High-Resolution Analysis of the Peptidoglycan Composition in Streptomyces coelicolor

Streptomycetes are bacteria with a complex lifestyle and are model organisms for bacterial multicellularity. From a single spore, a large multigenomic multicellular mycelium is formed, which differentiates to form spores. Programmed cell death is an important event during the onset of morphological differentiation. In this work, we provide new insights into the changes in the peptidoglycan composition and over time, highlighting changes over the course of development and between growing mycelia and spores. This revealed dynamic changes in the peptidoglycan when the mycelia aged, with extensive peptidoglycan hydrolysis and, in particular, an increase in the proportion of 3-3 cross-links. Additionally, we identified a muropeptide that accumulates predominantly in the spores and may provide clues toward spore development.

ABSTRACT

The bacterial cell wall maintains cell shape and protects against bursting by turgor. A major constituent of the cell wall is peptidoglycan (PG), which is continuously modified to enable cell growth and differentiation through the concerted activity of biosynthetic and hydrolytic enzymes. Streptomycetes are Gram-positive bacteria with a complex multicellular life style alternating between mycelial growth and the formation of reproductive spores. This involves cell wall remodeling at apical sites of the hyphae during cell elongation and autolytic degradation of the vegetative mycelium during the onset of development and antibiotic production. Here, we show that there are distinct differences in the cross-linking and maturation of the PGs between exponentially growing vegetative hyphae and the aerial hyphae that undergo sporulation. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis identified over 80 different muropeptides, revealing that major PG hydrolysis takes place over the course of mycelial growth. Half of the dimers lacked one of the disaccharide units in transition-phase cells, most likely due to autolytic activity. The deacetylation of MurNAc to MurN was particularly pronounced in spores and strongly reduced in sporulation mutants with a deletion of bldD or whiG, suggesting that MurN is developmentally regulated. Altogether, our work highlights the dynamic and growth phase-dependent changes in the composition of the PG in Streptomyces.

IMPORTANCE Streptomycetes are bacteria with a complex lifestyle and are model organisms for bacterial multicellularity. From a single spore, a large multigenomic multicellular mycelium is formed, which differentiates to form spores. Programmed cell death is an important event during the onset of morphological differentiation. In this work, we provide new insights into the changes in the peptidoglycan composition and over time, highlighting changes over the course of development and between growing mycelia and spores. This revealed dynamic changes in the peptidoglycan when the mycelia aged, with extensive peptidoglycan hydrolysis and, in particular, an increase in the proportion of 3-3 cross-links. Additionally, we identified a muropeptide that accumulates predominantly in the spores and may provide clues toward spore development.

INTRODUCTION

Peptidoglycan (PG) is a major component of the bacterial cell wall. It forms a physical boundary that maintains cell shape, protects cellular integrity against osmotic pressure, and acts as a scaffold for large protein assemblies and exopolymers (1). The cell wall is a highly dynamic macromolecule that is continuously constructed and deconstructed to enable cell growth and to meet environmental demands (2). PG is built up of glycan strands of alternating N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) residues that are connected by short peptides to form a mesh-like polymer. PG biosynthesis starts with the synthesis of PG precursors by the Mur enzymes in the cytoplasm and cell membrane, resulting in a lipid II precursor, undecaprenyl-pyrophosphoryl-MurNAc(GlcNAc)-pentapeptide. Lipid II is transported across the cell membrane by MurJ and/or FtsW/SEDS proteins, and the PG is polymerized and incorporated into the existing cell wall by the activities of glycosyltransferases and transpeptidases (3–5).

The Gram-positive model bacterium Bacillus subtilis grows via lateral cell wall synthesis followed by binary fission; in addition, B. subtilis forms heat- and desiccation-resistant spores (6, 7). In contrast, the vegetative hyphae of the mycelial Streptomyces grow by the extension of the hyphal apex, and cell division results in connected compartments separated by cross walls (8–10). This makes Streptomyces a model taxon for bacterial multicellularity (11). Multicellular vegetative growth poses different challenges to Streptomyces, including the synthesis of many chromosomes during vegetative growth and the separation of the nucleoids in the large multigenomic compartments during cross wall formation (12, 13). In submerged cultures, streptomycetes typically form complex mycelial networks or pellets (14). As surface-grown cultures, such as on agar plates, these bacteria develop a so-called aerial mycelium, whereby the vegetative or substrate mycelium is used as a substrate. The aerial hyphae eventually differentiate into chains of spores, a process whereby many spores are formed almost simultaneously, requiring highly complex coordination of nucleoid segregation and condensation and multiple cell divisions (12, 15, 16). Streptomycetes have an unusually complex cytoskeleton, which plays a role in polar growth and cell wall stability (17, 18). Mutants that are blocked in the vegetative growth phase are referred to as bald or bld, for their lack of the fluffy aerial hyphae (19), while those producing aerial hyphae but no spores are referred to as white (whi), as they fail to produce gray-pigmented spores (20).

