133
FgPiews
Metabolites of Pseudomonas
involved in the biocontrol
of plant disease
David N. Dowling and Fergal O'Gara
There is increasing commercial and environmental interest in the use of microbebased agents as alternatives to, or in combination with, chemicals for controlling
the spread and severity of a range of crop diseases. The identification of specific
microbial metabolites that are able to control certain plant diseases has led to the
development of strategies for improving the performance and predictability of the
microbial strains that produce these metabolites for application in the agricultural
industry. This article focuses on antimicrobial metabolites produced by fluorescent
pseudomonads, discusses their role in suppressing fungal diseases of important
crops and reviews the prospects of genetically manipulating the producer
organisms to improve the efficacy of these biocontrol agents.
The ability of some soils to suppress plant disease has
been well documented 1. An important example is the
development of soils suppressive to Gaeumannomyces
graminis var. tritici (Ggt), the causative agent of 'takeall' of wheat 2. Such natural examples of biological control have been attributed to the indigenous beneficial
rhizosphere microflora (see Glossary), especially fluorescent pseudomonads. Numerous mechanisms may
account for these biocontrol properties (see Glossary),
including the production of inhibitory compounds
or metabolites 3-8. Microbial metabolites such as
siderophores (see Glossary) and secondary metabolites
with antimicrobial properties are considered to play a
major role in disease suppression (see Table 1).
Metabolites with biocontrol properties have been
reported from diverse members of the beneficial
rhizosphere flora; however, those produced by the
fluorescent pseudomonads have received the most
attention. This is probably due to the abundance of
this diverse group of bacteria and their obvious
importance in the rhizosphere, coupled with the
relative ease with which they can be genetically
manipulated.
Fluorescent pseudomonads produce a variety o f
metabolites 22 (see Fig. 1), many of which are
inhibitory to other microorganisms and some of
which are implicated in the biological control of plant
D. N. Douding and F. O'Gara are at the Department of Microbiology,
University College Cork, Ireland.
© 1994, Elsevier Science Ltd
pathogens. The identification of a link between a
metabolite(s) and the suppression o f a particular disease is an on-going goal of many research groups,
which the use of recombinant D N A (rDNA) techniques has facilitated. Some of the criteria used to
indicate that a particular metabolite has a primary role
in biological control are shown below.
• Mutants defective in metabolite(s) are unable to
show inhibition of the pathogen in the laboratory.
• The biocontrol ability of the mutants is reduced in
the field.
• Complementation of the mutant with wild-type
D N A sequences restores biocontrol ability.
• The purified metabolite shows fungicidal or
antimicrobial properties.
• The metabolite may be detected in situ (i.e. in the
rhizosphere) when producing strains are present.
The ability to fulfil some or all of these points provides good evidence for the involvement of a particular metabolite in biocontrol. A selection of plant
diseases controlled by specific metabolites are summarized in Table 1 and the structures of a range of
these compounds are shown in Fig. 2.
Role of metabolites in biological control
Some of the metabolites implicated in biocontrol
appear to be broad-ranging in their inhibitory action.
For example, phloroglucinols and phenazines have
TIBTECHAPRIL1994 (VOL 12)
134
reviews
Table 1. Examples of specific microbial metabolites implicated in the control of crop diseases
Disease
Pathogen
Effective metabolite
Refs
Take-all of wheat
Phenazines
c-Acetyl phloroglucinols
9
10,11
Tan spot of wheat
Gaeumannomyces graminis
var. tritici (Ggt)
Pyrenophora triticirepentis
Pyrrolnitrin
12
Pre-emergent damping-off of:
cotton;
sugarbeet
Pythium spp.
Pythium ultimum
Oomycin A
Pyoluteorin
2,4-Diacetylphloroglucinol
13
Hydrogen cyanide
2,4-Diacetylphloroglucinol
16
Agrocin 84
17-19
Pseudobactin B10
20
Ammonia
21
Black root-rot of tobacco
Crown gall of fruit trees
Flax wilt
Damping-off
Thielabiopsis basicola
Agrobacterium
tumefaciens
Fusarium oxysporum
Pythium spp.
been shown to inhibit a wide range offungal pathogens
in the laboratory. Siderophores exhibit both fungistatic
and bacteriostatic effects in the laboratory under conditions of low iron. In the field, these iron-chelating
compounds are thought to deprive the pathogen of
iron, a limiting essential nutrient.
Other metabolites are known to have very specific
effects and to target particular pathogens; for example,
agrocin 84, produced by Agrobacterium radiobacter, is
specific for virulent strains ofAgrobacterium tumefaciens.
At the molecular level, agrocin 84 (a di-substituted
nucleotide) is thought to act by chain termination of
DNA synthesis17. However, the precise mode of actions
of many other metabolites is poorly understood.
Glossary
Biological control - The use of living organisms to
modify the agricultural ecosystem to control a crop
disease or prevent the establishment of a pest.
Rhizosphere - The root surface and that part of the
soil directly influenced by the root system.
Plant-growth-promoting rhizobacteria (PGPR}Bacteria isolated from the rhizosphere, which have
been shown to improve plant health or increase yield.
They may exert their effects directly through the
production of metabolites that can act as plant hormones and/or indirectly by the suppression of plant
pathogens (biological control).
Induced systemic acquired resistance -The ability,
of an agent, for example, a bacterium, chemical,
virus, nematode, etc., to induce plant- defence mechanisms that lead to systemic resistance to a number
of pathogens.
