الاثنين، 13 أبريل 2009

The Family Streptomycetaceae, Part I: Taxonomy

Phylogeny and Taxonomy
The family Streptomycetaceae was created by
Waksman and Henrici (1943). Originally this
family harbored only the type genus Streptomyces.
Zhang et al. (1997) proposed that the genus
Kitasatospora be included, and recently, a third
genus, Streptacidiphilus, was added (Kim et al.,
2003).
Description of the family Streptomycetaceae
Waksman and Henrici 1943 emend, Kim et al.
(2003) (Strep.to.my.ce.ta’ce.ae. ending to denote
a family; M.L. masc. n. Streptomyces, type genus
of the family) is based on data taken from Williams
et al. (1989), Zhang et al. (1997) and Kim
et al. (2003). These aerobic, Gram-positive, nonacid-
alcohol fast actinomycetes form an extensively
branched substrate mycelium that rarely
fragments. The aerial mycelium forms chains of
three to many spores. Members of a few species
bear short chains of spores on the substrate
mycelium. The organisms produce a wide range
of pigments responsible for the color of the
substrate and aerial mycelium. The organisms
grow within different pH ranges, namely 5.5–9
(Kitasatospora), 5–11.5 (Streptomyces), and 3.5–
6.0 (Streptacidiphilus). They are chemoorganotrophic
with an oxidative type of metabolism.
The substrate mycelium contains either LL-
(Streptacidiphilus and Streptomyces) or meso-
(Kitasatospora) diaminopimelic acid as the
predominant diamino acid; aerial or submerged
spores contain LL-diaminopimelic acid. In
whole-organism sugar profiles, either major
amounts of galactose or galactose and rhamnose
(Kitasatospora and Streptacidiphilus) can be
detected. Lipid profiles typically contain hexaand
octa-hydrogenated menaquinones with nine
isoprene units as the predominant isoprenologues.
The polar lipid profiles are composed
of diphosphatidylglycerol, phosphatidylethanolamine,
phosphatidylinositol, and phosphatidylinositol
mannosides. Fatty acids are complex
mixtures of saturated, iso- and anteiso-fatty
acids. Mycolic acids are not present.
The mol% G + C of the DNA ranges generally between 66
and 74%. Members of all three taxa are widely
distributed in terrestrial habitats, especially soil.
Very few species are pathogens for animals
(including man) and plants.
A phylogenetic tree showing selected representatives
of all three genera (all species of Streptacidiphilus
and Kitasatospora and selected
Streptomyces “species”) representing the clusters
of the numerical taxonomic study of Williams et
al. (1983a) is shown in Fig. 1. The genera are
difficult to differentiate on the basis of phenotypic
features (including chemotaxonomic markers).
Some characteristic features are shown in Table 1.
History
Early investigations of actinomycetes, including
streptomycetes, were dominated by a strong
emphasis of morphology and the high degree of
morphological diversity was subsequently considered
to be sufficient for their assignment to
genera and families (Waksman, 1961; Cross and
Goodfellow, 1973). A short summary of early
classification systems of actinomycetes is given in
Introduction to the Classification of the Actinomyces
in this Volume. Streptomycetes are the
producers of more than 5000 known bioactive
compounds (Anderson and Wellington, 2001),
and estimates of the total number of antimicrobial
compounds produced by representatives of
Streptomyces screened for new antibiotics are of
the order of 100,000 (Watve et al., 2001). In addition,
not only has the overall versatility of these
compounds been studied in great detail, but also
a high proportion of them have known biological
effects, which is unparalleled in the living world
(Kieser et al., 2000).
The family Streptomycetaceae was originally
proposed by Waksman and Henrici (1943) and
contained at that time only two genera: the genus
Streptomyces and the genus Micromonospora.
Streptomyces was described as “Streptomycetaceae,”
forming spores in chains on aerial hyphae.
Spores are apparantly endogeneous in origin,
formed by a segregation of protoplasm within
the hyphae into a series of round oval or cylindrical
bodies. Chains of spores are often spirally
coiled. Sporophores may be simple or branched
(Waksman and Henrici, 1943). Figure 2 shows
the morphology of the aerial mycelium of
three streptomycetes. Although they are important
(see Table 2 of The Family Nocardiopsaceae
in this Volume), morphological differences
between members of the actinomycete genera do
not represent the extensive diversity of genera
and species of the sporoactinomycetes.
Beginning in the 1950s, new developments in
the application of numerical phenetics, numerical
taxonomy, chemosystematics, and finally
molecular systematics revolutionized the classification
of actinomycetes. These developments
have been excellently reviewed for actinomycete
systematics by Goodfellow et al. (1999).
The determination of major cell wall sugars
and peptidoglycan composition (Lechevalier
and Lechevalier, 1970; Schleifer and Kandler,
1972) led to a classification system of wellcharacterized
chemotypes and peptidoglycan
types. The pioneering work of Lechevalier and
coworkers (Becker et al., 1964; Lechevalier and
Lechevalier, 1970) clearly showed that Streptomyces
and the other genera of the family Strep-
tomycetaceae, proposed by then, contained LLdiaminopimelic
acid (LL-A2pm) in its peptidoglycan
(cell wall type I), whereas meso-A2pm
was found in most of the other actinomycetes
described at that time. The genera containing
LL-A2pm in their peptidoglycan contained an
interpeptide bridge composed of a glycine residue
(type A3γ of Schleifer and Kandler, 1972).
In addition to these chemical traits, which had
the advantage of higher genetic stability in
comparison with morphological features, the
pattern of sugars in whole-cell hydrolysates
(Lechevalier and Lechevalier, 1970), phospholipids
(Lechevalier et al., 1977a), fatty acids
(Kroppenstedt, 1985), menaquinones (Alderson
et al., 1985; Kroppenstedt, 1985), and acetylated
muramic acid residues (Uchida and Aida, 1977)
were shown to be of essential importance for the
classification of actinomycetes. Major chemotaxonomic
markers of the genus Streptomyces and
other actinomycete genera are given in Table 2.
In combination with other phenotypic properties,
like physiological and biochemical characteristics,
these traits were also helpful in defining
genera within the family Streptomycetaceae.
This resulted in the reclassification of six
additional genera (Actinopycnidium, Actinosporangium,
Chainia, Elytrosporangium, Kitasatoa
and Microellobosporia), described mainly on the
basis of morphological features, to the genus
Streptomyces (Williams et al., 1983a; Goodfellow
et al., 1986b; Goodfellow et al., 1986c; Goodfellow
et al., 1986d; Goodfellow et al., 1986e).
The application of 16S rRNA oligonucleotide
cataloguing (Stackebrandt and Woese, 1981) and
subsequently the sequencing of the 16S rRNA
genes provided a basis for studies of the natural
relationships among actinomycetes and related
organisms (for details, see Stackebrandt et al.,
[1997] and Introduction to the Classification of
the Actinomyces in this Volume). On the basis of
these data, the description of the family Streptomycetaceae
was emended by Wellington et al.
(1992) and Witt and Stackebrandt (1990), who
proposed the unification of the genera Kitasatosporia
and Streptoverticillium with the genus
Streptomyces, and more recently by Stackebrandt
et al. (1997), who excluded the genus
Sporichthya. Zhang et al. (1997) demonstrated,
however, that the genus Kitasatosporia formed
a stable subbranch in Streptomyces, when
sequences from the almost complete 16S rRNA
genes were compared. In addition, members of
the genus Kitasatosporia can be distinguished
from Streptomyces by the ratio of meso-DAP to
LL-DAP and the presence of galactose in wholecell
hydrolysates (Zhang et al., 1997; Table 1).
The genera Kineosporia and Sporichthya, both
sharing chemotaxonomic similarities with members
of the genus Streptomyces and considered
to be members of this genus (Logan, 1994), have
been shown by 16S rRNA sequencing to be independent
genera: Sporichthya is a member of the
family Sporichthyaceae of the suborder Frankineae
(Stackebrandt et al., 1997), and the genus
Kineosporia is grouped together with Kineococcus
(Kudo et al., 1998) into the tentative family
“Kineococcaceae” (see Introduction to the Classification
of the Actinomyces in this Volume).
Recently the genus Streptacidiphilus has been
proposed by Kim et al. (2003) to accommodate
acidophilic actinomycetes forming a distinct
clade within the family Streptomycetaceae.
Note that although 16S rRNA sequence analyses
have provided a framework for prokaryotic
classification, the current classification system
based on this molecule has not yet solved the
taxonomic problems within the genera (especially
within the genus Streptomyces). Several
studies have attempted to use sequence data
from variable regions of 16S rRNA to establish
taxonomic structure within the genus, but the
variation is too limited to resolve problems of
species differentiation (see Witt and Stackebrandt
[1990], Stackebrandt et al., [1991], Stackebrandt
et al., [1992], Anderson and Wellington
[2001], and the references therein ).
The discovery of antibiotics produced by
streptomycetes in the 1940s, which led to extensive
screening for novel bioactive compounds,
and the subsequent need for patenting, which
led to an extreme overclassification of the genus,
complicated the situation. Producers of novel
natural products were described as new species
and patented. Species described within the
genus Streptomyces increased from approximately
40 to over 3000 (Trejo, 1970). The current
status of streptomycete taxonomy including phylogeny
has been summarized by Anderson and
Wellington (2001) and will be treated briefly in
the next sections. Of the 539 species and subspecies
listed under the List of Bacterial Names
with Standing in Nomenclature as of December
9, 2003, 376 are on the Approved lists (Tables 3
and 4).
Genus Kitasatospora
Zhang et al. (1997) revived the genus Kitasatospora
to accommodate actinomycete strains
forming a stable, separate subbranch on the basis
of phylogenetic analyses within the family Streptomycetaceae
and containing major amounts of
meso-DAP in their whole-cell hydrolysates. Phylogenetic
trees were also constructed by using
16S-23S rRNA gene spacers, leading to groupings
similar to those based on 16S rRNA
sequence data (Zhang et al., 1997).
The substrate mycelium of members of Kitasatospora
is as well developed as the Streptomy-
ces substrate mycelium. The aerial mycelium
bears long spore chains with more than 20
spores. Galactose is present in whole-cell
hydrolysates of Kitasatospora. Specific nucleotide
signatures in the sequences of both 16S
rRNA and 16S-23S rRNA gene spacers can differentiate
Kitasatospora from Streptomyces (for
details, see Zhang et al., 1997); however, phenotypic
differences between Streptomyces and Kitasatospora
are not pronounced so that the
separate genus status of Kitasatospora may be
questioned.
To date, eleven species of the genus Kitasatospora
have been recognized: Kitasatospora
setae (Omura et al. 1982), Kitasatospora phosalacinea
(Takahashi et al., 1984a), Kitasatospora
griseola (Takahashi et al., 1984a), Kitasatospora
mediocidica (Labeda, 1988), Kitasatospora
cystarginea (Kusakabe and Isono, 1988), Kitasatospora
cochleata (Nakagaito et al., 1992b;
Zhang et al., 1997), Kitasalospora paracochleata
(Nakagaito et al., 1992b; Zhang et al., 1997),
Kitasatospora azatica (Nakagaito et al., 1992b;
Zhang et al., 1997), Kitasalospora cheerisanensis
(Chung et al., 1999), Kitasatospora cineracea
(Tajima et al., 2001) and Kitasatospora niigatensis
(Tajima et al., 2001).
The designations of some species to this genus
are open to discussion. Kitasatospora cystarginea,
Kitasatospora griseola, Kitasatospora
mediocidica, Kitasatospora phosalacinea and
Kitasatospora setae are synonyms of Streptomyces
cystargineus, Streptomyces griseolosporeus,
Streptomyces mediocidicus, Streptomyces phosalacineus
and Streptomyces setae, respectively.
For these species, and according to scientific
opinion, an author may use Kitasatospora or
Streptomyces. See the List of Bacterial Names
with Standing in Nomenclature for detailed
comments.
Genus Streptacidophilus
The genus Streptacidophilus was proposed by
Kim et al. (2003) to accommodate acidophilic
actinomycetes isolated from acidic soils and litter.
On the basis of 16S rRNA sequence analysis, it
could be shown that the 11 isolates formed a
stable clade within the family Streptomycetaceae.
These organisms showed a distinctive pH profile,
showed a unique 16S rDNA signature, and contained
major amounts of LL-diaminopimelic
acid, galactose and rhamnose in whole-cell
hydrolysates (Kim et al., 2003). The members of
the genus form an extensively branched, nonfragmenting
mycelium carrying long chains of spores
in aerial mycelia at maturity (Kim et al., 2003).
To date, three species have been recognized:
Streptacidophilus albus, Streptacidophilus neutrinimicus
and Streptacidophilus carbonis.
Similar to Kitasatospora, Streptomyces and
Streptacidophilus have no pronounced phenotypic
differences, so that a separate genus status
also of Streptacidophilus may be questioned.
Genus Streptomyces
The genus Streptomyces Waksman and Henrici
(1943) is the type genus of the family. Most of
the general characteristics described below also
apply to members of the genera Kitasatospora
and Streptacidiphilus, unless stated otherwise.
General Characteristics Streptomycetes are
Gram-positive aerobic members of the order
Actinomycetales within the class Actinobacteria
(Stackebrandt et al., 1997) and have a DNA G+C
content of 69 ± 78 mol%. The vegetative hyphae
(0.5–2.0 μm in diameter) produce an extensively
branched mycelium that rarely fragments. The
aerial mycelium at maturity forms chains of
three to many spores. Some species may bear
short chains of spores on the substrate mycelium.
Sclerotia, pycnidial-, sporangia-, and synnematalike
structures may be formed by some species.
The spores are nonmotile. On complex agar
media, discrete and lichenoid, leathery or
butyrous colonies are formed. Colonies are initially
relatively smooth surfaced, but later they
develop an aerial mycelium that may appear
floccose, granular, powdery or velvety.
Members of the genus Streptomyces undergo
a complex life cyle, which has been studied most
intensively for strain “S. coelicolor” A2(3). Streptomyces
colonies are multicellular, differentiated
organisms exhibiting temporal and spatial control
of gene expression, morphogenesis, metabolism
and the flux of metabolites (see chapter 2 of
Kieser et al., [2000] for more details).
