Hopanoids are formed during transition from substrate to aerial hyphae in Streptomyces coelicolor A3(2)


  • Karl Poralla,

    Corresponding author
    1. Microbiological Institute, Microbiology/Biotechnology, University of Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen, Germany
      *Corresponding author. Tel.: +49 (7071) 2972080; Fax: +49 (7071) 294634, E-mail address: poralla@uni-tuebingen.de
    Search for more papers by this author
  • Günther Muth,

    1. Microbiological Institute, Microbiology/Biotechnology, University of Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen, Germany
    Search for more papers by this author
  • Thomas Härtner

    1. Microbiological Institute, Microbiology/Biotechnology, University of Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen, Germany
    Search for more papers by this author

*Corresponding author. Tel.: +49 (7071) 2972080; Fax: +49 (7071) 294634, E-mail address: poralla@uni-tuebingen.de


Streptomyces coelicolor A3(2) contains a cluster of putative isoprenoid and hopanoid biosynthetic genes. The strain does not produce the pentacyclic hopanoids in liquid culture but produces them on solid medium when sporulating. Mutants defective in the formation of aerial mycelium and spores (bld), with the exception of bldB, do not synthesize hopanoids, whereas mutants, which form aerial mycelium but no spores (whi), do. The membrane condensing hopanoids possibly may alleviate stress in aerial mycelium by diminishing water permeability across the membrane.


The triterpenoic, pentacyclic hopanoids condense the lipid part of biological membranes in a similar fashion to cholesterol (for review see [1,2]). Hopanoids occur in a considerable number of Gram-positive and Gram-negative Bacteria, but not in Archaea[1]. Recently, different genome projects have revealed the potential occurrence of hopanoids before their actual detection in the corresponding bacteria. This is also the case for Rhizobium NGR234, where a cluster of open reading frames (ORFs) with significant similarity to hopanoid biosynthetic genes in Zymomonas mobilis and Bradyrhizobium japonicum[3], including the gene for squalene–hopene cyclase, are encoded by the symbiosis plasmid [4]. Yet, no hopanoids have been detected in laboratory cultures [5], indicating that hopanoids are formed under specific environmental or developmental conditions.

In the genome project for Streptomyces coelicolor A3(2), a cluster of genes for isoprenoid biosynthesis has been detected on the annotated cosmid SC6A5 [6]. This cluster comprises 13 ORFs from which seven by their similarity are connected to isoprenoid and hopanoid biosynthesis. A scheme for this cluster is shown in Fig. 1. This cluster contains a pair of phythoene (squalene) synthases and a corresponding dehydrogenase in the same order as in the above mentioned Gram-negative bacteria. The first synthase in the B. japonicum hopanoid biosynthesis cluster was demonstrated to be a squalene synthase [3]. In addition, the cluster of S. coelicolor contains a putative gene for 1-deoxy-D-xylulose 5-phosphate synthase, which is an enzyme in the recently discovered non-mevalonate pathway for isoprenoids [7]. The presence of this gene in the cluster suggests that hopanoid biosynthesis may depend on the alternative pathway.

Figure 1.

Isoprenoid (hopanoid) biosynthetic genes on cosmid SC6A5 from S. coelicolor A3(2). Gray arrows, uncertain or no annotation. For the different ORFs a putative function is assigned where possible, by the S. coelicolor A3(2) Genome Project and/or by the authors. 08, phythoene or squalene synthase; 09, phythoene or squalene synthase; 10, unknown function; 11, phythoene dehydrogenase; 12, polyprenyl diphosphate synthase or farnesyl diphosphate synthase; 13, significant similarity also in the conserved motifs to other squalene–hopene cyclases; 14, lipoprotein; 15, unknown function; 16, unknown function; 17, putative 1-deoxy-D-xylulose 5-phosphate synthase; 18, putative aminotransferase; 19, DNA-binding protein; 20, unknown function. The putative genes (08, 09, 12, and 13) are sufficient to explain the synthesis of hopene from farnesyl diphosphate.

2Materials and methods

S. coelicolor A3(2) and Streptomyces griseus (Tü19; ‘Waksman strain’) were used in this study. bld and whi mutants of S. coelicolor A3(2) were obtained from Dr. Keith F. Chater and are listed below.

  • J1700: bldA39 hisA1 uraA1 strA1 SCP1 SCP2 Pg1

  • J669: bldB43 mthB2 cysD18 agaA7 NF(SCP2*?)

  • J660: bldC18 mthB2 cysD18 agaA7 NF(SCP2*)

  • J774: bldD53 cysA15 pheA1 mthB2 strA1 NF(SCP2*)

  • 166: bldF hisD3 pheA1 strA1 SCP1+ SCP2+

  • WC103: bldG103 hisA1 uraA1 strA1 SCP1 SCP2 Pgl

  • WC109: bldH109 hisA1 uraA1 strA1 SCP1 SCP2 Pgl

  • J2400: M145 whiG::hyg

  • J2402: M145 whiB::hyg

  • J2452: M145 whiJ::hyg

Wild-type S. coelicolor A3(2) and mutants were always grown on R2YE [8]. Agar cultures were grown on cellophane discs to facilitate harvesting of mycelium for the extraction of lipids. Cells from liquid cultures and agar cultures were washed once with saline and freeze-dried. Lipids were extracted according to [9]. A defined amount of the lipids was used for the quantitative determination of elongated hopanoids by periodate treatment, reduction with sodium boronhydride and acetylation [1]. Most elongated hopanoids were thereby transformed to 29-(2′-hydroxyethyl)-hopane acetate (Fig. 2). This compound was quantified by GLC with the hydrocarbon C32H66 as internal standard. For the qualitative determination of the elongated hopanoid, a TLC method on silica gel (solvent: chloroform:methanol:acetic acid:water 25:15:8:4 by vol.) was used. By comparison to an authentical aminotrihydroxybacteriohopane (provided by M. Rohmer) the identity was demonstrated using ninhydrin as reagent. The ninhydrin-positive spot was shown to represent the only elongated hopanoid in preparative TLC by derivatization (see above) and gas chromatographic/mass spectroscopic analysis. Thereby, the 29-(2′-hydroxyethyl)-hopane acetate was identified by its molecular peak at m/e 498 and peaks of typical fragments with characteristic intensity at m/e 369, 277, and 191.

