil/liquid culture mixtures, the lowest concentration detectable by the

ATP Photometer was 1x106 cfu/g soil (Table 3). This concentration is well
within the range of actinomycete populations found in some soils; therefore,
the lux fusion approach was not rejected due to this experiment. Comparison
of detection limits in liquid with those in liquid/soil mixtures indicated that
100-fold greater bacterial concentrations are required for detection in soil
due to light blockage by soil particles. Signiﬁcant signals were not obtained

using the Photon Camera and Fiber Optic System, which thus do not appear

to possess the sensitivity required for this purpose.

107

Table 3. Light emission from soil dilutions of tux-bearing
R. meliloti strain CV1 cells as detected by three instruments.

 

 

 

 

 

 

 

 

 

CONC. LIGHT DETECTED (RLU)
(CFUIGRAM
son.) OF R, ATP PHOTON FIBER OPTIC
MELILOT, PHOTOMETER CAMERA SYSTEM
STRAIN 0V1
0a 3b 104 432
1 x 10‘ 3 ND ND
1 x 10‘ 4 ND ND
1 x 103 181 79 ND
1 x 107 968 195 191
1 x 10“ ND° 548 ND

 

 

 

 

 

alR. meliloti strain 1021, the lax-minus parental strain of CV1, was
used here as a control at 2x10° cfu/g soil.

°Values are in Relative Light Units (RLU) that are speciﬁc to each
instrument. Each value is the mean of two independent experiments
each consisting of three replicate assays.

°Not Determined

Development of an S. coelicolor strain expressing a lux fusion. An
S. coelicolor strain that strongly expressed a lux fusion was required for
preliminary tests of detection in soil cultures. One candidate, provided by
Alan Schauer (U. Texas-Austin), was a strain of S. coelicolorJ1501 carrying
the low-copy plasmid (1-3 copies/chromosome) le342. This plasmid carries
V. harveyi luxAB transcribed from the S. coelicolor sporulation-related
promotor sapA (84, 191); an uncharacterized mutation results in earlier-than-
norrnal expression of the fusion. As an alternative, I constructed (Figure 13)
a transcriptional fusion between V. harveyi luxAB and the promoter of the S.

coelicolor actinorhodin-speciﬁc positive regulatory gene (actll-ORF4) in the

108

actll-ORFS (truncated) actll-ORF4 (truncated)
Bgl II CL Bgl II

I
°\ . r 4
‘ 191 ORFtru
lJ2925MCS i ‘9 °° ‘ mm

 

 

r

J
I
I

 

  

le4801

.HK 8650 bp

1“

\

\‘

"t ‘ .
t
l .
-- /
N. ,’
‘ \_¥ . 4,’
°‘\., V.‘ ' '
..—"”_/

Figure 13. Construction of pDA7. A 1.6 kb Bglll fragment containing the
S. coelicolor actll-ORF4 promotor (previously described in Chapter 2)

was ligated into the BamHI site of le4801 in the orientation shown

so that transcription of the V. harveyi luxAB genes were driven by that
promotor. "fd' is the major transcriptional terminator of coliphage fd.

"tsr' is the thiostrepton resistance gene.

109

moderate copy-number vector (30-300/Chromosome) le4801 (29). This
struct, pDA7, was transformed into the prototrophic S. coelicolor strain M600.
Expression from strains J1501/le342 and M600/pDA7 grown in complex
liquid medium were compared. Assays were performed using a Turner
Designs Luminometer (Model TD-20e, Turner Designs, Inc., Sunnyvale, CA),
an instrument similar to the ATP Photometer used in earlier experiments.
Light emission from J1501/pHI342 was easily detectable, peaking at 34 RLU
compared to background values of 0-1 RLU (Figure 14). However,
expression from strain M600/pDA7 peaked at 1954 RLU/sec, approximately
60-fold higher; therefore, strain M600/pDA7 was used in further experiments.
S. coelicolor strain M600/pDA7 was also compared to R. meliloti strain 0V1.
Cultures grown in rich liquid media to OD600= 1.0 were assayed by
luminemeter. In this experiment, signals were normalized to dry cell weight
Since optical density measurements of Rhizobium and Streptomyces are not
comparable (luciferase and dry weight measurements were performed as
described in Materials and Methods). Approximately 5-fold greater light
emission was detected from S. coelicolor M600/pDA7 (1.8x104 RLU/sec/mg
dry weight) compared to R. meliloti (3.7x10° RLU/seC/mg dry weight).
However, normalizing for lux fusion copy number per chromosome
(approximately 100-fold higher in the streptomycete) results in an
approximately 20-feld greater expression level/mg dry weight for R. meliloti

CV1.

110

 

2000 a

1500 2

      

Strain M600lpDA7

 

Light Detected (Relative Light Units)
8
O

500-

Strain J1501/le342

   

 

 

 

 

0 ‘ll'. I I r I i r I r .
40 50 60 70 80 90 100 1 10 120 1 30
Culture Age (hours)

Figure 14. Timecourse of luciferase expression from S. coelicolor
strains M600lpDA7 and J1501/pHI342 in liquid culture. Cultures were
grown in liquid YEME medium. At intervals, four replicate samples were
removed from each culture and assayed. Cell densities were comparable
between cultures as determined by plate count. Each value is the mean of
four assays of a single culture. Error bars indicate standard deviation from
the mean.

111

The limits of detection of S. coelicolor M600/pDA7 in liquid culture and in
liquid culture/soil mixtures were determined (experiments were performed
essentially as previously done with R. meliloti). Signiﬁcant signals were
detected by luminemeter at concentrations as low as 5x103 cfu/mL in liquid
culture and 5x104 chgram of soil/liquid mixture (Table 4; values deemed
signiﬁcantly above background are in bold type). Since detectable
concentrations in soil were well within the range of streptomycete populations

found in nature, these results did not rule out the lux reporter gene approach.

Table 4. Detection of light emission from S. coelicolor
strain M600/pDA7 in liquid and in liquid/soil mixtures.

 

 

 

 

 

 

 

CFU/mLa Signal (RLU)° Signal (RLU)
Or In Liquid In Soil
CFU/g Soil°
0c 0.1 0.00
5 x 102 . 0.1 0.03
5 x 103 1.7 0.10
5 x 10‘ 11.2 0.34
5 x 105 109.0 2.50
5 x 106 957.0 26.80

 

 

 

 

 

aCultures grown in liquid YEME medium were serially
diluted with fresh medium.

°Prepared by mixing liquid medium dilutions into soil.
°Uninoculated medium or soil.

°Relative Light Units. Each value is the mean of four
assays of a Single culture.

Detection of qu fusion expression from S. coelicolor M600lpDA7
during soil growth. S. coelicolor strain M600/pDA7 was grown in sterile soil

amended with either glucose or a starch/chitin mixture. Soils were inoculated

112

with 1x109 spores per gram of soil and incubated for 48 hours in humidiﬁed
conditions. Signiﬁcant signals were obtained from soil microcosms with
either glucose or starch/chitin (Figure 15). No signal was detected when
humidity was not maintained (data not Shown).

Stronger signals were attained through several modiﬁcations to the
assay procedure. Combining n-decanal solution and phosphate-buffered
saline before addition to soil (instead of adding these sequentially as done
previously) increased light emission, presumably through more
homogeneous distribution of n-decanal. More vigorous mixing after addition
of n-decanal solution, additional mixing before subsequent assays of the
same sample, and determination of the time of optimal emission following
addition of n-decanal (approximately 4 minutes) also resulted in stronger
signals. These modiﬁcations were combined in the timecourse experiment
shown in Figure 16 in which the peak signal is nearly 20-fold higher than in
the previous experiment. A signal-to-noise ratio of 188 was determined for

this signal.

113

0.10

 

 

 

 

 

 

Figure 15. Detection of tax fusion expression from S. coelicolor
M600lpDA7 growing in sterilized, nutrient-amended soil microcosms.
Soil microcosms consisted of 8x50 mm luminemeter assay tubes containing
0.1 g sterilized soil amended with glucose or with a starch/chitin mixture.
Each microcosm was inoculated by adding 1x10° spores of S. coelicolor
M600/pDA7 suspended in 20 pl water and mixing, followed by incubation at
30°C in a humidiﬁed chamber for 48 hours. Humidity was provided by a layer
of water at the bottom of the sealed container containing the microcosms.
Luciferase assays were initiated by adding 10 pl 0.1% n-decanal solution and
100 pl Phosphate-Buffered Saline (the latter providing additional liquid
volume to promote homogenous mixing of the n-decanal into soil) and mixing
vigorously. After 1.5 min incubation at room temperature, light emission was
assayed by luminemeter. Each point represents the mean of assays of 6
microcosms. Error bars indicate standard deviation from the mean.

114

 

 

 

 

215

 

-O- S. coelicolor M600/pDA7
+ Unineculated soil

 

2.0 *

 

 

1.5 ~

 

1.04

 

 

 

0.5 n 4°

Light Detected (Relative Light Units)

 

 

0.0 LI

.4

I 1 1 t I I F
4 6 8 10 12 14 16

Time after decanal addition (min)

Figure 16. Timecourse of light emission from S. coelicolor M600lpDA7
following addition of n-decanal. Microcosms were incubated and assays
performed essentially as described in Figure 15, but with the alterations
described in the text. Values are the mean of assays of 3 microcosms. Error
bars indicate standard deviation.

115

 

18

 

 

DISCUSSION

Signiﬁcant steps were taken towards the development of a lax-based
system for detecting S. coelicolor gene expression in soil. A rigorous
demonstration of in situ antibiotic synthesis would require additional steps,
most importantly; (i) The development of a qu fusion that would be stably
maintained without selection (i.e., chromosomaIly-integrated) and
demonstrated to be subject to normal regulation, (ii) The use of natural,
unsterilized soil in microcosms, and (iii) lnoculum concentrations limited to
streptomycete concentrations found in nature. Each of these steps would
necessarily result in decreased signal strength compared to that from the
microcosms described in this study. Extrapolation from current data suggests
that, at best, a marginally detectable signal will result. Therefore, it is likely
that signiﬁcant improvements in methodology giving increased signal
strength or detection sensitivity or both will be necessary for the qu fusion
approach to succeed.

It was shown that the time required for extraction and visualization of
mycelia from soil is not a barrier to the microscopic visualization approach.

S. coelicolor mycelia were extracted from soil and located microscopically
within several minutes of disturbance of the soil microcosm. Exposure to the
luciferase substrate n-decanal could be accomplished concurrently with

location by microscope by placing a drop of the volatile liquid into the

116

chamber created by the microscope slide, cover Slip, and petroleum jelly;
mycelia would be quickly exposed to vapors. Cal and Welk (38) successfully
monitored lux expression from individual cells of Anabaena strain PCC 7120
by this approach; however, a very strong light-emitting strain was required for
detection. It is unknown whether any S. coelicolor antibiotic genes would be
expressed to the same degree as the Anabaena nitrogen uptake-related
genes studied by Cal and Welk, particularly under relatively nutrient-poor soil
conditions. Improvements to the Photon Camera (most obviously, the
addition of a close-up Iense to the video camera) would increase sensitivity.
However, it must be concluded that the microscopic imaging approach, while
appealing, is less likely to succeed than assays of bulk light emission from
soil samples due to sensitivity limitations. Other reporter genes or reporter
gene versions might be considered as a solution to this problem, e.g.,
genetically engineered versions of qu or gfp with enhanced signal strength.
Recently, confocal laser scanning microscopy has been used to detect
expression in individual mycelia of single-copy chromosomal fusions between
enhanced gfp and redD (209), and lat and ccaR (antibiotic biosynthetic and
regulatory genes, respectively in S. clavuligen's (86, 131).