Siderophore - Low-molecular-weight compound that is
produced as a specific iron chelator by a microorganism.
IBTECHAPRIL1994(VOL12)
14
15
10
Effects such as direct killing or pathogen inhibition
exhibited by a particular metabolite in the laboratory
cannot always be translated directly into biological
control in the field, as the conditions experienced in
the field are much more complex. However, there
are a number of studies demonstrating a direct effect
of particular antimicrobial metab-olites on pathogen
numbers in soil. One such example is the suppression
of root pathogens by plant-growth-promoting rhizobacteria (PGPR) (see Glossary)23. In another case, the
population size of antibiotic-producing Pseudomonas
fluorescens 2-79 was shown to correlate positively with
a reduction in the number of lesions on wheat roots
caused by Ggt (Ref. 24).
Evidence is accumulating to support the theory
that metabolite production has beneficial effects
on the ecological competence of the producer
strain. Production of these compounds is thought
to pro-vide the producing strain with a selective
advantage in the highly competitive environment
of the plant rhizosphere (see R.ef. 25 for review).
This idea has been substantiated by a recent report 26,
which demonstrated that phenazine (Phz) antibiotics contributed to the persistence of the producer
strains (P. fluorescens 2-79 and P. aureofaciens 30-84)
compared with Phz-deficient mutants in a simulated wheat-rotation microcosm. Persistence may
lead to improved competition and colonization of
the producer strain on the plant surface, leading
to niche-exclusion of the pathogen. The role of
siderophores and other metabolite(s) in biocontrol
may not be simply a direct antagonism of the pathogenic fungus, but may also involve more subtle
ecological effects.
Alternative, more indirect, modes of action
could involve stimulation of the plant's own defence mechanisms by induced systemic acquired
resistance27-29 or, possibly, by direct uptake and
translocation of the metabolite within the plant.
135
reviews
Siderophores*
Pterines*
Pyroles*
Pyochelin
Pyoverdine
Pseudobactin- B10, M114,
A214, 7SR1, A112, B117
Ferribactin
PhY~:hid:r°$h°res
~ ~
Pterine
Am inopterine
Ribilyllumazine
Putidolumazine
Pyoluteorin
Pyrrolnitrin
Phenylpyrolles
Isopyrrolnitrin
Aminopyrrolnitrin
- . ~ ~ ' k
Indoles
3-chloroindole
Indole-3-carboxaldehyde
6-bromoindole-3-carboxaldehyde
7-chloroindoleacetic acid
Indoleacryloisonitrile
/
Lipids/Pyocompounds
Pseudanes
Rhamnolipids
Pyolipids
Compound B
Jarvis rhamnolipid
Compound A
Ferroxamine B
Alginate
Antibiotic properties
miscellaneous*
Cyanhydric acid
Aeruginoic acid
Magnesidin
Pseudomonic acid
Pseudomonic acid a
Pseudomonic acid b
Antibiotic P2563
P2563 a
P2563 b
,Amino-2 acetophenone
c-Acetyl phloroglucinols
Antibiotic DB-2073
Fluopsin C+F
Sorbistin A1 +B
Salicylic acid
Indole-3-acetic acid
Amino acids and peptides
L-2-amino-L-methoxybuteonic acid
o-ethylhomoserine
Phaseotoxin A
Phaseolotoxins
Tabtoxins
Isotabtoxins
Tabtoxinine
Coronatine
Proferrosamine A
Pyfimine
Viscosin
Phenazines*
Phenazine-1-carboxylic acid
Pyovanine
Hemipyocanine
Idoinin
Chlororaphin
Oxychlororaphin
Aeruginosin A+B
Figure 1
The range of secondary metabolites and other compounds produced by fluorescent Pseudomonas spp, (Data largely extracted from
Ref. 22.) The compoundsindicatedby an asteriskhavebeenimplicatedin the biocontrolabilityof the producerstrain.
Some key metabolites implicated in biocontrol
Siderophores
Siderophores are produced by many microorganisms
as a means of sequestering limiting iron 3°. Pseudomonads produce a range of iron-chelating compounds
including salicylic acid, pyochelins and fluorescent
pseudobactins and pyoverdines (see Fig. 2 for structures). Fluorescent siderophores are unique to pseudomonads - a trait that has implicated these organisms
as P G P R (1Kefs 3,4). Fluorescent siderophores have
been isolated from soiP 1,32 and there is considerable
genetic and biochemical evidence that demonstrates
their role in the promotion of plant growth and in biocontrol20,33,34.
Fluorescent siderophores such as pseudobactins have
a very high affinity for ferric iron (iron in the trivalent
state). The molecules comprise a fluorescent chromophore and a quinoline moiety with two hydroxyl
groups, which is involved in binding iron in conjunction with two hydroxymate groups from the peptide
backbone of the molecule (for review see IKef. 35).
The structures of two fluorescent siderophores,
pseudobactin and pyoverdine, are shown in Fig. 2.
Variations in the peptide moiety of the molecule are
responsible for the strain specificity often observed
with pseudobactins.
Uptake of the fluorescent iron-siderophore complex
by an organism requires a specific outer-membrane
receptor. Although an individual Pseudomonas strain
usually produces only one type ofpseudobactin, strains
that can utilize a range of pseudobactin siderophores
by harbouring a number of different outer-membrane
siderophore receptors can be isolated36-38.