Figure 2.

Hopanoids mentioned in the text. 1, hopene; 2, tetrahydroxybacteriohopane; 3, 29-(2′-hydroxyethyl)-hopane acetate; 4, aminotrihydroxybacteriohopane. Compounds 1 and 4 occur in differentiated S. coelicolor A3(2) cells.

3Results and discussion

After growth of S. coelicolor A3(2) in liquid medium for different periods and at 28°C or 37°C no or only traces of hopanoids were detected in the lipid fraction (Table 1). After 5 days of growth, the strain showed an intense pigment formation indicating antibiotic production. The simultaneous absence of hopanoids indicates decoupling of antibiotic and hopanoid biosynthesis.

Table 1.  Occurrence of hopanoids in liquid cultures and after sporulation on agar in S. coelicolor A3(2) and in different whi and bld mutants
StrainGrowth conditionHopeneMonolacetate
Wild-typeLiquid culture, 2 days, 28°C
Wild-typeLiquid culture, 5 days, 28°C
Wild-typeLiquid culture, 2 days, 37°C
Wild-typeSpore culture, 7 days, 28°C++
bldA7 days, 28°C
bldB7 days, 28°C++
bldC7 days, 28°C
bldD7 days, 28°C
bldF7 days, 28°C
bldG7 days, 28°C
bldH7 days, 28°C
whiG7 days, 28°C++
whiB7 days, 28°C++
whiJ7 days, 28°C++
Sporulating wild-type cultures and all mutants were grown on cellophane discs. Monolacetate (=29-(2′-hydroxyethyl)-hopane acetate) is a common derivative of elongated hopanoids in gas chromatographic analysis.

In contrast, when grown on solid medium allowing differentiation, hopanoids were formed by S. coelicolor A3(2) (for hopanoid structures see Fig. 1). The content of 0.23 to 2.5% of the extracted lipid seems low in comparison to some other bacteria which have amounts of 5 to 10 times higher. However, the varying content of the storage lipid triacylglycerol could explain the sometimes relative low amounts of hopanoids [10]. Besides minor amounts of hopene, the only elongated hopanoid detected was aminotrihydroxybacteriohopane (see structures in Fig. 2). This hopanoid is conceivably the biosynthetic derivative of the oxidized form of the most common tetrahydroxybacteriohopane [2]. The last biosynthetic reaction may correlate to the occurrence of gene 18 (Fig. 1) which encodes a putative aminotransferase.

The differentiation behavior of S. coelicolor A3(2) differs from that of S. griseus, which produces spores also in liquid medium [11]. The hopanoid content of S. griseus was measured after growth in liquid medium for 2 and 5 days. For both incubation periods, we detected hopanoids in a similar amount as compared to a differentiated culture of S. coelicolor A3(2). These results are an indication that hopanoid synthesis is coupled to differentiation.

Two sets of differentiation mutants of S. coelicolor A3(2) exist which are defective in either the formation of aerial hyphae and antibiotics (bld), when grown on glucose, or of spores (whi) [12]. A selection of these mutants was tested for the production of hopanoids when grown on solid medium. The tested whiB, G and J mutants produce hopanoids in a similar amount (0.3 to 1.7% of the lipids) to the sporulated culture of the wild-type. In the case of bld mutants only bldB was a producer (see Table 1).

It may be interesting to note that bldB, alone among the bld mutants tested, does not fit into the extracellular complementation cascade of Willey et al. [13]. Also, the bldB mutant is the most pleiotropic bld mutant. They even show the bld phenotype in media with poor carbon sources [12,14]. In the case of hopanoids, an intact BldB protein may not be essential for biosynthesis. This is in contrast to the production of antibiotics, which depends on BldB [12,14] and demonstrates regulatory decoupling of hopanoid and antibiotic formation.

The occurrence of hopanoids in the bldB mutant demonstrates that the genes for hopanoid biosynthesis in the wild-type are expressed before aerial mycelium formation begins or, alternatively, are derepressed in this mutant. Aerial growth is accompanied by osmotic stress (water stress). This stress may be counteracted by two different strategies. On the one hand, synthesis or uptake of compatible solutes may be initiated and on the other, the condensing effect of hopanoids may help to minimize the diffusion of water out of the cytoplasm [15] and protects against desiccation. Hopanoids induce a higher packaging of lipids in the membrane thereby reducing diffusion for small molecules and even water [15,16].

Future substitution mutagenesis experiments in the isoprenoid gene cluster will demonstrate if the hopanoid biosynthesis is essential for aerial mycelium formation. Also, we will test the influence of the bld mutations on hopanoid formation when grown on a poor carbon source (mannitol), where the wild-type phenotype is restored.


We thank especially Dr. Keith Chater providing us with bld and whi mutants and Dr. M. Rohmer for aminotrihydroxybacteriohopane. The work was funded by Deutsche Forschungsgemeinschaft (SFB 323). We thank Dolly Fink for critical reading of the manuscript.