The Streptomyces genome encodes a large number of cell wall-modifying enzymes, such as cell wall hydrolases (autolysins), carboxypeptidases, and penicillin-binding proteins (PBPs), a complexity that suggests strong heterogeneity of the PGs of these organisms (21, 22). Several concepts that were originally regarded as specific to eukaryotes also occur in bacteria, such as multicellularity (11, 23, 24) and programmed cell death (25, 26). Programmed cell death (PCD) likely plays a major role in the onset of morphological development, required to lyse part of the vegetative mycelium to provide the nutrients for the aerial hyphae (27, 28). PCD and cell wall recycling are linked to antibiotic production in Streptomyces (29).

All disaccharide peptide subunits (muropeptides) in the PG are variations on the basic building block present in lipid II, which in Streptomyces, typically consists of GlcNAc-MurNAc-l-Ala-d-Glu-ll-diaminopimelate(Gly)-d-Ala-d-Ala (30, 31). Here, we analyzed the cell wall composition of vegetative mycelium and mature spores of Streptomyces coelicolor by liquid chromatography-mass spectrometry (LC-MS) to obtain a detailed inventory of the monomers and dimers in the cell wall. This revealed extensive cell wall hydrolysis and remodeling during vegetative growth and highlights the difference in cell wall compositions between vegetative hyphae and spores.

RESULTS

To assess how growth and development translate to variations in the PG composition, we isolated PGs of S. coelicolor and analyzed the muropeptide profiles during different phases in liquid-grown cultures and of spores. In submerged cultures, S. coelicolor does not sporulate, while it forms aerial hyphae and spores on solid medium. Vegetative mycelia of S. coelicolor M145 were harvested from cultures grown in liquid minimal medium (NMM+). Samples taken after 18 h and 24 h represented exponential growth, while samples taken after 36 h and 48 h represented mycelia in transition phase (Fig. 1A and B). Samples from solid-grown cultures were taken at 24 h to represent vegetative growth, at 48 h, representing growth of aerial hyphae, and at 72 h, when the strain has formed spores (Fig. 1C). Spores were harvested from soy flour-mannitol (SFM) agar plates and filtered to exclude mycelial fragments.

FIG 1.

Growth of S. coelicolor in liquid medium (top) and on solid mediim (bottom). (A) Growth curve on NMM+ medium based on triplicate dry weight measurements. (B) Pellet morphology in liquid medium. After spore germination, hyphae emerge through top growth and branching that form an intricate network or pellet. The center of the pellets eventually lyses due to PCD (gray). (C) Growth on solid medium, starting with the development of vegetative mycelium from a single spore; after the onset of development, the vegetative hyphae differentiate into aerial hyphae that grow into the air, coinciding with lysis of the vegetative mycelium (zone of lysis represented in gray). Chains of spores are generated by septation of the aerial hyphae.

To enable the analysis of a large number of samples simultaneously and in a reasonable time frame, we adapted a method for PG purification (32) for use in S. coelicolor. The advantage of this method is that it requires only a small amount of input biomass and much faster sample handling. For this, 10 mg of lyophilized cell wall material was isolated by boiling cells in 0.25% SDS in 2-ml microcentrifuge tubes, and secondary cell wall polymers such as teichoic acids were removed by treatment for 4 h with hydrochloric acid (HCl) (see Materials and Methods for details). As a control for the validity of the method, it was compared to a more elaborate method that is used more routinely (33). In the latter method, biomass from a 1-liter culture of S. coelicolor was boiled in 5% SDS and subsequently treated for 48 h with hydrofluoric acid (HF) to remove teichoic acids. A comparison of the two methods revealed comparable outcomes between the two methods in peak detection (see Table S5 in the supplemental material). This validated the rapid method based on 0.25% SDS and HCl, which was therefore used in this study.

