A surface active protein involved in aerial hyphae formation in the filamentous fungus Schizophillum commune restores the capacity of a bald mutant of the filamentous bacterium Streptomyces coelicolor to erect aerial structures


J. M. Willey. E-mail biojmw@office.hofstra.edu; Tel. (516) 463 6542; Fax (516) 463 5112.


The filamentous bacterium Streptomyces coelicolor undergoes a complex process of morphological differentiation involving the formation of a dense lawn of aerial hyphae that grow away from the colony surface into the air to form an aerial mycelium. Bald mutants of S. coelicolor, which are blocked in aerial mycelium formation, regain the capacity to erect aerial structures when exposed to a small hydrophobic protein called SapB, whose synthesis is temporally and spatially correlated with morphological differentiation. We now report that SapB is a surfactant that is capable of reducing the surface tension of water from 72 mJ m−2 to 30 mJ m−2 at a concentration of 50 μg ml−1. We also report that SapB, like the surface-active peptide streptofactin produced by the species S. tendae, was capable of restoring the capacity of bald mutants of S. tendae to erect aerial structures. Strikingly, a member (SC3) of the hydrophobin family of fungal proteins involved in the erection of aerial hyphae in the filamentous fungus Schizophyllum commune was also capable of restoring the capacity of S. coelicolor and S. tendae bald mutants to erect aerial structures. SC3 is unrelated in structure to SapB and streptofactin but, like the streptomycetes proteins, the fungal protein is a surface active agent. Scanning electron microscopy revealed that aerial structures produced in response to both the bacterial or the fungal proteins were undifferentiated vegetative hyphae that had grown away from the colony surface but had not commenced the process of spore formation. We conclude that the production of SapB and streptofactin at the start of morphological differentiation contributes to the erection of aerial hyphae by decreasing the surface tension at the colony surface but that subsequent morphogenesis requires additional developmentally regulated events under the control of bald genes.


The streptomycetes are a morphologically and developmentally complex group of prokaryotes. These Gram-positive soil bacteria, well known for their capacity to produce a vast repertoire of antibiotics and other secondary metabolites, exhibit a multicellular life cycle similar to that of the filamentous fungi. This cycle is characterized by the germination of a uninucleate spore, which sends downwardly growing vegetative hyphae into the substratum to create a substrate mycelium. Upon receipt of as yet unidentified cues, these multinucleate vegetative hyphae differentiate into aerial hyphae that grow away from the colony surface into the air. The resulting aerial mycelium imparts a distinct white, fuzzy appearance to the colony surface. As growth continues, aerial hyphae begin to septate, and chains of spores are generated. Ultimately, the spores pinch off and, when triggered to germinate, begin the cycle anew (Chater, 1993).

One striking aspect of morphological differentiation in the genetically well-defined species Streptomyces coelicolor is the apparent involvement of intercellular communication. In this species, aerial mycelium formation is governed by the bald (bld ) genes, so called because bld mutants produce smooth colonies that are shiny and ‘hairless’. Aerial mycelium formation can be restored to a bld mutant either by growth of the mutant near a wild-type strain of S. coelicolor (Willey et al., 1991) or, in the case of specific bld mutants, by growth near certain other bld mutants (Willey et al., 1993). This phenomenon is called extracellular complementation and, in both cases, complemented bld mutants regain the capacity to form aerial hyphae.

Extracellular complementation of bld mutants when plated near wild-type S. coelicolor on complex medium is thought to result, at least in part, from the diffusion of a small morphogenetic protein, SapB, from the wild-type colony to the bld mutant colony. This conclusion is based on several lines of evidence, the most compelling of which is the observation that, when purified SapB is applied exogenously to any one of a wide variety of bld mutants, extracellular complementation occurs (Willey et al., 1993). A role for SapB in aerial mycelium formation during growth on complex medium is further suggested by the pattern of temporal, spatial and genetic regulation of the small protein. Thus, SapB biosynthesis coincides with the onset of aerial mycelium formation and is localized to the aerial mycelium and spores. It is thus only made when spores are germinated on solid medium. Also, all known bld mutants are impaired in their capacity to produce SapB. Aerial mycelium formation during growth on minimal medium evidently occurs by a pathway that does not involve SapB, as the small morphogenetic protein cannot be detected during aerial mycelium formation on medium with glucose or mannitol as a carbon source (Willey et al., 1991).

