Taxonomic and functional diversity of Streptomyces in a forest soil


Correspondence: Cyril Bontemps, Dynamique des Génomes et Adaptation Microbienne, UMR1128, Université de Lorraine, Vandoeuvre-lès-Nancy, F-54506, France. Tel.: +33 3 83684190; fax: +33 3 83684499; e-mail:


In this work we report the isolation and the characterization of 79 Streptomyces isolates from a French forest soil. The 16S rRNA gene phylogeny indicated that a great diversity of Streptomyces was present in this soil, with at least nine different and potentially new species. Growth plate assays showed that most Streptomyces lineages exhibit cellulolytic and hemicellulolytic capacities and potentially participate in wood decomposition. Molecular screening for a specific hydrogenase also indicated a widespread potential for atmospheric H2 uptake. Co-culture experiments with representative strains showed antagonistic effects between Streptomyces of the same population and between Streptomyces and various fungi. Interestingly, in certain conditions, growth promotion of some fungi also occurred. We conclude that in forest soil, Streptomyces populations exhibit many important functions involved in different biogeochemical cycles and also influence the structure of soil microbial communities.


Streptomyces are Gram-positive filamentous bacteria of the actinomycetal order with more than 600 described species ( ubiquitous in soils, sediments and seawater (Labeda et al., 2012). Thanks to an important carbohydrate-active enzyme reservoir, some Streptomyces isolates participate in the degradation of recalcitrant biopolymers such as cellulose, hemicelluloses or lignocellulose found in wood (Gottschalk et al., 1999; Bruce et al., 2010; Adams et al., 2011; Amore et al., 2012; Li et al., 2012b). These capacities give them a trophic advantage when amendments are added into the soil. Indeed, Streptomyces can dominate certain cellulolytic microbial communities after long-term manure application (Ulrich et al., 2008). In a mesocosm, cellulose and lignin inputs increased Streptomyces density more than glucose (Schlatter et al., 2009). Such observations highlight the potential importance of Streptomyces for nutrient turnover in forests. It could also partly explain why in forest soils, where wood is the major carbon source (Martínez et al., 2005), Streptomyces are always among the most represented bacterial genera (Bruce et al., 2010; Nacke et al., 2011; Montaña et al., 2012). In addition to carbon cycling, Constant et al. (2010) showed in Streptomyces the presence of a high affinity hydrogenase involved in dihydrogen uptake. Despite a low number of isolates, these results suggested that this enzyme may be widespread in the genus Streptomyces and may actively participate in dihydrogen cycling in forests.

The important secondary metabolism of Streptomyces also represents a great tool kit to interact with other microorganisms. Antibiotic production confers Streptomyces crucial advantages in ecological niche occupancy by exerting a high selective pressure on a wide range of organisms from insects (Liao et al., 2009; Wang et al., 2011), fungi (Meschke & Schrempf, 2010; Li et al., 2012a; Zhang et al., 2013) and other bacteria species (Park et al., 2011; Poulsen et al., 2011; Lawson & Stevenson, 2012) as well as closely related Streptomyces strains (Neeno-Eckwall et al., 2001; Vetsigian et al., 2011). In other cases, the Streptomyces-secreted secondary metabolites are beneficial to other organisms. They can symbiotically help to degrade cellulose for xylophagous insects (Schäfer et al., 1996; Adams et al., 2011), protect fungal gardens of ants (Seipke et al., 2011), suppress phytopathogenic fungi (Quecine et al., 2008; De Oliveira et al., 2010; Shi et al., 2010), elicit plant defense (Lerat et al., 2009; Lin et al., 2012) and promote mycorrhizal symbiosis (for review Frey-Klett et al., 2007).

Many Streptomyces isolates have been described as having functions that influence their environment, e.g. biogeochemical cycles or in the community structure of other microbes. However, studies generally focus on a single or few isolates considered the most efficient after a single function screening (Semêdo et al., 2004; Xu & Yang, 2010; Da Vinha et al., 2011). Thus, information is sparse regarding function variation between Streptomyces isolates of the same natural population. To better understand the ecological roles of Streptomyces as a population, we here assessed the Streptomyces taxonomic diversity in a forest soil, and studied the distribution of already known functions of interest among isolates.