The isolated PGs were digested with mutanolysin (32, 34) and the muropeptide composition was analyzed by LC-MS. Peaks were identified in the m/z range from 500 to 3,000 Da, whereby different m/z′s in coeluting peaks were further characterized by tandem mass spectrometry (MS/MS). The eluted m/z values were compared to a data set of theoretical masses of predicted muropeptides. Table 1 shows a summary of the monomers and dimers that were detected during growth in liquid medium, and Table 2 shows a summary of muropeptides during growth on solid medium. The full data sets are given in Tables S1 to S4. We identified several modifications, including the amidation of d-iGlu to d-iGln at position 2 of the stem peptide, deacetylation of MurNAc to MurN, removal of amino acids to generate mono-, di-, tri-, and tetrapeptides, loss of ll-diaminopimelate (DAP)-bound glycine, and the presence of Gly (instead of Ala) at position 1, 4, or 5. The loss of GlcNAc or GlcNAc-MurNAc indicates hydrolysis (Fig. 2).

TABLE 1.

Relative abundance of muropeptides in vegetative cells from liquid NMM+

Peptidea	Abundance (%)b in S. coelicolor M145
	18 h	24 h	36 h	48 h
Monomers
Mono	1.6	2.1	3.3	3.3
Di	14.2	15.5	14.5	13.2
Tri	27.4	32.2	35.1	35.8
Tetra	26.7	24.4	23.9	23.9
Tetra[Gly4]c	3.5	5.3	6.9	8.2
Penta	22.7	16.9	13.1	12.9
Penta[Gly5]c	4.7	4.8	4.7	4.4
d-Glutamine	67.3	62.3	61.1	63.7
Deacetylated	3.9	6.0	7.9	8.0
MurN-Tri	0.1	0.7	1.2	2.3
GlcNAc-MurN-Tri	1.8	2.2	2.6	2.1
Dimers
TriTri(3-3)	4.1	4.8	6.5	7.0
TriTri-MurNAcGlcNAc	8.7	14.8	23.7	34.3
TriTetra(3-3)	23.9	24.2	22.3	16.9
TriTetra(3-4)	1.0	8.7	8.2	6.1
TriTetra-MurNAcGlcNAc	9.6	15.1	16.1	16.2
TetraTetra(3-4)	23.3	13.5	10.1	8.6
TetraTetra-MurNAcGlcNAc	6.0	7.3	4.8	5.6
TetraPenta(3-4)	24.6	9.1	5.6	3.0
MurN	1.8	1.2	1.5	1.2
Missing GlcNac	0.3	0.6	1.1	1.2
Missing MurNAcGlcNAc	24.3	37.2	44.6	56.1
Proportion (%) of 3-3 cross-links	36.5	48.0	54.5	57.3

^aMonomers and dimers were treated as separate data sets.

^bRelative abundance was calculated as the ratio of the peak area over the sum of all peak areas recognized in the chromatogram.

^cGly detected instead of Ala.

TABLE 2.

Relative abundance of muropeptides in mycelia and spores of S. coelicolor M145 harvested after growth on SFM agar plates

Peptidea	Abundance (%)b in S. coelicolor M145
	24 h	48 h	72 h	Spores
Monomers
Mono	3.6	4.3	4.1	4.5
Di	21.6	17.6	17.9	13.1
Tri	29.6	34.3	34.2	28.1
Tetra	25.4	29.5	32.0	48.3
Tetra[Gly4]c	0.9	1.1	1.0	2.3
Penta	16.8	9.9	7.2	5.3
Penta[Gly5]c	1.2	1.4	1.3	4.0
Deacatylated	3.7	4.4	6.1	4.5
d-Glutamine	76.2	80.3	82.9	74.0
Missing GlcNAc	1.5	3.4	5.0	4.8
MurN-Tri	0.6	1.7	3.1	3.5
GlcNAc-MurN-Tri	1.9	1.4	1.6	0.1
Dimers
TriTri(3-3)	7.4	10.5	12.6	4.9
TriTri-MurNAcGlcNAc	0.6	0.6	0.3	7.1
TriTetra(3-3)	20.4	22.2	21.8	19.1
TriTetra(3-4)	9.7	12.7	11.8	4.7
TriTetra-MurNAcGlcNAc	13.3	14.5	13.0	6.3
TetraTetra(3-4)	13.3	15.8	15.7	38.9
TetraTetra-MurNAcGlcNAc	17.3	13.7	13.2	17.1
TetraPenta(3-4)	12.7	7.3	5.4	0.7
MurN	1.0	0.3	1.2	0.4
Missing GlcNAc	0.4	0.2	0.4	0.1
Missing MurNAcGlcNAc	31.1	28.7	26.5	30.4
Proportion (%) of 3-3 cross-links	43.8	47.8	51.1	35.1

^aMonomers and dimers were treated as separate data sets.