Strains belonging to the genus Streptomyces
may produce a wide variety of pigments responsible
for the color of the vegetative and aerial
mycelia (Figs. 3 and 4). In addition, colored diffusible
pigments may also be formed. Note that
the production of pigments largely depends on
the medium composition and cultivation conditions
(Figs. 3 and 4). Many strains produce one
or more antibiotics (more details are given in
The Family Streptomycetaceae, Part II: Moleular
Biology in this Volume). The metabolism is oxidative
and chemoorganotrophic. The catalase
reaction is positive, and generally, nitrates are
reduced to nitrites. Most representatives can
degrade polymeric substrates like casein, gelatin,
hypoxanthine, starch and also cellulose. In addition,
a wide range of organic compounds is used
as sole sources of carbon for energy and growth
(Williams et al., 1983a; Kämpfer et al., 1991b;
Korn-Wendisch and Kutzner, 1992a). The optimum
temperature for most species is 25–35°C;
however, several thermophilic and psychrophilic
species are known. The optimum pH range for
growth is 6.5–8.0.
The Embden-Meyerhof-Parnas (glycolysis)
pathway of glucose catabolism has been found in
many streptomycetes (Cochrane, 1961), but also
the hexose monophosphate shunt (Salas et al.,
1984) was detected in S. antobioticus. Several
streptomycetes are able to switch from glycolysis
to the hexose monophosphate shunt during
secondary metabolism (Kieser et al., 2000). At
present, no streptomycete is known to use the
Entner-Doudoroff pathway. Sugar transport is
mediated in connection with phosphorylation by
specific kinases (Sabater et al., 1972; lkeda et al.,
1984). The phosphoenolpyruvate:fructose phosphotransferase
system (PTS) for the transport
and phosphorylation of fructose has recently
been detected in S. coelicolor, S. lividans and
S. griseofuscus (Titgemeier et al., 1995). More
details about specific metabolic pathways,
including nitrogen metabolism and the regulation
processes involved, are given in chapter 1 of
Kieser et al. (2000) and the references therein.
On the basis of 16S rRNA/DNA sequence
comparisons, members of the genus Streptomyces
form a separate line of descent, and Stackebrandt
et al. (1997) proposed the emendation of the
family Streptomycetaceae in the suborder Streptomycinae
and the order Actinomycetales. The
intrageneric phylogenetic relationships of many
of the 346 recognized species in Bergey’s Manual
of Systematic Bacteriology (Williams et al., 1989)
inferred from the 350 complete 16S rRNA
sequences, however, are clearly restricted by the
limited resolving power of the method to discriminate
between related species and are often in
contrast with a morphologically and physiologically
based classification. Though about 350
almost complete 16S rRNA sequences are available
to date, the high degree of conservation
within 16S rRNA genes causes problems for
resolving phylogenetic relationships at the intergeneric
level.
Notably, the different methods used for grouping
of the Streptomyces species often lead
to contradictory results. In Table 4, all 376
Streptomyces species and subspecies with valid
names (as of December 9, 2003; taken from the
List of Bacterial Names with Standing in
Nomenclature); and some additional species
with names not validly published (but included
in taxonomic studies) are given with their grouping
according to different studies.
The chemotaxonomic features for the identification
of strains at the genus level (for details see
below) are of high value and can be summarized
as follows: The cell wall peptidoglycan contains
major amounts of LL-diaminopimelic acid
(LL-A2pm). Genus members lack mycolic acids,
contain major amounts of saturated, iso- and
anteiso-fatty acids, possess either hexa- or
octahydrogenated menaquinones with nine isoprene
units as the predominant isoprenolog, and
have complex polar lipid patterns that typically
contain diphosphatidylglycerol, phosphatidylethanolamine,
phosphatidylinositol, and phosphatidylinositol
mannosides (Table 2; Figs. 5 and
6). In addition to these traits, the acyl type of the
muramyl residues in the cell-wall peptidoglycans
is acetyl (Uchida and Seino, 1997). Strains are
widely distributed and abundant in soil, including
composts (see detailed description below). A
few species are pathogenic for animals and man,
and others are phytopathogens. The type species
is Streptomyces albus (Rossi-Doria 1891) Waksman
and Henrici (1943).
Cell Wall Composition
Peptidoglycan The cell walls of streptomycetes
show the typical ultrastructure and chemical
composition of Gram-positive bacteria
(Schleifer and Kandler, 1972). They appear
under the electron microscope as homogeneous
less electron dense layers of about 16–35 nm.
The cell walls have a multilayered structure of
peptidoglycan strands. The peptidoglycan is a
heteropolymer consisting of heteropolysaccharide
chains cross-linked through short peptide
units. The so-called “sugar back bone” of the
peptidoglycan is constructed of alternating β-
1,4-linked units of N-acetylglucosamine and Nacetylmuramic
acid. The carboxyl group of
muramic acid is substituted by an oligopeptide
of alternating D- and L-amino acids (Schleifer
and Kandler, 1972). Streptomyces is characterized
by the tetrapeptide L-Ala–D-Glu–LLA2pm–
D-Ala. This tetrapeptide is crosslinked by
a pentaglycine bridge which extends from the Cterminal
D-alanine of the peptide unit to the
amino group located on the D carbon of LLA2pm,
resulting in the macromolecule structure
forming the cell envelope. This LL-A2pm-Gly5,
or A3γ peptidoglycan type (Schleifer and
Kandler, 1972), is diagnostic for streptomycetes
and some other combined-wall chemotype I actinomycetes
(Lechevalier and Lechevalier, 1970).
The aerobic actinomycetes were grouped
(using specific amino acids in purified cell walls)
into four so-called “wall chemotypes” by
Lechevalier and coworkers. Cell walls with
meso-DAP and LL-DAP were detected early. In
another study, Takahashi et al. (1984b) reported
that strains belonging to this group change cell
wall composition during sporulation. They found
in submerged mycelium LL-DAP and glycine
(wall chemotype I) whereas in spores, only meso-
DAP could be detected (wall chemotype III
according to Lechevalier and Lechevalier, 1970).
In 11 streptomycetes, the cell wall compositions
of aerial, substrate and submerged mycelium
differed in the quantitative distribution of
cell wall amino acids and cell wall sugars. NAcetylmuramic
acid is found in the glycolyl type
of cell wall of Streptomyces, as in all other actinomycetes
(Uchida and Aida, 1977). Muramic
acid phosphate residues are the attachment
points to teichoic acids, which are of diagnostic
value for Gram-positive bacteria. The cell wall
teichoic acids (polymeric substances containing
repeating phosphodiester groups) consist of
polyols (i.e., the sugar alcohols glycerol and ribitol)
or N-acetylamino sugars or both. The
teichoic acids of streptomyces are of the same
structure as those of other Gram-positive bacteria,
containing either ribitol phosphate or
glycerol phosphate polymers; significantly, the
teichoic acid of actinomycetes does not contain
ester-bound D-alanine but does have esterlinked
acetic acid and sometimes succinic acid
residues (Naumova et al., 1980).
For streptomycetes, the synthesis of either ribitol
phosphate (S. streptomycinii and S. violaceus)
or glycerol phosphate polymers (S. thermovulgaris,
S. levoris, S. rimosus and S. antibioticus) has
been reported (Naumova et al., 1980). In ribitol
teichoic acids, positions 1 and 5 of ribitols are
connected to the phosphates; in glycerol teichoic
acids, position 1 is commonly connected to 3; and
in other types, position 1 connected to 2 (as in S.
antibioticus) is not common. The polyol phosphates
can be substituted with various combinations
of sugars or amino sugars or both, which
are linked to glycerols or ribitols via glycosidic
bonds. At present, few strains (species) have
been investigated in detail and so the role of
teichoic acid in the taxonomy of Streptomyces is
not clear (Naumova et al., 1980).
Cell Wall Polysaccharides These compounds
seem to be of no diagnostic value in those strains
where LL-DAP is found in whole cell hydrolysates
(Lechevalier et al., 1971). Some of the diagnostic
sugars found in other actinomycetes, like
xylose, galactose and arabinose, were reported
occasionally in streptomycetes. Hundreds of
streptomyces were analyzed for the presence
of diagnostic sugars (Kroppenstedt, 1977), and
mainly ribose, mannose and glucose were usually
found in small amounts.
Phospho- and Glycolipids The lipids of streptomycetes
comprise mainly diphosphatidylglycerol
(DPG), phosphatidylethanolamine (PE), phosphatidylinositol
(PI), and phosphatidylinositolmannosides
(PIMs). Lipid composition has been
extensively investigated and summarized by
Lechevalier et al. (1977b). Glycolipids do not
occur consistently in streptomycetes, and their
qualitative and quantitative lipid composition
depends largely on culture conditions. Under
phosphate limiting conditions, the amount of glycolipids
increase, significantly.
The taxonomic significance of polar lipids in
actinomycetes was demonstrated by Lechevalier
et al. (1977b). From the phospholipid results of
97 actinomycete strains representing 20 genera,
Lechevalier et al. (1977b) proposed a classification
of five phospholipid types. These five groups
are based on the presence or absence of certain
nitrogenous phospholipids. The marker lipids
of type II (PII) are phosphatidylethanolamine
(PE), methyl-PE, hydroxy-PE, and lyso-PE. In
addition to various other families, members of
the family Streptomycetaceae contain the lipids
of phospholipid type II. Additional lipids (e.g.,
phosphomonoester [PME] and OH-PE) and the
presence or absence of PI and PG allow further
differentiation (Fig. 5).
Menaquinones Streptomycetes contain only
menaquinones (Collins and Jones, 1981), and
like the majority of actinomycetes, they synthesize
quinones that have a partly saturated
isoprenoid side chain at position 3 of the naphthoquinone
ring. Menaquinone composition is
very useful for differentiation of actinomycetes
because of the different numbers of isoprene
units, the different degree of hydrogenation, and
the position of hydrogenated isoprene units
(Table 2). These three variations are useful for
classification and identification. Streptomycetes
synthesize menaquinones with a highly hydrogenated
isoprenoid chain. Three to four (rarely
five) isoprene units are saturated. The actinomycetes
which belong to this type synthesize
menaquinones with the same chain length but
different degree of saturation (Fig. 6).
Phenotypic Methods for
Classification within the Genus
Streptomyces
Phenotypic methods comprise all those that
are not directed towards DNA or RNA. They
include also chemotaxonomic techniques.
Between 1916 and 1943, most of the studies on
streptomyces were published by soil microbiologists,
who were mainly interested in ecological
questions. Only few species were described at
that time, mainly on the basis of morphological
criteria, pigmentation and ecological requirements
(Waksman and Curtis, 1916; Waksman,
1919; Jensen, 1930).
The discovery of actinomycin from S. antibioticus
(Waksman and Woodruf, 1940) was the starting
point of the investigations of antibiotics and
other bioactive substances produced by streptomycetes
in the 1940s and this led to extensive
screening approaches for novel bioactive compounds
in the following two decades.
The description of each producer of a novel
natural product as a new species (often patented)
led to an explosion of species descriptions
and resulted in an overclassification of the genus.
In the 1970s, the number of species increased to
over 3000 (Trejo, 1970).
Reduction in the number of species names was
first attempted in 1964 by the International
Streptomyces Project (ISP), which introduced
standard criteria for determining species
(described in Shirling and Gottlieb, 1968a, Shirling
and Gottlieb, 1968b, Shirling and Gottlieb,
1969, and Shirling and Gottlieb, 1972) to reduce
the number of poorly described synonymous
species. The major drawback of these descriptions
was that they were based mainly on morphology
(i.e., spore chain morphology, spore
surface ornamentation, color of spores, substrate
mycelium, soluble pigments, and production of
melanin pigment), in addition to a few physiological
properties, which were mainly restricted
to utilization tests of different carbon sources.
In these classical papers, more than 450 Streptomyces
species were redescribed, and type
strains were deposited in internationally recognized
culture collections. Although intended, the
efforts of the ISP did not result in an applicable
identification scheme.
A first step in this direction was the development
of numerical taxonomic methods in the
1960s including the methods of numerical identification.
The first numerical taxonomic studies
of streptomycetes by Silvestri et al. (1962) found
considerable diversity within the genus but also
groups that corresponded to the initial morphological
descriptions. These studies did not result
in nomenclatural changes, and despite the development
of other small databases for identification
of streptomycetes (Kurylowicz et al., 1975;
Gyllenberg, 1976), these studies had no impact
on streptomyces systematics (Table 3). Data
from a large-scale numerical taxonomic study by
Williams et al. (1983a) of 475 strains (including
394 Streptomyces type cultures from the ISP) for
139 unit characters were analyzed with simple
matching, the Jaccard coefficient, and the average
linkage algorithm. Consequently, the genus
Streptomyces was subdivided into species groups.
Streptomyces type strains (394) were clustered
according to similarities obtained from the phenetic
tests. At the 77 ± 5% simple matching coefficient
(SSM) level, 19 major, 40 minor and 18
single strain clusters were recovered. Many of
the minor clusters consisted of less than five
strains. Major clusters varied in size from 6 to 71
strains. Each cluster was addressed as a single
“species” despite the high diversity observed
within some clusters, and these therefore were
addressed as “species groups.” The largest species
group is Streptomyces albidoflavus (cluster
1), containing 71 strains, including 44 type
strains, 15 invalidly published species, and 12
unnamed strains. This cluster is further subdivided
into three clusters: cluster 1a, Streptomyces
albidoflavus subsp. albidoflavus (20 strains), cluster
1b, Streptomyces albidoflavus subsp. anulatus
(38 strains), and cluster 1c, Streptomyces albidoflavus
subsp. halstedii (13 strains; Williams et
al., 1989).
The high phenotypic diversity of this cluster is
obvious from the different test pattern. All
strains produced yellow gray colonies, produced
smooth spores in straight chains and no melanin,
and exhibited resistance to a number of antibiotics
including penicillin, lincomycin and
cephaloridine. Many of the strains showed
also antimicrobial activity; 39% produced compounds
with antifungal activity, 32% produced
compounds active against Gram-positive microorganisms
and 10% against Gram-negative
microorganisms (Williams et al., 1983b), showing
the large diversity within one cluster and exemplifying
clearly the problems with streptomycete
systematics (Anderson and Wellington, 2001).