We show that Streptomyces were taxonomically diverse in our sample, and that they had a great potential as a population to participate in decomposition of wood components and dihydrogen uptake. Confrontation tests indicated that they frequently inhibited growth of other Streptomyces from the same sample and modulated the growth of different fungi in either a beneficial or an antagonistic way.

Material and methods

Sampling and Streptomyces collection

The upper organic layer of soil (0–5 cm depth) was collected at nine points of the Montiers Forest experimental site. This site belongs to the long-term environmental research monitoring and testing system Andra OPE and to the French network SOERE F-ORE-T (France, ANDRA/INRA/ONF site). Beech is the dominant tree species of this forest. For each sample, 1 g of soil was suspended in 3 mL of sterile water, thoroughly vortexed for 10 min and incubated for 1 h at 50 °C. Several dilutions from 10−3 to 10−6 were spread on actinomycetal selective glycerol–arginine plates (Suutari et al., 2002) supplemented with 5 μg mL−1 carboxin and 10 μg mL−1 nalidixic acid. Streptomyces-like isolates were randomly selected based on morphology after 10 days of incubation at 30 °C. After three purifications on SFM plates (Kieser et al., 2000), isolates were stored at −80 °C as a mycelial or spore suspension.

Gene sequencing and phylogenetic analysis

PCR amplifications were performed using the ThermoPol kit® (New England Biolabs) with genomic DNA extracted as described by Kieser et al. (2000). For all strains, a nearly full-length 16S rRNA gene (1450 bp) was amplified with the universal primers 16S_FD1 (5′-AGAGTTTGATCCTGGCTCAG) and 16S_rP2 (5′-AAGGAGGTGATCCAGCC) as described by Weisburg et al. (1991) and sequenced with primer 16S_FD1. The almost full length hhyL gene (1396 bp) was amplified and sequenced with primers Nife-244f (5′-GGGATCTGCGGGGACAACCA) and Nife-1640r (5′-TGCACGGCGTCCTCGTACGG) as described by Constant et al. (2010). All sequenced genes were deposited in GenBank under the accession numbers HF570959 to HF571037 for 16S rRNA gene and HF677110 to HF677117 for hhyL. Seaview software (Gouy et al., 2010) was used to construct nucleotide alignments with the muscle algorithm and to build neighbor-joining trees using Kimura's 2-parameter distance correction with 100 bootstrap replicates.

Individual wood cell component decomposition

To test the individual wood component decomposition capacity of bacterial isolates, growth assays were performed on minimum media plates [0.2 g MgSO4, 1 g (NH4)2SO4, 0.5 g K2HPO4, 0.001 g Fe2SO4, 0.01 g Fe2(SO4)3·6H2O, 0.01 g CuSO4·5H2O, 0.01 g ZnSO4·7H2O, 0.01 g MnSO4·H2O, and 15 g agar per liter] with 5 g of pure lignin (kindly provided by Nicolas Brosse, LERMAB lab), xylan from beechwood (Sigma) or carboxymethylcellulose (CMC) as sole carbon sources. For each strain, four velvet replica platings were done starting from mass streaking on an SFM plate and using a fabric of approximately 1 cm in diameter. For lignin and xylan, decomposition ability was considered positive when the isolate showed visually significant growth in comparison with positive control plates supplemented with 5 g L−1 of mannitol (MMM medium) and with negative control plates lacking a carbon source. Cellulolytic activities of the isolates were estimated after 10 days by Congo-red coloration, as described by Bruce et al. (2010). Cellulase activity was indexed as the diameter of the colony plus the surrounding clear zone divided by the diameter of the colony (Huang et al., 2012) (Table 1).

Table 1. Table of strains and wood component degradation characteristics
Strain16S clusterpH7pH7pH 6pH 7pH 8
  1. For each isolate, the 16S rRNA gene cluster determined in Fig. 1 is indicated. Wood component decomposition was tested by plate growth assays using pure lignin, carboxymethylcellulose (CMC) or xylan (hemicellulose) as sole carbon source.