Of three light-detecting Instruments tested in “bulk” soil assays, the Lab-
Line ATP Photometer was the most sensitive. Detection by the Fiber Optic
System or Photon Camera required bacterial concentrations several orders
of magnitude higher. The Fiber Optic System potentially allows fast,

location-targeted, and convenient assays, but its sensitivity may be limited by

117

the small area sampled by the tip of the ﬁber optic cable. As with
microscopic imaging, the addition of a close-up lense to the Photon Camera
would improve its sensitivity for bulk soil assays.

The S. coelicolor strain M600/pDA7 constructed in this study expressed
lux at approximately 60-fold higher levels than did S. coelicolor carrying a
low-copy sapAzzlux fusion; this factor roughly correlates with fusion copy
number, suggesting Similar promotor strengths. S. coelicolor M600/pDA7
was also compared to R. meliloti CV1, which carries a single-copy qu fusion.
Expression from the Rhizobium strain was approximately 20-fold higher per
mg cell dry weight when adjusted for fusion copy number. The reason for
this difference is unknown; obvious possibilities lie in the strength of the
respective promoters and in physiological differences between the species
(the latter perhaps resulting in reduced FMNHz, 02, or n—decanal substrate
availability in the streptomycete). Since promotor strength is probably the
easiest factor to manipulate, the identiﬁcation of an antibiotic-related
promotor with greater transcriptional activity than actll-ORF4 might be a good
ﬁrst step to signal improvement.

Given this comparison of the Streptomyces and Rhizobium strains, it may
seem surprising that minimum detectable concentrations in liquid were very
Similar (5x103 cfulmL and 1x104 cfulmL, respectively) and favored the
streptomycete in soil/liquid culture mixtures (5x10‘ cfu/gram and 1x10°
cfu/gram, respectively). The explanation undoubtedly lies in the differing

properties of a “colony-forming unit” for each organism. Streptomycetes grow

118

 

in liquid culture as pellets of intermeshed mycelia that may reach
considerable cell mass, and each pellet is likely to appear as a Single colony
in a plate count. In contrast, a colony-fenning unit of Rhizobium will consist
of one or a few cells. Therefore, at comparable cfulmL or cfulgram
concentrations, it is likely that much more streptomycete cell mass was
assayed for luciferase activity.

What can be inferred about the detectability of soil-grown Streptomyces
from the values obtained by mixing liquid-grown culture with soil? When
inoculated into sterilized soil containing appropriate nutrients and moisture,
most spores will germinate and grow to form branching networks of mycelia
(141); however, these will undoubtedly consist of signiﬁcantly less mass than
those grown in rich liquid media. Moreover, in unsterilized soil many spores
remain dormant due to unspeciﬁed interactions with other soil microbes
(141). Therefore, it can be expected that detection of mycelia grown in
natural soil will require populations several orders of magnitude higher than
the 5x10‘ cfulgram detected when a liquid—grown culture was mixed into soil.

Expression from S. coelicolor M600/pDA7 was easily detected after
spores were inoculated at 1x109 cfulgram soil in sterile, nutrient-amended
soil microcosms that were then incubated for several days. Method
improvements resulted in increased light emission with a signal-to-noise
ration of 188.

S. coelicolor M600/pDA7 is unsuited for attempts to detect expression

under natural conditions because the qu fusion is carried on a multi-copy

119

 

plasmid. It is unlikely that this plasmid will be stably maintained during soil
growth in the absence of selection. Therefore, a strain expressing a single-
copy, chromosomally-integrated fusion construct would be highly preferable.
Based on available data, would expression from such a fusion be detectable?
An approximately 100—fold decrease in fusion copy number would likely result
in a corresponding decrease in light emission. Given the signal-to-noise ratio
of 188 achieved in the experiment depicted in Figure 16, it seems probable
that a marginally detectable signal would result. Greater signal strength
might be achieved by using an act biosynthetic gene promoter, at least some
of which are likely to be expressed to a greater degree than the actll-ORF4
promoter (in addition, the use of a biosynthetic gene promoter would have
the beneﬁt of more directly reporting antibiotic synthesis). Furthermore, it is
likely that assay sensitivity can be further increased by experimenting with
such factors as n-decanal concentration and method of addition, nutrient
amendment, and by determining the time period during incubation of a soil
microcosm during which fusions are optimally expressed. However, these
improvements must be balanced against requirements for “natural”
conditions. First, inocula must be reduced approximately 100-feld to reﬂect
streptomycete concentrations found in natural soil. Actinomycetes have been
reported at up to 1x10° cfulgram of natural sell (152, 221) and

streptomycetes often make up 10% of the total actinomycetes recovered from
soil (130); therefore, a concentration of 1x107 cfu/ gram soil would seem to be

the maximum allowable. Since this number corresponds to recoverable

120

concentrations of streptomycetes, it would be permissable to inoculate with
higher concentrations if (11) predation and competition lower recoverable
numbers. However, it has been found that a high percentage (approximately
90%) of spores inoculated into unsterilized soil do not germinate (141). The
combination of these factors will likely result in a signal reduction of
approximately 1000—feld. Considering both factors that will likely increase
and decrease signal strength, in balance it appears unlikely that antibiotic
synthesis in soil can be monitored using luxAB without additional
improvements in expression or detection.

As mentioned above, rather impressive results have been published
recently in the detection of gene expression through fusions with the gfp
reporter gene. Although cultures used in those studies were grown in rich
laboratory medium, nevertheless the use of enhanced versions of gfp would

appear to be a promising approach to in situ detection.

121

 

 

CHAPTER 4

CONCLUSIONS AND PROSPECTS

A primary goal of this work is to determine whether the absA and absB
loci affect Act and Red synthesis through transcriptional control of the
antibiotic-speciﬁc regulators (ASRs) in the act and red gene clusters. It is
hoped that this knowledge will help to deﬁne the pathways by which the abs
loci globally inﬂuence S. coelicolor antibiotic biosynthesis. In the following
discussion, two models are considered in light of the data generated in this
study and the ﬁndings of other researchers. In one, ASR (i.e., actll-ORF4
and redD) transcripts are, in fact, targeted by the abs loci (note that control of
either transcription rate or transcript stability Is possible in this model). In the
alternative “downstream target“ model, a process downstream of ASR
transcription/transcript stability (for example, biosynthetic gene transcription)
is targeted. It should be noted that control of transcript abundance does not
necessarily imply regulation; the possibility exists that absB, in particular,
performs a “housekeeping” and not a true regulatory function.

Do the abs loci control actll-ORF4 and redD transcript abundance?
It has been shown through S1 nuclease protection assays that mutations in
absA and absB result in altered levels of actll-ORF4 and redD transcripts.
Compared to an abSA+ strain, plate cultures of an absA“ mutant (in which
absA1 is believed to be locked into phosphorylation mode) showed 3 to 6-

fold less actll-ORF4 transcripts, 7-fold less redD transcripts, and between 4-

122

fold and greater than 50-fold less acNI-ORFI biosynthetic gene transcripts.
Disruption of absA2, in contrast, resulted in approximately 2-feld increases in
each of these transcripts. An absB' mutation resulted in 2 to 12-fold less
actll-ORF4 transcripts and 7 to 13-fold less redD transcripts. It should be
noted here that the absB-120 mutation is now known to only partially
inactivate absB activity; thus, different results might be obtained if other
mutations that completely inactivate absB (resulting in a tight, absA*-like
antibiotic-minus phenotype) are studied.

The simplest interpretation of the results described above is that the abs
genes control ASR transcript abundance which, in turn, affects biosynthetic
gene expression. In contrast, a “downstream target” model must incorporate
a speculative mechanism to account for the observed decrease in ASR
transcripts. One such possibility is a positive feedback loop in which Act,
Red, or their precursors serve to stimulate ASR transcription; in an absB”
mutant, for example, reduced levels of antibiotic precursors resulting from
decreased biosynthetic gene expression might feed back to reduce actll-
ORF4 and redD transcription. A possible precedent for such a feedback
mechanism can be found in the tylosin pathway of Streptomyces fradiae (33).
Production of the core polyketide component of tylosin is severely reduced
when a deoxyhexose component, also synthesized by tyl gene-encoded
enzymes, is unavailable for addition to the polyketide. A positive feedback
mechanism involving expression of the regulator Tle (analogous to Actll-

ORF4 and RedD) has been suggested to account for this effect. In addition

123

to a feedback mechanism, it is not difﬁcult to imagine the existence of other
indirect effects of the abs genes on ASR transcription, given the pleiotropic
effects of the abs mutations.

An ASR target is also favored by the discovery that introduction of actll-
ORF4 and redD in high copy-number overcomes the antibiotic-minus effects
of both absA and absB mutations (3, 4, 42). The most straightfonlvard
interpretation of this result is that deﬁciencies in ASR transcripts have been
repaired by the introduction of the high-copy genes. However, this is not a
particularly strong argument since it is quite possible that deﬁciencies
“downstream” might also be overcome by overexpression of ASRs.

Both of these models must account for two apparent discrepancies in the
S1 nuclease protection data: (a) The disproportionate effects of the abs
mutations on ASR transcripts and on biosynthetic gene transcripts, and (b)
The disproportionate effects on biosynthetic gene transcripts and on
antibiotic production. The ﬁrst discrepancy is illustrated in Table 1, which
shows (in three separate cultures) mean decreases in actll-ORF4 transcripts
of 4-fold in an absA* mutant and 6-fold in an absB mutant, compared to mean
reductions of approximately 22-feld for acNI-ORF1 in each of the mutants.
This phenomenon is easily accounted for by the “downstream target” model,
Since ASR transcripts would be affected indirectly, if at all, by alteration of
downstream activities. In the “ASR target” model, however, abs mutations
inﬂuence biosynthetic gene transcription through levels of ASR transcripts.

In a simple system where actll-ORF4 and redD transcripts are limiting factors

124

in Act and Red synthesis, alterations in levels of those transcripts Should
translate to approximately proportional changes in biosynthetic gene
transcription. On the other hand, if ASR transcripts are not a limiting factor,
alterations should result in less acute changes in. downstream processes,
contrary to what is observed.