Nevertheless, the comprehensive survey of
Williams et al. (1983a) resulted subsequently in
a reduction of the number of described Streptomyces
species; however, the problem of overspeciation
remained. Numerous species and
subspecies were described and many natural
isolates did not match the reference strains
used to construct the identification matrices
(Goodfellow and Dickenson, 1985). Although
probability matrices for identification purposes
were published (Williams et al., 1983b; Langham
et al., 1989), these matrices were not widely
adopted by the scientific community.
This study was the basis of the taxonomic
scheme for streptomyces presented in the 1989
edition of Bergey’s Manual of Systematic Bacteriology,
in which 142 species are listed (Williams
et al., 1989), in contrast to 463 species described
in the 1974 edition of Bergey’s Manual of Determinative
Bacteriology (Pridham and Tresner,
1974). A further numerical taxonomic analysis
by Kämpfer et al. (1991b) included more strains
and more than one strain of each species when
available. A total of 821 strains were tested for
329 physiological properties, and the resulting
cluster analysis was compared with the data published
by Williams et al. (1983a) in addition to
published genetic and chemotaxonomic data.
Many of the clusters defined by Williams et al.
(1983a) were again recognized; for example the
S. albidoflavus, S. anulatus, S. griseus, S. halstedii
group appeared as cluster 1 in both studies, in
which 28 of the S. griseus strains were grouped.
Interestingly, most of the strains sharing the
same specific epithet were grouped together,
indicating previous identification was reliable,
but some exceptions were also observed. For
example S. hygroscopicus strains were recovered
in cluster 1 but also in several other clusters and
subclusters.
On the basis of this study, a probability matrix
was constructed (Kämpfer and Kroppenstedt,
1991a), but this matrix was likewise not widely
used by other reseach groups.
Parallel with the numerical taxonomic studies,
additional chemotaxonomic and also molecular
methods were developed that are now often used
together with (often) few physiological tests to
study streptomycetes; however, a clear species
concept is still pending. In Table 4, the cluster
allocation of the species is given in comparison.
Other phenotypic methods include cell wall
analysis (Lechevalier and Lechevalier, 1970),
fatty acid profiling (Hofheinz and Grisbach,
1965; Lechevalier, 1977a; Saddler et al., 1986;
Saddler et al., 1987; Kroppenstedt, 1992), rapid
biochemical assay for utilization of 4-methylumbelliferone-
linked substrates (Goodfellow et
al., 1987c), serological assay (Ridell et al., 1986),
phage typing (Wellington and Williams, 1981a;
Korn-Wendisch and Schneider, 1992b), and protein
profiling (Manchester et al., 1990; Goodfellow
and O’Donnell, 1993; Lanoot et al., 2002),
including comparison of ribosomal protein patterns
(Ochi, 1989; Ochi, 1992; Ochi, 1995).
Fatty acids
The initial studies on actinomycete fatty acids
were carried out by Hofheinz and Grisebach
(1965) on Saccharopolyspora erythraeus (formerly
“Streptomyces erythraeus”) and Streptomyces
halstedii to elucidate the biosynthetic pathway
of branched fatty acids. It was shown that streptomyces
synthesize terminally branched fatty
acids. Anteiso-branched fatty acids are synthesized
from 2-methylbutyrate, leading to anteiso
fatty acids with an odd number of carbon atoms.
In contrast, isovalerate and isobutyrate as starting
compounds lead to the formation of iso-branched
fatty acids with even and odd numbers of Catoms,
respectively. For this reason iso- and
anteiso-branched fatty acids appear in pairs with
odd numbers of C-atoms only.
In their early studies, Hofheinz and Grisebach
(1965) separated the fatty acids as their methyl
esters by gas chromatography on different stationary
phases. Identification of the individual
fatty acids was obtained by comparing the equivalent
chain lengths of unknown fatty acids with
those of standard mixtures. The results were confirmed
by preparative gas-chromatography and
by physical methods such as mass spectrometry
and nuclear magnetic resonance (NMR) spectrometry.
In both species, iso- and anteisobranched
fatty acids with chain lengths of 15 and
17 carbon atoms were detected. High amounts of
14-methyl pentadecanoic acid (iso-C16:0) were
found in addition, while minor amounts of
unbranched fatty acids, tuberculostearic acid and
their homologues, could be detected in “S. erythraeus”
(now Saccharopolyspora erythraea) but
not in Streptomyces halstedii. These results are
congruent with those of several other studies
(Lechevalier et al., 1977a; Saddler et al., 1985;
Saddler et al., 1987) in which 10-methyl
branched fatty acids could not be detected
among streptomycetes. Usually only small
amounts of hydroxy fatty acids are synthesized
by a limited number of streptomycetes under
optimal oxygen supply. The hydroxy fatty acids
are easily destroyed in a non-deactivated injection
port of capillary gas chromatography
system. Therefore hydroxy fatty acids in streptomycetes
often go unnoticed. If streptomyces are
grown under reproducible culture conditions, the
hydroxy fatty acids they produce are highly diagnostic
for some streptomyces species. Hydroxy
fatty acids were detected in all strains of S. coelicolor
(30), S. rimosus (14), and S. violaceusniger
(18) and in 20 of 27 S. hygroscopicus strains but
not in S. violaceoruber (16), S. lavendulae (18), S.
griseus (22), S. fradiae (25), S. viridochromogenes
(25), S. glaucescens (8) and S. albus (33; Kroppenstedt,
1992; R. M. Kroppenstedt, unpublished
observation). Standardized growth and
cultivation conditions are a general prerequisite
for the use of fatty acid patterns below the genus
level (Saddler et al., 1986). Saddler et al. (1987)
used fatty acid profiles to investigate the taxonomy
of Streptomyces cyaneus strains and soil isolates
showing also blue spores. The S. cyaneus
cluster harbors 13 of 19 blue-spored strains of
streptomycetes (Hütter, 1962; Pridham and
Tresner, 1974; Korn et al., 1978). In the study of
Saddler et al. (1987), 8 of their 10 blue-spored
isolates clustered, while 17 of the 34 S. cyaneus
strains were assigned to a separate cluster. The
conclusions of the fatty acid study (Saddler et al.,
1987) and Williams et al. (1983a) agree that conventional
features like spore chain morphology,
color and ornamentation of spores may be helpful
for presumptive identification but are not
definitive for classification of streptomycetes.
The same combination of features may be found
in different clusters, yet one cluster may have
members with different features. The study of
Saddler et al. (1987), however, demonstrated
also the heterogeneity of the Streptomyces cyaneus
taxon as defined by Williams et al. (1983a).
Fatty acid patterns in general cannot delimit
Streptomyces species (Phillips, 1992; R. M. Kroppenstedt,
unpublished observation), but using
standardized conditions, they are still of high
value for the rapid characterization (independent
of the taxonomic status) of large numbers
of wild-type streptomycetes isolated from the
environment (Saddler et al., 1987). By using the
automated commercially available MIDI system
consisting of a Hewlett-Packard model 5890 capillary
gas chromatograph and a computer with
specific software (Microbial ID, Inc., Newark,
DE), the fatty acids are automatically identified
and quantified by the computer using fatty acid
standard mixtures for comparison. In their chapter
on The Family Nocardiopsaceae in this Volume,
Kroppenstedt and Evtushenko give a table
of types and the fatty acids diagnostic for different
genera of Actinomycetales, including Streptomyces.
The comparison of different methods
revealed that the use of numerical methods to
determine taxonomy lumped too many strains
together into some clusters (Williams et al.,
1983a; Kämpfer et al., 1991b).
Curie-point Pyrolysis Mass
Spectrometry (PyMS)
This method has also been applied to the classification
and identification of actinomycetes (Sanglier
et al., 1992). Similar to fatty acid profiling,
highly standardized conditions are necessary.
Whole cells are subject to high temperatures and
subsequent nonoxidative thermal degradation.
The resulting pyrolysate is then analyzed using
mass spectrometry, resulting in a fingerprint for
each organism.
Sanglier et al. (1992) applied this method to
strains belonging to the largest Streptomyces
species group, Streptomyces albidoflavus. Interestingly,
Streptomyces albidoflavus and Streptomyces
anulatus strains could be separated into
distinct groups. Three of the six Streptomyces halstedii
strains investigated also clustered into a
distinct group, whereas the remaining strains
clustered into two other groups. The study of
Kämpfer et al. (1991b) also found that Streptomyces
albidoflavus strains and Streptomyces anulatus
strains grouped separately. Interestingly, it
was confirmed that Streptomyces anulatus ISP
5361T, the strain used to name the Streptomyces
anulatus cluster, formed also a single-member
cluster (Table 4).
Serology
Few results using serological methods have been
published. Antisera against the mycelia from
streptomycetes, streptoverticillia and Nocardiopsis
species (Ridell et al., 1986) were used to
confirm the high similarity between Streptomyces
lavendulae and the streptoverticillia (Witt and
Stackebrandt, 1990; Kämpfer et al., 1991b). The
antisera of Kirby and Rybick (1986) raised
against Streptomyces griseus (Streptomyces anu-
latus, cluster 1B of Williams et al., 1983a) and
“Streptomyces cattleya” (cluster 47) were shown
to be genus-specific and to a certain degree also
group-specific. The monoclonal antibody produced
by Wipat et al. (1994) to “Streptomyces
lividans” 1326 was shown to be specific for
“Streptomyces lividans” strain 1326. Interestingly,
antigenic reaction was observed also for
strains grouped into cluster 21 of Williams et al.
(1983a).
Phage Typing
Phage typing can be used for host identification
at the genus and the species level (Welsch et al.,
1957; Kutzner, 1961a; Kutzner, 1961b; Korn et al.,
1978; Wellington and Williams, 1981a). Many
actinophages (most of them virulent) for phage
typing have been described. Streptomycete
phages can be either polyvalent (e.g., C31; Chater
et al., 1986b) or species-specific (Anderson and
Wellington, 2001; Table 5). The specificity of actinophages
at the genus level (e.g., Wellington and
Williams, 1981a; Korn-Wendisch, 1982; Prauser,
1984) was an additional feature that justified the
transfer of Actinopycnidium, Actinosporangium,
Chainia, Elytrosporangium, Microellobosporia,
Kitasatoa and Streptoverticillium to the
genus Streptomyces (Goodfellow et al., 1986b;
Goodfellow et al., 1986c; Goodfellow et al.,
1986d; Goodfellow et al., 1986e; Witt and Stackebrandt,
1990). Other transfers justified by phage
specificity include Actinoplanes armeniacus to
the genus Streptomyces (Kroppenstedt et al.,
1981; Wellington et al., 1981b) and “S. erythraeus”
to the genus Saccharopolyspora (Labeda,
1987). Species or group identification of Streptomyces
using phage typing has been less successful,
but there are a few exceptions (Table 5).
Phages are also useful in industrial microbiology
studies (Carvajal, 1953; Ogata, 1980) and in
genetic studies (for review, see Chater [1986a]
and chapter 12 of Kieser et al. [2000]). One of
the best-investigated actinophages is C31, a temperate
phage with a broad host range within the
genus Streptomyces (Lomovskaya et al., 1980).
This phage has become the subject of extensive
studies and has been employed for many purposes
(e.g., transfection, transduction, detection
of transposon-like elements of host DNA, and
cloning). Details are given in chapter 12 of
Kieser et al. (2000).
Protein Profiling
Polyacrylamide gel electrophoresis (PAGE) of
total protein extracts generate more or less
complex banding patterns. These patterns can be
used to differentiate species and subspecies
within various bacterial genera. Protein patterns
can be determined using one-dimensional (1-D)
or two-dimensional (2-D) protein electrophoresis.
The protein profiles of streptomycetes were
first analyzed by Manchester et al. (1990), who
investigated 37 Streptomyces strains (among
them 5 streptoverticillia). Some taxonomic correlations
were found between these profiles and
the phenotypic groupings observed by Williams
et al. (1983a) and Kämpfer et al. (1991b) in addition
to some DNA hybridization groupings
(Table 4). But Lanoot et al. (2002) confirmed
only a few of these correlations. For Streptomyces
isolates that are the causal agent of common
potato scab, Paradis et al. (1994) used both
PAGE and DNA-DNA hybridizations to elucidate
the taxonomy of their strain. Isolates
obtained from potato tubers were divided into
two groups with a correlation coefficient of 0.75
using sodium dodecylsulfate (SDS)-PAGE analysis.
The same two groups were resolved at
approximately 44% similarity using DNA-DNA
hybridization analysis. The fatty acid analysis
results of the same study did not correlate with
the SDS-PAGE and the DNA-DNA hybridization
groupings, which can be explained by the
influence of growth conditions on the profiles
obtained (Saddler et al., 1986; Saddler et al.,
1987). Protein profiling was not able to differentiate
pathogenic from nonpathogenic strains.
Other more specific patterns are obtained with
multilocus enzyme electrophoresis (MLEE) and
depend on the relative mobilities of cellular
enzymes in a gel matrix. Oh et al. (1996) studied
24 Streptomyces strains and demonstrated how
MLEE could be used for both inter- and
intraspecific characterization of streptomycetes,
provided the appropriate enzymes were used.
However, because only a restricted set of strains
was studied, no general recommendations can be
made for the usefulness of this method. In a
more comprehensive study, 93 Streptomyces
reference strains were investigated using SDSPAGE
of whole-cell proteins (Lanoot et al.,
2002). Subsequent computer-assisted numerical
analysis revealed 24 clusters encompassing
strains with very similar protein profiles. Five of
them included several type strains with visually
identical patterns. DNA-DNA hybridizations
revealed similarities higher than 70% among
these type strains. On the basis of these results,
consideration of Streptomyces albosporeus
subsp. albosporeus LMG 19403T as a subjective
synonym of Streptomyces aurantiacus LMG
19358T, Streptomyces aminophilus LMG 19319T
as a subjective synonym of Streptomyces cacaoi
subsp. cacaoi LMG 19320T, Streptomyces niveus
LMG 19395T and Streptomyces spheroides LMG
19392T as subjective synonyms of Streptomyces
caeruleus LMG 19399T, and Streptomyces violatus
LMG 19397T as a subjective synonym of
Streptomyces violaceus LMG 19360T was proposed
(Table 4).
