Purification of an Extracellular Signaling Molecule Involved in Production of Aerial Mycelium by Streptomyces coelicolor

Abstract

We have extensively purified a factor from conditioned medium that restores aerial mycelium formation to a mutant of Streptomyces coelicolor that is defective in morphological differentiation. Response to this factor is shown to depend on the presence of the BldK oligopeptide import system. We suggest that this substance acts at the first step in a putative cascade of developmental regulatory signals.

Multicellular organisms use intercellular signaling molecules to control many developmental processes (5). The filamentous, fungus-like bacterium Streptomyces coelicolor has a multicellular lifestyle in which extensive cell-cell signaling controls the formation of a spore-forming cell type called the aerial hyphae (11). The life cycle of S. coelicolor commences with spore germination and the formation of a branched network of vegetative hyphae called the substrate mycelium. Because cell division occurs only sporadically in the substrate mycelium, individual filaments typically contain tens or hundreds of chromosomes. Twenty-four to 36 h after germination, the colony begins to erect the aerial hyphae, in which a massive round of synchronous cell division leads to the formation of uninucleoidal spores over the next few days. A small hydrophobic molecule called SapB coats the aerial filaments and evidently contributes to their capacity to break with surface tension and stand up into the air (9a, 10). The resulting white fuzzy layer of aerial hyphae on the surface of the colony acts like a fruiting body, allowing the dispersal of the mature spores, thereby completing the life cycle (2).

The bld (from “bald”) genes (1, 6, 7, 11) are believed to be involved, at least in part, in a cascade of extracellular signals that controls events like SapB synthesis and that leads to aerial mycelium formation (11). Mutations in bld261, for example, appear to block the production of the first signal of the cascade (signal 1 [8]). Cells containing mutations in bldK, which encodes an oligopeptide importer, are believed to be blocked after the release of, but before the response to, signal 1. It has been proposed that wild-type colonies produce and export signal 1, perhaps constitutively, and that after a period during which signal 1 accumulates extracellularly, import it through the BldK oligopeptide permease, triggering the next step in the cascade. Cells containing the mutations bldA, -H, -G, -C, and -D are believed to be blocked at later steps in this cascade (8, 11).

Many of the bld mutants show a marked defect in the catabolite repression of metabolic operons by rich carbon sources that is a characteristic of wild-type S. coelicolor physiology (9). This and the fact that the developmental phenotype of some of the bld mutants can be suppressed by growth on poor carbon sources have led to the suggestion that a role of the bld gene products, and perhaps therefore the signaling cascade, is to couple morphological differentiation to the nutritional state of the substrate mycelium (1, 7, 9).

To explore this putative cell-cell signaling mechanism, we used bld mutant strains of S. coelicolor (Table 1) to condition Streptomyces growth medium and examined the effects of this conditioning on the developmental phenotype of the bld261 mutant strain HU261. These experiments were carried out by using a previously described procedure (11) with the modification that recipient cells were grown directly on conditioned medium rather than on filter discs. When equivalent amounts of medium conditioned by either a bldK mutant (NS40) or a bldA mutant (LS17) were added to fresh R2YE growth medium (4), both stimulated some degree of aerial mycelium formation in HU261, indicating that both strains produced a signal molecule(s) whose production is blocked by the bld261 mutation. Whereas medium conditioned by the bldK mutant induced HU261 to produce an abundant lawn of aerial mycelium (reminiscent of wild-type development), medium conditioned by the bldA mutant stimulated only a thin layer of aerial mycelium that was not visible unless viewed under magnification (Fig. 1). Likewise, medium conditioned by mutants believed to be blocked at later steps in the signaling cascade (bldH, bldG, bldC, or bldD) was able to stimulate only a small amount of aerial mycelium formation in HU261 (data not shown). We interpret this to mean that a signaling molecule(s) that induces aerial mycelium formation in a bld261 mutant is produced by the bldA, bldH, bldG, bldC, and bldD mutants but that most of it is then imported into the donor cells by the BldK oligopeptide importer. In contrast, because the bldK mutant is defective in oligopeptide import, the signaling molecule accumulates to a higher concentration in medium conditioned by this mutant than in medium conditioned by the mutants that are bldK⁺. The stimulation of aerial mycelium formation in HU261 was not due to a spurious effect of growing the cells on partially depleted medium because HU261-conditioned medium did not elicit any effect on the any of the bld mutants (Fig. 1 and data not shown).