Siderophore production and uptake of the ironsiderophore complex are considered to be significant
for an organism as they may provide it with a competitive advantage under conditions of low iron, which
would contribute directly or indirectly to preventing
the establishment of a pathogen in the rhizosphere.
Indeed, in laboratory experiments, the introduction
of an additional receptor gene may improve the
performance of the host strain39,4°.
Antimicrobial metabolites
Soil bacteria, in particular Pseudomonas spp., produce
an array of low-molecular-weight metabolites, some of
which are potent antifungal agents. Apart from the
direct use of microbial metabolites as lead or precursor
molecules by the chemical industry for production of
new disease-control agents<, metabolites are normally
produced in situ (i.e. on the root surface) by the producing organism or biocontrol agent. Table 1 lists some
of the metabolites involved in disease control in soil.
Important antimicrobial metabolites include the
phenazines, such as phenazine-l-carboxylate (PCA)
(Kef. 9) and the c-acetyl-phloroglucinols1°,11, which
TIBTECHAPRIL1994(VOL12:
136
reviews
OH
COOH
H
O
I
H
Pyoluteorin
Phenazine-1 -carboxylate
N.
.0 9'
\ /o ,r-~.oH
S.
o
/CH3
Z
o
H
Pyochelin
al°
R8~'-~N~
7
'"
R6
O
D-allo-Thr
\ L.Ala/
O
R2
3
R5
CH2OH t
4
OHO ~ H O
Phenazines
-OH
--O~ //(3
/
/- P\
NH
O-
I
O
I
Ho~ ~
L-Thr
OH
_
2,4-Diacetylphloroglucinol
HO
H
HN-P-O-CHe
J
H O-
H-O-OH
./N
L-Lys
I
OH L-Thr
\D_Ser--NH~-ThS
o
"</
NH
NH
D-Ser
oJl
oJl
OH
/C.. ~
..C...
HO
NH ~
Pseudobactin
(fluorescent)
N-(3-oxohexanoyl)homoserine lactone
R1
NH
/
NH2
\
O
S..~C02H
9
/
,_~;~o
I
Pyoverdine
(fluorescent)
o
I ~
COOH
c,
Agrocin 84
cI
Pyrrolnitrin
Salicylic acid
Figure 2
Structures of some of the bacterial metabolites implicated in the biological control of plant disease in the field.
are effective against 'take-all' of wheat. Both of these
comlSounds have been isolated t~om microcosm rhizospheres colonized by the producing strain, but not from
roots colonized by mutants defective in the metabolitesynthesis pathway m,42.
Phloroglucinol and pyoluteorin have been shown to
be largely responsible for the prevention of'dampingoff' of sugar-beet is (caused by Pythium ultimum) and
cotton 14 (caused by Pythium spp.), whereas phloroglucinol and hydrogen cyanide (HCN) are responsible
tbr the control of black root-rot of tobacco (caused by
Thielalfiwsis basicola) by P. fluorescens CHA0 (P,.efs
10,1B). Putative biochemical pathways for the syn-
TIBTECHAPRIL1994(VOL12)
thesis of many of these compounds have been
reviewed (see P-.ef. 22).
Volatile compounds such as ammonia 21 and H C N
are produced by many rhizosphere strains and have
been implicated as important metabolites in biocontrol. For example, some species of Pseudomonas can
produce levels of H C N in vitro that are toxic to certain
pathogenic fungi, e.g. Thielabiopsis basicola, and, thus,
prevent black root-rot of tobacco 1('. High concentrations of H C N are toxic to some plants and it has
been suggested that fluorescent pseudomonads producing H C N may be responsible for a reduction in
the yield of certain crops, e.g. potato 34.
137
reviews
Plant hormones
Salicylic acid, a precursor of pyochelin43 and a
siderophore in its own right 44, is also a plant hormone 4s and is implicated in the induction of systemic
acquired resistance in plants 46. Another plant hormone, indole acetic acid (IAA), is also produced by
many strains that exhibit biocontrol properties.
Although IAA has not been directly implicated as
a metabolite in disease control, it is bioactive and
stimulates root elongation (by stimulating the growth
of root hairs) 47. Indole acetic acid is produced by the
nitrogen-fixing bacteria, AzospMllum, and is thought
to play a key role in the plant-growth-promoting
effect that these bacteria have on graminaceous
plants 48.
Factors affecting metabolite production
Antimicrobial-metabolite production and its effect
on pathogens is dependent on environmental factors
such as soil chemistry, temperature and water potential
- factors that will be beyond the control of the biotechnologist. In addition to active and aggressive colonization of the producing strain, the organism must synthesize the metabolite at the correct time and in the
appropriate location. In the laboratory, adequate production ofmetabolites requires the producing strain to
be grown in synthetic media containing a suitable carbon and nutrient source. In the field, these nutrients
will be supplied in the form of root or seed exudates.
A number of key factors have been shown to influence the metabolite production of some biocontrol
strains. These include:
• Zinc: increases the production ofphenazine (PCA)
(Ref. 49).
• Temperature: 12°C is optimum for the production
of 2,4-diacetylphloroglucinol (DAPG) (Ref. 15).
• Surface contact: DAPG production by strain
Fl13 is enhanced if the bacteria attach to solid
surfaces is.
• Water potential: oomycinA (Re£ 13).