Two-dimensional PAGE of the total cellular
proteins allows a finer resolution of the individual
gene products. This technique results in very
complex patterns and seems to be too sensitive
to investigate proteins with high rates of evolution
(Hori and Osawa, 1987). Mikulik et al.
(1982) and later Ochi (1989) were the first to
apply 2-D PAGE to determine the variability of
ribosomal proteins for use in streptomycete taxonomy.
These studies were later extended by
focusing on AT-L30 proteins, which give genusspecific
profiles (Ochi, 1992). In an even more
specific analysis, Ochi (1995) correlated the N
termini sequences of the ribosomal AT-L30 protein
of 81 streptomycete strains from different
taxonomic groups to phylogenetic groupings
within the genus and pointed out that on this
basis the genus Streptomyces seems to be well
described. However, Ochi’s groupings did not
correlate with those of Williams et al. (1983a)
and Kämpfer et al. (1991b). For details of these
groupings, see Table 4.
More detailed taxonomic studies of Streptomyces
have been performed by isolating and
sequencing specific proteins. For example, Taguchi
et al. (1996) used the Streptomyces subtilisin
inhibitor protein (SSI), which plays unidentified
role(s) in physiological or morphological regulation,
to investigate the taxonomic status of the
Streptomyces coelicolor strains. The amino acid
sequence, of SSI from “Streptomyces lividans”
66, Streptomyces coelicolorMüller ISP 5233T and
“Streptomyces coelicolor”A3(2) were compared.
The alignments supported ribosomal sequence
comparisons, indicating that “Streptomyces coelicolor”
A3(2) is more closely related to “Streptomyces
lividans” 66 (cluster 21 of Williams et al.,
1983a) than to the type strain, Streptomyces
coelicolor Müller ISP 5233T (cluster 1).
Genotypic Methods for
Classification within the Genus
Streptomyces
Genotypic methods comprise all those that are
directed towards DNA or RNA molecules
(Schleifer and Stackebrandt, 1983; Vandamme et
al., 1996). The development of molecular methods
to analyze bacterial genomes has provided a
new basis for studying bacterial taxonomy and in
some cases phylogenetic relationships of the
prokaryotes at the genus, species and subspecies
level. These methods are widely used in modern
taxonomic studies. The general taxonomic values
of different molecular techniques are given by
Vandamme et al. (1996) and in Prokaryote Characterization
and Identification in Volume 1. Their
usefulness in taxonomic studies to delimit species
within the genus Streptomyces is discussed
briefly in the following chapters and has also
been reviewed by Anderson and Wellington
(2001).
Problems in assigning new strains to existing
species on the basis of this molecule alone persist
and doing this will not be possible in the future.
As already pointed out by Stackebrandt and
Schumann in Introduction to the Classification
of the Actinomyces in this Volume, comparative
analysis of sequences of homologous and genetically
stable semantides has demonstrated that
several classification systems based on morphology
and physiology do not reflect the natural
relationships among actinomycetes and related
organisms. In this respect, rRNA sequence com-
parison is a powerful tool in modern taxonomy
and has revolutionized our insight in phylogenetic
lineages of major taxonomic groups. However,
the resolving power of 16S rRNA sequences
is not sufficient to delimit species. Nevertheless,
rRNA sequence comparisons are important in
the taxonomy of Streptomyces and the results
have also been used to answer questions about
horizontal transfer of genes within the genus
(Huddleston et al., 1997). Genes for 16S RNAs
are highly conserved within bacteria. Within the
genus Streptomyces, three regions within the
gene have enough sequence variation to be useful
as genus-specific (a and b regions) and species-
specific (c regions) probes (see Stackebrandt
et al., 1991a; Stackebrandt et al., 1991b; Stackebrandt
et al., 1992; Anderson and Wellington,
2001). Not only 16S rRNA genes, but also 23S
rRNA and 5S rRNA genes (Mehling et al.,
1995), 16S-23S rRNA internally transcribed
spacer (ITS) sequences (Song et al., 2003), and
ribosomal protein sequences have also been used
to investigate species relationships within the
genus Streptomyces (Liao and Dennis, 1994;
Ochi, 1995). Wenner et al. (2002) studied the
nucleotide composition of the ITS sequences of
the six rDNA operons of two Streptomyces
ambofaciens strains. Their findings suggested that
recombination frequently occurs between the
rDNA loci, leading to the exchange of nucleotide
blocks, and confirmed that a high degree of ITS
variability is a common characteristic among
Streptomyces spp. Note that rRNA sequences
cannot be used alone because of the intraspecific
variation and intragenomic heterogeneity.
DNA-DNA Hybridizations
The percent DNA-DNA hybridization and the
decrease in thermal stability of the hybrid are at
present the “gold standard” methods of species
delineation in bacteriology (Wayne et al., 1987).
An ad hoc committee for the re-evaluation of the
species definition in bacteriology (Stackebrandt
et al., 2003) has encouraged investigators to
verify the species concept with other methods;
however, DNA-DNA similarity and change in
melting temperature (ΔTm; Wayne et al., 1987)
remain the acknowledged standard for definition
of species.
DNA-DNA hybridizations of total chromosomal
DNA have also been used within the
genus Streptomyces. In an initial study, Mordarski
et al. (1986) compared the numerical and
phenetic groupings of strains of the Streptomyces
albidoflavus cluster 1 of Williams et al. (1983a)
with the results of DNA-DNA hybridizations
(reassociation of labeled DNA on nitrocellulose
filters). In congruence with numerical studies,
DNA-DNA hybridization studies showed this
cluster could be subdivided into three subclusters
and confirmed the homogeneity of the Streptomyces
albidoflavus subcluster albidoflavus.
Two further subclusters obtained by DNA-DNA
hybridizations were not congruent with the
groupings of S. anulatus or halstedii by Williams
et al. (1983a); however, some correlations were
found with the groupings of Kämpfer et al.
(1991b). Unification using DNA-DNA similarity
(Witt and Stackebrandt, 1990) of Streptoverticillium
strains with the genus Streptomyces (reassociation
of labeled DNA on filters) was confirmed
by phenotypic data in the numerical study of
Kämpfer et al. (1991b).
The most extensive application of DNA-DNA
hybridizations (thermal renaturation method) to
the study of the major streptomycete phenetic
groups of Williams et al. (1983a) was by Labeda
and coworkers. In studies using Streptomyces
cyaneus (Labeda and Lyons, 1991a), Streptomyces
violaceusniger (Labeda and Lyons, 1991b),
Streptomyces lavendulae (Labeda, 1993), the verticil-
forming streptomycetes (formerly Streptoverticillium
species; Labeda, 1996; Hatano et
al., 2003), and S. fulvissimus and S. griseoviridis
phenotypic clusters (Labeda, 1998), the degree
of DNA similarities was often not congruent
with the phenetic groupings of Williams et al.
(1983a). But again, more correlations were
found with the groupings of Kämpfer et al.
(Kämpfer et al., 1991b; Table 4). The usefulness
of the DNA-DNA hybridization technique to
delineate species within the genus Streptomyces
has been questioned. In support of the usefulness
of this technique is the considerable genetic
instability of certain regions within the Streptomyces
chromosome (Redenbach et al.,
1993). The two complete Streptomyces genome
sequences available at this time (Omura et al.,
2001; Ikeda et al., 2003) indicate that the central
core region contains mostly the essential housekeeping
genes, while the chromosome arms
comprise laterally acquired contingency genes.
Another issue is the presence of large plasmids
in strains of Streptomyces, which can considerably
influence the results of DNA-DNA hybridizations.
General properties of Streptomyces
plasmids and their use for gene cloning are given
in chapter 11 of Kieser et al. (2000).
Fingerprinting Techniques
Randomly Amplified Polymorphic DNA
(RAPD) Polymerase Chain Reaction (PCR)
RAPD-PCR is used as a rapid screening method
to detect similarity among streptomycete strains.
Single primers with arbitrary nucleotide
sequences to amplify DNA are used in addition
to a low annealing temperature so that polymorphisms
can be detected. Stringent standardiza-
tion of the reaction parameters is required.
These include primer sequence, annealing temperatures,
buffer components, concentration and
quality of template DNA. The resulting characteristic
fingerprint of PCR products enables
detection of chromosomal differences between
individual isolates without having any prior
knowledge of the chromosomal sequence (Williams
et al., 1990). Applying this technique,
Mehling et al. (1995) could not detect any characteristic
banding patterns for closely related
species unless a highly specific actinomycete
primer was used. As a major disadvantage, the
resulting fingerprints contained only few bands
(one to four), reducing the effectiveness of this
method. Huddleston et al. (1995) reported similar
results. They evaluated the use of RAPDPCR
for the resolution of interspecific relationships
among members of the Streptomyces albidoflavus
cluster of Williams et al. (1983a). Anzai
et al. (1994) demonstrated the variation in fingerprints
obtained when a single base was substituted
on the arbitrary primer; 11 primers were
investigated and the number of bands ranged
from zero to 20. The most significant differences
were observed when the sequence at the 3′ end
was altered. Anzai et al. (1994) also investigated
the relationship of Streptomyces virginiae strains
to Streptomyces lavendulae strains by RAPDPCR.
Williams et al. (1983a) found that Streptomyces
virginiae was a synonym of Streptomyces
lavendulae. These strains were also grouped into
the same cluster in the study of Kämpfer et al.
(1991b). In addition to RAPD-PCR, DNA-DNA
hybridization, low-frequency restriction fragment
analysis (LFRFA), and cultural and physiological
tests were performed. Consistent results
were obtained using all these methods after
RAPD PCR optimization. It was however not
possible to clarify the interspecific relationship
of Streptomyces lavendulae and Streptomyces
virginiae.
Restriction Digests of Total Chromosomal
DNA Similar to the RAPD PCR techniques,
low-frequency restriction fragment analysis
(LFRFA) uses the entire bacterial chromosome,
which is digested with restriction endonucleases
that cut infrequently. Because streptomycetes
belong to bacteria with a high DNA G + C content,
rare AT cutters are used. The fragments
obtained were separated by pulsed-field gel electrophoresis
(PFGE). In a first study, Beyazova
and Lechevalier (1993) included 59 strains from
eight species groups and found the method useful
for the clustering of related strains. However,
some discrepancies were found, for example for
strains grouped into the Streptomyces cyaneus
cluster of Williams et al. (1983a). Again this
method seems to be useful for the detection of
very closely related strains, but similar to RAPDPCR,
it cannot resolve interspecific relationships.
In addition, Rauland et al. (1995) found
that large chromosomal amplifications or deletions
may also lead to misinterpretations of
resulting banding patterns.
Nucleic Acid Sequence Comparisons of 16S
rRNA and Other Genes The application of
16S rRNA gene sequence analysis to the study
of the taxonomy of streptomycetes is reviewed
by Stackebrandt et al. (1992), who highlighted
the importance of the region selected for comparison.
Sequence analysis of rRNA genes has
already been applied at the genus, species and
strain level. The relationships obtained differed
considerably depending on the variable region
(a, b or c). Kataoka et al. (1997) subsequently
investigated the c region from 89 streptomycete
type strains representing several clusters of
Williams et al. (1983a). Though these variable
regions were useful for resolving inter- and
intraspecies relationships within the streptomycetes,
they were too variable for determining
generic relationships. Among the 89 strains studied,
57 variants were detected and 42 strains
were found to have unique sequences. After this
publication, the c regions from 485 streptomycete
strains were sequenced and deposited in
GenBank by this group. At present, this is the
largest publically available set of streptomycete
16S rDNA sequence data.
Anderson and Wellington (2001) published a
phylogenetic tree based on comparison of the c
regions from representatives of the major cluster
groups defined by Williams et al. (1983a). The
taxonomic status of the phenotypic groups was
confirmed, although they did not cluster
together. Only Streptomyces olivaceoviridis and
Streptomyces griseoruber strains (which are
found in clusters 20 and 21 of Williams et al.
[1983a], but in a single cluster 9 of Kämpfer et
al. [1991b]) had identical c regions. The Streptomyces
albidoflavus group, which was previously
divided into three species groups by Williams et
al. (1983a) and contained more than 60 strains
(Williams et al., 1989), was now divided into six
groups using sequence comparisons of the c
region (Kataoka et al., 1997). The three phenotypic
subgroups of Williams et al. (1983a) were
maintained but did not cluster together.
Hain et al. (1997) designed 16S rRNA oligonucleotide
probes to determine intraspecific
relationships within the Streptomyces albidoflavus
group. As a result, sequence comparisons
were found to be useful for species delimitation
but of no value for strain differentiation.
Also the intergenic 16S-23S rRNA spacer
regions were investigated in detail and they were
obviously more suitable for delineation of the
intraspecific relationships within that cluster.
Genus specific probes based on 23S rRNA gene
sequences (Mehling et al., 1995) and 5S rRNA
gene sequences (Park et al., 1991) have been
developed. The reclassification of the genera
Chainia, Elytrosporangium, Kitasatoa, Microellobosporia
and Streptoverticillium into the genus
Streptomyces (Park et al., 1991) was confirmed
using 5S rRNA gene sequence evaluation.
At present about 350 complete 16S rRNA
sequences are available from public databases.
Although 16S rRNA sequence analyses have
provided a framework for prokaryotic classification,
note that the current classification system
based on this molecule has not yet solved the
taxonomy within the genera (especially within
the genus Streptomyces). Several studies have
attempted to use sequence data from variable
regions of 16S rRNA to establish taxonomic
structure within the genus, but the variation is
too limited to help resolve problems of species
differentiation (Witt and Stackebrandt, 1990;
Stackebrandt et al., 1991b; Stackebrandt et al.,
1992; Anderson and Wellington, 2001).