TABLE 1.

Strains and plasmids used in this work

Strain or plasmid	Genotype or features	Background	Source
Streptomyces coelicolor
HU261	bld261 hisA1 uraA1 strA1 NF SCP2*		J. Willey
LS17	bldA39 hisA1 uraA1 strA1 Pgl−		K. Chater
NS17	bldK1::aadA		Lab collection
NS40	bldK1::aad hisA1 uraA1 strA1		This study
NS47	bld261 hisA1 uraA1 strA1 NF SCP2* pJRM10		This study
NS48	bld261 hisA1 uraA1 strA1 NF SCP2* pbldK22		This study
NS49	bld261 bldK1::aadA hisA1 uraA1 strA1 NF SCP2* pJRM10		This study
NS50	bld261 bldK1::aadA hisA1 uraA1 strA1 NF SCP2* pbldK22		This study
Escherichia coli
DH5α	supE44 ΔlacU169 (φ80lacZΔM15)hsdR17 recA1 endA1 gyrA96 thi-1 relA1		Lab collection
ER-2	F′ lacIq leuB6 thi-1 fhuA31 lacY1 tsx-78 galK2 galT22 supE44 hisG4 rpsL136 (Strr) xyl-5 mtl-1 dam13::Tn9(Cmr) dcm-6 mcrB1 mcrA hsdR2(rK− mK+)		J. McCormick
Plasmids
pJRM10	bla tsr	pIISK+ and pIJ911	Lab collection
pbldK22	bldKABCDE tsr	pJRM-10	Lab collection

FIG. 1.

Stimulation of aerial mycelium formation in a bld261 mutant by conditioned media. bld261 recipient cells were grown on control medium or a 1:4 mixture of bld261-, bldK-, or bldA-conditioned medium and fresh R2YE agar (4) as indicated.

To determine whether the capacity of a bld261 mutant to respond to the factor present in conditioned medium requires the BldK oligopeptide importer, we created strains doubly mutant for bld261 and bldK. NS47 and NS48 contain the bld261 mutation and, respectively, the control plasmid pJRM10 and the BldK-expressing plasmid pbldK22. NS49 and NS50 contain the bld261 mutation, the bldK1 mutation, and plasmids pJRM10 and pbldK22, respectively. When these strains were grown on medium conditioned by a bldK mutant, only those bld261 mutant cells with an active BldK oligopeptide importer were able to produce the aerial mycelium (Table 2). This suggests that the signaling molecule present in this medium might be an oligopeptide and that it must be imported into cells through the BldK oligopeptide importer to bring about aerial mycelium formation. The dependence of the activity of this factor on uptake by BldK further suggests that it is the first signal of the cascade that has previously been designated signal 1 (8).

TABLE 2.

Requirement of bldK for signal response

Assay strain	Genotype	Aerial mycelium in response to bldK-conditioned mediuma
NS47	bld261(pJRM10)	+
NS48	bld261(pbldK22)	+
NS49	bld261 bldK(pJRM10)	−
NS50	bld261 bldK(pbldK22)	+

^aAerial mycelium was (+) or was not (−) produced. 