• Carbon sources: glucose stimulates the production
ofoomycin A (Ref. 50), but represses the production
of DAPG (Ref. 15). Sucrose stimulates the production o f D A P G (Ref. 15).
• Iron: represses the production o f siderophores >
and stimulates the production of H C N (Ref. 16).
Carbon sources usually have a major influence on
the type and levels of antibiotics produced, and
laboratory studies on the production of metabolites
with respect to carbon sources are yielding interesting
results. In combination with a knowledge of the
relative amounts and types of sugars, amino acids, etc.
present in the rhizosphere, it may be possible to
extrapolate this information to the field in order to
predict whether or not metabolites will be produced
in situ.
Table 2. Genetic-engineering approaches to modify the production of microbial metabolites for improved
biological control of plant pathogens
Strategy
Modification a
Altered expression Inserted alternative promoter (ptac)
upstream of biosynthesis
genes
Synthesized regulatory mutants with
constitutive siderophore production
Gene dosage
Heterologous
expression
Deletion
Effect
Refs
Constitutive expression of
oomycin A. Enhanced
biocontrol
13
Siderophore-mediated
biocontrol in the laboratory
under conditions of high iron
52
Over- production of
2,4-diacetylphloroglucinol
and pyoluteorin. Enhanced
biocontrol on cucumber,
phytotoxic to cress
53
Introduced phl, hcn,phz
genes into a non-producer strain
Non-producing strains converted
to producing strains. Enhanced
biocontrol ability in soil
7,11,16,54
Added pseudobactinreceptor genes
Potential to utilize a larger
range of indigenous
siderophores. More competitive?
36-39
Increased copy number of
gene(s) in wild type
Specifically deleted tra
genes from plasmid
encoding agrocin 84
Eliminates potential transfer of
agrocin 84 resistance gene to
target pathogen. Safer strain
18,19
aAbbreviations: ptac, a syntheticpromoter made from the lac and trp promoters; phi, gene for the synthesisof phloroglucinols;hcn,
gene for synthesisof hydrogencyanide; phz, gene for synthesisof phenazines;tra, transfer gene.
TIBTECHAPRIL1994(VOL12}
138
reviews
Current strategies to improve biocontrol agents
Unmodified strains and consortia
Most of the commercial biocontrol strains that are
currently available are not genetically manipulated,
and the use of wild-type strains as a primary source of
material for biological control is likely to continue.
The use o f more rational screening systems that simulate field conditions are likely to generate many new
isolates. For example, in our laboratory, a screen of
natural rhizosphere pseudomonads for strains that
produced fluorescent siderophores under conditions
of high levels of iron, produced one isolate that also
showed good biocontrol ability (D. J. O'Sullivan and
E O'Gara, unpublished).
The use of mixtures o f microbial strains as biocontrol agents is an important development in the use
of unmodified inoculants. Many reports illustrate the
synergistic effect of mixed cultures in the control of
certain pathogens 7,51. This may be due to the production of an increased range o f metabolites by the
consortia and/or other mechanisms. By choosing
strains with diverse mechanisms for biocontrol, e.g.
siderophores and phloroglucinols, it may be possible
to design consortia that have improved biocontrol
properties in the field.
Much o f the information on the role ofmetabolites
in biocontrol has come from the use ofrDNA-based
technology. The use o f such technology can also be a
powerful tool for manipulating the production and
range o f metabolites produced by a single strain.
Genetically modified strains
Standard rDNA and genetic approaches are currently being used to manipulate strains for improved
production (and uptake) o f important metabolites. A
summary of the different strategies, with examples, are
shown in Table 2. Some examples are discussed in
more detail below.
The genetics and biochemistry o f siderophore
biosynthesis is complex, and it is obvious that
phlX
I
b
oii
H3c/C~
HO
oii
OH
OH
MAPG
oii
~, HaG/C ~ C . . .
',:~
~/
OH
MAPG
acetyltransferase
HO
~
OH3
OH
DAPG
Figure 3
Genetic engineering of a biocontrol strain. Introduction of gene(s) (phlX) encoding a
monoacetylphloroglucinol (MAPG) acetyltransferase (constitutively expressed from
the tet promoter) into a suitable heterologous recipient strain confers the ability to
synthesize the more active (lOOx) antifungal metabolite, 2,4-diacetylphloroglucinol
(DAPG). Such an engineered derivative was found to show enhanced biocontrol activity in soil experiments54. 58.
TIBTECH
APRIL1994(VOL12i
siderophores will only be effective in soils with low
levels o f iron. Engineering strains to produce
siderophores at higher levels of soil iron may lead to
improved performance of an inoculant strain. Studies
on iron-mediated regulation of siderophores indicate
that siderophore biosynthesis is under both positive
and negative regulation ss 57, and that the generation
o f deregulated mutants 52 with a constitutive phenotype may offer a possible approach to improving this
important trait. As discussed previously, broadening
the ferric iron-siderophore uptake potential o f a strain
may also increase its rhizosphere and biocontrol competence.
Engineering over-production of metabolites has
been a strategy used by many researchers. The first
antibiotic genes to be cloned and manipulated were
from P.jquorescensHV37a, which produces oomycin A.
This antibiotic is primarily responsible for biocontrol
(70% o f disease control) by this strain o f Pythiuminduced root infection of cotton seedlings. The loci
responsible for biosynthesis, AfuE, AfuR, AfuP, and
regulation, AfuAB, were identified and cloned.