The regulation of SapB production appears to be governed partially by the extracellular signals produced by various bld mutants. Evidence for this stems from the other, more complex form of extracellular complementation. In this case, complete morphological differentiation of some bld mutants can be triggered simply by growing ‘recipient’bld strains near ‘donor’bld strains. This intercellular communication between bld mutants is invariably unidirectional, and the formation of aerial hyphae by the recipient strains is always accompanied by the regained ability to produce SapB. The exchange of intercellular signals among specific bld mutants is such that it is possible to arrange many bld mutants in a manner that predicts the hierarchical cascade of at least five signalling molecules (Willey et al., 1993; Nodwell et al., 1997). It has been proposed that this cascade of extracellular signals culminates with the synthesis and secretion of SapB (Chater, 1993; Willey et al., 1993; Nodwell et al., 1996). Alternatively, the production of SapB may be only one of several events involved in morphogenesis that are directly or indirectly dependent upon the bld genes (Pope et al., 1996).

The precise structure and function of SapB has remained obscure. It is known to be a hydrophobic peptide with a molecular mass of 2027 Da, and it has been hypothesized that SapB might function in one of two capacities. The first hypothesis suggests that SapB might serve as a scaffold-like molecule, coating the surface of the nascent aerial mycelium and providing structural rigidity for upward growth. Alternatively, SapB might reduce the surface tension at the colony–air interface, thereby facilitating the upward emergence of the aerial hyphae as they are released from the colony surface (Willey et al., 1991). If the latter hypothesis were the case, SapB would function in a manner analogous to the fungal hydrophobins. These developmental proteins are produced and secreted by a variety of fungi. Notably, this structurally and functionally conserved class of proteins is characterized by their high surface activity and the capacity to self-assemble into amphipathic films at hydrophobic–hydrophilic interfaces, thereby changing the wettability of the surface. Such self-assembly of hydrophobin monomers at the medium–air interface reduces the surface tension to such a degree that hyphae can breach the medium–air interface to grow into the air. In addition, the extracellular presence of hydrophobins appears to mediate the interactions of fungi with the environment and with other organisms (Wösten et al., 1993; 1994; 1995; Wösten and Wessels, 1997). Structurally, however, SapB bears no resemblance to the much larger hydrophobins, which share a common hydropathy profile and a conserved pattern of cysteine residues (Willey et al., 1991; Wessels, 1997).

We sought to investigate the notion that SapB might function like a hydrophobin in reducing the surface tension at the colony surface, thereby facilitating the upward growth of aerial hyphae. We show here that SapB is highly surface active, capable of reducing the surface tension to a level similar to that of other microbial surfactants (Jenny et al., 1991), including the fungal hydrophobins (van der Vegt et al., 1996; Wösten and Wessels, 1997). In addition, the ability of SapB to restore to bld mutants the capacity to erect aerial hyphae was not species specific, because the bald phenotype of a mutant of S. tendae, a species that produces the surface active peptide streptofactin, was corrected by SapB as well as by streptofactin (Richter et al., 1998). Importantly, the capacity of bld mutants of S. coelicolor and S. tendae to erect aerial structures could be restored by a hydrophobin (SC3) purified from the fungus Schizophyllum commune (Wösten et al., 1993; Wessels, 1997). Scanning electron microscopy revealed that these aerial structures were undifferentiated vegetative hyphae that had grown into the air, rather than the coiled aerial hyphae characteristic of wild-type colonies undergoing spore formation. These results are consistent with the idea that SapB and streptofactin contribute to the erection of aerial hyphae by reducing the surface tension at the air–water interface, thereby allowing the release and growth of hyphae away from the aqueous surface of the colony. Because the bacterial surfactants were insufficient to promote spore formation, however, morphological differentiation on complex medium must involve other processes under the control of bld genes in addition to the production of SapB.