Mechanisms that could account for this phenomenon and that ﬁt into the
“ASR target” model can be imagined and are supported by some
circumstantial evidence. For example, the Actll-ORF4 protein might be
bound by various act biosynthetic gene promoters with signiﬁcantly different
afﬁnities; a modest decrease in ActIl-ORF4 (as in the abs-542 mutant) might
then result in titration of the protein by stronger-binding act promoters and a
disproportionate scarcity at weaker binding sites. If the achl-ORF4 promoter
were a weak binding site, transcription from that gene could be severely
reduced in this situation. Differential regulation of act biosynthetic genes is
supported by a comparison of the often substantial acIVl-ORF1 transcripts
observed in the abs mutants in this work with the complete lack of detectable
expression from an actl::xylE fusion in these mutants (the latter experiment
described in Aceti and Champness, 1998). In addition, Arias et al. (12) found
different ActlI-ORF4 binding efﬁciencies in vitro for DNA fragments containing
act biosynthetic gene promoters (one with the divergent promoters for ach-
ORF1 and aclVI-ORFA, the other with actlll and actl-ORF1 divergent
promoters). In a second scenario, disproportionate expression of ach-

ORF1 could be caused by a cooperative regulatory mechanism. This idea is

125

based on the multiple Actll-ORF4 binding sites discovered in that gene (12);
footprinting studies identiﬁed a presumed primary site located within the
promoter region and a secondary site approximately 20 bp downstream of the
ﬁrst and within the transcribed region. A cooperative mechanism could result
in a sigmoidal expression curve such that changes in Actll-ORF4
concentrations over a certain range would have disproportionately large
effects. Such a mechanism is exempliﬁed by the E. coli deo operon. Binding
of a regulatory protein to either the deoP1 or deoP2 operator leads to a 2-
fold effect on transcription while binding to both Sites results in a 20 to 30-
fold effect (58); in this way, a modest decrease in regulatory protein
concentration could severely reduce transcription from the regulated gene.

The second discrepancy (lack of detectable Act in cultures with
signiﬁcant levels of actVl-ORF1 biosynthetic gene transcript) can also be
explained by differential regulation of the act biosynthetic genes. A more
severe reduction in expression of m act biosynthetic genes could block
Act production. Again, such a reduction appears to be demonstrated in both
abs mutants by the actl::xylE fusion. The difference between the models on
this issue is the identity of the differentially regulating effector: Actll-ORF4 in
the “ASR target” model, AbsA2~P and AbSB or their downstream effectors in
the “downstream target” model.

In Chapter 3, the behavior of acﬂl-ORF4::xyIE fusions in absA* and absB
mutant strains is described. It should be noted that, in recent years,

conﬁdence in the accuracy of xyIE as a reporter gene has declined

126

considerably among many researchers studying streptomycetes (44). The
importance attached to these results should be adjusted accordingly.

The actll-ORF4::xylE fusions exhibited an approximately 2-fold decrease
in activity in the absB strain, a result consistent with results of S1 nuclease
protection assays. However, the fusions showed essentially no response to
the absA* mutation; i.e., Act synthesis was completely blocked while actll-
ORF4 transcript levels were unaffected. If accurate, this result poses a
problem for the “ASR target” model. In balancing the reliability of the xylE
fusion assays versus S1 nuclease protection assays, however, it must be
concluded that this result is most likely an artifact.

It is not clear why the fusion would respond inaccurately to an absA*
mutation while reporting with (apparent) accuracy in J1501 and absB mutant
backgrounds. In one scenario, if absA targeted actll-ORF4 transcript
stability, the replacement of the 3-prime end of that transcript with xylE mRNA
could render it unresponsive to regulation by absA. Alternatively, a
regulatory site targeted by absA could be missing from the fusion construct
while sites for absB (and possibly other regulators) were present. Since
greater than 200 bp of sequence downstream and about 1400 bp of upstream
of the transcriptional start site are included in the construct, this would
require an absA target distant from the actll-ORF4 gene. Regulatory
mechanisms operating at such a distance are not unknown (58). In either
case, lack of regulation by absA would be expected to result in above normal

levels of the transcript reminiscent of a Pha-type mutant; however, it is

127

possible that the approximately 2-fold increase expected would not be
distinguishable in this assay.

These “ASR target” and “downstream target' models can also be
compared with respect the degree of precedent for regulation at each level.
Many examples of apparent regulation at the ASR level exist. Few
indications of regulation at the downstream level exist, one being the
disruption of ast, which resulted in a severe decrease in actlll transcripts
but no effect on actll-ORF4 transcripts (69). However, the most relevant
precedent was recently provided by Ryding and Champness (187), who have
discovered that cda biosynthetic genes are apparently controlled directly by
the absA regulatory pathway, not through the putative ASR for that cluster
(cdaR). While cdaR has not been conclusively proven to be an ASR, other
possible regulatory genes in the cda cluster were tested and no evidence
was found for regulation of those genes by absA. Given the likelihood that
absA evolved to regulate CDA production, it can be argued that its easiest
evolutionary course to regulation of Act and Red would be through analogous
routes in those systems. These recent discoveries regarding regulation of the
cda cluster by absA, therefore, tend to favor the “downstream targets” model
in the act and red systems.

In conclusion, neither the “ASR target” nor the “downstream target”
model can be ruled out at present. Arguments discussed early in this chapter
are relevant to both absA and absB and tend to favor ASR targets. The ﬁnal

two arguments (regarding the behavior of actll-ORF4zzxylE in absA mutants

128

 

and the precedent of cda regulation) are speciﬁc to absA. The ﬁrst must be
given relatively little weight due to the reputation of the xylE reporter for
unreliability, and the second is speculative. At this time, therefore,
preference must be given to the theory that the abs loci target ASR
transcripts.

These questions could be further explored in a number of ways. The
ability of act and red biosynthetic and regulatory gene promoters to bind
AbsA~P can be tested; although it is not known whether AbsA~P is the
terminal represser in the absA pathway, this is the simplest case and the only
testable one at this time. If AbsA~P is not the terminal represser,
identiﬁcation of the other component(s) of this regulatory pathway is a very
high priority. Assay of transcripts from a greater number of act biosynthetic
genes (and from red biosynthetic genes, which are now better characterized)
would address the hypothesis of differential regulation. The possible
existence of a feedback mechanism could be tested relatively easily by
determining whether actll-ORF4 and redD transcripts are reduced in various
biosynthetic gene mutants that already exist.

Do the bld A, B, D, G, H, I, or A95 loci control ASR transcript
abundance? Expression from actll-ORF4zzxylE fusions in a bldA mutant
strain was not Signiﬁcantly altered compared to expression in the parental
type; this result was expected, as the bldA mutation is known to affect
translation, not transcription, of actll-ORF4. The bldD and bldH-181 mutants

produced signiﬁcant and minor quantities of Act, respectively, under the

129

conditions of the experiment; therefore, it was not surprising that the actll-
ORF4::xyIE fusion was expressed in these strains.

In contrast, no prior information was available on the remaining bld
mutants tested in this study (bids B, G, I, and A95) to enable prediction of the
results. Expression from the actll-ORF4 fusion was not signiﬁcantly different
in these strains when compared to the parental type. Again, suspicions
about the accuracy of the xylE fusions must be considered; however, the only
available data at this time suggests that the effects of the bld loci on Act
synthesis takes place downstream from actll-ORF4 transcription, perhaps at
translation or activity of the ActIl-ORF4 protein or biosynthetic gene
expression. The bldB gene has been predicted to encode a regulatory
protein; therefore, if it directly targets antibiotic synthesis, it is likely to be at
the level of biosynthetic gene transcription. However, because the bldB
mutation has pleiotropic effects including loss of sporulation, it is likely to
target higher order regulators. Mutations in bldB and in bldH are known to
affect ADP-ribosylation of a number of proteins (197) and their effects on
antibiotic synthesis might be moderated through this mechanism. The bldG
gene is predicted to encode an anti-sigma or anti-anti-sigma factor (23); the
pleiotropic effects caused by the bldG mutation leads to the prediction that it
too is likely to target higher order regulators instead of antibiotic biosynthetic
genes directly.

Additional insights into role of absA. What is the role in S. coelicolor

of the signal transduction system encoded by absA? Does absA regulate the

130

timing (onset and/or shutdown) of antibiotic synthesis, or does it modulate
quantities of antibiotics produced, or both? As discussed in Chapter 2, S1
nuclease protection experiments suggest that the timing of transcriptional
onset of neither actll-ORF4 and redD is greatly affected by abs mutations.
Additional data bearing on this question was presented in Chapter 3, in which
actll-ORF4 transcription was monitored via xylE reporter gene fusions. The
most relevant of these are comparisons of early timepoints in Figures 10B
and 10D. Although one cannot be certain that the onset of transcription has
been captured in these experiments (particularly that shown in Figure 10B),
three of the four time courses appear to Show the beginning of major
transcription in abs strains within the same 12 hour period as in the parental
strain. The exception is the absB C120 strain in Figure 103, in which major
transcription may be delayed.

Brian at al. (26) presented evidence from reporter gene fusion studies for
a 6-12 hour shift in expression of the actl biosynthetic gene in an absA
mutant background (in this case, a disruption of absA1 resulting in
precocious hyperproduction of Act and an apparent shift to an earlier onset of
actl transcription). Recently, Ryding and Champness (187) studied
transcription of cda (53) genes as well as of absA itself using both S1
nuclease protection and RT-PCR. It was found that major transcription of
absA does not occur until after transcription from cda biosynthetic gene
promoters; thus, absA is apparently uninvolved in the timing of onset of cda

biosynthesis unless this involvement occurs in late-expressing cda genes

131

that were not assayed. Mutations in absA do appear to affect both the
quantities of transcript from cda biosynthetic gene transcripts as well as the
shutdown of transcription at later timepoints. Transcription from redD and
absA were also compared in these studies and found to reach signiﬁcant
levels within the same 6-hour period. Although a more precise study is
needed in the case of redD, these results tend to support the idea that absA
functions to modulate quantity, and possibly Shutdown of synthesis, of both
CDA and Red. Although there is ﬁrm evidence that both Act and CDA
production are detectable earlier in absA Pha mutants (10, 26), this may be
due to stronger (not earlier) expression, resulting in earlier detection of the
antibiotics. The preponderance of evidence at this time, then, suggests that
the abs genes do not signiﬁcantly affect the timing of transcriptional onset
from act, red, and cda genes.

If absA does function to modulate quantities of antibiotics produced, its
ultimate purpose may be to balance the ﬂow of energy or building blocks
between antibiotic production and sporulation. The formation of spores is
presumably more essential than antibiotic production to the survival of the
organism; therefore, absA may repress antibiotic synthesis when energy or
nutrients are in short supply. The behavior of Pha mutants, which
overexpress antibiotics and are somewhat deﬁcient in sporulation, supports

this hypothesis.

132

APPENDIX

Reproductions of Figures 1A, 2, 3A, 4A, and 5.

Figure 1A.
Time (h) 39 48 54 SM (n!)

actl l-ORF4 D

glk (2+3) D

glk (3) D

 

Figure 2

11501 C542 C [20 C430
Time(h) 394854 394854 394854 394854 SM(nt)
Masai,
ORF4

 
  
  
  
 

404

actll-ORF4.
307

glk (2+3) D
242
238
2”

81k (3) D
20I

 

 

 

 

 

J1 501 0542 0120 c430
Figure3A. _
Time(h) 39 48 54 65 39 48 5465 3948 54 65 48 54 5M ("I
9M3) >- -. _- . 2‘7
O 201
acw"m:1}w - ‘ o 190
g 180
r -.- 1O
. 160
Figure 4 A.
11501 I C542 0120 I
SM(nt) 39 54 60 39 48 54 60 39 48 S4 60 Time(h)
404
dredD
307
Cgllt(2+3)
242
238
2” ultra)

 

134

 

Figure 5

135

10.

11.

REFERENCES

. Aceti, D. J. and W. C. Champness. 1998. Transcriptional regulation of

Streptomyces coelicolor pathway-speciﬁc antibiotic regulators by the absA
and absB loci. J.Bacteriol. 180:3100-3106.

Adamidis, T. 1994. Michigan State University, East Lansing, MI. Doctoral
Dissertation.