Expression of the AfuE gene has been increased by
cloning it downstream from the tac promoter (a synthetic promoter made from the lac and trp promoters).
This led to increased production of oomycin A by the
engineered strain 13. Preliminary experiments have
suggested that over-production o f oomycin leads to
improved inhibition of Pythium ultimum.
Careful evaluation o f the host plant, target pathogen and inoculant is required when over-producing
metabolites in the absence of appropriate regulation.
A derivative of the wild-type biocontrol strain CHA0
harbours a recombinant plasmid that confers the
ability to over-produce both DAPG and pyoluteorin 53.
This strain showed enhanced biocontrol activity with
cucumber but was phytotoxic, compared with the
unmodified parent strain, to cress. This indicates that
metabolite production may have to be regulated for
optimum disease control on susceptible host plants.
Another strategy is the introduction of a new trait
into heterologous hosts with a view to improving their
biocontrol potential. The production o f DAPG by
Pseudomonas sp. strain F] 13 is a key trait in the biocontrol of P. ultimum 'damping-off' o f sugar-beet
seedlings 15. The genes involved in the biosynthesis of
DAPG have been cloned 54 and the final step in the
pathway, monoacetylphloroglucinol (MAPG) acetyltransferase activity, is encoded by a 6kb D N A fragment ss. This fragment was cloned upstream o f a constitutive tetracycline resistance promoter (tet) in the
cloning vector pSUP106 (see Fig. 3) and introduced
into a wild-type strain (Ml14) that is unable to synthesize DAPG. The engineered strain expresses the
enzyme activity that converts MAPG to DAPG and,
thus, is able to synthesize DAPG. The presence of this
new trait has enhanced the biocontrol ability of strain
M l 1 4 against P. ultimum, both in the laboratory and
in greenhouse experiments 54. Introduction of the
D N A fragment into other pseudomonads and E. coli
does not lead to DAPG production, suggesting that
139
reviews
these organisms may lack other components that are
necessary for production of the antibiotic.
Other examples of this strategy include the heterologous expression of the genes for H C N biosynthesis is, which leads to enhanced biocontrol of
tobacco black root-rot in the recipient strain, and the
introduction of the genes for the synthesis of DAPG
([Zef. 11) and phenazine into other pseudomonads 7.
This strategy may offer a simple way to construct
strains that produce multiple bioactive metabolites.
Antimicrobial compounds can be considered as secondary metabolites that preferentially accumulate late
in the growth cycle (i.e. in stationary phase) in laboratory media. This is consistent with the notion that
antimicrobial compounds have evolved as microbial
defence mechanisms 59 to inhibit competitors under
stress situations such as nutrient depletion.
The regulation of the synthesis of important antimicrobials from fluorescent pseudomonads is summarized in Fig. 4. Antibiotics are regulated by a global
regulator, GacA (Ref. 60). The environmental stimulus that triggers this response is not yet known;
however, there is physiological and genetic evidence
that a cell-density-dependent autoinduction mech
anism, analogous to the prototype luxR/I system
[the regulatory genes for bacterial (Vibrio sp.) lightproduction enzymes], may also be involved. The lux
genes are regulated in a cell-density-dependent
manner that requires the accumulation of a
small-molecular-weight autoinducing metabolite,
N-(3-oxohexanoyl)homoserine lactone (see Fig. 2 for
structure), in combination with the luxR activator <.
Recently, the phzA gene, a homologue of luxR, was
shown to be required for the regulation ofphenazine
biosynthesis (Fig. 4). The production ofphenazines is
also enhanced by the presence of a small-molecularweight compound isolated from culture supernatants
of the producing strain62. Such a cell-densitydependent requirement for efficient production of
antimicrobial metabolites may explain the apparent
necessity for efficient root- or seed-colonization by
the producer strain for efficient and reproducible
biocontrol in the field.
The luxR/I type of regulatory system has been
implicated in the control of a range of important
microbial processes, including plasmid transfer66,
antibiotic 67,68 and exoenzyme 69,7° production. Exoenzymes such as chitinases may be useful factors in the
biological control of pests, including some chitincontaining fungi, nematodes and insects. It will be of
interest to determine if such enzymes are also under
the control of autoinducing metabolites.
The presence of autoinducing metabolites and
pheromone-mediated regulation is a new and exciting development that may prove to be of major
importance in understanding and manipulating mechanism(s) of biocontr01.
Future research d i r e c t i o n s
The major bottle-neck to the commercial use of
microbe-based biocontrol agents is their unpre-
Q
® []/~
®
®
Fe2+
/~
'
I
fur
I
[]
Antifungal
metabolites
~
®
[]
~_+_~
_l
gacA
~-~-g~
pit
D
I
I
prfA
--I
anr
~
' pbrA
m
--
+
I pbsC
,-.