^bRelative abundance was calculated as the ratio of the peak area over the sum of all peak areas recognized in the chromatogram.

^cGly detected instead of Ala.

FIG 2.

Summary of structures of main monomers and dimers observed in PGs from S. coelicolor. Modifications to the PG include the following: alteration of the length of the amino acid chain; [Gly1], l-Ala is replaced by Gly; [Glu], where glutamic acid (Glu) is present instead of d-glutamine (Gln); [Gly4], where d-Ala (4) is replaced by Gly; and [Gly5], where d-Ala (5) is replaced by Gly. Specific for dimers: (3-3) shows a cross-link between ll-DAP[3] to ll-DAP[3] with a Gly bridge; (3-4) shows a cross-link between ll-DAP[3] and d-Ala[4] with a Gly bridge; (-MurNAcGlcNAc) shows hydrolysis of a set of sugars.

For all amino acid positions in the pentapeptide chain, the position is indicated as [n], whereby n is the number in the chain (with [1] the position closest to the PG backbone, i.e., the MurNAc residue, and [5] the last amino acid residue).

Growth phase-dependent changes in PG composition.

The muropeptide that is incorporated from lipid II by glycosyltransferases contains a pentapeptide with a Gly residue linked to ll-DAP at amino acid position 3 (ll-DAP[3]). In many bacteria, pentapeptides are short-lived muropeptides that occur mostly at sites where de novo cell wall synthesis takes place, i.e., during growth and division (35, 36). This is reflected by the high abundance of pentapeptides in the samples obtained from exponentially growing cells, with a pentapeptide content of 21% during early exponential growth (18 h) compared to 14% and 11% during late exponential growth (24 h) and transition phase (36 h) and stationary phase (48 h), respectively. Conversely, tripeptides increased over time, from 24% during early exponential phase to 32% in transition-phase cultures.

The addition of Gly to the medium and, in consequence, the incorporation of Gly in the PGs can cause changes in morphology (37, 38). This property has been applied to facilitate the lysozyme-mediated formation of protoplasts in Streptomyces, used for protoplast transformation methods (39–41). In S. coelicolor, Gly can be found instead of l-Ala[1], d-Ala[4], or d-Ala[5] in the pentapeptide chain. During liquid growth, tetrapeptides carrying Gly[4] increased from 3% during early growth to 8% during the latest time points. The relative abundance of pentapeptides carrying Gly at position 5 (4% to 5%) did not vary over time. On solid-grown cultures, the Gly content of the peptidoglycans was around 1%, which is significantly lower than in liquid-grown cultures.

Abundance of 3-3 cross-links increases over time.

Two types of cross-links are formed via separate mechanisms, namely, the canonical d,d-transpeptidases (PBPs) producing 3-4 (d,d) cross-links between ll-DAP[3] and d-Ala[4] and l,d-transpeptidases that form 3-3 (l,d) cross-links between two ll-DAP[3] residues (Fig. 2). These types of peptidoglycan cross-linking can be distinguished on the basis of differences in retention time and their MS/MS fragmentation patterns. Dimers containing a tripeptide and a tetrapeptide (TetraTri) may have either cross-link, giving rise to isomeric forms that elute at different retention times, enabling assessment by MS/MS (Fig. 3A and B). In S. coelicolor, the ratio of 3-3 cross-linking increased over time toward transition phase; the relative abundance increased from 37% of the total amount of dimers at 18 h (exponential phase) to 57% of all dimers at 48 h (Fig. 3A and B).

FIG 3.

MS/MS fragmentations of TetraTri dimers with either 3-3 cross-link (A) or 3-4 cross-link (B). Differentiation between these two types of cross-links is possible at the point of asymmetry, at Gly attached to ll-DAP. (A) The 3-3 cross-linked dimer fragments into masses of m/z 966.0 and m/z 941.3, which can be found in the respective MS/MS spectrum. (B) The 3-4 cross-linked dimer fragments into masses of m/z 1037.4 and m/z 870.5. These masses are found in the MS/MS spectrum. Boxed MS/MS spectra show a magnification of masses between m/z 850 and 1,050 to show masses present in lower abundance. (C) A TriTri dimer lacking GlcNAc-MurNAc with an M+H of 1,355.6. Diagnostic fragments are given in the proposed structures.