To isolate the signal 1 molecule, NS17-conditioned medium was converted to a form in which it could be applied to a C₁₈ column. This involved melting the medium at 80°C for 1 h and spinning it at 35,000 rpm and 15°C, allowing the agar to resolidify and form a pellet. In preparative-scale experiments, 4 liters of signal-containing supernatant was applied to a series of C₁₈ columns and eluted with increasingly shallow gradients of acetonitrile (Table 3). Fractions collected from the eluates were lyophilized, resuspended in 0.01 Tris (pH 7.0)–0.01 M NH₄Cl, added to fresh R2YE growth medium, and assayed for aerial mycelium-stimulating activity by using strain HU261. Interestingly, during the second chromatographic step of the purification, factors having the capacity to induce aerial mycelium formation in the bld261 mutant were detected in two discrete peaks (peaks A and B) (Fig. 2), both of which eluted later than the majority of the UV-absorbing material that had been loaded on the column. Fractions containing the factors corresponding to peak A (fractions 38 to 41) and peak B (fractions 46 to 48) were pooled separately and subjected to high-performance liquid chromatography with a C₁₈ column, where they eluted at 28 and 42%, respectively. It proved impossible to isolate sufficient amounts of the factor corresponding to peak A for further chromatographic or biophysical analysis. The factor present in peak B was purified to apparent homogeneity in one further chromatography step (Fig. 3). Each of the UV-absorbing peaks in this final elution profile was assayed for the stimulation of aerial mycelium formation in HU261. Only the peak indicated in Fig. 3, eluting at 40% acetonitrile, showed any such activity. Fractions flanking either side of this peak showed no activity. It was concluded that this fraction contained an essentially pure signaling molecule.

TABLE 3.

Purification of peak A and peak B signal molecules

Step	Column	Flow rate (ml/min)	Elution method	Comment
1	150-ml C18 Waters 125 A bulk packing materiala	8	200 ml of 80% acetonitrile
2	50-ml C18 Waters 125 A bulk packing materiala	1	0 to 80% acetonitrile gradient; 0.8%/ml over 100 ml	Two activity peaks
3A	4.6- by 250-mm Vydac 218TP C18 (peak A)a	0.5	20 to 34% acetonitrile gradient; 0.18%/min over 75 min	Eluted at 28% acetonitrile
3B	4.6- by 250-mm Vydac 218TP C18 (peak B)a	0.5	35 to 48% acetonitrile gradient; 0.18%/min over 75 min	Eluted at 42% acetonitrile
4	1- by 150-mm Zorbax C18 (peak B only)b	0.05	0 to 100% acetonitrile gradient; 0.25%/min over 400 min	Eluted at 40% acetonitrile

^aRun and eluted in the presence of 0.02 M phosphate buffer at pH 7. 

^bRun and eluted in the presence of 0.1% trifluoroacetic acid. 

FIG. 2.

C₁₈ chromatography resolves two discrete factors in conditioned medium. Signal activity from the first step in the purification procedure was passed over a 50-ml C₁₈ column and eluted with a gradient of acetonitrile. Fractions were assayed for developmental activity against bld261 mutant recipient cells.

FIG. 3.

Purification of the signaling molecule contained in peak B to homogeneity. The column eluate from the third chromatography step in the purification that was highly enriched for the peak B factor was passed over a high-performance liquid chromatography C₁₈ column. Each UV-absorbing peak was assayed for the presence of a signal molecule that could induce aerial mycelium formation in bld261 mutant recipient cells. mAU, milliabsorbance units.

We anticipated that this signaling molecule would be an oligopeptide since it rescues the phenotype of bld261 mutants in a manner that depends on active oligopeptide uptake. Mass spectrometry and amino acid composition analyses suggested that the purified factor from peak B contains a molecule with a mass of 655 Da that contains serine and glycine, properties that would be consistent with import by an oligopeptide importer such as BldK (3). Unfortunately, however, repeated attempts at obtaining an amino acid sequence by sequential Edman degradation were unsuccessful, suggesting that if this factor is a peptide, its amino terminus is blocked by some sort of covalent modification.

Notably, very small amounts of the factors in peaks A and B (we presume nanomolar concentrations on the basis of UV absorbance) were able to reproducibly induce a large amount of aerial mycelium in bld261 mutants, supporting the idea that these are bona fide components of signal 1. Finally, there is at present no explanation for the existence of two factors, peaks A and B, having the activity of signal 1, and it remains possible that one of these molecules is a biologically active fragment of the other. We anticipate that future molecular genetic studies, including especially the cloning and sequencing of the bld261 locus, will shed further light on the nature of signal 1.