•1
.pbuA
,,,
pupR pffpB
Fe2+
el
/
O
~CN
/
Lional
receptor
,I
Fluorescent " - . .
siderophores
0
I
i
~+
\\
+ \., h c n - - - -
~
Heterologous
ferric-siderophores
Figure 4
Model of the regulatory circuits controllingthe production of metabolites important
in biological control by fluorescent Pseudomonas spp. Many strains produce one or
more antifungal metabolites. Genes responsible for the synthesis of these compounds, e.g. phenazines (phz), phloroglucinols (phi), pyrroles (pit) are coordinately
regulated in a positive fashion by the Global Antibiotic and Cyanide control gene product, GacA (Ref. 60). This is thought to be the regulator part of a two-component system of which the sensor (and signal) remain to be identified. The phzA gene is a homologue of luxR and is responsible for the regulation of the phenazine, PCA, in
combinationwith supernatant extract62. Cyanogenesis is positively regulated by both
the GacA protein and the Anaerobic Regulator, Anr (analogous to Fnr)6°,63. Synthesis of fluorescent siderophores are negativelyregulated by a ferric-uptake regulator Fur and Fe II co-factor% and transcriptionally activated by the product of the
pbrA gene (D. N. Dowling and F. O'Gara, unpublished). Another gene product, PrfA,
a homologue of AIgQ is also required for transcriptional activation and may function
in signal transduction sT. These regulatoryelements act on genes coding for biosynthesis of the siderophore, e,g. pbsC (Ref. 64), and genes involved in uptake of the
ferric-siderophore complex from the environment, e.g. pbuA (Ref. 65). Some strains
possess additional receptors and regulatory elements, e.g. PupB/PupR (Ref. 37)
that allow the utilization of heterologous ferric-siderophore complexes. These
siderophores act as inducers for their own uptake system.
dictability in the field, relative to conventional chemical treatments. This may be due to a variety of
reasons, including abiotic factors, poor and inconsistent colonization, and the failure to produce the
metabolite(s) at the appropriate time or levels.
Laboratory studies are increasing our knowledge
of the environmental factors that may influence the
production of metabolite(s) at the appropriate
microsites in the rhizosphere; however, a large gap
remains in our understanding of the interactions
between rhizosphere exudates and microbial metabolite production. The appropriate temporal and
spatial regulation of metabolite(s) synthesis by the
TIBTECHAPRIL 1994 (VOL 12)
140
reviews
biocontrol agent is essential tO improve the effectiveness o f biological control. Recently, sensor bacteria
with reporter genes fused to the promoters o f antibiotic synthesis genes have been used to measure
transcription of these genes in situ on the seed coat 6s.
The extensive use of appropriate sensor bacteria with
suitable reporter genes (e.g. lux, luc, ice, lacZ) fused
to the promoters of genes included in the synthesis of
antimicrobial metabolites will allow researchers to
monitor precisely the expression o f these genes in situ,
and will complement the direct measurement of
metabolites in soils and rhizospheres. However, the
final goal will be to regulate and fine-tune metabolite
synthesis such that the appropriate biocontrol genes
are expressed at the required level by a predetermined
environmental signal. This signal could be derived
from some appropriate environmental conditions
during crop cultivation, such as soil-nutrient level,
water potential or temperature, which are known to
favour the pathogen or disease. A more elegant 'trigger' for the expression of biocontrol genes in an
introduced inoculant strain would be for the genes
to be activated directly, or indirectly, by the presence
of the specific pathogen in the rhizosphere.
Conclusions
The evidence that some microbial metabolites play
a key role in the biological control o f certain key plant
pathogens provides a strong incentive to develop such
material as biocontrol agents for commercial use. The
use of microbial strains producing antibiotic metabolites in controlled environments, such as horticulture
in greenhouses, offers the best possibility for success
in the short term. The major challenges to be resolved
prior to widespread commercial exploitation of biocontrol strains lies in the ability to predict more confidently the behaviour o f such strains in the field.
Fortunately, as interest in these organisms grows, more
information on the genetics, physiology and ecology
ofmetabolite production is becoming available. Such
data are of immense importance for the selection of
wild-type strains with desirable traits from nature and
to provide a more rational framework for the choice
of strains for use in inoculant consortia. Recombinant
D N A methods have enabled genetic manipulation of
metabolite production with promising results. In this
context, research is also directed towards the evaluation o f possible risks associated with the large-scale
release of these genetically modified organisms
(GMOs). Based on the knowledge that is now being
accumulated, particularly in the area o f molecular
microbial ecology of the rhizosphere and metabolite
regulation, improved or manipulated biocontrol
agents should become more predictable and reliable
for use under field conditions.
Acknowledgements
We should like to thank the members of the
Pseudomonas biocontrol group at UCC, especially
Anne Fenton, Ray Sexton and Paul Gill, Jr for the
inclusion of unpublished data. Work in the authors'
IBTECHAPRIL 1994 (VOL 12)
laboratory was supported, in part, by contracts fi'om
the European Community (ECLAIR AGILE 0019C; B R I D G E BIOT-CT90-0166-C, B I O T - C T 9 1 0293, BIOT-CT91-0283; FLAIR AGP,.E-0019-C;
B I O T E C H BIO2-CT93-0053, BIO2-CT93-0196,
BIO2-CT92-0084).