PG hydrolysis increases as the culture ages.

PG hydrolysis is associated with processes such as the separation of daughter cells after cell division and autolysis, and mutants of bacteria that fail to produce PG amidases grow in chains of connected cells (42, 43). On solid medium, vegetative hyphae of Streptomyces undergo programmed cell death (PCD) and hydrolysis. In liquid-grown cultures, cell death occurs in the center of dense pellets. During spore maturation, spores are separated hydrolytically from one another. Some streptomycetes sporulate in submerged culture, but this is not the case for S. coelicolor (44). Our data show that as growth proceeds in submerged cultures, the S. coelicolor peptidoglycan progressively loses GlcNAc and GlcNAc-MurNAc moieties (Table 1) as a result of N-acetylglucosaminidase activity. The proportion of dimers lacking GlcNAc-MurNAc thereby increases in time from 24% at 18 h to 56% at 48 h. Figure 3C shows MS/MS profiles of a TriTri-dimer with a single set of glycans. During growth on solid medium, the trend was inversed. This may be due to the different developmental stages, whereby 24 h corresponds to early developmental events and PCD, 48 h to aerial growth, and 72 h to sporulation. This analysis shows the relative abundance of muropeptides to the total amount of biomass when hydrolysis has occurred at the vegetative mycelium. During later stages of growth on agar plates, a large amount of aerial hyphae is formed, and this can therefore not be compared directly to samples that only contain vegetative hyphae (Table 2).

Deacetylation of MurNAc is associated with mycelial aging and sporulation.

Modifications to the glycan strands are commonly linked to lysozyme resistance (45). In particular, N-deacetylation of PG strands is widespread among bacteria, which can occur both at GlcNAc and at MurNAc (46). In the case of S. coelicolor, the only glycan modification is the deacetylation of MurNAc to MurN. Our data show that this modification becomes more prominent as the vegetative mycelium ages, from 5% during early growth to 8% during later growth stages. On agar plates, 3.7% of the monomers was deacetylated at 24 h, 4.4% at 48 h, and 6.1% at 72 h.

The PG compositions of spores and vegetative mycelia were compared to get more insights into the possible correlations between PG composition and important processes such as dormancy and germination. Muropeptides in spores were strongly biased for tetrapeptides, making up 44% of the monomers, compared to 23% to 25% of the vegetative PGs. Conversely, pentapeptides were found in much smaller amounts in spores (5% of the monomers) compared to 10% to 22% in vegetative hyphae. The muropeptide that stood out in the analysis of the spore PGs was a tripeptide which lacks GlcNAc and contains a deacetylated MurNAc, called MurN-Tri (Fig. 2). In spores, MurN-Tri made up 3.5% of the monomers, whereas the less modified muropeptide, GlcNAcMurN-Tri, only made up 0.2% of the monomers.

To further investigate this interesting phenomenon and show the applicability of our work for the analysis of developmental mutants, we analyzed bldD and whiG mutants. The bldD gene product is a global transcription factor that controls the transcription of many developmental genes and is therefore blocked in an early stage of morphogenesis (47), while the whiG gene product is a σ factor that controls early events of aerial growth (48). The monomer profiles of S. coelicolor M145 and its bldD and whiG mutants are summarized in Table 3. For the wild-type strain M145, 24 h represents vegetative growth, 48 h aerial growth, and 72 h spore formation. In line with the notion that MurN-Tri accumulates particularly in spores, the bldD mutant accumulated hardly any MurN-Tri (0% to 0.2%) over time and the whiG mutant accumulated 0.4%, 0.6%, and 1.3% after 24 h, 48 h, and 72 h, respectively. In contrast, the wild-type strain M145 had 0.6%, 1.7%, and 3.1% MurN-Tri at these time points, respectively, strongly suggesting that MurN-Tri accumulates in a sporulation-specific manner.

TABLE 3.