Surface activity of SapB

In an effort to determine if SapB had the potential to function as a bacterial surfactant, its surface activity, i.e. its ability to reduce the water surface tension, was measured as a function of its concentration. This was performed by axi-symmetric drop shape analysis by profile (ADSA-P). This method is based on the fact that the shape of a 100 μl droplet is determined by the counteracting forces imparted by water surface tension and gravity. Surfactants cause a flattening of the water droplet, which is indicative of a decrease in water surface tension (Wessels, 1997). As shown in Fig. 1, in experiments performed over a period of 2 h, the water surface tension of 72 mJ m−2 was decreased gradually with increasing concentration of SapB to 30 mJ m−2 at 50 μg ml−1. At 150 μg ml−1, surface tension was reduced to 27 mJ m−2. These values are similar to those reported for the fungal hydrophobins (van der Vegt et al., 1996; Wösten and Wessels, 1997) and other bacterial surfactants (Jenny et al., 1991). We extrapolate that the concentration of SapB known to be sufficient to restore the capacity to form aerial hyphae in S. coelicolor bld mutants (1 μg delivered in a 20 μl drop to the surface of the colony; Willey et al., 1991) could decrease the surface tension at the colony surface to as low as approximately 30 mJ m−2.

Figure 1.

. Relationship between the concentration of SapB and its surface activity at the water–air interface. As SapB concentration increases, the surface activity is reduced from approximately 50 mJ m−2 to a minimum value of 30 mJ m−2. Note that the surface activity of water is 72 mJ m−2.

Interfacial assembly of SapB

One key feature of the fungal hydrophobins is their capacity to self-assemble at hydrophobic–hydrophilic interfaces, such as the hyphal surface in contact with air. This prompted us to explore the possibility that SapB might be capable of forming aggregates at water–air interfaces (Wösten et al., 1993; 1994; 1995). If this were the case, SapB might be enriched relative to other spore proteins by the introduction of air bubbles into a crude spore protein suspension, around which SapB might polymerize. Thus, a crude preparation of all proteins solubilized from the S. coelicolor spore coat (Guijarro et al., 1988) was prepared. This protein preparation was mixed vigorously in a Waring blender as described in Experimental procedures. After centrifugation, an insoluble hydrophobic film was found at the surface of the supernatant. This insoluble material was collected, solubilized and dissociated by the addition of 100% trifluoroacetic acid. Protein prepared in this way was significantly enriched for SapB, as judged by SDS–PAGE and Western blotting with anti-SapB (data not shown). In addition, HPLC/MS was performed, for which it was necessary to analyse roughly 100 times more crude protein (approximately 220 μg, Fig. 2A) than blender-enriched protein (approximately 2 μg, Fig. 2B and C) in order to identify the fraction containing SapB.

Figure 2.

. Analysis of SapB purification as performed by the introduction of an air–water interface. SapB was purified from a crude spore protein preparation using a Waring blender as described in Experimental procedures and analysed by HPLC coupled to a MS detector. HPLC analysis of the crude protein preparation, before blender treatment (A) and after enrichment resulted in SapB elution at 13.3 min (B). The MS signal from enriched material a with retention time of 13.3 min was obtained with an average of scans 191–201. SapB is shown in peaks 1013.7 [M + 2H] 2+ and 1024.2 [M + Na] 2+(C).

Effect of the hydrophobin SC3 on the formation of aerial structures

Evidence for functional analogy between SapB and the fungal hydrophobins was obtained in experiments in which SapB and SC3 were applied to colonies of the S. coelicolor mutants bldA and bld261 grown on rich sporulation agar. As shown in Fig. 3, after 2 days of growth in the presence of either protein, colonies of bld261 developed the white, fuzzy appearance typical of aerial mycelium formation only where the purified protein had been applied. In the case of SapB, the capacity to form the aerial mycelium is known to be dose dependent (Willey et al., 1991); this was also found to be true for SC3, with a minimum response visible at 10 μg and a maximum response between 20 μg and 30 μg (Fig. 3C and D). These results are equivalent to those seen with S. coelicolor bldA (data not shown) and similar to those obtained when the hydrophobic peptide streptofactin was applied exogenously to growing colonies of the S. tendae bld mutant strain Tü 901/S 2566-KM1 (Richter et al., 1998). In addition, SapB and SC3 could restore the formation of aerial structures to S. tendae Tü 901/S 2566-KM1 at concentrations similar to those used in rescuing S. coelicolor bld261 (2 μg and 20 μg respectively) (Fig. 4).