The situation is complicated by the fact that
Streptomyces species may harbor different 16S
rRNA gene sets. For example, S. coelicolor
A3(2), S. lividans and several Streptomyces species
harbor six ribosomal rRNA gene sets. Each
set comprises one gene copy for 16S, 26S, and 5S
rRNA (Van Wezel et al., 1991) and lacks tRNA
genes.
A comprehensive study of the phylogenetic
relationships between 64 whorl-forming streptomycetes
using partial gyrB sequences (the structural
gene of the B subunit of the DNA gyrase)
has been published by Hatano et al. (2003). The
strains in this study consisted of 46 species and
eight subspecies and in addition 13 species whose
names have not been validly published (including
10 strains examined by the International
Streptomyces Project [ISP]). Two major groups
were found. The typical whorl-forming species
(59 strains) were further divided into six major
clusters of three or more species, seven minor
clusters of two species, and five single-member
clusters on the basis of the threshold value of
97% gyrB sequence similarity. Major clusters
contained Streptomyces abikoensis, Streptomyces
cinnamoneus, Streptomyces distallicus, Streptomyces
griseocarneus, Streptomyces hiroshimensis
and Streptomyces netropsis. Phenotypically,
members of each cluster resembled each other
closely except for those in the S. distallicus cluster
(which was divided phenotypically into the S.
distallicus and Streptomyces stramineus subclusters)
and the S. netropsis cluster (which was
divided into the S. netropsis and Streptomyces
eurocidicus subclusters). Strains in each minor
cluster closely resembled each other phenotypically.
At the conclusion, 59 strains of typical
whorl-forming Streptomyces species were placed
into the following 18 species, including subjective
synonym(s): S. abikoensis, Streptomyces ardus,
Streptomyces blastmyceticus, S. cinnamoneus, S.
eurocidicus, S. griseocarneus, S. hiroshimensis,
Streptomyces lilacinus, “Streptomyces luteoreticuli,”
Streptomyces luteosporeus, Streptomyces
mashuensis, Streptomyces mobaraensis, Streptomyces
morookaense, S. netropsis, Streptomyces
orinoci, S. stramineus, Streptomyces thioluteus
and Streptomyces viridiflavus (Table 4). All
strains showing 98.5–100% gyrB sequence similarity
also had a high DNA-DNA similarity (70–
100%), showing better resolution with gyrB
sequences than with 16S rRNA sequences.
Other genes known to be conserved between
species, such as housekeeping genes (e.g., elongation
factors and ATPase subunits), may be
useful as primary target genes for study (Ludwig
and Schleifer, 1994). Huddleston et al.
(1997) used tryptophan synthase genes in addition
to 16S rRNA gene comparisons to determine
the phylogeny of streptomycin-producing
streptomycetes and provide evidence for the
horizontal transfer of antibiotic resistance
genes. Predictably, sequence information on
other genes (i.e., other housekeeping genes) will
lead to better insight into the intraspecific structure
of the genus Streptomyces (Stackebrandt
et al., 2003).
Rapid Methods for Gene Analysis in
Streptomycete Taxonomy
Several alternative methods have been described
that do not involve sequencing, using either
restriction analysis (Clarke et al., 1993; Fulton et
al., 1995) or specialized gel electrophoresis techniques
to monitor the mobility of the product
(Hain et al., 1997; Heuer et al., 1997). Restriction
fragment length polymorphism (RFLP) patterns
of purified rRNA were used by Clarke et al.
(1993) with strains from the Streptomyces albidoflavus
cluster (subgroups 1A and 1B of Williams
et al., 1983a). The following combination of
enzymes was used: BglI, EcoRI, PstI and PvuII.
The resulting RFLP profiles varied considerably
between species groups but were found to enable
differentiation of phenotypically similar strains.
Fulton et al. (1995) performed ribosomal restriction
analysis using MseI fingerprints of rRNA
operons (RiDiTS) to group 98 named streptomycete
strains including members of cluster
group A (comprising clusters 1–41) and F (comprising
clusters 55–67) of Williams et al. (1983a)
and other strains. The resulting RiDiTS belonged
to 11 pattern types with varying degrees of similarity
to the Williams subclusters. At low resolution
(70% similarity), cluster groups A and F
could be differentiated, but individual clusters
could not.
Further studies are based on genotypic variation
monitored using denaturing gradient gel
electrophoresis (DGGE; Muyzer et al., 1993)
with or without DNA-binding agents (Hain
et al., 1997). Anderson and Wellington (2001)
recommended DGGE in combination with other
techniques. Using the variable 16S rRNA
regions, this method enables delimitation of
genus groups and species-groups. Huddleston
et al. (1995) allocated isolates ASB33, ASB37
and ASSF22 to Streptomyces albidoflavus,
Streptomyces griseoruber and Streptomyces
albidoflavus, respectively, using a combination of
techniques including numerical taxonomy,
PFGE and sequence comparisons (Huddleston
et al., 1997).
Identification of Streptomycetaceae
at the Genus Level
Sequencing of 16S rRNA genes and comparison
of these sequences after careful alignment with
published sequences is currently the most reliable
method for assigning unknown organisms to
the different genera. The calculation of phylogenetic
trees at the subgeneric level should be done
very carefully and may lead to misinterpretations.
Notably, phylogenetic analysis on the basis
of 16S rRNA comparisons does not allow species
delineation.
Colony morphology (color of the aerial mycelium,
color of the substrate mycelium, and soluble
pigment) is (especially in the case of the
genus Streptomyces) very useful (Tables 6 and 7;
Figs. 3 and 4). Here the traditional methods
extensively described by Korn-Wendisch and
Kutzner (1992a) are highly recommended.
A microscopic characterization (particularly
the morphologies of the aerial mycelium,
arthrospores and vegetative mycelium) is of
high value (Figs. 7–11). See Korn-Wendisch and
Kutzner (1992a) and chapter 3 of Kieser et al.
(2000) for details of the methods on microscopy.
Furthermore, the detection of LL-A2pm in cell
wall or whole-cell hydrolysates, the lack of
mycolic acids, the predominance of mainly isoand
anteiso-methyl branched fatty acids, and the
16S rRNA sequence are well suited for genus
identification (Table 1).
Identification of Species
Kitasatospora Novel isolates can be readily
identified as members of the genus by 16S rRNA
gene and 16S-23S rRNA gene spacer analyses.
The presence of meso-A2pm in whole cell
hydrolysates is an important chemotaxonomic
character for differentiation from Streptomyces
(Table 1). Additional chemotaxonomic investigations
(polar lipids, fatty acid patterns, and
menaquinone type) are helpful for allocation of
an unknown isolate to the genus. The meso-
A2pm content is 49–89% in Kitasatospora strains
and 1–16% in Streptomyces strains (Zhang et al.,
1997). Strains belonging to the genus Streptacidophilus
contain (like Streptomyces) LLdiaminopimelic
acid as predominant diamino
acid (Kim et al., 2003). For further species identification,
the characters shown in Table 8 are
helpful.
Streptacidiphilus Novel isolates can be readily
identified as members of the genus by 16S rRNA
gene analysis. The presence of LL-A2pm in
whole cell hydrolysates and other chemotaxonomic
characters (e.g., polar lipids, fatty acid
patterns, and menaquinone type; see family
description) are shared by members of the genus
Streptomyces. The three species of the genus
Streptacidophilus can be differentiated on the
basis of some phenotypic properties (Table 9).
Streptomyces The identification of species poses
severe problems. Because of the high number of
validly published species (Table 4), most of
which are based on a single strain description, it
is at present not possible to recommend a single
method or even a set of methods for identification
at the species level. Because a clear species
concept within the genus Streptomyces is still
pending, investigators should be very careful
with species allocations on the basis of the results
of one or few methods. The ICSP Subcommittee
on the Systematics of Streptomycetaceae
(Kämpfer and Labeda, 2003) has recommended
that before a species concept of the genus Streptomyces
is formulated more genomic information
should be evaluated, and it was agreed “that
the proposal of new species should only be
accepted on the basis of very careful studies done
with sufficient practice and considering all other
species.” In recent years, only a few species have
been proposed on the basis of 16S rRNA
sequence analysis and phenotypic characterization
(most often restricted to those species
closely related by phylogenetic analysis [e.g.,
Bouchek-Mechine et al., 2000; Kim et al., 2000;
Kim and Goodfellow, 2002a, b; Li et al., 2002a;
Li et al., 2002b; Meyers et al., 2003; Petrosyan et
al., 2003; Zhang et al., 2003]). The ICSP Subcommittee
on the Systematics of Streptomycetaceae
(Kämpfer and Labeda, 2003) recommended a
careful look at synonymy as a first step to reduce
the number of “species” within the genus.
At present two complete Streptomyces
genomes are available (Omura et al., 2001; Ikeda
et al., 2003). Bentley et al. (2002) reported the
first complete genome sequence of S. coelicolor
and described a model for the evolution of the
large linear chromosome, where the central core
region contains mostly the essential housekeeping
genes and the “arms” contain laterally
acquired contingency genes. A comparison of the
S. averilitils and S. coelicolor genomes supports
this model (Bentley et al., 2003). Most of the
secondary metabolism gene clusters are located
in the arms. The detailed study of genomes of
different streptomycetes and housekeeping genes
may provide a more reliable basis for an intrageneric
subdivision (Stackebrandt et al., 2003).
Ecophysiology and Habitat
The following sections are largely based on the
information summarized by Korn-Wendisch and
Kutzner (1992a). Members of the family Streptomycetaceae
are ubiquitous in nature. Members
of the genus Kitasatospora have been predominantly
isolated from soils (Zhang et al., 1997),
and Streptacidiphilus species have been isolated
from acidic soils and litter (Kim et al., 2003).
Streptomycetes can be isolated in high
numbers in soil, which is their natural habitat.
Most streptomycetes can degrade complex and
recalcitrant plant and animal materials, often
polymeric residues including polysaccharides
(e.g., starch, pectin, cellulose and chitin), proteins
(e.g., keratin and elastin), lignocellulose,
and aromatic compounds.
Members of the genus Streptomyces are
involved in the biodegradation of various polymers
abundant in soil owing to their ability to
produce extracellular enzymes. The biodegradatve
activities of actinomycetes in general were
reviewed in the 1980s by Lechevalier (Lechevalier,
1981a; Lechevalier, 1988), Crawford (1988),
and Peczynska-Czoch and Mordarski (1988).
Streptomycetes are among the very few bacteria
able to degrade lignin which occurs in nature
closely associated with cellulose and xylan (hemicellulose),
i.e., in the lignocellulose complex.
Although fungi play the more important role in
lignin decomposition (Crawford, 1981; Crawford,
1988; Janshekar and Fiechter, 1983; Kirk and
Farell, 1987), evidence from experiments using
14C-labeled lignin now shows that streptomycetes
(Crawford, 1978; Antai and Crawford, 1981) and
also several other genera of actinomycetes are
involved in this process (McCarthy and Broda,
1984a; McCarthy et al., 1984b; McCarthy et al.,
1986). As a constituent of the lignocellulose complex,
cellulose can be degraded by the few ligninolytic
streptomycetes. More details are given by
Ramachandra et al. (1988), Wang et al. (1990),
Crawford et al. (1993), Chamberlain and Crawford
(2000), Kormanec et al. (2001), Gottschalk
et al. (2003), and Kaneko et al. (2003).
In addition, multicomponent cellulases consisting
of several endoglucanases and exoglucanases
have been found in thermophilic and mesophilic
streptomycetes (Enger and Sleeper, 1965; Crawford
and McCoy, 1972; MacKenzie et al., 1984;
Schrempf and Walter, 1995; Harchand and Singh,
1997; Marri et al., 1997; Ulrich and Wirth, 1999;
Wirth and Ulrich, 2002). Also, xylanases involved
in the decomposition of the lignocellulose complex
have been found in streptomycetes
(Kluepfel and Ishaque, 1982; Kluepfel et al., 1986;
Deobald and Crawford, 1987; Godden et al.,
1989; Schäfer et al., 1996; Morosoli et al., 1999).
Again, xylanases seem to be more widespread
among thermophilic actinomycetes (McCarthy et
al., 1985). Pectinolytic streptomycetes have been
reported (Sato and Kaji, 1975; Sato and Kaji,
1977; Sato and Kaji, 1980a; Sato and Kaji, 1980b)
and chitinolytic complexes consisting of chitinase
and chitobiase have been isolated from various
streptomycetes: Streptomyces griseus (Berger
and Reynolds, 1958), Streptomyces antibioticus
(Jeuniaux, 1966), and othe streptomycetes (Beyer
and Diekmann, 1985). For more details, see The
Family Streptomycetaceae, Part II: Molecular
Biology in this Volume).
The ability to decompose starch (the primary
material for the textile, paper and food industry)
is widespread among fungi and bacteria. The
involved enzymes, amylases, have been detected
in various streptomycetes (Mordarski et al.,
1970; Suganuma et al., 1980; Fairbairn et al.,
1986; McKillop et al., 1986).
In addition to degrading polymeric compounds
(as described above), streptomycetes can
play an important role in the destruction of other
organic materials used by humans for diverse
purposes, among them cotton and plant fibers
(Khan et al., 1978; Lacey and Lacey, 1987), wool
(Noval and Nickerson, 1959), hydrocarbons in
jet fuel and emulsions (Genner and Hill, 1981),
and rubber (Cundell and Mulcock, 1975; Hutchinson
et al., 1975). The biodeterioration of natural
and synthetic substances has been reviewed
in detail by Lacey (1988) and Behal (2000). More
details are given in The Family Streptomycetaceae,
Part II: Molecular Biology in this Volume.
Most isolated streptomycetes are nonfastidious;
they do not require organic nitrogen sources
or vitamins and other growth factors. Soil as a
habitat gives support to their mycelial growth.
Furthermore, spore formation enables streptomycetes
to adapt to various physical conditions
in soil (such as shifts in aeration, moisture tension,
and pH), periods of drought, frost, hydrostatic
pressure, and anaerobic conditions which
may change dramatically and quickly.