  2. a

    −, no growth on lignin and xylan.

  3. b

    Growth on CMC was tested at three different pH. CMC degradation was assessed by observation of halos after Congo red coloration. Cellulolytic activities of strains were compared with a ratio between degradation halo upon colony size (H/C). H/C value below 1; +, H/C from 1 to 1.5; ++, H/C 1.5 to 2; +++, H/C > 2; nd, not determined.


Bacteria–fungus co-cultures

Phanerochaete chrysosporium strain RP78 is a white rot-fungus able to degrade wood. The ectomycorrhizal Paxillus involutus strain ATCC 200175 was originally isolated under birch in Midlothian, Scotland. The Laccaria bicolor strain S238N was isolated under Tsuga mertensiana in Crater Lake National Park, Oregon. Hypomyces chrysospermus is a parasitic ascomycete of boletes isolated in a fir forest in Vosges, France. Confrontations were realized on HT or on MMM medium plates with 5–10 repetitions for each strain and results were compared with 10 control plates without bacteria. Streptomyces were taken from a single colony grown on SFM plate and spread in mass on the side of a Petri dish at a distance of approximately 2.5 cm. Plates were incubated 5 days at 30 °C and 4-mm plugs were realized from fungal colonies grown on either MMN medium (Paxillus, Laccaria, Hypomyces) (Blaudez et al., 2000) or malt agar (40 g malt extract, 20 g agar per liter, pH 5.5). Phanerochaete chrysosporium were placed on the opposite side of the plate at a distance of 5.5 cm from the Streptomyces. Plates were then incubated at 30 °C for P. chrysosporium and at 25 °C for the other fungi. The radial distance of mycelium growth was measured from the fungal plugs over 3 days for P. chrysosporium, 10 days for P. involutus and H. chrysospermus and 28 days for L. bicolor. For P. involutus, mycelia were recovered directly from the plate and dried at 60 °C for 2 days before dry mass determination.

StreptomycesStreptomyces confrontations

Tester Streptomyces were streaked from a single colony to form a line on an HT or SFM plate and were incubated at 30 °C for 5 days. The receiver strain was then applied perpendicularly to the tester and incubated for 3 days at 30 °C. Each pair of tester–receiver was tested twice. The tester strain was considered inhibitory when it prevented total growth or part of the receiver streak in its vicinity.

Results and discussion

The experimental forest hosts a large diversity of Streptomyces

After selective procedures for actinomycete isolation, 100 isolates were randomly selected among Streptomyces-like colonies. Their 16S rRNA gene sequences were obtained following DNA extraction and PCR amplification with universal primers. Blast searches against databases (NCBI) confirmed that 79 isolates belonged to the genus Streptomyces (Table 1). A more precise taxonomic position of these Streptomyces was assessed by phylogenetic analyses with reference sequences exhibiting the highest identities for each sequence after blast searches. Based on 514 nt sequences, 14 different 16S rRNA genotypes were obtained for the 79 Streptomyces isolates. In the phylogenetic tree, four of these genotypes were different from any reference or other isolate sequences and formed individual single gene lineages, respectively named 2, 3, 6 and 7 (Fig. 1). The other isolate sequences were grouped into different gene clusters. Twenty isolates were found in cluster 1, five in cluster 4, six in cluster 5, three in cluster 8 and 41 in cluster 9 (Table. 1). According to their taxonomic position, bootstrap values, position of reference strains and evolutionary distance, each gene lineage and gene cluster identified in this study represented separate species of Streptomyces. Because they excluded any other reference strain sequences, the gene lineages 2, 3, 6 and 7 and the gene clusters 4 and 8 could represent new species of Streptomyces. Closely related genotypes were found within gene clusters 1 and 9 (respectively genotypes 1a, 1b, 1c, 1d, Streptomyces olivochromogenes and Streptomyces mirabilis in cluster 1 and 9a, 9b, 9c, 9d, Streptomyces sanglieri and Streptomyces atratus in cluster 9). The different 16S rRNA genotypes grouped within these clusters might represent closely related, but different, species. Thus, considering the different gene lineages and gene clusters encompassing the isolates from this study, at least nine different species of Streptomyces can be identified, with a potential underestimation of this number. Polyphasic analyses are needed to confirm potentially new species (Vandamme et al., 1996). These results showed that a large diversity of Streptomyces exists in the investigated forest soil and that it constitutes an interesting reservoir for describing new species.