Figure 3.

. The effect of hydrophobic proteins on S. coelicolor bld261. S. coelicolor bld261 is blocked in the formation of aerial mycelium and thus displays the Bld (bald) phenotype as noted by a shiny colony surface (A). The application of either 1 μg of the streptomycete peptide SapB (B) or the fungal protein SC3 at a concentration of 10 μg (C) or 20 μg (D) resulted in the erection of hyphae, imparting a white, powdery appearance to the colony surface. Scale bar = 1 mm.

Figure 4.

. The effect of hydrophobic proteins on an S. tendae bld mutant. Like S. coelicolor bld mutants, S. tendae Tü 901/S 2566-KM1 appeared Bld even after prolonged growth (A). However, the application of either 2 μg of SapB (B) or 20 μg of SC3 (C) resulted in the formation of white, fuzzy, upwardly growing hyphae. Scale bar = 1 mm.

Ultrastructure of aerial structures induced by treatment of bld mutants with SapB, streptofactin and the hydrophobin SC3

We used scanning electron microscopy to investigate the nature of the aerial structures observed when colonies of bld mutant cells were treated with the bacterial surfactants and the fungal hydrophobin. Accordingly, colonies of S. coelicolor bld261 were treated with 1 μg of SapB or 20 μg of SC3 or PBS buffer alone as a negative control. In addition, colonies of S. tendae Tü 901/S 2566-KM1 were treated with streptofactin (3–5 μg). Streptofactin is actually a family of extracellular peptides made by S. tendae when grown on solid substrates. Like SapB, these peptides are small, consisting of nine amino acids, hydrophobic and bear a non-proteinogenic moiety (Richter et al., 1998). For comparison, the wild-type strain S. coelicolor J1501 was examined at 2, 3, 4 and 5 days of growth (Fig. 5). Aerial mycelium could be distinguished from substrate mycelium in this strain by the characteristic presence of curled hyphae undergoing spore formation in the aerial hyphae and the presence of branching hyphae in the substrate mycelium. This was in contrast to the morphology of S. coelicolor bld261, which formed only the long, relatively straight vegetative mycelium that lie flattened at the colony surface (Fig. 6A and B). Furthermore, despite the macroscopic powdery appearance of the bld261 colony surface after the addition of either purified SapB or SC3 (Fig. 3C and D), at the ultrastructural level, the upwardly growing filaments lack the curly morphology and spore chains characteristic of the wild-type aerial hyphae (Fig. 6C and D). Instead, the vegetative mycelium appear to have gained the capacity to grow away from the agar surface. This was also true of S. tendae Tü 901/S 2566-KM1 after the application of streptofactin. Again, in contrast to the wild type, streptofactin-treated mycelium were characterized by long, straight vegetative hyphae that were able to rise above the surface of the colony (Fig. 7).

Figure 5.

. Ultrastructural analysis of morphologically wild-type S. coelicolor J1501. Initially, vegetative hyphae grow into and along the surface of the substrate and begin to branch, as indicated by arrows (A). After 2–3 days of growth, aerial mycelium formation is marked by hyphal curling (B) and septation, which generates chains of spores (C). Scale bar = 5 μm.

Figure 6.

. Ultrastructural analysis of extracellularly complemented S. coelicolor HU261. In comparison to the normal Bld colony surface (A and B), S. coelicolor HU261 colonies to which either SapB (C) or SC3 (D) have been applied appear to have undifferentiated filaments that stand erect. Scale bar = 5 μm.

Figure 7.

. Ultrastructural analysis of extracellularly complemented S. tendae bld15. Wild-type S. tendae (A) undergoes cellular development like that seen in S. coelicolor, while the bld mutant strain S. tendae Tü 901/S 2566-KM1 (B) fails to form differentiated aerial mycelium. Upon the addition of streptofactin to S. tendae Tü 901/S 2566-KM1 (C), the undifferentiated filaments break away from the substrate. Scale bar = 5 μm.