Adamidis, T. and W. Champness. 1992. Genetic analysis of absB, a
Streptomyces coelicolor locus involved in global antibiotic regulation.
J.Bacteriol. 174:4622-4628.

Adamidis, T. and W. Champness. Personal Communication. 2001.

Adamidis, T., P. Riggle, and W. Champness. 1990. Mutations in a new
Streptomyces coelicolor locus which globally block antibiotic biosynthesis
but not sporulation. J.Bacteriol. 172:2962-2969.

Aigle, B., A. Wietzorrek, E. Takano, and M. J. Bibb. 2000. A single amino
acid substitution in region 1.2 of the principal sigma factor of Streptomyces
coelicolor A3(2) results in pleiotropic loss of antibiotic production.
Mol.Microbiol. 37:995-1004.

Al-Tai, A., 8. Kim, S. B. Kim, G. P. Manfio, and M. Goodfellow. 1999.
Streptomyces malaysiensis sp. nov., a new streptomycete species with
rugose, ornamented spores. lnt.J.Syst.Bacteriol. 49 Pt 4:1395-1402.

Anderson, T. 2000. Michigan State University, East Lansing, MI. Doctoral
Dissertation.

Anderson, T., P. Brian, P. Riggle, R. Kong, and W. Champness. 1999.
Genetic suppression analysis of non-antibiotic-producing mutants of the
Streptomyces coelicolor absA locus. Microbiology 145:2343-2353.

Anderson, T. 8., P. Brian, and W. C. Champness. 2001. Genetic and
transcriptional analysis of absA, an antibiotic gene cluster-linked two-
component system that regulates multiple antibiotics in Streptomyces
coelicolor. Mol.Microbiol. 39:553-566.

Angell, 8., C. G. Lewis, M. J. Buttner, and M. J. Bibb. 1994. Glucose
repression in Streptomyces coelicolor A3(2): a likely regulatory role for
glucose kinase. Mol.Gen.Genet. 244:135-143.

136

12.

13.

14.

15.

16.

17.

18.
19.

20.

21.

22.

23.

Arias, P., M. A. Femandez-Moreno, and F. Malpartida. 1999.
Characterization of the pathway-speciﬁc positive transcriptional regulator for
actinorhodin biosynthesis in Streptomyces coelicolor A3(2) as a DNA-
binding protein. J.Bacteriol. 181:6958-6968.

Bainton, N. J., P. Stead, S. R. Chhabra, B. W. Bycroft, G. P. Salmond, G.
S. Stewart, and P. Williams. 1992. N-(3-oxohexanoyl)-L-homoserine

lactone regulates carbapenem antibiotic production in Erwinia carotovora.
Biochem.J. 288 ( Pt 3):997-1004.

Baltz, R. H. 1978. Genetic recombination in Streptomyces fradiae by
protoplast fusion and cell regeneration. J.Gen.Microbiol. 107:93—102.

Bao, W., E. Wendt-Pienkowski, and C. R. Hutchinson. 1998.
Reconstitution of the iterative type II polyketide synthase for tetracenomycin
F2 biosynthesis. Biochemistry 37:8132-8138.

Beringer, J. E. 1974. R factor transfer in Rhizobium Ieguminosarum.
J.Gen.Microbiol. 84:188-198.

Bibb, M. 1996. 1995 Colworth Prize Lecture. The regulation of antibiotic
production in Streptomyces coelicolor A3(2). Microbiology 142 ( Pt 6):1335-
1344.

Bibb, M. Personal Communication. 2001.

Bibb, M. J. and S. N. Cohen. 1982. Gene expression in Streptomyces:
construction and application of promotor-probe plasmid vectors in
Streptomyces Iividans. Mol.Gen.Genet. 187:265-277.

Bibb, M. J., P. R. Findlay, and M. W. Johnson. 1984. The relationship
between base composition and codon usage in bacterial genes and its use

for the simple and reliable identiﬁcation of protein- coding sequences. Gene
30:157-166.

Bibb, M. J., R. F. Freeman, and D. A. Hopwood. 1977. Physical and
genetical characterisation of a second sex factor, SCP2, for Streptomyces
coelicolor A3(2). Mol.Gen.Genet. 154: 1 55-1 66.

Bibb, M. J., J. M. Ward, and D. A. Hopwood. 1978. Transformation of
plasmid DNA into Streptomyces at high frequency. Nature 274:398-400.

Bignell, D. R., J. L. Warawa, J. L. Strap, K. F. Chater, and B. K. Leskiw.
2000. Study of the bldG locus suggests that an anti-anti-sigma factor and an
anti-sigma factor may be involved in Streptomyces coelicolor antibiotic
production and sporulation. Microbiology 146:2161-2173.

137

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Bradford, M. M. 1976. A rapid and sensitive method for quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal.Biochem. 72:248-254.

Brasch, M. A., G. S. Pettis, S. C. Lee, and S. N. Cohen. 1993. Localization
and nucleotide sequences of genes mediating site-speciﬁc recombination of
the SLP1 element in Streptomyces Iividans. J.Bacteriol. 175:3067-3074.

Brian, P., P. J. Riggle, R. A. Santos, and W. C. Champness. 1996. Global
negative regulation of Streptomyces coelicolor antibiotic synthesis mediated
by an absA-encoded putative signal transduction system. J.Bacteriol.
178:3221-3231.

Brian, P. W. 1957. The ecological signiﬁcance of antibiotic production, p.
168-190. In Seventh Symposium of the Society for General Microbiology
Held at Royal Institution, London April 1957. The University Press,
Cambridge.

Brockmann, H., A.Zeeck, K.Van Der Merwe, and W.Muller. 1966. Die
Konstitution des Actinorhodins. Justus Liebigs Annalen der Chemie
698:209-229.

Brown, G. Unpublished Work. 1993.

Brown, K. L., S. Wood, and M. J. Buttner. 1992. Isolation and
characterization of the major vegetative RNA polymerase of Streptomyces

coelicolor A3(2); renaturation of a sigma subunit using GroEL. Mol.Microbiol.
6:1 133-1 139.

Bruton, C. J., E. P. Guthrie, and K. F. Chater. 1991. Phage vectors that
allow monitoring of transcription of secondary metabolism genes in
Streptomyces. Biotechnology (N.Y.) 9:652-656.

Burke, J., D. Schneider, and J. Westpheling. 2001. Generalized
transduction in Streptomyces coelicolor. Proc.Natl.Acad.Sci.U.S.A 9826289-
6294.

Butler, A. R., S. A. Flint, and E. Cundliffe. 2001. Feedback control of
polyketide metabolism during tylosin production. Microbiology 147:795—806.

Buttner, M. J., K. F. Chater, and M. J. Bibb. 1990. Cloning, disruption, and
transcriptional analysis of three RNA polymerase sigma factor genes of
Streptomyces coelicolor A3(2). J.Bacteriol. 172:3367-3378.

138

35.

36.

37.

38.

39.

40.

41.

42.

43.

45.

46.

Buttner, M. J. and C. G. Lewis . 1992. Construction and characterization of
Streptomyces coelicolor A3(2) mutants that are multiply deﬁcient in the
nonessential hrd-encoded RNA polymerase sigma factors. J.Bacteriol.
174:5165-5167.

Buttner, M. J., A. M. Smith, and M. J. Bibb. 1988. At least three different
RNA polymerase holoenzymes direct transcription of the agarase gene
(dagA) of Streptomyces coelicolor A3(2). Cell 52:599-607.

Bystrykh, L. V., M. A. Femandez-Moreno, J. K. Herrema, F. Malpartida,
D. A. Hopwood, and L. Dijkhuizen. 1996. Production of actinorhodin-
related "blue pigments" by Streptomyces coelicolor A3(2). J.Bacteriol.
178:2238-2244.

Cai, Y. and C. P. Walk. 1997. Nitrogen deprivation of Anabaena sp. strain
PCC 7120 elicits rapid activation of a gene cluster that is essential for
uptake and utilization of nitrate. J.Bacteriol. 179:258-266.

Cane, D. E., C. T. Walsh, and C. Khosla. 1998. Harnessing the
biosynthetic code: combinations, permutations, and mutations. Science
282263-68.

Chakraburtty, R. and M. Bibb. 1997. The ppGpp synthetase gene (relA) of
Streptomyces coelicolor A3(2) plays a conditional role in antibiotic
production and morphological differentiation. J.Bacteriol. 179:5854-5861.

Chakraburtty, R., J. White, E. Takano, and M. Bibb. 1996. Cloning,
characterization and disruption of a (p)ppGpp synthetase gene (relA) of
Streptomyces coelicolor A3(2). Mol.Microbiol. 19:357-368.

Champness, W., P. Riggle, T. Adamidis, and P. Vandervere. 1992.
identiﬁcation of Streptomyces coelicolor genes involved in regulation of
antibiotic synthesis. Gene 115:55-60.

Champness, W. C. 1988. New loci required for Streptomyces coelicolor
morphological and physiological differentiation. J.Bacteriol. 170:1 168-1 174.

Champness, W. C. Personal Communication. 2001.

Chang, H. M, M. Y. Chen, Y. T. Shieh, M. J. Bibb, and C. W. Chen. 1996.
The cutRS signal transduction system of Streptomyces Iividans represses
the biosynthesis of the polyketide antibiotic actinorhodin. Mol.Microbiol.
21:1075-1085.

Chater, K. F. 1998. 1997 Fred Grifﬁth Review Lecture: Taking a genetic
scalpel to the Streptomyces colony. Microbiology 144:1465-1478.

139

47.

48.

49.

50.

51.

52.

53.

55.

56.

57.

Chater, K. F. and M. J. Bibb. 1997. Regulation of bacterial antibiotic
production, p. 57-105. In H. Kleinkauf and H. von Dohren (eds.), Products of
secondary metabolism. VCH, Weinheim.

Chater, K. F. and C. J. Bruton . 1985. Resistance, regulatory and
production genes for the antibiotic methylenomycin are clustered. EMBO J.
4:1893-1897.

Chater, K. F., C. J. Bruton, A. A. King, and J. E. Suarez. 1982. The
expression of Streptomyces and Escherichia coli drug-resistance
determinants cloned into the Streptomyces phage phi C31. Gene 19:21-32.

Chater, K. F., C. J. Bruton, K. A. Plaskitt, M. J. Buttner, C. Mendez, and
J. D. Helmann. 1989. The developmental fate of S. coelicolor hyphae
depends upon a gene product homologous with the motility sigma factor of
B. subtilis. Cell 59:133-143.

Chater, K. F. and M. J. Merrick. 1979. Streptomycetes, p. 93-114. In J. H.
Parish (ed.), Developmental Biology of Prokaryotes. Univ. of California
Press, Berkeley and Los Angeles.

Cholodny, N. 1930. Uber eine neue methode zur untersuchung der
bodenmikroﬂora. Archiv fur Mikrobiologie 12620-652.

Chong, P. P., S. M. Podmore, H. M. Kieser, M. Redenbach, K. Turgay, M.
Marahiel, D. A. Hopwood, and C. P. Smith. 1998. Physical identiﬁcation of
a chromosomal locus encoding biosynthetic genes for the lipopeptide
calcium-dependent antibiotic (CDA) of Streptomyces coelicolor A3(2).
Microbiology 144 ( Pt 1):193-199.