Figure 1.

16S rRNA gene phylogenetic relationships of Streptomyces strains isolated in this study (shown in bold) and reference strains based on neighbor-joining analysis. Only percentage bootstrap support (100 replicates) above 50% is indicated and branch length represents substitutions per site. Sequences of Kitasatospora species were used as outgroups to root the tree. Braces and arrow respectively denote clusters and single gene lineages formed by our isolates. Accession numbers of reference sequences are indicated in brackets. Accession numbers of isolates from this study were deposited as HF570959HF571037. K, Kitasatospora; S, Streptomyces.

Streptomyces can degrade wood components

In forests, wood represents the main carbon source (Hendricks et al., 1995) in the form of a hard matrix of carbohydrate biopolymers of lignin, cellulose and hemicelluloses. For instance, cellulose can constitute between 20% and 30% of the litter mass (Berg & Laskowski, 2006).

During the present work, screenings were done after collection to assess the presence/absence of functions within the population. Many Streptomyces are able to grow on lignocellulose, which is a complex of lignin, cellulose and hemicelluloses. However, it is sometimes not clear whether isolates can metabolize lignin or only break it up and feed on cellulose and hemicelluloses (Chamberlain & Crawford, 2000). Thus, the tests were done with CMC, xylan (a main component of hemicelluloses) or pure lignin as individual wood components (Table 1). Pure lignin did not sustain growth of our isolates, indicating that lignin by itself cannot be used as a carbon source for our isolates. With the exception of most of the isolates from cluster 9, all the tested isolates showed similar or better growth on xylan plates than on positive mannitol control plates. Thus, all the taxonomic clusters possess hemicellulolytic isolates, suggesting a widespread ability to decompose this compound despite taxonomic distance and confirming that Streptomyces species harbored a large reservoir of hemicellulolytic enzymes (Deesukon et al., 2012; Liu et al., 2012; Saito et al., 2013). CMC utilization assays were carried out at different pH values. CMC degradation is easy to visualize by Congo red coloration, revealing degradation halos. Thus, it is often used to detect cellulolytic bacteria (Florencio et al., 2012). Apart from few exceptions, strains from clusters 5, 6 and 9 did not show CMC degradation, whereas the other genotypes did. Quantification and comparison of cellulase activities by the measurement of degradation halo/colony size ratio indicated that cellulolytic strains were generally efficient, with a ratio greater than 1.5. Isolates from clusters 4 and 8 have the highest potential to degrade CMC (ratio > 2). CMC is an easy form of cellulose to degrade, but Streptomyces are known also to use more recalcitrant forms of crystalline cellulose (Schlösser et al., 2000; Takasuka et al., 2013), which we will be investigating for our isolates. The pH variations did not influence the cellulolytic activities except for strains S3n28, S4n2 and S9n14, indicating that cellulolytic activities can occur in the tested range that overlapped acidic, neutral and alkaline conditions. In soil, Streptomyces can thus remain cellulolytic despite pH variations that can occur seasonally. The rare isolates from cluster 5, 6 or 9 growing on xylan or CMC plates show that functional heterogeneity could also exist among closely related strains. All in all, these results confirmed that despite their taxonomic distance, hemicellulolytic and cellulolytic activities are common and widespread in our isolates.