References
1 Weller, D. M. (1988) Atom. Rev. Phytapathol. 26, 379-407
2 Cook, R.J. and Weller, D. M. (1986) in Im,watiueApp~,aches to Plant
Disease Control (Chet, I., ed.), pp. 41 76, John Wiley & Sons
3 Loper, J. E. and Buyer,J. S. (1991) Mol. PIantMicrabe I, teracl. 4, 5-13
4 O'Sullivan, D.J. and O'Gara, F. (1992) BacterM. Rev. 56, 662-676
5 Cook, R.J. (1993) Annu. Rev. Phytopathol. 31, 53-80
6 Voisard, C. et al. in Malecular Ecolo,W ~" Rhizosphere Microo~;¢anisms
(O'Gara, F., Dowling, 1). N. and Boesten, B., eds) (in press)
7 Weller, D. M. and Thomashow, L. S. in Molecular Ecol~w o/'Rhi~.o~here Miovo~,anisms (O'Gara, F., Dowling, D. N. and Boesten, B.,
eds) VCH (in press)
8 Defago, G. and Haas, D. (1990) in Sail Biochemistry Vol. 6 (Bollag,
J-M. and Stozky, G., eds), pp.249~91, Marcel Dekker
9 Thonrashow, L. S. and Weller, D. M. (1988) J. Bacteriol. 170,
3499-3508
10 Keel, C. et al. (1992) Mol. Plant Mien,be Interact. 5, 4 13
11 Vincent, M. N. et al. (1991) AppI. Enviran. Microbiol. 57, 2928~934
12 Pfender, W. F., Kraus, J. and Loper, J. E. (1993) Phytopathalaw 83,
1223-1228
13 Gutterson, N. (1990) Oft. Rev. Biotechnol. 10, 69 9/
14 Howell, C. R. and Stipanovic, R. D. (1980) Ph)nopathology 70,
712-715
15 Shanahan, P., O'Sullivan, D. J., Simpson, P., Glennon, J. D. and
O'Gara, F. (1992) AppI. Environ. Mirn+ial. 58, 353-358
16 Voisard, C., Keel, C., Haas, D. and Defago, G. (1989) EMBOJ.
8, 351-358
17 Clare, B. G. (i993) in Biochemistry and Genetics of Antibiotic Biosynthesis (Vining, L. C. and Stuttard, C., eds), pp. 71-83,
Butterworth Heinemann
18 Kerr, A. (1989) Aui. ScL 2, 41-44
19 Jones, D. A., Ryder, M. H., Clare, B. G., Farrand, S. K. and Kerr,
A. (1988) Mol. Gen. Genet. 212, 207 214
20 Kloepper, J. W., Leong, J., Teintze, T. and Schroth, M. N. (1980)
Nature 286, 885-886
21 Howell, C. R., Beier, R. C. and Stipanovic, R. D. (1988) Philopathology 78, 1075-1078
22 Leisinger, T. and Margraff, R. (1979) Microbiol. Rev. 43, 422-442
23 Kloepper, J. W. and Schroth, M. N. (1981) Phytapathology 71,
1020-1024
24 Bull, C. T., Weller, D. M. and Thomashow, L. S. (1991) Phytopathology 81,954-959
25 Vining, L. C. (1990) Annu. Reu. Microbiol. 44, 395 427
26 Mazzola, M., Cook, R. J., Tbomashow, L. S., Weller, D. M. and
Pierson, L. S., III (1992) Appl. Environ. Micrabiol. 58, 2616-2624
27 Sequeira, L. (1979) in Recognitioi~ and Specificity in Plant Hast-Parasite
Interactions (Daly, J. M. and Uritani, I., eds), pp. 231-251, University
Park Press
28 van Peer, R., Nieman, G.J., Schippers, B. and van Peer, R. (1991)