Coco, E. A., K. E. Narva, and J. S. Feitelson. 1991. New classes of
Streptomyces coelicolor A3(2) mutants blocked in undecylprodigiosin (Red)
biosynthesis. Mol.Gen.Genet. 227:28-32.

Craster, H. L., C. A. Potter, and S. Baumberg. 1999. End-product control
of expression of branched-chain amino acid biosynthesis genes in
Streptomyces coelicolor A3(2): paradoxical relationships between DNA
sequence and regulatory phenotype. Microbiology 145 ( Pt 9):2375-2384.

Crecy-Lagard, V., P. Servant-Moisson, J. Viala, C. Grandvalet, and P.
Mazodier. 1999. Alteration of the synthesis of the Clp ATP-dependent
protease affects morphological and physiological differentiation in
Streptomyces. Mol.Microbiol. 32:505-517.

Cundliffe, E. 2000. Antibiotic synthesis: Some thoughts on "Why?" and
"How?", p. 409-417. In R. A. Garrett, S. R. Douthwaite, A. Liljas, A. T.

140

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

Matheson, P. B. Moore, and H. F. Noller (eds.), The Ribosome: Structure,
Function, Antibiotics, and Cellular Interactions. ASM Press, Washington,
DC.

Dandanell, G., P. Valentin-Hansen, J. E. Larsen, and K. Hammer. 1987.
Long-range cooperativity between gene regulatory sequences in a
prokaryote. Nature 325:823-826.

Denis, F. and R. Brzezinski. 1992. A versatile shuttle cosmid vector for use
in Escherichia coli and actinomycetes. Gene 111:115-118.

Doull, J. L. and L. C. Vining. 1990. Nutritional control of actinorhodin
production by Streptomyces coelicolor A3(2): suppressive effects of nitrogen
and phosphate. Appl.Microbiol.Biotechnol. 32:449-454.

Elliot, M., F. Damji, R. Passantino, K. Chater, and B. Leskiw. 1998. The
bldD gene of Streptomyces coelicolor A3(2): a regulatory gene involved in
morphogenesis and antibiotic production. J.Bacteriol. 180:1549-1555.

Elliot, M. A., M. J. Bibb, M. J. Buttner, and B. K. Leskiw. 2001. BIdD is a
direct regulator of key developmental genes in Streptomyces coelicolor
A3(2). Mol.Microbiol. 40:257-269.

Embley, T. M. and E. Stackebrandt. 1994. The molecular phylogeny and
systematics of the actinomycetes. Ann.Rev.Microbiol. 48:257-289.

Feitelson. J. S., F. Malpartida. and D. A. Hopwood. 1985. Genetic and
biochemical characterization of the red gene cluster of Streptomyces
coelicolor A3(2). J.Gen.Microbiol. 131:2431-2441.

Femandez-Moreno, M. A., J. L. Caballero, D. A. Hopwood, and F.
Malpartida. 1991. The act cluster contains regulatory and antibiotic export
genes, direct targets for translational control by the bldA tRNA gene of
Streptomyces. Cell 66:769-780.

Femandez-Moreno, M. A., A. J. Martin-Triana, E. Martinez, J. Niemi, H.
M. Kieser, D. A. Hopwood, and F. Malpartida. 1992. abaA, a new
pleiotropic regulatory locus for antibiotic production in Streptomyces
coelicolor. J.Bacteriol. 174:2958-2967.

Femandez-Moreno, M. A., E. Martinez, L. Boto, D. A. Hopwood, and F.
Malpartida. 1992. Nucleotide sequence and deduced functions of a set of
cotranscribed genes of Streptomyces coelicolor A3(2) including the
polyketide synthase for the antibiotic actinorhodin. J.Biol.Chem. 267:19278-
19290.

141

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

Flett, F. and J. Cullum. 1987. DNA deletions in spontaneous
chloramphenicol-sensitive mutants of Streptomyces coelicolor A 3(2) and
Streptomyces Iividans 66. Mol.Gen.Genet. 207:499-502.

Floriano, B. and M. Bibb. 1996. ast is a pleiotropic but conditionally
required regulatory gene for antibiotic production in Streptomyces coelicolor
A3(2). Mol.Microbiol. 21:385-396.

Forsberg, A. J., G. D. Pavitt, and C. F. Higgins. 1994. Use of
transcriptional fusions to monitor gene expression: a cautionary tale.
J.Bacteriol. 176:2128-2132.

Forsman, M. and B. Jaurin. 1987. Chromogenic identiﬁcation of promoters
in Streptomyces Iividans by using an ampC beta-lactamase promoter-probe
vector. Mol.Gen.Genet. 210:23-32.

Fravel, D. R. 1988. Role of antibiotics in the biocontrol of plant diseases.
Ann.Rev.Phypathol. 25:75-91.

Fujii, T., H. C. Gramajo, E. Takano, and M. J. Bibb. 1996. redD and actll-
ORF4, pathway-speciﬁc regulatory genes for antibiotic production in
Streptomyces coelicolor A3(2), are transcribed in vitro by an RNA
polymerase holoenzyme containing sigma hrdD. J.Bacteriol. 178:3402-3405.

Furuya, K. and C. R. Hutchinson. 1996. The InrN protein of Streptomyces
peucetius, a pseudo-response regulator, is a DNA-binding protein involved
in the regulation of daunorubicin biosynthesis. J.Bacteriol. 178:6310—6318.

Gehring, A. M., J. R. Nodwell, S. M. Beverley, and R. Losick. 2000.
Genomewide insertional mutagenesis in Streptomyces coelicolor reveals
additional genes involved in morphological differentiation.
Proc.Natl.Acad.Sci.U.S.A 97:9642-9647.

Gerber, N. N., A.G.Mclnnes, D.G.Smith, J.A.Walter, and J.L.C.Wright.
1978. Biosynthesis of prodiginines. 13C resonance assignments and
enrichments patterns in nonyI-, cyclononyl-, methylcyclodecyI-, and
butylcycloheptyl-prodiginine produced by actinomycete cultures
supplemented with 13C-Iabeled and 15N-labeled nitrate. Can.J.Chem.
56:1155-1163.

Gladek, A. and J.Zakrzewska. 1984. Genome size of Streptomyces. FEMS
Microbiol.Lett 24:72-76.

Gonzalez-Ceron, G., P. Licona, and Servin-Gonzalez.L. 2001. Modiﬁed

xylE and xyITE reporter genes for use in Streptomyces: analysis of the effect
of xle. FEMS Microbiol.Lett 196:229-234.

142

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

Gottleib, D. and P. Siminoff. 1952. The production and role of antibiotics in
the soil. ll. Choromycetin. Phytopathology 42:91-98.

Gottlieb, D. and E.B.Shirling. 1967. Cooperative description of type
cultures of Streptomyces; I. The International Streptomyces Project.
Int.J.Syst.Bacteriol. 17:315-322.

Gramajo, H. C., E. Takano, and M. J. Bibb. 1993. Stationary-phase
production of the antibiotic actinorhodin in Streptomyces coelicolor A3(2) is
transcriptionally regulated. Mol.Microbiol. 72837-845.

Granozzi, C., R.Billetta, R.Passantino, M.Sollazzo, and A.M.Puglia. 1990.
A breakdown in macromolecular synthesis preceding differentiation in
Streptomyces coelicolor A3(2). J.Gen.Microbiol. 136:713-716.

Gregory, K. F., O. N. Allen, A. J. Riker, and W. H. Perterson. 1952.
Antibiotics as agents for the control of certain damping-off fungi.
Amer.J.Botany 39:405-418.

Guijarro, J., R. Santamaria, A. Schauer, and R. Losick. 1988. Promoter
determining the timing and spatial localization of transcription of a cloned
Streptomyces coelicolor gene encoding a spore- associated polypeptide.

J.Bacteriol. 170:1895-1901.

Guthrie, E. P., C. S. Flaxman, J. White, D. A. Hodgson, M. J. Bibb, and
K. F. Chater. 1998. A response-regulator-like activator of antibiotic
synthesis from Streptomyces coelicolor A3(2) with an amino-terminal domain
that lacks a phosphorylation pocket [published erratum appears in
Microbiology 1998 Jul;144(Pt 7):2007]. Microbiology 144 ( Pt 3):727-738.

Han, L., A. Khetan, W. S. Hu, and D. H. Sherman. 1999. Time-lapsed
confocal microscopy reveals temporal and spatial expression of the lysine

epsilon-aminotransferase gene in Streptomyces clavuligerus. Mol.Microbiol.
34:878-886.

Hara, 0., S.Horinouchi, T.Uozumi, and T.Beppu. 1983. Genetic analysis
of A-factor synthesis in Streptomyces coelicolor A3(2) and Streptomyces
griseus. J.Gen.Microbiol. 129:2939—2944.

Hara, 0., T. Murakami, S. lmai, H. Anzai, R. ltoh, Y. Kumada, E. Takano,
E. Satoh, A. Satoh, and K. Nagaoka. 1991. The bialaphos biosynthetic
genes of Streptomyces viridochromogenes: cloning, heterospeciﬁc

expression, and comparison with the genes of Streptomyces hygroscopicus.
J.Gen.Microbiol. 137 ( Pt 2):351-359.

143

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

Hassett, D. J., U. A. Ochsner, S. L. Groce, K. Parvatiyar, J. F. Ma, and J.
Lipscomb. 2000. Hydrogen peroxide sensitivity of catechol-2.3-
dioxygenase: a cautionary note on use of xylE reporter fusions under
aerobic conditions. AppI.Environ.Microbiol. 6624123.

Hayes, A., G. Hobbs, C. P. Smith, S. G. Oliver, and P. R. Butler. 1997.
Environmental signals triggering methylenomycin production by
Streptomyces coelicolor A3(2). J.Bacteriol. 179:5511-5515.

Hesketh, A., J. Sun, and M. Bibb. 2001. Induction of ppGpp synthesis in
Streptomyces coelicolor A3(2) grown under conditions of nutritional
sufﬁciency elicits actll-ORF4 transcription and actinorhodin biosynthesis.
Mol.Microbiol. 39:136-144.

Hobbs, G., C.M.Frazer, D.C.J.Gardner, F.FIett, and S.G.Oliver. 1990.
Pigmented antibiotic production by Streptomyces coelicolor A3(2): kinetics
and the inﬂuence of nutrients. J.Gen.Microbiol. 136:2291-2296.

Hobbs, G., A. l. Obanye, J. Petty, J. C. Mason, E. Barratt, D. C. Gardner,
F. Flett, C. P. Smith, P. Broda, and S. G. Oliver. 1992. An integrated
approach to studying regulation of production of the antibiotic
methylenomycin by Streptomyces coelicolor A3(2). J.Bacteriol. 17421487-
1494.

Hong, S. K., M. Kito, T. Beppu, and S. Horinouchi. 1991. Phosphorylation
of the Ast product, a global regulatory protein for secondary-metabolite
formation in Streptomyces coelicolor A3(2). J.Bacteriol. 173:2311-2318.

Hopwood, D. A. 1967. Genetic analysis and genome structure in
Streptomyces coelicolor. Bacteriol.Rev. 31:373-403.

Hopwood, D. A. 1999. Forty years of genetics with Streptomyces: from in
vivo through in vitro to in silico. Microbiology 145 ( Pt 9):2183-2202.