Two lines of evidence confirm and extend the idea (Willey et al., 1991; 1993; Nodwell et al., 1996) that SapB contributes to the erection of aerial hyphae by lowering the surface tension at the colony–air interface. First, we have shown that SapB is a surfactant that can apparently aggregate at an air–water interface. Its surface activity was similar to that described for other microbial surfactants including the fungal hydrophobins (van der Vegt et al., 1996; Wösten and Wessels, 1997) and surfactin (Jenny et al., 1991), although surfactin could not substitute for SapB or streptofactin (Richter et al., 1998). Secondly, heterologous proteins implicated in the formation of aerial hyphae in a streptomycete, streptofactin, and in a filamentous fungus, the amphipathic hydrophobin SC3, could substitute for SapB, albeit at higher concentrations, in restoring the formation of aerial structures in bld mutants. Because SC3 is unrelated to SapB and streptofactin in structure, it seems simplest to propose that all three agents are morphogenetic proteins whose capacity to prompt the erection of aerial structures is the result of their hydrophobicity and their ability to associate with the outer surface of the growing hyphae where they reduce the surface tension at the colony–air interface. Reinforcing this conclusion is the observation that, reciprocally, streptofactin can restore the ability of an S. commune SC3 null mutant to form aerial hyphae (H. Wösten, unpublished results). Thus, the involvement of hydrophobin-like proteins in the emergence of nascent aerial hyphae appears to be a common feature of prokaryotic and eukaryotic filamentous organisms. The capacity to stimulate the erection of aerial structures is not, however, a general feature of any surfactant; in other work, we have found that the Bacillus subtilis surfactant surfactin could not restore aerial hyphae formation to a bld mutant (Richter et al., 1998).

Our ultrastructural analysis demonstrated that the microbial surfactants function at the extracellular level because their application did not restore the formation of wild-type aerial hyphae, characterized by the formation of curling hyphae that form spore chains. This suggests that, during morphogenesis, hydrophobin-like proteins serve only to reduce the surface tension at the colony surface, resulting in the capacity of the nascent aerial hyphae to break away from the aqueous colony surface and grow into the air. Although it is possible that the bld mutants' failure to produce SapB at hyphal tips could result in the observed phenotype, the absence of further morphogenesis even after prolonged incubation following the exogenous application of SapB (Willey et al., 1991) suggests that complete morphological differentiation is dependent on other intracellular events involving the products of the bld genes. The fact that bld mutants undergo morphological differentiation (i.e. sporulate) when grown in juxtaposition with SapB-producing strains or other bld strains (Willey et al., 1991; 1993) suggests the involvement of other extracellular factors that serve in a regulatory capacity. This is in agreement with the hypothesis proposed that differentiation is mediated in part by extracellular regulatory signals culminating in the production of SapB (Willey et al., 1991; 1993; Nodwell et al., 1996). If this is the case, the present study suggests that the intermediates in this cascade are crucial for the development of the aerial mycelium and spores. It will thus be of interest to determine the nature of these extracellular regulatory signals.

Experimental procedures

Bacterial strains and media

S. coelicolor J1501 (hisA1, uraA1, strA1, SCP1, SCP2, Pgl) and J1700 (bldA39 hisA1 uraA1 strA1 SCP1 SCP2 Pgl) (Piret and Chater, 1985) were provided by K. Chater. HU261 (bld261 hisA1 uraA1 strA1 NF SCP2* Pgl) was isolated by J. Willey and has been described previously (Willey et al., 1993). S. lividans TK64 (pro-2 str-6 SLP2 SLP3) transformed with the plasmid pS2 (Ma and Kendall, 1994) was used as a source of SapB and was a gift from K. Kendall. S. tendae wild type and bld strain Tü 901/S 2566-KM1 (Richter et al., 1998) were provided by H. P. Fiedler. All S. coelicolor and S. lividans cultures were grown on R2YE (Hopwood et al., 1985), while S. tendae was grown on HA (Shüz and Zähner, 1993) medium. All cultures were maintained at 30°C.

Axi-symmetric drop shape analysis by profile and hydrophobicity measurements

Surface tension was determined by axi-symmetric drop shape analysis by profile (ADSA-P) as described previously (Noordmans and Busscher, 1991). SapB was dissolved in water, and 100 μl was placed on a cleaned fluoroethylene–propylene–Teflon surface (FEP; Norton Fluoroplast). The droplet profile was digitized with a contour monitor, and the data were used to calculate the liquid surface tension. All measurements were performed at room temperature at least in triplicate taking 52 profiles of water droplets during a period of 2 h. To prevent evaporation, measurements were carried out with the droplet sitting in a water vapour-saturated chamber.