The spores can be regarded as a semi-dormant
stage in the life cycle that facilitates survival in
soil for long periods (Mayfield et al., 1972;
Ensign, 1978). Morita (1985) reported viable cultures
from 70-year-old soil samples. A relatively
high number of streptomycetes in soil is almost
always present as inactive spores. The very low
germination efficiencies often obtained may be
caused by competition with other indigenous
microorganisms, but pre-germinated spores are
found to grow for a short time and then resporulate
(Lloyd, 1969a). Germination may
depend on the presence of special signaling
factors, and there is evidence that exogenous
nutrients, water and Ca2+ are required (Ensign,
1978). Furthermore the nutrient status of the
germination site influences the extent of hyphal
growth and the time to differentiation into aerial
hyphae. Many other “habitats” may come into
contact with soil owing to human or other activities.
As listed by Korn-Wendisch and Kutzner
(1992a), these are 1) fodders and other organic
material and 2) freshwater and marine habitats
as well as potable water systems. Mesophilic
and especially thermophilic streptomycetes are
involved in the degradation of many natural substrates
(e.g., hay, fodder, grain, and wood) and
can degrade synthetic products (e.g., cotton textiles,
fabric, paper, rubber, plastics and plasticizers).
Drainage after heavy rainfalls causes creeks
and rivers to become contaminated with soil
streptomycetes that find their way into the sediments
of freshwater lakes and even to marine
biotopes. According to some reports, drinking
water supplies may also become contaminated
with streptomycetes; some of them produce
odorous compounds leading to the spoilage of
the water. Streptomycetes play only a minor role
as plant pathogens, and although very few streptomycetes
have been isolated from pathological
material so far, their role as agents of infectious
diseases cannot be ignored (more details are
given below).
Soil as Habitat
A number of biotic and abiotic properties characterize
any habitat and determine the current
composition of the community and also the numbers
of microorganisms. These are: 1) vegetation
and content and kind of organic matter, 2) soil
type, 3) season and climate, 4) temperature, 5)
circulation of water and air, and 6) pH. As already
pointed out by Korn-Wendisch and Kutzner
(1992a), the reports published on the occurrence
of streptomycetes in soil are numerous and too
extensive to be treated quantitatively. The reader
is referred to reviews by Lechevalier (Lechevalier,
1981a; Lechevalier, 1988), Williams (1982a),
Goodfellow and Williams (1983), Williams et al.
(1984b), Goodfellow and Simpson (1987a), Korn-
Wendisch and Kutzner (1992a) and to a series of
papers by Williams and coworkers dealing specifically
with the ecology of actinomycetes in various
soils and under specific conditions: Davies
and Williams (1970), Williams and Mayfield
(1971b), Williams et al. (Williams et al., 1971c;
Williams et al., 1972; Williams et al., 1977), Mayfield
et al., (1972), Ruddick and Williams (1972),
Watson and Williams (1974), Khan and Williams
(1975), Flowers and Williams (1977), Williams
and Robinson (1981). By direct observation of
the soil microflora, several authors have shown
that streptomycetes perform their typical life
cycle in this natural habitat. For details of this life
cycle and its genetic control, the reader is referred
to Kieser et al. (2000). As summarized by Korn-
Wendisch and Kutzner (1992a), in most soils, 104
to 107 colony forming units (CFU) per g can be
expected, accounting for about 1–20% of the total
viable count; in some soils however streptomycetes
dominate. The number of streptomycetes
and also the number of subgroups vary
under different conditions. Details can be found
in the studies of Flaig and Kutzner (1960a), Misiek
(1955), Szabó and Marton (1964), and Küster
(1976).
Note that the detection and localization of different
Streptomyces species or types in their
natural habitat are based mainly on cultivation
dependent techniques. In addition, the difficulties
in the intrageneric classification of the genus
Streptomyces often do not allow a comparison of
these ecophysiological studies. With respect to
moisture, it has been shown by Williams et al.
(1972) that streptomycetes resist desiccation
because they form arthrospores. In addition they
need a lower water tension for growth than other
bacteria need, but they may be very sensitive to
water-logged conditions.
Most streptomycetes prefer neutral to alkaline
soils as a natural habitat (e.g., Flaig and Kutzner,
1960a). But several studies of acidic soils employing
media adjusted to acid pH and supplemented
with antifungal agents found numerous acidophilic
as well as acido-tolerant streptomycetes.
Khan and Williams (1975), Hagedorn (1976), and
Williams et al. (1977) showed in detail that acidic
forest soils and other acidic habitats contained
different groups or species of streptomycetes.
These streptomycetes also have unusual properties
when compared with neutrophilic strains,
e.g., production of specific amylases (Williams
and Flowers, 1978b) and chitinases (Williams and
Robinson, 1981). Only a few reports of alkalophilic,
acid-sensitive actinomycetes have been
published (Taber, 1959; Taber, 1960; Mikami et
al., 1982; Mikami et al., 1985).
Streptomycetes, like other soil bacteria, may
also be found in the intestinal tract of earthworms
(Brüsewitz, 1959; Parle, 1963b; Parle,
1963a), the gut of arthropods (Szabó et al., 1967;
Bignell et al., 1980; Bignell et al., 1981; Bignell,
1984), and the pellets produced by millipedes
and woodlice (Márialigeti et al., 1984).
Studies on the occurence of streptomycetes in
the rhizosphere have been published by various
authors (for a review, see Goodfellow and
Williams, 1983). The competitive advantage of
antibiotic-producing organisms over nonproducing
microbes in soil has been suggested since the
time these compounds were first discovered.
However, evidence for the in situ production of
antibiotics in soil is still not clear (Williams,
1982a). This may be due to 1) the instability and
low concentrations in soil (Brian, 1957; Williams,
1982a) and 2) possible adsorption to soil colloids
in combination with inadequate and insensitive
detection methods (Williams, 1982a) and 3) the
ephemeral growth of producers in response
to nutrient shortage (Williams and Khan, 1974;
Williams, 1982a).
However, Rothrock and Gottlieb (1984)
reported antibiotic production occurs in sterilized
soil supplemented with nutrients and
inoculated with a potent producer. Many soil
microbiologists support the assumption that
streptomycetes play an important role in the
control of fungal root pathogens (Williams,
1978a; Williams, 1982a; Sing and Mehrotra, 1980;
Rothrock and Gottlieb, 1981). In addition many
streptomycetes are often successful in competition
with other rhizosphere bacteria such as
pseudomonads and bacilli, especially in relatively
dry soil.
Thermophilic Streptomycetes
The genus Streptomyces contains mainly mesophilic
species in addition to some thermotolerant
(growing up to 45°C) and a few thermophilic
species. A detailed taxonomic study of thermophilic
streptomycetes has been published by Kim
et al. (1999). Additional thermophilic species (S.
thermocoprophilus and S. thermospinisporus)
have been described by Kim et al. (2000) and
Kim and Goodfellow (2002b). The thermophilic
streptomycetes described so far grow at 28–55°C,
and several grow at even higher temperatures.
Thermophilic streptomycetes go through an
interesting cycle in nature in regard to their dispersal:
active growth takes place at sites of high
temperature such as in compost, manure, and selfheating
hay or grain. The vegetative phase ends
with the formation of a large number of spores.
These are returned with the compost or manure
to the fields and pastures where they infect plant
material and hay directly or via soil dust (Korn-
Wendisch and Kutzner, 1992a). Not surprisingly,
the majority of actinomycetes isolated from
bioaerosols in the surroundings of composting
facilities belong to the genus Streptomyces (P.
Kämpfer et al., unpublished observation). For
this reason thermophilic actinomycetes are widespread
and can be isolated from various sources
like soils (Tendler and Burkholder, 1961; Craveri
and Pagani, 1962), pig feces (Ohta and Ikeda,
1978), sewage-sludge compost (Millner, 1982),
and freshwater habitats (Cross, 1981).
Freshwater Environments, Water Supplies
and Marine Environments
Actinomycetes can easily be isolated from fresh
water and especially from sediments of rivers
and lakes. Cross (1981) stated, however, that
most of these organisms may not be active at
these sites. Nevertheless, these wash-in forms
(“aliens”) from surrounding terrestrial environments
can survive as dormant spores in aquatic
habitats for a long time (Al-Diwany and Cross,
1978), and especially rivers carry a load of
various actinomycetes, among them also streptomycetes.
In a study on the occurence of
actinomycetes in the river Thames, Burman
(1973) found 59–200 streptomycetes per ml and
10–20 micromonosporae per ml. These organisms
were found to grow on decaying vegetation
on riverbanks and mud flats at low water or on
floating mats of decaying algae or other vegetation.
Under these conditions odorous substances
are produced, and subsequent increase in river
levels washes them into the water, thus giving
rise to “earthy taste” complaints. Of these odorous
compounds, geosmin and methyl iso-borneol
are most often detected (Gerber, 1979a; Gerber,
1979b). As summarized by Wood et al. (1983),
the prevention of earthy tastes in reservoirs and
water supply systems depends on locating the
production sites and determining the patterns of
distribution of these compounds (Silvey and
Roach, 1975; Lechevalier et al., 1980).
Burman (1973) studied the fate of actinomycetes
of river water in the course of production
of drinking water. Filtration processes
reduce the number of streptomycetes considerably.
In the distribution system, a new type
named “aquatic strains of Streptomyces” has
appeared, which can be enriched (for details, see
Burman, 1973).
Okazaki and Okami (1976), Cross (1981), Weyland
(1981b), Weyland (1981a), Weyland and
Helmke (1988), and Goodfellow and Haynes
(1984) have reviewed the occurrence of streptomycetes
in marine habitats including sediments.
Two localities can be distinguished: 1) the littoral
and inshore zone and 2) deep-sea sediments. From
both localities streptomycetes can be isolated;
however, similarly to streptomycetes in freshwater
environments, most of these organisms are
“survivors” rather than constituents of the
autochthonous microflora. From the littoral zone,
streptomycetes have been isolated both from sediments
(Roach and Silvey, 1959) and from decaying
seaweed (Siebert and Schwartz, 1956). These
isolates were able to grow on polymeric substances,
e.g., agar and chitin (Humm and Shepard,
1946), alginate and laminarin (Chesters et al.,
1956), and cellulose (Chandramohan et al., 1972),
characteristic for these microsites.
In sediments, both the depth and location of
sample sites play important roles in determining
the ratio of different taxa of actinomycetes
(Weyland, 1981b; Weyland and Helmke, 1988).
Sites in the open sea generally contain only low
numbers of actinomycetes (viable counts were
about 100 CFU per ml of wet sediment). The
distribution (horizontal as well as vertical) is
assumed to correlate with the physiological
properties of the three taxa Streptomycetes,
Micromonosporae and Rhodococci barotolerance
(Helmke, 1981), halotolerance, and psychrophilism
(Weyland, 1981a). Goodfellow and
Haynes (1984), however, were not able to find a
correlation between salinity, pH, or depth and
the number of actinomycetes recovered from
marine sediments. In this study numerous isolates
are described in detail, and from a total of
732 organisms, 250 belonged to Streptomyces,
250 to Micromonospora, 140 to Rhodococcus,
and 92 to Thermoactinomyces. A selected number
of isolates belonging to the genus Streptomyces
was subsequently identified using 41
diagnostic tests and applying the MATIDEN
program and the Streptomyces probability matrix
(Williams et al., 1983b). Around 50% of these
streptomycetes were similar to those grouped
into cluster 1 of Williams et al. (1983a).
Okami and Okazaki (1978) found streptomycetes
(300–1270 colonies per cm3) mainly
in the sediments of shallow seas (70–520 m
deep), whereas in samples 700–1600 m deep,
Micromonospora dominated. No actinomycetes
were obtained from depths of 2800 and 5000 m
in the Pacific Ocean. A higher salt tolerance of
marine streptomycetes than of their terrestrial
counterparts was observed. However, already
Tresner et al. (1968) found that salt tolerance
among streptomycetes is widespread and this
feature may be due to selection of the more tolerant
organisms in marine habitats. Few streptomycetes
isolated from marine environments
were found to be obligate halophiles (Okazaki
and Okami, 1976).
Notably, marine habitats are essential for
screening programs in the search of new antimicrobial
and anti-insecticidal compound producers.
In early studies, Nissen (1963) found a high
percentage of antibiotic-producing streptomycetes
from decaying seaweed, and subsequent
investigations confirmed these findings (Okami
and Okazaki, 1972; Okami et al., 1976; Hotta et
al., 1980).
Isolation
The procedures of isolation of streptomycetes
(extensively summarized by Korn-Wendisch and
Kutzner, 1992a) are partly summarized in the
next paragraphs. Additional information about
isolation for special purposes, growth of streptomycetes,
and preservation of streptomycetes can
be obtained from the excellent textbook Practical
Streptomyces Genetics (Kieser et al., 2000).
Generally, all procedures for the isolation of
microorganisms are influenced by the nature
of the microorganism and the number of
propagules relative to the number of other
microbes within the habitat (Stolp and Starr,
1981). If the organism to be isolated is best
adapted to the selected isolation conditions, then
direct plating of a serial dilution on a nutrient
agar medium can readily lead to a pure culture.
This procedure does not work well for isolation
of streptomycetes. These actinomycetes are isolated
usually by enrichment or use of selective
media and specific isolation conditions or both.
As already pointed out by Korn-Wendisch and
Kutzner (1992a), Streptomycetaceae members
can be isolated using general selective isolation
procedures (Williams and Wellington, 1982b;
Williams and Wellington, 1982c; Williams et al.,
1984a). These procedures require 1) choice of
the material containing the selected microorganisms,
2) pretreatment of the sample, and in some
cases, enrichment of the chosen microbial
groups, 3) use of selective media or selective
incubation conditions or both, and 4) colony
selection on the basis of colony morphology and
purification.
Streptomycetes can be isolated from a wide
variety of habitats, and most isolation procedures
involve extraction from soil or another
environmental sample followed by dilution of
the cells (cell aggregates) to allow cultivation on
solid media.
Isolation and Enrichment from Soil
Because the vegetative mycelium and spore
chains are often closely associated with the mineral
and organic particles of the soil, vigorous
shaking of the sample with the diluent is often
needed to suspend the spores of mycelial fragments.