Hopwood, D. A., M. J. Bibb, K. F. Chater, T. Kieser, C. J. Bruton, H. M.
Kieser, D. J. Lydiate, C. P. Smith, J. M. Ward, and H. Schrempf. 1985.
Genetic Manipulation of Streptomyces: A Laboratory Manual. The John
Innes Foundation, Norwich, United Kingdom.

Hopwood, D. A., K. F. Chater, and M. J. Bibb. 1995. Genetics of antibiotic
production in Streptomyces coelicolor A3(2), a model streptomycete.
Biotechnology 28265-102.

Hopwood, D. A., D.J.Lydiate, F.Malpartida, and H.M.Wright. 1985.
Conjugative sex plasmids of Streptomyces, p. 615-633. In D. R. Helinski,

144

 

100.

101.

102.

103.

104.

105.

106.

107.

108.

S.N.Cohen, D.B.Clewell, D.A.Jackson, and A.Hollaender (eds), Plasmids in
Bacteria. Plenum Press, New York.

Hopwood, D. A. and H. M. Wright. 1983. CDA is a new chromosomally-
determined antibiotic from Streptomyces coelicolor A3(2). J.Gen.Microbiol.
129 ( Pt 12):3575-3579.

Horinouchi, S. and T. Beppu. 1992. Regulation of secondary metabolism
and cell differentiation in Streptomyces: A-factor as a microbial hormone and
the Ast protein as a component of a two-component regulatory system.
Gene 115:167-172.

Horinouchi, S. and T. Beppu. 1994. A-factor as a microbial hormone that
controls cellular differentiation and secondary metabolism in Streptomyces
griseus. Mol.Microbiol. 12:859—864.

Horinouchi, 8., O. Hara, and T. Beppu. 1983. Cloning of a pleiotropic gene
that positively controls biosynthesis of A-factor, actinorhodin, and
prodigiosin in Streptomyces coelicolor A3(2) and Streptomyces Iividans.
J.Bacteriol. 155:1238-1248.

Horinouchi, 8., M. Kito, M. Nishiyama, K. Furuya, 8. K. Hong, K. Miyake,
and T. Beppu. 1990. Primary structure of Ast, a global regulatory protein
for secondary metabolite formation in Streptomyces coelicolor A3(2). Gene
95:49-56.

Horinouchi, 8., F. Malpartida. D. A. Hopwood, and T. Beppu. 1989. afsB
stimulates transcription of the actinorhodin biosynthetic pathway in

Streptomyces coelicolor A3(2) and Streptomyces Iividans. Mol.Gen.Genet.
2152355—357.

Hu, D. 8., D. W. Hood, R. Heidstra, and D. A. Hodgson. 1999. The
expression of the trpD, trpC and trpBA genes of Streptomyces coelicolor
A3(2) is regulated by growth rate and growth phase but not by feedback
repression. Mol.Microbiol. 32:869-880.

Huisman, G. W. and R. Kolter. 1994. Sensing starvation: a homoserine
lactoneudependent signaling pathway in Escherichia coli. Science 2652537-
539.

lngram, C., M. Brawner, P. Youngman, and J. Westpheling. 1989. xylE

functions as an efﬁcient reporter gene in Streptomyces spp.2 use for the
study of galP1, a catabolite-controlled promoter. J.Bacteriol. 17126617-6624.

145

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

Ishizuka, H., S. Horinouchi, H. M. Kieser, D. A. Hopwood, and T. Beppu.
1992. A putative two-component regulatory system involved in secondary
metabolism in Streptomyces spp. J.Bacteriol. 17427585-7594.

Jacobsen, J. R., C. R. Hutchinson, D. E. Cane, and C. Khosla. 1997.
Precursor-directed biosynthesis of erythromycin analogs by an engineered
polyketide synthase [see comments]. Science 277:367-369.

Janssen, G. R. and M. J. Bibb . 1990. Tandem promoters, tsrp1 and tsrp2,
direct transcription of the thiostrepton resistance gene (tsr) of Streptomyces
azureus: transcriptional initiation from tsrp2 occurs after deletion of the -35
region. Mol.Gen.Genet. 221:339-346.

Janssen, G. R. and M. J. Bibb . 1993. Derivatives of pUC18 that have Bglll
sites ﬂanking a modiﬁed multiple cloning site and that retain the ability to
identify recombinant clones by visual screening of Escherichia coli colonies.

Gene 124:133—134.

Janssen, G. R., J. M. Ward, and M. J. Bibb. 1989. Unusual transcriptional
and translational features of the aminoglycoside phosphotransferase gene
(aph) from Streptomyces fradiae. Genes Dev. 32415-429.

Kang, J. G., M. Y. Hahn, A. lshihama, and J. H. Roe. 1997. Identiﬁcation of
sigma factors for growth phase-related promoter selectivity of RNA

polymerases from Streptomyces coelicolor A3(2). Nucleic Acids Res.
25:2566-2573.

Kang, J. G., M. 8. Paget, Y. J. Seok, M. Y. Hahn, J. B. Bae, J. 8. Hahn, C.
Kleanthous, M. J. Buttner, and J. H. Roe. 1999. RsrA, an anti-sigma factor
regulated by redox change. EMBO J. 18:4292-4298.

Kang, 8. G., W. Jin, M. Bibb, and K. J. Lee. 1998. Actinorhodin and
undecylprodigiosin production in wild-type and relA mutant strains of

Streptomyces coelicolor A3(2) grown in continuous culture. FEMS
Microbiol.Lett. 168:221-226.

Kawabuchi, M., Y.Hara, T.Nihira, and Y.Yamada. 1997. Production of
butyrolactone autoregulators by Streptomyces coelicolor A3(2). FEMS
Microbiol.Lett. 157281-85.

Kieser, H. M., T. Kieser, and D. A. Hopwood. 1992. A combined genetic

and physical map of the Streptomyces coelicolor A3(2) chromosome.
J.Bacteriol. 174:5496-5507.

146

 

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

Kieser, T., M. J. Bibb, M. J. Buttner, K. F. Chater, and D. A. Hopwood.
2000. Practical Streptomyces Genetics. John Innes Foundation, NonNich, U.
K.

Kieser, T., D. A. Hopwood, H. M. Wright, and C. J. Thompson. 1982.
le101, a multi-copy broad host-range Streptomyces plasmid: functional

analysis and development of DNA cloning vectors. Mol.Gen.Genet. 1852223-
228.

Killham, K. and M. K. Firestone. 1984. Proline transport increases growth
efﬁciency in salt-stressed Streptomyces griseus. AppI.Environ.Microbiol.
48:239-241.

Kinashi, H. 1994. Linear plasmids from actinomycetes. Actinomycetologia
8:87-96.

Kinashi, H. and M. Shimaji-Murayama. 1991. Physical characterization of
SCP1, a giant linear plasmid from Streptomyces coelicolor. J.Bacteriol.
173:1523-1529.

Kinashi, H., M. Shimaji-Murayama, and T. Hanafusa. 1992. Integration of
SCP1, a giant linear plasmid, into the Streptomyces coelicolor chromosome.
Gene 115235-41.

King, J. M. H., P. M. DiGrazia, B. Applegate, R. Burlage, J. Sanseverino,
F. Dunbar, F. Larimer, and G. 8. Sayler. 1990. Rapid, sensitive
bioluminescent reporter technology for naphthalene exposure and
biodegradation. Science 249:778-781.

Kirby, R. and D. A. Hopwood. 1977. Genetic determination of
methylenomycin synthesis by the SCP1 plasmid of Streptomyces coelicolor
A3(2). J.Gen.Microbiol. 98:239-252.

Kluepfel, D. 1991. Protein biosynthesis and secretion, In S.Baumberg et al.

(ed.), Genetics and Product Formation in Streptomyces. Plenum Press, New
York.

Kramer, P. J., R.J.X.Zawada, R.McDaniel, C.R.Hutchinson,
D.A.Hopwood, and C.Khosla. 1997. Rational design and engineered

biosynthesis of a novel 18-carbon aromatic polyketide. J.Am.Chem.Soc.
119:635-639.

Kuhstoss, 8., M. A. Richardson, and R. N. Rao. 1991. Plasmid cloning

vectors that integrate site-speciﬁcally in Steptomyces spp. Gene 97 2143-
146.

147

 

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

Kutzner, H. J. 1981. The Family Streptomycetaceae, p. 2031-2090. In
Mortimer P.8tarr et.a|. (ed.), The Prokaryotes 2 a handbook on habitats,
isolation, and identiﬁcation of bacteria. Springer-Verlag, Berlin.

Kyung, Y. 8., W.-S. Hu, and D. H. Sherman. 2001. Analysis of temporal
and spatial expression of the CcaR regulatory element in the cephamycin C
biosynthetic pathway using green ﬂuorescent protein. Molecular
Microbiology 40:530-541.

Lakey, J. H., E. J. Lea, B. A. Rudd, H. M. Wright, and D. A. Hopwood.
1983. A new channel-forming antibiotic from Streptomyces coelicolor A3(2)
which requires calcium for its activity. J.Gen.Microbiol. 129 ( Pt 12):3565—
3573.

Lawlor, E. J., H. A. Baylis, and K. F. Chater. 1987. Pleiotropic
morphological and antibiotic deﬁciencies result from mutations in a gene
encoding a tRNA-like product in Streptomyces coelicolor A3(2). Genes Dev.
121305-1310.

Leblond, P. and B. Decaris. 1994. New insights into the genetic instability
of streptomyces. FEMS Microbiol.Lett1232225-232.

Leblond, P., P. Demuyter, J. M. Simonet, and B. Decaris. 1990. Genetic
instability and hypervariability in Streptomyces ambofaciens: towards an

understanding of a mechanism of genome plasticity. Mol.Microbiol. 4:707-
714.

Leblond, P., M. Redenbach, and J. Cullum. 1993. Physical map of the
Streptomyces Iividans 66 genome and comparison with that of the related
strain Streptomyces coelicolor A3(2). J.Bacteriol. 175:3422-3429.

Leskiw, B. K., M. J. Bibb, and K. F. Chater. 1991. The use of a rare codon
speciﬁcally during development? Mol.Microbiol. 5:2861-2867.

Leskiw, B. K., E. J. Lawlor, J. M. Femandez-Abalos, and K. F. Chater.
1991. TTA codons in some genes prevent their expression in a class of
developmental, antibiotic-negative, Streptomyces mutants.
Proc.Natl.Acad.Sci.U.S.A 8822461 -2465.

Leskiw, B. K., R. Mah, E. J. Lawlor, and K. F. Chater. 1993. Accumulation
of bldA-speciﬁed tRNA is temporally regulated in Streptomyces coelicolor
A3(2). J.Bacteriol. 175:1995-2005.

Lindley, H. K., V. J. Deeble, U. Peschke, M. O'Neill, 8. Baumberg, and J.
Cove. 1995. Dependence on reporter gene of apparent activity in gene

148

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

fusions of a Streptomyces griseus streptomycin biosynthesis promoter.
Can.J.Microbiol. 41:407-417.

Lloyd, A. B. 1969. Behavior of Streptomycetes in soil. J.Gen.Microbiol.
562165-170.

Lonetto, M. A., K. L. Brown, K. E. Rudd, and M. J. Buttner. 1994. Analysis
of the Streptomyces coelicolor sigE gene reveals the existence of a
subfamily of eubacterial RNA polymerase sigma factors involved in the
regulation of extracytoplasmic functions. Proc.Natl.Acad.Sci.U.S.A 9127573-
7577.