Some Streptomyces have the molecular potential to fix atmospheric H2

The biogeochemical cycle of molecular hydrogen (H2) relies on its biological uptake within the upper soil layers. Recently, Constant et al. (2010) claimed that this function might be related to the presence of a specific high affinity [NiFe]-hydrogenase conferring the unique ability to oxidize atmospheric H2. This enzyme was first identified in Streptomyces and only found in a few phyla by genome data mining (Constant et al., 2011). Constant et al. (2010) suspected that this enzyme might be widespread within the Streptomyces genus. However, only a few natural isolates were investigated for its presence. Almost full-length PCR amplifications of the hhyL gene that encodes the large subunit of this enzyme were performed with specific primers for 14 strains chosen as representative of the taxonomic 16S rRNA gene diversity of the whole collection. PCR amplification was successful in eight isolates and failed for strains from clusters 2, 3 and 7. Strains from clusters 1a, 1b and 1d only gave faint bands that could not be sequenced and will require further PCR optimization. As expected according to PCR primer specificity, phylogenetic analysis of the hhyL sequences showed that the sequences of our isolates were grouped within the monophyletic group 5 of [NiFe]-hydrogenases, considered by Constant et al. (2011) to be involved potentially in atmospheric H2 oxidation (Fig. 2). Moreover, sequences of strains S6n14, S6n19, S4n11 and S6n6 were closely related to sequences of other Streptomyces empirically proven to be involved in atmospheric molecular hydrogen soil uptake (Constant et al., 2010). Further biochemical analyses will be required to conclusively prove that our isolates are indeed able to perform this oxidative reaction. However, this molecular screening increases the number and the diversity of Streptomyces potentially involved in atmospheric H2 oxidation. It reinforces the idea that Streptomyces are important for H2 uptake in forest.

Figure 2.

hhyL gene phylogenetic relationships of reference Streptomyces isolated in this study (shown in bold) and reference strains based on neighbor-joining analysis. Only percentage bootstrap support (100 replicates) above 50% is indicated and branch length represents substitutions per site. Sequences of the hydrogenase gene hydB were used as outgroups. Brace denotes the representative sequences forming the Group 5 [NiFe]-hydrogenase potentially involved in atmospheric H2 uptake identified by genome data mining in Constant et al. (2011). Asterisks indicate bacteria in which high-affinity H2 oxidation was demonstrated. Accession numbers of sequences are indicated in brackets. Accession numbers of isolates from this study were deposited as HF677110 to HF677117.

Streptomyces modulate fungal growth

The influence of the presence of representative Streptomyces on the growth of four different fungi that play different roles in forest ecosystems, was tested in co-culture bioassays. P. involutus and L. bicolor are both symbiotic ectomycorrhizal fungi that promote tree growth, P. chrysosporium is a white-rot fungus particularly efficient in degrading wood and H. chrysospermus is a parasite of boletes. Tests were performed on both minimum MMM or rich HT media to assess the potential influence of growth conditions on secondary metabolism of Streptomyces (Berdy, 2005) (Fig. 3). On HT medium, almost all tested Streptomyces demonstrated an antagonistic effect against each fungus. On MMM medium, results were more strain- and fungus-specific. For instance, depending on the Streptomyces, no effect or antagonistic effects were observed against H. chrysospermus. For L. bicolor, antagonistic, beneficial or no effects were observed. Interestingly, for P. involutus and P. chrysosporium, all Streptomyces promoted fungal growth (Fig. 4). For P. involutus in co-culture with S9n29 (chosen as a representative strain), mycelial extension was correlated with dry biomass increase (data not shown). This indicated that it was a genuine growth promotion and not a medium prospecting due to a stress response. In summary, these experiments indicate that Streptomyces from forest soil generally influence the growth of fungi, as already observed with mycorrhiza-associated Streptomyces (Schrey et al., 2012). Under certain conditions, some effects, either beneficial or deleterious, were shared by almost all strains on the same fungus. This suggests non-strain-specific but widespread mechanisms that may rely on changes in the composition of the medium (i.e. acidification) (Viollier et al., 2001) or in the production of similar secondary metabolites such as vitamins, chitinase and antibiotics (Chater et al., 2010; Nagpure & Gupta, 2012). Depending on their sensitivity to Streptomyces metabolites, the growth of different fungi can be promoted or inhibited by the same Streptomyces (Riedlinger et al., 2006). Interestingly, we show here that, depending on growth conditions, the same Streptomyces can have opposite effects on a given fungus. This statement is important in field use as the tuning of the Streptomyces effect can be affected by fluctuating soil conditions. All in all, these results reveal a great potential of our isolates to structure fungal population within soils and could find applications in the field to suppress parasitic fungi (Intra et al., 2011) or improve mycorrhization (Frey-Klett et al., 2007).