Use of glass beads and agitation on a
shaker may aid suspension. In the literature,
methods involving mechanical devices such as
the Ultrasonics sonicator-disrupter, Ultra-Turrax
homogenizer, Turmix blender, Waring blender,
or a mortar and pestle are described, but a
detailed comparison of the dilution efficiency of
these pretreatments is still missing. Chemical disruption
methods are also reported in literature.
Gently shaking soil samples with an ionexchange
resin Chelex-100 (Biorad) followed by
differential centrifugation and filtration was used
to separate the mycelium from the spores
(Herron and Wellington, 1990).
A subsequent treatment of samples (i.e., preparing
dilutions and plating) differs little from
general bacteriological practice. Most often the
coarse particles of the soil suspension are
allowed to settle before dilutions are made. But
soil particles may also be used directly for incu-
bation of “soil plates” (Warcup, 1950), which are
also used to isolate fungi. The addition of lime to
soil can enrich for streptomycetes (see chapter 2
of Kieser et al. [2000] and references therein).
Isolation plates may be surface-inoculated with
a sterile glass rod (or Drigalski spatula).
Spread of motile bacteria via water films can
be avoided by drying the plates at 45°C before
incubation or mixing the soil suspension with
the molten agar, which is highly recommended
(Korn-Wendisch and Kutzner, 1992a). The addition
of CaCO3 to air-dried soil samples (10:1 w/
w) and subsequent incubation at 26°C for 7–9
days in a water-saturated atmosphere can lead to
a 100-fold increase of streptomycete colonies on
isolation plates (Tsao et al., 1960; El-Nakeeb and
Lechevalier, 1963).
The enrichment of streptomycetes by soil
amendment with keratin was first carried out by
Jensen (1930). Later the addition of chitin to soil
was found to stimulate growth of actinomycetes
(Williams and Mayfield, 1971b). Williams and
Robinson (1981) similarly obtained the enrichment
of acidophilic and neutrophilic streptomycetes
in acidic soil and litter containing pure
and fungal chitin. Another isolation strategy
using chitin in the form of insect wings has been
used as a baiting method (Veldkamp, 1955; Jagnow,
1957; Okafor, 1966). Porter and Wilhelm
(1961) studied isolation using various other
organic materials, like salmon viscera meal, peanut
meal, cottonseed meal, and dried blood flour
(15 mg/g of soil). They found that incubation of
the enrichment cultures under moist conditions
led to an up to 1000-fold increase in the number
of streptomycetes.
Besides the frequently used arginine glycerol
agar (El-Nakeeb and Lechevalier, 1963), the
following media (details below) are also often
applied for selective isolation of streptomycetes:
HV agar (Hayakawa and Nonamura, 1987a,
1987b), colloidal chitin agar (Hsu and Lockwood,
1975), and reduced arginine starch salts agar.
Several physical, chemical and biological
methods have been studied to reduce or inhibit
other microbes (for review, see Goodfellow and
Williams, 1986a). Nüesch (1965) centrifuged soil
suspensions for 20 min at 1600 × g to separate
the spores of streptomycetes (in the supernatant)
from other bacteria and spores of fungi (in the
sediment); however, the method has not been
very successful. Using a similar approach, El-
Nakeeb and Lechevalier (1963) obtained a significantly
smaller number of streptomycete colonies
as compared with the control. Voelskow
(1988/1989) described a simple sedimentation
method in which 1 g of soil was suspended in
15 ml of salt solution, vigorously shaken, and
then treated with ultrasonic vibrations. After 1,
2, and 4 h of sedimentation, samples were taken
from different levels of this solution, further
diluted, and plated on agar surfaces.
Initial drying and heating procedures were
applied because arthrospores have a relatively
high resistance to low moisture tension. Drying
of the sample or prolonged storage at ambient
temperatures for mesophiles and at 50–60°C for
thermophiles led to a relative increase in streptomycete
concentrations. Williams et al. (1972)
showed that heat treatment of soil (40–50°C, 2–
16 h) leads to a significant reduction of the vegetative
bacterial proportion without affecting the
colony counts of streptomycetes.
Membrane filtration has been mainly used for
the enrichment of streptomycetes from water
samples (Burman et al., 1969) and from seawater
and mud (Okami and Okazaki, 1972), but it has
also been a first step in the isolation of streptomycetes
from soil (Trolldenier, 1966). These
authors used 1 ml of a series of 10-fold dilutions,
which were membrane-filtered (0.3-μm pore
size). This filter was then placed upside down on
a suitable agar medium, which was supplemented
with 10% compost soil. Colonies
developed between the agar surface and the
membrane filter, and the streptomycetes (but not
other bacteria or fungi) were able to grow
through the pores. Using this method, the selective
effects of the physical barrier (the membrane)
and the soil lead to a 3–5-fold increase in
the number of streptomycete colonies in comparison
with poured plates without soil.
Hirsch and Christensen (1983) introduced the
use of cellulose ester membrane filters (pore
size 0.01–3.0 μm), which were placed onto
nutrient agar containing antifungal antibiotics
(cycloheximide and candicidin). These plates
were inoculated with different samples from
soil, water and vegetable materials. After 4
days, the hyphae of actinomycetes penetrated
the filter pores and grew on the underlying agar
medium, whereas the growth of the other bacteria
was restricted to the surface of the filter.
To allow further development of actinomycete
colonies, the membrane filter was removed and
the plates were reincubated. Filters (0.22–
0.45 μm) were also found to be suitable for the
exclusive recovery of actinomycetes. Polsinelli
and Mazza (1984) and Hanka et al. (1985) used
this approach independently.
Several authors have added chemicals to
improve the isolation efficiency. Phenol treatment
of a dense soil suspension (1.4% for
10 min) was recommended to eliminate bacteria
and fungi, but El-Nakeeb and Lechevalier (1963)
obtained less favorable results with this method.
Chloramine, ammonia, and sodium hypochlorite
were mainly employed for the treatment of
water samples, because it has been found that
streptomycetes and other actinomycetes are
slightly more resistant to these agents than other
bacteria are (Burman et al., 1969).
Isolation of Airborne Spores
For the isolation of Streptomyces spores from
self-heating material such as hay or compost,
samples can be agitated in a wind channel (Lacey
and Dutkiewicz, 1976b) or sedimentation chamber
(see below; Lacey and Dutkiewicz, 1976a).
Plates are then inoculated with the aerosol using
an Andersen sampler (Goodfellow and Williams,
1986a). This method, widely employed for the
isolation of thermophilic actinomycetes, may
also be used for the isolation of mesophilic streptomycetes
from soil.
In addition, other devices like filtration samplers
(e.g., Sartorius MD 8) are suitable for the
sampling of airborne streptomycetes.
Use of Selective Media and Incubation
Selective media always play an important role in
the isolation of the desired microorganisms.
A number of factors can be varied: 1) nutrient
composition and concentration of the isolation
medium, i.e., choice of carbon and nitrogen
sources preferred by the organisms; 2) addition
of chemical substances to inhibit selectively the
accompanying flora of the natural habitat or those
which are stimulating the desired organisms; 3)
pH, for acidophilic, neutrophilic, and alkalophilic
organisms; and 4) temperature, e.g., for the isolation
of thermophiles or psychrophiles.
Many different media have been formulated
empirically and proposed for the isolation of
streptomycetes. Selected carbon and nitrogen
compounds listed in Table 10 are especially suitable
for the isolation of these organisms. The
most frequently used media with their formulas
are listed in Tables 11 and 12. Alternatively,
streptomycetes can be grown on very poor media
such as water agar.
Different Carbon and Nitrogen Sources for
Enrichment Because it was early recognized
that streptomycetes can degrade chitin (Veldkamp,
1955; Jagnow, 1957), a chitin medium was
devised by Lingappa and Lockwood (1962) for
selective isolation. However, the authors and
later also El-Nakeeb and Lechevalier (1963)
found that their chitin agar was only a little better
than water agar. Hsu and Lockwood (1975)
added mineral salts to this medium (Table 11),
which was shown to be useful for the isolation
of actinomycetes (Streptomyces, Nocardia and
Micromonospora) from water samples but had
little effect when isolating actinomycetes from
soil. Note that chitinolytic activity is not a genus
specific feature for streptomyces. Williams et al.
(1983a) found that only 25% of over 300 strains
were strongly chitinolytic, so this widely used
medium selects the chitinolytic streptomycete
strains, which may not be the most abundant
strains in soil. Starch is degraded by the vast
majority of streptomycetes and can therefore
be used as a selective carbon source. An early
finding was that the combination of starch with
nitrate, which is utilized by many streptomycetes
(in contrast to other bacteria) as nitrogen source,
is very useful for the selective isolation of streptomycetes
(Flaig and Kutzner, 1960a). Küster
and Williams (1964), who improved this medium,
stated: “The three best media, allowing good
development of streptomycetes while suppressing
bacterial growth, were those containing
starch or glycerol as the carbon source with
casein, arginine or nitrate as the nitrogen source.”
Benedict et al. (1955) were the first to report
that the combination of glycerol and arginine
favored streptomycete isolation. El-Nakeeb and
Lechevalier (1963) found that this medium
(Tables 11 and 12) was superior to nine other
media, resulting in higher number and proportion
of streptomycete colonies.
Other compounds (e.g., pectin [Wieringa,
1955], poly-ß-hydroxybutyrate [Delafield et al.,
1965], rubber [Nette et al., 1959], cholesterol
[Brown and Peterson, 1966], elemental sulfur
[Wieringa, 1966], and natural and artificial humic
acids [Hayakawa and Nonomura, 1987a, 1987b])
have been used successfully, and most of them
strongly select certain organisms producing
visible zones of clearing or other changes in the
medium.
The use of compounds with antifungal activity
(antibiotics) as supplements to isolation media
has also been widely used to suppress fungal
growth (Table 10). The most frequently used
compound is cycloheximide (actidione, 50–
100 μg/ml) by Williams and Davies (1965). Pimaricin
and nystatin (each 10–50 μg/ml) were found
to be even more effective (Williams and Davies,
1965).
The use of compounds with antibacterial activity
is more restricted because actinomycetes are
often also sensitive to them. Williams and Davies
(1965) found polymyxin (5 μg/ml) and penicillin
(1 μg/ml) suppressed bacterial flora; however,
they also inhibited streptomycetes. Preobrazhenskaya
et al. (1978) showed that the genera of
Actinomycetales may differ significantly in their
sensitivity to antibacterial antibiotics and that
streptomycetes were the most sensitive group.
Thus, the use of antibacterial compounds may be
more helpful in isolating other genera of this
order (Cross, 1982).
In contrast, some antibiotics may facilitate the
isolation of certain species or groups of Streptomyces.
For example, the selective isolation of
members of the Streptomyces diastaticus cluster
sensu Williams et al. (1983a) was achieved on
starch casein medium containing rifampicin
(50 μg/ml; Vickers et al., 1984). Wellington et al.
(1987) found a similar effect with several media
containing different C and N sources as well as
with media supplemented with inhibitors.
Hanka et al. (1985) described a selective
isolation medium for streptoverticil-producing
Streptomyces species containing cycloheximide
and nystatin (each 50 g/ml, to control fungal
growth) and oxytetracycline (25 μg/ml, to suppress
other actinomycete genera and other Streptomyces
groups). Hanka and Schaadt (1988)
enhanced selectivity of this medium by adding
lysozyme (1000 μg/ml). Also, sodium propionate
can suppress the competing fungi (Crook et al.,
1950; Table 11), and Rose Bengal (35 mg/liter)
in starch casein nitrate agar (Ottow, 1972) can
suppress most of the bacteria and inhibit the
spreading growth of fungi.
pH of the Isolation Medium and Incubation
Temperature Most streptomycetes grow optimally
at neutral pH values, i.e., are neutrophilic.
Therefore, the pH of most isolation media is 7.0–
7.5. However, for the isolation of acidophilic
streptomycetes, the medium pH is 4.5 (Khan and
Williams, 1975), and for alkalophilic strains, it
is 10–11 (Mikami et al., 1982). Most streptomycetes
isolated from soils are mesophilic, and
therefore plates are most often incubated at 22–
37°C (mostly at 28°C). Psychrophilic strains (e.g.,
from marine environments) grow at 15–20°C. In
contrast, thermotolerant and thermophilic representatives
can be isolated at higher temperatures
(40, 45, 50, or 55°C). Note that thermophiles
often form colonies after a short period of incubation
(within 2–5 days), and mesophilic members
produce visible colonies within 7–14 days.
However, marine and other psychrophilic organisms
often need several weeks (up to 10) for the
formation of visible colonies.
Colonies of Streptomyces are in most cases
readily recognized by their macroscopic and
microscopic appearance. In most cases, Streptomyces
are easily purified by picking colonies and
transfer to a nonselective medium. According to
Williams and Wellington (1982b), purification
is “undoubtedly the most time-consuming and
often the most frustrating stage of the isolation
procedure.” Acidiphilic members of the Streptomycetaceae
can be isolated on acidified starch
casein agar containing cycloheximide and nystatin
(Kim et al., 2003).
Isolation of Antibiotic Producers and
Strains for Genetic Studies
The isolation of antibiotic producing streptomycetes
follows the same procedures as given
above. Most often, the activity is tested after isolation
of pure cultures, but this can also be combined
with the isolation procedure. Thus, strains
exhibiting antibiotic activity can be recognized
even on the initial dilution plates if they are
treated with an appropriate test organism, either
by flooding or by spraying. Zones of inhibition
can be detected after further incubation (Lindner
and Wallhäusser, 1955; Wilde, 1964). As
an alternative, a simple replication procedure
allows the examination of the antibiotic activity
of the colonies against selected sensitive organisms
(Lechevalier and Corke, 1953). Selective
techniques for isolation and screening of actinomycetes
that produce antibiotics and other secondary
metabolites of clinical relevance have
been summarized by Nolan and Cross (1988). In
addition, Kieser et al. (2000) have provided a
set of protocols for selective isolation of streptomycetes,
generating spore suspensions, and several
other more sophisticated procedures.