Lydiate, D. J., H. Ikeda, and D. A. Hopwood. 1986. A 2.6 kb DNA
sequence of Streptomyces coelicolor A3(2) which functions as a
transposable element. Mol.Gen.Genet. 203279-88.

Makins, J. F. and G. Holt. 1981. Liposome-mediated transformation of
streptomycetes by chromosomal DNA. Nature 293:671-673.

Malpartida, F. and D. A. Hopwood. 1984. Molecular cloning of the whole
biosynthetic pathway of a Streptomyces antibiotic and its expression in a
heterologous host. Nature 309:462-464.

Malpartida, F. and D. A. Hopwood. 1986. Physical and genetic
characterisation of the gene cluster for the antibiotic actinorhodin in
Streptomyces coelicolor A3(2). Mol.Gen.Genet. 205:66-73.

Malpartida, F., J. Niemi, R. Navarrete, and D. A. Hopwood. 1990. Cloning
and expression in a heterologous host of the complete set of genes for

biosynthesis of the Streptomyces coelicolor antibiotic undecylprodigiosin.
Gene 93:91-99.

Martinez-Costa, O. H., P. Arias, N. M. Romero, V. Parro, R. P. Mellado,
and F. Malpartida. 1996. A relA/spoT homologous gene from Streptomyces
coelicolor A3(2) controls antibiotic biosynthetic genes. J.Biol.Chem.
271:10627-10634.

Matsumoto, A., 8. K. Hong, H. Ishizuka, 8. Horinouchi, and T. Beppu.
1994. Phosphorylation of the Ast protein involved in secondary metabolism
in Streptomyces species by a eukaryotic-type protein kinase. Gene 146247-
56.

Matsumoto, A., H. Ishizuka, T. Beppu, and S. Horinouchi. 1995.
Involvement of a small ORF downstream of the ast gene in the regulation
of secondary metabolism in Streptomyces coelicolor A3(2).
Actinomycetologica 9237-43.

149

 

151.

152.

153.

154.

155.

156.

157.

1,58.

159.

160.

161.

162.

Mazodier, P., R. Petter, and C. Thompson. 1989. lntergeneric conjugation
between Escherichia coli and Streptomyces species. J.Bacteriol. 17123583-
3585.

McCarthy, A. J. and 8. T. Williams. 1992. Actinomycetes as agents of
biodegradation in the environment--a review. Gene 115:189-192.

McDaniel, R., 8. Ebert-Khosla, D. A. Hopwood, and C. Khosla. 1993.
Engineered biosynthesis of novel polyketides. Science 26221546-1550.

McDaniel, R., C. M. Kao, 8. J. Hwang, and C. Khosla. 1997. Engineered
intermodular and intramodular polyketide synthase fusions. Chem.Biol.
42667-674.

Meade, H. M., 8. R. Long, G. B. Ruvkun, S. E. Brown, and F. M. Ausubel.
1982. Physical and genetic characterizaton of symbiotic and auxotrophic
mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis.
J.Bacteriol. 149:114-122.

Merrick, M. J. 1976. A morphological and genetic mapping study of bald
colony mutants of Streptomyces coelicolor. J.Gen.Microbiol. 96:299-315.

Meurer, G. and C. R. Hutchinson. 1999. Genes for the Biosynthesis of
Microbial Secondary Metabolites, p. 740-758. In A. L. Demain and J. E.
Davies (eds), Manual of Industrial Microbiology and Biotechnology. ASM
Press, Washington, DC.

Mogk, A. and R. S. W. Hayward . 1996. Integrative vectors for constructing
single-copy transcriptional fusions between Bacillus subtilis promotors and
various reporter genes encoding heat-stable enzymes. Gene 182233-36.

Molnar, l. and Y. Murooka. 1993. Helix-turn-helix DNA-binding motifs of
Streptomyces-~a cautionary note [letter]. Mol.Microbiol. 8:783-784.

Moran, M. A., L. T. Rutherford, and R. E. Hodson. 1995. Evidence for
indigenous Streptomyces populations in a marine environment determined
with a 168 rRNA probe. Appl.Environ.Microbiol. 61:3695—3700.

Narva, K. E. and J. 8. Feitelson. 1990. Nucleotide sequence and
transcriptional analysis of the redD locus of Streptomyces coelicolor A3(2).
J.Bacteriol. 172:326—333.

Neal, R. J. and K. F. Chater. 1987. Nucleotide sequence analysis reveals

similarities between proteins determining methylenomycin A resistance in
Streptomyces and tetracycline resistance in eubacteria. Gene 58:229-241.

150

163.

164.

165.

166.

167.
168.

169.

170.

171.

172.

Nemetz, C., R. Reichhuber, R. Schweizer, P. Hloch, and M. Watzele.
2001. Reliable quantiﬁcation of in vitro synthesized green ﬂuorescent
proteins: Comparison of ﬂuorescence activity and total protein levels.
Electrophoresis 22:966-969.

Neumann, T., W.Pipersberg, and J.Distler. 1996. Decision phase
regulation in Streptomyces griseus. Microbiology 142:1953-1963.

Nodwell, J. R. and R. Losick. 1998. Puriﬁcation of an extracellular
signaling molecule involved in production of aerial mycelium by
Streptomyces coelicolor. J.Bacteriol. 180: 1 334-1 337.

Nodwell, J. R., K. McGovern, and R. Losick. 1996. An oligopeptide
permease responsible for the import of an extracellular signal governing
aerial mycelium formation in Streptomyces coelicolor. Mol.Microbiol. 22:881-
893.

Ochi, K. 1990. A relaxed (rel) mutant of Streptomyces coelicolor A3(2) with
a missing ribosomal protein lacks the ability to accumulate ppGpp, A-factor
and prodigiosin. J.Gen.Microbiol. 136 ( Pt 12):2405-2412.

Ochi, K., D. Zhang, 8. Kawamoto, and A. Hesketh. 1997. Molecular and
functional analysis of the ribosomal L11 and 812 protein genes (rle and
rpsL) of Streptomyces coelicolor A3(2). Mol.Gen.Genet. 2562488-498.

Onaka, H., T. Nakagawa, and 8. Horinouchi. 1998. Involvement of two A-
factor receptor homologues in Streptomyces coelicolor A3(2) in the
regulation of secondary metabolism and morphogenesis. Mol.Microbiol.
28:743-753.

Otten, 8. L., C. Olano, and C. R. Hutchinson. 2000. The dnrO gene
encodes a DNA-binding protein that regulates daunorubicin production in
Streptomyces peucetius by controlling expression of the dnrN pseudo
response regulator gene. Microbiology 146 ( Pt 6):1457-1468.

Paget, M. 8., G. Hinterrnann, and C. P. Smith. 1994. Construction and
application of streptomycete promoter probe vectors which employ the

Streptomyces glaucescens tyrosinase-encoding gene as reporter. Gene
146:105-110.

Paget, M. 8., J. G. Kang, J. H. Roe, and M. J. Buttner. 1998. sigmaR, an
RNA polymerase sigma factor that modulates expression of the thioredoxin
system in response to oxidative stress in Streptomyces coelicolor A3(2).
EMBO J. 17:5776-5782.

151

173.

174.

175.

176.

177.

178.

179.
180.

181.

182.

183.

Paget, M. 8., E. Leibovitz, and M. J. Buttner. 1999. A putative two-
component signal transduction system regulates sigmaE, a sigma factor
required for normal cell wall integrity in Streptomyces coelicolor A3(2).
Mol.Microbiol. 33297-107.

Pope, M. K., B. Green, and J. Westpheling. 1998. The bldB gene encodes
a small protein required for morphogenesis, antibiotic production, and
catabolite control in Streptomyces coelicolor. J.Bacteriol. 180:1556-1562.

Pope, M. K., B. D. Green, and J. Westpheling. 1996. The bld mutants of
Streptomyces coelicolor are defective in the regulation of carbon utilization,
morphogenesis and cell-cell signalling. Mol.Microbiol. 19:747-756.

Potter, C. A. and S. Baumberg . 1996. End-product control of enzymes of
branched-chain amino acid biosynthesis in Streptomyces coelicolor.
Microbiology 142 ( Pt 8)21945-195.2

Potuckova, L., G. H. Kelemen, K. C. Findlay, M. A. Lonetto, M. J.
Buttner, and J. Kormanec. 1995. A new RNA polymerase sigma factor,
sigma F, is required for the late stages of morphological differentiation in
Streptomyces spp. Mol.Microbiol. 17:37-48.

Price, B., T. Adamidis, R. Kong, and W. Champness. 1999. A
Streptomyces coelicolor antibiotic regulatory gene, absB, encodes an
RNase llI homolog. J.Bacteriol. 181 :6142-6151.

Ragatz, D. and F. de Bruijn. Personal Communication. 1993.

Redenbach, M., H. Illl. Kieser, D. Denapaite, A. Eichner, J. Cullum. H.
Kinashi, and D. A. Hopwood. 1996. A set of ordered cosmids and a

detailed genetic and physical map for the 8 Mb Streptomyces coelicolor
A3(2) chromosome. Mol.Microbiol. 21:77-96.

Redenbach, M., E. Kleinert, and A. Stoll. 2000. Identiﬁcation of DNA
ampliﬁcations near the center of the Streptomyces coelicolor M145
chromosome. FEMS Microbiol.Lett 191:123-129.

Rodicio, M. R., C. J. Bruton, and K. F. Chater. 1985. New derivatives of
the Streptomyces temperate phage phi C31 useful for the cloning and
functional analysis of Streptomyces DNA. Gene 34:283-292.

Rowe, C. J., J. Cortes, S. Gaisser, J. Staunton, and P. F. Leadlay. 1998.
Construction of new vectors for high-level expression in actinomycetes.
Gene 216:215-223.

152

184.

185.

186.

187.

188.

189.

190.
191.

192.

193.

194.

195.

Rudd, B. A. and D. A. Hopwood . 1979. Genetics of actinorhodin
biosynthesis by Streptomyces coelicolor A3(2). J.Gen.Microbiol. 1 14:35-43.

Rudd, B. A. and D. A. Hopwood . 1980. A pigmented mycelial antibiotic in
Streptomyces coelicolor: control by a chromosomal gene cluster.
J.Gen.Microbiol. 1192333-340.

Ruddick, S. M. and 8. T. Williams. 1972. Studies on the ecology of
actinomycetes in soil V. Some factors inﬂuencing the dispersal and
adsorption of spores in soil. Soil Biol.Biochem. 4:93-103.

Ryding, N. J., T. B. Anderson, and W. C. Champness. 2001. Regulation of
the Streptomyces coelicolor antibiotic CDA by absA, a cluster-linked two-
component system. submitted .

Ryding, N. J., T. B. Anderson, and W. C. Champness. Personal
Communication. 2001.

Sauer, U. and P. Durre. 1992. Possible function of tRNA(Thr)ACG in
regulation of solvent formation in Clostridium acetobutylicum. FEMS
Microbiol.Lett. 792147-153.

Schauer, A. Personal Communication. 1993.

Schauer, A., M. Ranes, R. Santamaria, J. Guijarro, E. Lawlor, C. Mendez,
K. Chater, and R. Losick. 1988. Visualizing gene expression in time and
space in the ﬁlamentous bacterium Streptomyces coelicolor. Science
240:768-772.