Figure 3.

Growth modulation of fungi by Streptomyces. The influence of plate co-culture of Streptomyces representative for different clusters on the growth of (a) Paxillus involutus (mycorrhizal fungus) and Phanerochaete chrysosporium (white-rot fungus) and (b) Hypomyces chrysospermus (fungus parasite) and Laccaria bicolor (mycorrhizal fungus) was tested on HT and MMM plates. No result was obtained for P. involutus on HT plate as the medium did not sustain its growth. Because of different growth speed, mycelium extension was measured at day 3 for P. chrysosporium, day 10 for P. involutus and H. chrysospermus and day 28 for L. bicolor and related to the treatment without bacteria (None = value 100). Mean and standard error of each experiment with at least five replicates are indicated. Significant difference in mycelial growth in comparison with control without bacterial inoculation was determined with a Student's t-test (< 0.05) and is indicated by asterisks.

Figure 4.

Paxillus involutus growth promotion by a Streptomyces isolate in coculture on MMM plate. (a) Control P. involutus without bacteria and (b) P. involutus in coculture with Streptomyces sp.S9n29. Streptomyces S9n29 was streaked from a single colony and incubated for 5 days at 30 °C. Fresh P. involutus plugs were then placed at the opposite side of the Streptomyces streak on the plate. Pictures were taken after 19 days of fungal growth at 25 °C and show clear growth promotion of P. involutus in the presence of the bacteria.

Streptomyces can inhibit growth of other Streptomyces

To assess the potential of StreptomycesStreptomyces interactions, confrontations in co-culture of representative strains were performed on HT and SFM media (Fig. 5). Some extreme cases were observed, with strains inhibiting the growth of none (i.e. S2n2 or S6n19) or all (i.e. S6n6) other strains. Culture media only influenced strain S4n22, which inhibited all but one strain on HT, and had no influence on SFM. All strains were inhibited by at least two and up to six strains, and generally had specific sensitivity patterns and were not inhibited by the same testers. Thus, as previously shown within the same niche (Neeno-Eckwall et al., 2001; Vetsigian et al., 2011; Schrey et al., 2012), Streptomyces are likely to inhibit the growth of their counterparts, exhibit complex interaction networks via their secondary metabolism and participate to make the soil a competitive environment.

Figure 5.

Bioassay evaluation of respective growth inhibition of representative Streptomyces. The tester strain was incubated on (a) SFM or (b) HT medium plate for 5 days before streaking of the receiver strain. Black-colored box stands for growth inhibition of the receiver strain by the tester and white-colored boxes for no inhibition. Numbers in brackets indicate the 16S rRNA gene cluster of the tester strain.


We show that Streptomyces are taxonomically diverse in forest soils and that several different species can be easily isolated from one soil. Functional screenings highlighted roles of Streptomyces in (1) wood component decomposition; (2) H2 uptake; and (3) growth modulation of fungi and other Streptomyces. These different roles are widespread and shared by many isolates and highlight the important roles of Streptomyces as a bacterial population in forest ecosystems.


This work was funded by the Région Lorraine, the French certified environmental monitoring and research system SOERE-OPE (Système d'Observation et d'Experimentation au long terme pour la Recherche en Environnement, Observatoire Pérenne de l'Environnement,,, and the French National Research Agency through the Laboratory of Excellence ARBRE (ANR-12- LABXARBRE-01). We would like to thank Eric Gelhaye for providing Phanerochaete chrysosporium, Nicolas Brosse for providing lignin, Laurence Lacercat for helpful discussions and Yoann Perrin for his technical help. The authors have declared that no conflict of interest exists.