Phytapatholagy 81,728-734
29 Smith, J. A., Hammerschmidt, IL. and Fulbright, D. W. (1991)
Physiol. Mol. Plant Pathol. 38, 223-235
30 Neilands, J. B. (1984) Micrabiol. Sci. 1, 9 14
31 Powell, P. E., Cline, G. R., Reid, C. P. P. and Szaniszlo, P.J. (1980)
Nature 287, 833-834
32 Powell, P. E., Szaniszlo, P. J. and Reid, C. P. P. (1993) Appl.
Environ. Microbiol. 46, 1080-1083
33 Bakker, P. A. H. M., Laurers, J. G., Bakker, A. W., Marugg, J. D.,
Weisbeek, P. J. and Schippers, B. (1986) Neth. 1. Plant Pathol. 92,
249-256
34 Bakker, A. W. and 5chippers, B. (1987) Soil Biol. BiochenL 19,451-470
35 Leong, J. (1986) Altair. Rev. Phytopathol. 24, 187-209
141
reviews
36 Marugg, J. D. et al. (1989)J. BacterioI. 171, 2819-2826
37 Koster, M., van de Vossenberg, J., Leong, J. and Weisbeek, P. J.
(1993) Mol. Microbiot. 8, 591 600
38 Morris, J., O'Sullivan, D.J., Koster, M., Leong, J., Weisbeek, P. and
O'Gara, F. (1992) Appl. Era, iron. Microbiol. 58, 630-635
39 Dowling, D. N., Boesten, B., O'Sullivan, D.J., Stephens, P., Morns,
J. and O'Gara, F. (1992) in Pseudomonas: Molecular Biology and
Biotechnology (Galli, E., Silver, 8. and Witholt, B., eds), pp. 408-414,
American Society for Microbiology
40 Dowling, D. N., Boesten, B., Gill, P. R., Jr and O'Gara, F. in
Molecular Ecology ~f Rhizosphere Microorganisms (O'Gara, F., Dowling,
D. N. and Boesten, 13., eds) VCH (in press)
41 Lang, L., Breimholt, F. W., Rasmussen, R. I. and Nielsen, R. I.
(1993) Pesticide Sci. 39, 155-160
42 Thomashow, L. S., Weller, D. M., Bonsall, R. F. and Pierson,
L. S., IIl (1990) AppL Environ. Microbiol. 56, 908-912
43 Ankenbauer, R. G. and Cox, C. D. (i988)J. Bacteriol. 170, 5364-5367
44 Visca, P., Ciervo, A., Sanfilippo, V. and Orsi, N. (1993)/. Gen.
MicrobioI. 139, 1995-2001
45 Raskin, I. (1992) Annu. Rev. Plant Mol. Biol. 43, 43%463
46 Gaffney, T. et al. (1993) Science 261,754-756
47 Loper, J. E. and Schroth, M. N. (1986) Phytopathology 76,
386-389
48 Michiels, K., Vanderleyden, J. and Van Gool, A. (1989) Biol. Fertil.
Soils 8, 356 368
49 Sliniger, P. J. and Jackson, M. A. (1992) AppI. MicrobioL Biotechnol.
37, 388-392
50 Gutterson, N., Ziegle, J. S., Warren, G.J. and Layton, T.J. (1992)
J. BacterioL 170, 380-385
51 Weller, D. M. and Cook, R.J. (1983) Phytopathology 73, 463-469
52 O'Sullivan, D. J. and O'Gara, F. (1991) in The Rhizosphere and Plant
Growth (Keister, D. L. and Cregan, P. B., eds), p. 310, Kluwer
Academic Publishers
53 Maurhofer, M., Keel, C., Schnider, U., Voisard, C., Haas, D. and
Defago, D. (1992) Phytopathology 82, 190-195
54 Fenton, A. M., Stephens, P. M., Crowley, L., O'Callaghan, M.
and O'Gara, F. (1992) Appl. Environ. Microbiol. 58, 3873-3878
55 Prince, R. W., Cox, C. D. and Vasil, M. L. (1993) J. BacterioI. 175,
1589 1598
56 O'Sullivan, D. J. and O'Gara, F. (1990) Mol. Plant Mio'obe interact. 3,
86-93
57 Venturi, V., Ottevanger, C., Leong, J. and Weisbeek, P. J. (1993)
Mol. Microbiol. 10, 63-73
58 Shanahan, P., Glennon, J. D., Crowley, J. J., Donnelly, D. F. and
O'Gara, F. (1993) Analytica Ckimica Acta 272, 271-277
59 Stone, M.J. and Williams, D. H. (1992) Mol. Microbiol. 6, 29-34
60 Laville, J., Voisard, C., Keel, C., Maurhof'er, M., Defago, G. and
Haas, D. (1992) Proc. Natl Acad. Sci. USA 89, 1562-1566
61 Silvm~nan, M., Martin, M. and Engebrecht, J. (1989) in Genetics of
Bacterial Diversity (Hopwood, D. A. and Chater, K. F., eds),
pp. 71 85, Academic Press
62 Pierson, L. S., III (1993) in Pseudomonas i993; Fourth International
Symposium on Pseudomonas: Biotechnology and Molecular Biology,
p. 218
63 Zimmermann, A., Reimmann, C., Galimand, M. and Haas, D.
(1991) Mol. Microbiol. 5, 1483 1490
64 Adalns, C., Dowling, D. N., O'Sullivan, D. J. and O'Gara, 1a. MaL
Gen. Genet. (in press)
65 Morris, J , Donnelly, D. F., O'Neill, E., McConnell, F. and O'Gara,
F. (1994) Mol. Gem Genet. 242, 9-16
66 Zhang, L., Murphy, P. J., Kerr, A. and Tate, M. E. (1993) Nature
362, 446-448
67 Bainton, N.J. et al. (1992) Biochem. I. 288, 997-1004
68 Horinouchi, S. and Beppu, T. (1990) Crit. Rev. Biotechnol. 10,
191~04
69 Jones, S. et al. (1993) EMBOJ. 12, 2477~482
70 Passador, L., Cook, J. M., Gambello, M. J., Rust, L. and Igleweski,
B. H. (1993) Science 260, 1127-1130
71 Howie, W. J. and Suslow, T. V. (1991) Mol. Plant Microbe Interact.
4, 393-399
Errata
We apologize to the authors, and to the readers who may have been misled by the publication errors which
appeared in the article "Retinal-protein complexes as optoelectronic components" by N. N. Vsevolodov
and T. V. Dyukova, published in the March 1994 issue of TIBTECH. The corrected information appears
below:
Table
2.T h e correct structure and information for 4-keto retinal is:
Retinal and
its analogs
4-Keto
retinal
Absorption spectra
for ground (-) and
photoinduced (- -)
BR forms
Structural
formula
of retinals
410
~/x~x0
510
8
, &
=~ ,,. ~:
0
Half life
time for
photoinduced
forms
From
minutes
to tens
of minutes
Wavelength (nm)
Table 3. T h e correct reference for senso W rhodopsins SR-1 and S R - 2 should have been R e £ 26 [SpLidich,
J. L. and
Bogomolni, R. A. (1992)/. Bioenergetics Biomembratles 24, 193-200].
R e f e r e n c e s : T h e correct page numbers for R e £ 4 are Oesterhelt, D. and Stoeckenius, W. (197 i) Nature N e w Biol. 223, 149-152.
TIBTECHAPRIL1994 (VOL121