Schneider, 0., C. J. Bruton, and K. F. Chater. 1996. Characterization of
spaA, a Streptomyces coelicolor gene homologous to a gene involved in
sensing starvation in Escherichia coli. Gene 177:243—251.

Scholz, O., A. Thiel, W. Hillen, and M. Niederweis. 2000. Quantitative
analysis of gene expression with an improved green ﬂuorescent protein.
Eur.J.Biochem. 267:1565-1570.

Schrempf, H. 1983. Deletion and ampliﬁcation of DNA sequences in
melanin-negative variants of Streptomyces reticuli. Mol.Gen.Genet.
189:501-505.

Schrempf, H., H. Iujard, D. A. Hopwood, and W. Goebel. 1975. Isolation
of covalently closed circular deoxyribonucleic acid from Streptomyces
coelicolor A3(2). J.Bacteriol. 121:416-421.

153

196.

197.

198.

199.

200.

201.

202.

203.

204.

205.

206.

207.

Shelton, D. R., S. Khader, J. S. Kams, and B. M. Pogell. 1996.
Metabolism of twelve herbicides by Streptomyces. Biodegradation 72129-
136.

Shima, J., A. Penyige, and K. Ochi. 1996. Changes in patterns of ADP-
ribosylated proteins during differentiation of Streptomyces coelicolor A3(2)
and its development mutants. J.Bacteriol. 17823785-3790.

Sohaskey, C. D., H. lm, A. D. Nelson, and A. T. Schauer. 1992. Tn4556
and luciferase: synergistic tools for visualizing transcription in Streptomyces.
Gene 115267-71.

Sohaskey, C. D., H. lm, and A. T. Schauer. 1992. Construction and
application of plasmid- and transposon-based promoter-probe vectors for
Streptomyces spp. that employ a Vibrio harvei lucerifase reporter cassette.
J.Bacteriol. 1742367-376.

Sosio, M., F. Giusino, C. Cappellano, E. Bossi, A. M. Puglia, and S.
Donadio. 2000. Artiﬁcial chromosomes for antibiotic-producing
actinomycetes. Nat.Biotechnol. 182343-345.

Stackebrandt, E., W. Liesack, and D. Witt. 1992. Ribosomal RNA and
rDNA sequence analyses. Gene 1152255-260.

Stackebrandt, E. F., A.Rainey, and N.L.Ward-Rainey. 1997. Proposal for a
new hierarchic classifcation system, Actinobacteria classis nov.
lnt.J.Syst.BacterioI. 47:479-491.

Stein, D. and S. N. Cohen. 1989. A cloned regulatory gene of Streptomyces
Iividans can suppress the pigment deﬁciency phenotype of different
developmental mutants. J.Bacteriol. 17122258-2261.

Strauch, E., E. Takano, H. A. Baylis, and M. J. Bibb. 1991. The stringent
response in Streptomyces coelicolor A3(2). Mol.Microbiol. 5:289-298.

Strohl, W. R. 1992. Compilation and analysis of DNA sequences associated
with apparent streptomycete promoters. Nucleic Acids Res. 20:961-974.

Strohl, W. R. 2001. Industrial antibiotics: today and the future, p. 1-48. In
W.R.Strohl (ed.), Biotechnology of Antibiotics. Marcel Dekker, Inc., New
York.

Stutzman-Engwall, K. J., S. L. Otten, and C. R. Hutchinson. 1992.
Regulation of secondary metabolism in Streptomyces spp. and

overproduction of daunorubicin in Streptomyces peucetius. J.Bacteriol.
174:144-154.

154

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

Sun, J., A. Hesketh, and M. Bibb. 2001. Functional analysis of relA and
rshA, two relA/spoT homologues of Streptomyces coelicolor A3(2).
J.Bacteriol. 183:3488-3498.

Sun, J., G. H. Kelemen, J. M. Femandez-Abalos, and M. J. Bibb. 1999.
Green ﬂuorescent protein as a reporter for spatial and temporal gene
expression in Streptomyces coelicolor A3(2). Microbiology 145 ( Pt 9):2221-
2227.

Takano, E., H. C. Gramajo, E. Strauch, N. Andres, J. White, and M. J.
Bibb. 1992. Transcriptional regulation of the redD transcriptional activator
gene accounts for growth-phase-dependent production of the antibiotic
undecylprodigiosin in Streptomyces coelicolor A3(2). Mol.Microbiol. 622797-
2804.

Takano, E., T. Nihira, Y. Hara, J. J. Jones, C. J. Gershater, Y. Yamada,
and M. Bibb. 2000. Puriﬁcation and structural determination of SCB1, a
gamma- butyrolactone that elicits antibiotic production in Streptomyces
coelicolor A3(2). J.Biol.Chem. 27521 1010-11016.

Tanaka, K., T. Shiina, and H. Takahashi. 1988. Multiple principal sigma
factor homologs in eubacteria: identiﬁcation of the "rpoD box". Science
24221 040-1042.

Tang, L., A. Grimm, Y. X. Zhang, and C. R. Hutchinson. 1996. Puriﬁcation
and characterization of the DNA-binding protein Dnrl, a transcriptional factor
of daunorubicin biosynthesis in Streptomyces peucetius. Mol.Microbiol.
222801-813.

Tang, L., S. Shah, L. Chung, J. Carney, L. Katz, C. Khosla. and B. Julien.
2000. Cloning and heterologous expression of the epothilone gene cluster.
Science 287:640-642.

Thomashow, L. 8., D. M. Weller, R. F. Bonsall, and L. S. Pierson Ill.
1990. Production of the antibiotic phenazine-1-carboxylic acid by ﬂuorescent

Pseudomonas species in the rhizosphere of wheat. Appl.Environ.Microbiol.
56:908-912.

Tsao, S. W., B. A. Rudd, X. G. He, C. J. Chang, and H. G. Floss. 1985.
Identiﬁcation of a red pigment from Streptomyces coelicolor A3(2) as a
mixture of prodigiosin derivatives. J.Antibiot.(Tokyo) 38:128-131.

Tseng, H. C. and C. W. Chen. 1991. A cloned ompR-like gene of
Streptomyces Iividans 66 suppresses defective melC1, a putative copper-
transfer gene. Mol.Microbiol. 521187-1196.

155

 

218.

219.

220.

221.

222.

223.

224.

225.
226.

227.

228.

229.

Viollier, P. H., W. Minas, G. E. Dale, M. Folcher, and C. J. Thompson.
2001. Role of acid metabolism in Streptomyces coelicolor morphological
differentiation and antibiotic biosynthesis. J.Bacteriol. 183:3184-3192.

Vogtli, M., P. C. Chang, and S. N. Cohen. 1994. ast2: a previously
undetected gene encoding a 63-amino-acid protein that stimulates antibiotic
production in Streptomyces Iividans. Mol.Microbiol. 14:643-653.

Volff, J. N. and J. Altenbuchner. 1997. High frequency transposition of the
Tn5 derivative Tn5493 in Streptomyces Iividans. Gene 194281-86.

Waksman, S. A. 1950. The Actinomycetes: Their Nature, Occurrence,
Activities, and Importance. Chronica Botanica Company, Waltham, MA.

Waksman, S. A. and A.T.Henrici. 1943. The nomenclature and
classiﬁcation of the actinomycetes. J.Bacteriol. 46:337-341.

Ward, J. M., G. R. Janssen, T. Kieser, M. J. Bibb, M. J. Buttner, and M. J.
Bibb. 1986. Construction and characterisation of a series of multi-copy
promoter- probe plasmid vectors for Streptomyces using the aminoglycoside
phosphotransferase gene from Tn5 as indicator. Mol.Gen.Genet. 2032468-
478.

Weigel, B. J., S. G. Burgett, V. J. Chen, P. L. Skatrud, C. A. Frolik, S. W.
Queener, and T. D. lngolia. 1988. Cloning and expression in Escherichia
coli of isopenicillin N synthetase genes from Streptomyces Iipmanii and
Aspergillus nidulans. J.Bacteriol. 170:3817-3826.

Wellington, E. M. Personal Communication. 1995.

Wellington, E. M. H., P.Marsh, I.Toth, N.Cresswell, L.Huddleston, and
M.B.Schilhabel. 1993. The selective effects of antibiotics in soil, p. 331-
336. In R.Guerrero and C.Pedros-Alio (eds), Trends in Microbial Ecology.
Spanish Society for Microbiology.

Westpheling, J., M. Ranes, and R. Losick. 1985. RNA polymerase
heterogeneity in Streptomyces coelicolor. Nature 313222-27.

White, J. and M. Bibb. 1997. bldA dependence of undecylprodigiosin
production in Streptomyces coelicolor A3(2) involves a pathway-speciﬁc
regulatory cascade. J.Bacteriol. 179:627-633.

Vinetzorrek, A. and M. Bibb. 1997. A novel family of proteins that regulates

antibiotic production in streptomycetes appears to contain an OmpR-like
DNA-binding fold [letter]. Mol.Microbiol. 25:1181-1184.

156

230.

231.

232.

233.

234.

235.

236.

237.

Willey, J., R. Santamaria, J. Guijarro, M. Geistlich, and R. Losick. 1991.
Extracellular complementation of a developmental mutation implicates a
small sporulation protein in aerial mycelium formation by S. coelicolor. Cell
65:641-650.

Willey, J., J. Schwedock, and R. Losick. 1993. Multiple extracellular
signals govern the production of a morphogenetic protein involved in aerial
mycelium formation by Streptomyces coelicolor. Genes Dev. 72895-903.

Williams, S. T. 1985. Oligotrophy in soil: Fact or ﬁction?, p. 81-104. In M.
Fletcher and G. D. Floodgate (eds), Bacteria in their natural environments.
Academic Press, Harcourt Brace Jovanovich, London.

Williams, 8. T., M. Goodfellow, G. Alderson, E. M. Wellington, P. H.
Sneath, and M. J. Sackin. 1983. Numerical classiﬁcation of Streptomyces
and related genera. J.Gen.Microbiol. 129 (Pt 6):1743-1813.

Wright, J. M. 1956. The production of antibiotics in soil. IV. Production of
antibiotics in coats of seeds sown in soil. Ann.Appl.BioI. 44:561-566.

Wright, L. F. and D. A. Hopwood. 1976. Identiﬁcation of the antibiotic
determined by the SCP1 plasmid of Streptomyces coelicolor A3(2).
J.Gen.Microbiol. 95296-106.

Yuan, W. M. and D.L Crawford. 1995. Characterization of Streptomyces
lydicus WYEC108 as a potiential biocontrol agent against fungal root and
seed rots. Appl.Environ.Microbiol. 61:3119-3128.

Zaborina, 0., B. Baskunov, L. Baryshnikova, and L. Golovleva. 1997.
Degradation of pentachlorophenol in soil by Streptomyces rochei 303.
Journal of Environmental Science and Health 832255-70.

238. Zukowski, M. M., D.F.Gaffney, D.Speck, M.Kauffman, A.Findeli,

A.Wisecup, and J.-P.Lecocq. 1982. Chromogenic identiﬁcation of genetic
regulatory signals in Bacillus subtilis based on expression of a cloned
Psuedomonas gene. Proc.Natl.Acad.Sci.U.S.A 80:1101-1105.

157

IIIIIIIIIIIIIIIIIIIII

III[Illijjljljjij[13]]II