Partial purification of SapB

Spores of S. lividans/pS2 were plated on rich sporulation agar, R2YE, prepared in 40–60 150 mm × 15 mm Petri plates. After complete sporulation, spores were harvested, and spore-associated proteins were stripped from the spore cell wall as described previously (Guijarro et al., 1988). The spore proteins were then blended in a Waring blender for 30 min, followed by centrifugation at 13 000 × g for 30 min at 4°C. A hydrophobic tan, waxy material floating on the surface of the supernatant was collected and centrifuged at 14 000 × g for an additional 30 min. The liquid beneath this resulting hydrophobic layer was carefully removed, and the hydrophobic material allowed to air dry. This dried material was then solubilized in 100% trifluoroacetic acid (TFA), and the contents transferred to a glass tube to prevent sticking during subsequent purification steps. The TFA was next evaporated by bubbling with N2. Most of the resulting residue could then be solubilized in either water or PBS buffer (Willey et al., 1991); that which was not readily soluble was assumed to be contaminating protein and was removed by centrifugation. Protein was quantified using the Bio-Rad protein assay.

High-pressure liquid chromatography (HPLC) and mass spectrometry (MS) was performed on both crude spore protein and protein enriched for SapB. The mass spectra and tandem mass spectra were recorded on an API III Taga 6000 equipped with an ion spray source (Sciex). The orifice voltage was set to 80 V. Argon was used as the collision gas for collision-induced dissociation. The HPLC-MS coupling was performed with an Applied Biosystems ABI 140A HPLC system and narrow-bore column (Nucleosil C-1, 2 × 100 mm; 5 μm; Grom) with a gradient from 10% to 100% buffer B in 25 min. Buffer A was 0.1% TFA and buffer B was 0.1% TFA in acetonitrile.

Application of purified proteins to bld mutant strains

The potential for the purified proteins to restore the capacity to form aerial structures in bld mutants of S. coelicolor and S. tendae was performed essentially as described previously (Willey et al., 1993; Richter et al., 1998). SapB was gel purified from spores as described previously (Willey et al., 1991). After 3 days of growth, SapB (1 μg in a volume of 20 μl) or SC3 (20 μg in a volume of 20 μl) was pipetted directly onto growing colonies of S. coelicolor. Similarly, streptofactin (3–5 μg in a volume of 30 μl) was added to freshly plated cells of the bld mutant S. tendae Tü 901/S 2566-KM1. Plates were then returned to 30°C, and growth was monitored daily.

Scanning electron microscopy

Cells were fixed for 1 h at room temperature or overnight at 4°C in 2.5% glutaraldehyde and 0.5% paraformaldehyde, in 0.1 M sodium cacodylate buffer containing 10.3% sucrose (SCBS). After aldehyde fixation, specimens were washed twice for 10 min in SCBS and post-fixed for 30 min at room temperature in 1% osmium tetroxide in SCBS. After two 10 min washes in 0.1 M sodium cacodylate buffer, samples were dehydrated using a graded series of aqueous ethanol followed by freon substitution. Specimens were next placed in a CO2 bomb for critical point drying and coated in a vacuum evaporator with a thin layer of Au/Pd. Observations were made with an Hitachi S-2460N scanning electron microscope with secondary electron mode operating at 20 kV.

Purification of SC3 and streptofactin

SC3 was purified as described previously (Wösten et al., 1993; Wessels, 1997), and streptofactin was prepared as detailed by Richter et al. (1998).

The purity of both proteins was verified by SDS–PAGE and mass spectroscopy (de Vocht et al., 1998; Richter et al., 1998).


Gel electrophoresis and Western blot analysis was performed as described previously (Guijarro et al., 1988).


We thank R. Süßmuth at the University of Tübingen, Germany, for HPLC-MS analysis, Professor G. Grimes for assistance with the scanning electron microscopy, Ms L. Cacioppo and B. Frost of Hofstra University for their technical expertise with figures and photography. We thank Professor J. Wessels of the University of Groningen and Professor R. Losick and Dr J. Nodwell of Harvard University for critical reading of this manuscript.