Isolation of Thermophilic Streptomycetes
As already stated, thermophilic streptomycetes
and other thermophilic actinomycetes are often
isolated from samples derived from high temperature
environments (e.g., compost materials,
manure heaps, and fodders). The high temperatures
(45–60°C) can serve as a selective condition
that favors enrichment of the desired organisms
(Festenstein et al., 1965).
As pointed out by Greiner-Mai et al. (1987), a
very important requisite for the isolation of thermophilic
streptomycetes is a humid atmosphere,
which can be achieved by incubating plates in
large jars with water in the bottom. Alternatively,
the sealing of Petri dishes with masking tape is
also effective.
Interestingly, the media recommended for the
isolation of thermophilic actinomycetes, including
streptomycetes, contain higher nutrient
concentrations than those used for mesophilic
strains. In addition, antifungal agents and
antibacterial agents are sometimes added as supplements
(Lacey and Dutkiewicz, 1976b; Goodfellow
et al., 1987c). For some special isolation
techniques, the reader is referred to the papers
of Uridil and Tetrault (1959), Fergus (1964), Gregory
and Lacey (1963), and Cross (1968). Additional
information can be obtained from the
publications of Kim et al. (Kim et al., 1996; Kim
et al., 1998; Kim et al., 2000).
Isolation from Aquatic Habitats
For the isolation of streptomycetes from water,
the media listed in Table 11 can be used. In a
comparative study on the suitability of media,
Hsu and Lockwood (1975) found that chitin-agar
was superior to the other four (egg albumin,
glycerol arginine, starch casein, and Actinomyces
isolation agar; see also Table 11).
Water samples can be directly streaked onto
the solid medium after dilution. When low numbers
are expected, the samples can be concentrated
by membrane filtration (for details, see
Burman et al. [1969] or Trolldenier [1967]).
Streptomycetes from marine habitats were
successfully isolated on media containing 25
or 75% seawater (Weyland, 1981a; Weyland,
1981b), artificial seawater (Goodfellow and
Haynes, 1984), or deionized water supplemented
with 3.0% NaCl (Okami and Okazaki, 1978). For
further details, see also Weyland (1981b) and
Goodfellow and Haynes (1984).
Isolation from Infected Plants
For the isolation of streptomycetes from diseased
plant tissue (i.e., from scabby potato or
beet surface layers), three general steps have
been recommended (see also Korn-Wendisch
and Kutzner, 1992a): 1) sterilization of the surfaces
of tubers, beets, or roots; 2) maceration of
the plant tissues; and 3) use of appropriate media
for plating.
Methods for isolating Streptomyces scabies
from potatoes have been described in detail by
several authors (Taylor, 1936; KenKnight and
Munzie, 1939; Menzies and Dade, 1959; Adams
and Lapwood, 1978; Archuleta and Easton,
1981).
Cultivation
Nutritional Requirements and Media
for Sporulation
The vast majority of streptomycetes are nonfastidious
organisms, having a chemoorganotrophic
metabolism. The nutritional requirements are
(in most cases) restricted to an organic carbon
source (e.g., starch, glucose, glycerol and lactate)
and an inorganic nitrogen source (NH4
+ or NO3
–),
in addition to the essential mineral salts for
growth. The necessity to amend media with
specific trace elements has not been studied in
detail. Many of the early used media (even the
“synthetic” media) were not supplemented with
trace elements, although the positive effect of
trace elements in soil on the growth of streptomycetes
has been reported (Spicher, 1955).
Other authors used quite different recipes
(Tables 11 and 12), each containing only a select
number of metal ions. A rather complete mixture
(SPV-4; Table 12) has been found to be optimal
for actinomycetes and other bacteria (Voelskow,
1988/1989).
Because no specific requirement for vitamins
or organic growth factors has been described,
“synthetic media” can be used for their cultivation.
Notably, complex organic substrates (e.g.,
oatmeal, yeast extract, or malt extract) are well
utilized and enhance growth rates and biomass
production. A combination of a complex organic
carbon source with a single amino acid as nitrogen
source (e.g., glutamic acid, arginine or asparagine)
is also suitable.
Several authors have proposed “general
media” for streptomycetes that allow the completion
of the streptomyces’ life cycle, i.e. germination
of spores, growth of substrate and aerial
mycelium, and formation of spores (visible
because of the typical color of the spores). Some
of these media have been used in the International
Streptomyces Project (ISP). Of the great
number of useful media (Pridham et al., 1956/
1957), four are of particular value and most often
used (Table 12). Additional media are listed by
Waksman (1961) and by Williams and Cross
(1971a). CaCO3 added to some media not only
supplies Ca2+ for growth but also neutralizes
acids produced by many streptomycetes. These
media also allow good sporulation. Because
macroscopically heavy aerial mycelia may contain
very few spores and aerial mycelia hardly
detectable by the naked eye may be a good
source of spores, it is advisable to check these
cultures microscopically. Specialized media,
especially for genetic studies, are given by Kieser
et al. (2000).
Media Containing Soil, Clay, Minerals and
Calcium Humate
Addition of soil to isolation media increases the
number of colonies as it promotes growth, sporulation
and pigmentation (Trolldenier, 1966).
Martin et al. (1976) observed the stimulation of
both growth and metabolic activity of some actinomycetes
when montmorillonite or Ca-humate
was added to a liquid medium. A similar effect
was observed for clay in dialysis tubes after a
short lag period, which was explained by a possible
adsorption of one or more inhibitory substances
produced during growth. Martin et al.
(1976) also reported a positive effect of these
adsorbing materials on the genetic stability of
other bacteria and on fungi.
Temperature and Oxygen
As pointed out above, most streptomycetes are
mesophilic organisms; however, psychrophilic as
well as thermotolerant and thermophilic species
are also known. Note that in many instances, the
optimum temperature for fast growth or maximal
yield may not be the best choice for studying
the production of secondary metabolites (e.g.,
antibiotics and pigments). Most streptomycetes
have an obligately aerobic metabolism, but many
streptomycetes are able to reduce nitrate to
nitrite under strictly anaerobic conditions.
In semisolid agar with a nutrient medium, they
grow at the surface of the agar column; however,
in semisolid agars with a poor medium or nonutilizable
carbon source, they grow microaerophilically.
A stationary liquid culture grows as a
pellicle at the surface, and the medium remains
completely clear.
Cultivation and Preparation of Inoculum
For subcultivation and maintenance, and for
most diagnostic tests, streptomycetes are preferably
cultivated on solid media in dishes or slants.
On solid media many strains produce aerial
mycelia and spores when the entire surface is
covered by confluent growth. Since Streptomyces
colonies, in contrast to most molds, spread over
a limited distance, a point inoculation will often
not lead to a confluent growth.
Some strains show sporulation only when the
plates are cross-hatched inoculated, i.e., by a
method which leaves empty spaces between
the streaks. Sporulation generally occurs better
under dry conditions. For this reason, slants
should be incubated horizontally for the first two
days to allow the liquid to soak into the surface
of the agar (Hopwood et al., 1985). The starting
material should be a suspension of inoculum in
liquid (Kieser et al., 2000).
The propagation of cultures by successive
rounds of mass culture should be avoided, but
instead a single colony should be picked and
streaked to start the next culture. Especially in
genetic studies, this reduces the accumulation of
revertants or the gradual loss of selected plasmids
or both (Kieser et al., 2000).
Note that morphological heterogeneity is
often observed when streptomycetes are cultivated
on solid media (for more details, see
Kieser et al., 2000).
Because it may be difficult to produce a homogeneous
suspension from the grown colonies (a
prerequisite for inoculation of some diagnostic
tests; Kämpfer et al., 1991b), precultivation in
liquid media is sometimes useful. This is the case
for certain physiological studies (e.g., degradation
tests), for the provision of cell material for
biochemical analysis, and for the production
of secondary metabolites (e.g., antibiotics) or
enzymes. For these purposes, streptomycetes can
be cultivated in liquid medium with agitation.
The use of liquid cultures started from an inoculum
of spores is also recommended for many
detailed studies, e.g., for preparation of protoplasts
for fusion, transformation or transfection.
Notably, the multicellular lifestyle of streptomycetes
complicates the study of metabolic
properties, where all cells of the initial suspension
should be in the same physiological condition.
In general, streptomycetes grow by mycelial
elongation and branching. But when central
parts of the colony become nutrient limited,
physiological homogeneity cannot be sustained.
To overcome these problems, spore germlings
are used for physiological studies, even though a
large amount of spores is needed. Other solutions
include liquid cultures supplemented with
dispersants, like sucrose, polyethylene glycol,
Junlon®, starch, agar and carboxymethylcellulose.
Chapter 2 of Kieser et al. (2000) summarizes
the advantages and disadvantages. Since
streptomycetes are highly aerobic, the cultures
need to be shaken during incubation. Use of
Erlenmeyer flasks with indentations or stainless
steel springs is recommended, but for small
quantities (3–5 ml being enough for some physiological
tests), tubes in a slanted position on a
shaker or roller also allow excellent supply of
oxygen. Note that some secondary metabolites
(e.g., antibiotics and pigments, which are produced
on solid media) may fail to be synthesized
under these conditions.
Korn-Wendisch and Kutzner (1992a) recommend
two media that have been widely used for
the submerged cultivation of streptomycetes (g/
liter): 1) GPYB broth (glucose, 10.0; peptone
from casein, 5.0; yeast extract, 5.0; beef extract,
5.0; CaCl2 · 2H 2O, 0.74; pH 7.2) and 2) soybean
meal-mannitol nutrient medium (soybean meal,
20.0; mannitol, 20.0; pH 7.2). Two kinds of noculation
material can be employed for subculturing
streptomycetes: 1) arthrospores and 2)
vegetative mycelium, occasionally including
“submerged spores” (Wilkin and Rhodes,
1955). Kieser et al. (2000) recommend similar
procedures.
Spore suspensions can be used over a period
of several weeks when stored at 4°C, but since
the spores tend to settle and clump, the addition
of a few glass beads to the screw-cap tube helps
to resuspend the spores before use. Chapters 8
and 9 of Kieser et al. (2000) describe the preparation
of mycelia for detailed DNA or RNA
studies.
Preservation
Several different procedures have been
employed for the short- and long-term preservation
of microorganisms (Kirsop and Snell, 1984).
Three short-term preservation methods have
been described by Korn-Wendisch and Kutzner
(1992a). First, agar slope cultures may be stored
at 4°C for few months. Second, spore suspensions
can be mixed with soft water agar and kept
at 4°C (Kutzner, 1972). And third, glycerol can
be added to spore suspensions (final concentration,
10%, v/v) and stored at –20°C (Wellington
and Williams, 1978); these cultures (after thawing)
can serve as inoculum for most diagnostic
tests except carbon utilization (Williams et al.,
1983a). Kieser et al. (2000) recommended for
long-term preservation the preparation of a
spore suspension in 20% glycerol and freezing
at –20°C. Another procedure is the growth of
strains in complex media (like trypticase soy
broth [TSB] agar), addition of 20% glycerol plus
10% lactose, and storage in the vapor phase of
liquid nitrogen. In addition, drying on unglazed
porcelain beads (Lange and Boyd, 1968), followed
by soil culture (Pridham et al., 1973), and
lyophilization (Hopwood and Ferguson, 1969)
are used. For longer preservation (see The Family
Nocardiopsaceae in this Volume), spore suspensions
or homogenized mycelia are mixed
with glycerol to a final concentration of 25% and
kept at 25°C (Wellington and Williams, 1978).
Alternatively, spores and mycelia suspended in
10% skim milk are lyophilized. A very simple,
reliable, and time-saving method is liquid nitrogen
cryopreservation of living cells in small polyvinyl
chloride (PVC) tubes (“straws”) at –196°C.
The procedure has been tested for various actinomycetes.
The strains are harvested from wellsporulated
cultures grown on suitable agar
media in Petri dishes. A 2 × 25-mm piece of sterile
PVC tubing is pressed into the mycelial mat
and agar and carefully raised to excise the agar
plug. This is repeated until the tube is filled with
agar. The filled tubes are placed in a sterile cryovial
(the screw cap marked with the strain
accession number). A 1.8-ml vial will hold up to
13 tubes. Two vials are prepared for each strain
and then fixed to a metal clamp for freezing in
the gas phase of a liquid nitrogen container.
After 10–15 min, when temperature falls below
–130°C, the clamp can be immersed in the liquid
phase at –196°C. A container with a capacity of
250 liters will hold at least 8000 vials or 4000
strains. For viability testing, one tube is removed
from the vial within the nitrogen gas atmosphere
of the container and placed directly and thawed
on a suitable agar medium. After a few days
incubation, the mycelium will be visible. For
those strains that do not produce an abundant
mycelium, the plugs may be pushed out of the
tubes by a sterile needle.
Detection of Streptomyces Using
Cultivation-independent Methods
Microorganisms in the environment can be
identified without cultivation by retrieving and
sequencing macromolecules and using oligonucleotide
probes (largely based on small subunit
rRNA; Amann et al., 1995; Rappé and Giovannoni,
2003). Stackebrandt et al. (1991b) developed
16S rRNA-targeted oligonucleotide probes
specific for certain Streptomyces species and
subsequently studied bacterial diversity in a soil
sample from a subtropical Australian environment
(Stackebrandt et al., 1993). They found that
most sequences were from alpha subclass Proteobacteria
and only a few from streptomycetes.
Hahn et al. (1992), using the in situ hybridization
approach, were unable to analyze bacterial populations
in soil without prior activation by adding
nutrients. Growing cells, e.g., Streptomyces scabies
hyphae, were easily detected. The use of
specific primers in connection with environmental
clone libraries is a powerful approach for
study of the microbial diversity in soil (Felske et
al., 1997; Rheims et al., 1999; Rintala et al., 2001;
Courtois et al., 2003) and discovery of novel bioactive
metabolites (Donadio et al., 2002). More
details are given in The Family Streptomycetaceae,
Part II: Molecular Biology in this Volume.
Acknowledgments. The basis of the sections on
ecophysiology, isolation and habitats has been
the excellent and comprehensive treatise of
Korn-Wendisch and Kutzner (1992a) from the
second edition of The Prokaryotes, which is still
recommended for deeper study of classical
approaches in Streptomyces biology.