Relative abundances of monomers from developmental bldD and whiG mutants and the wild-type strain of S. coelicolor M145

Strain or type	Relative abundance (%)a
	Mono	Di	Tri	Tetra	Penta	Deacetylated	MurN-Tri	GlcNAc-MurN-Tri
ΔbldD mutant
24 h	4.5	25.7	28.0	23.0	10.8	6.5	0.0	5.3
48 h	4.3	26.3	38.3	23.4	11.1	8.5	0.2	6.6
72 h	4.3	27.2	40.9	19.9	9.5	7.6	0.2	5.8
ΔwhiG mutant
24 h	3.5	23.2	27.0	32.5	15.2	3.0	0.4	1.3
48 h	3.6	17.5	44.3	25.5	7.9	5.0	0.6	3.2
72 h	4.1	18.5	48.8	20.9	6.9	6.2	1.3	3.8
M145 (wild type)
24 h	3.6	21.6	29.6	25.4	16.8	3.7	0.6	1.9
48 h	4.3	17.6	34.3	29.5	9.9	4.4	1.7	1.4
72 h	4.1	17.9	34.2	32.0	7.2	6.1	3.1	1.6
Spores	4.5	13.1	28.1	48.3	5.3	4.5	3.5	0.1

^aRelative abundance was calculated as the ratio of the peak area over the sum of all peak areas recognized in the chromatogram.

DISCUSSION

In this study, we analyzed the changes in the compositions of peptidoglycans during the growth and development of Streptomyces coelicolor. The different masses were thereby identified by MS and MS/MS analyses, which enabled detailed identification of the subunits, including dimers that are cross-linked by either 3-3 or 3-4 cross-links between the peptide moieties. Our data show that the Streptomyces peptidoglycan composition dynamically changes, whereby major peptidoglycan recycling was seen, with over half of all GlcNAc-MurNAc dimers hydrolyzed in late exponential cultures.

l,d-Transpeptidases (LDTs) are especially prevalent in the actinobacterial genera Mycobacterium, Corynebacterium, and Streptomyces. These bacteria have a much higher percentage of 3-3 cross-links, with an abundance of at least 30% 3-3 cross-links in investigated actinobacterial peptidoglycans compared to that of bacteria with lateral cell wall growth, such as Escherichia coli (<10%) and Enterococcus faecium (3%) (30, 49, 50). LDTs attach to d-Ala[4] and form a cross-link between glycine and ll-DAP[3]. d-Ala[4] is considered a donor for this type of cross-link (51). An interesting feature of these two mechanisms is that 3-4 cross-links can only be formed when a pentapeptide is present to display the d-Ala[5] donor, whereas 3-3 cross-links can be formed with a tetrapeptide as a donor strand. Dimers in vegetative (liquid-grown) cells carry 36.5% 3-3 cross-links at 18 h of growth, increasing to 48% at 24 h, 54.5% at 36 h, and 57.3% at 48 h. Between these stages of growth, the main structural difference is the length of hyphae compared to that of growing tips. The data agree with the idea that 3-3 cross-links may be required to remodel the cell wall beyond the tip complex, using available tetrapeptides rather than newly constructed pentapeptides (52–55).

A major event associated with the lytic degradation of the cells is programmed cell death (PCD). PCD is likely a major hallmark of multicellularity (11) and has been described in the biofilm-forming Streptococcus (56) and Bacillus (57), in myxobacteria that form fruiting bodies (58), in the filamentous cyanobacteria (59, 60), and in the branching Streptomyces (28, 61, 62). In streptomycetes, cell wall hydrolases support developmental processes such as branching and germination (21). Additionally, PCD and the autolytic release of GlcNAc from the cell wall are important signals for the onset of morphological differentiation and antibiotic production in streptomycetes (29, 63). Our data show an exceptionally large amount of dimers which carry a cross-linked set of peptides but a single set of glycans, from 25% of dimers in 18-h-old liquid cultures to 56% in 48-h-old cultures. The increase in abundance of dimers lacking a set of glycans is especially prevalent in liquid-grown mycelia, while the overall increase in hydrolyzed dimers is not as high in mycelia grown on solid medium. It should be noted that on agar plates, aerial hyphae are also formed, which are not subject to the extensive lysis seen in vegetative hyphae, and this may reduce the relative content of these glycan-less peptides.

We also analyzed changes in the PGs that correlate with sporulation. One question that remains to be answered is how future sites of branching in the hyphae or germination in the spores are marked; one interesting possibility is that the cell wall is changed as a marker for the start of future de novo PG synthesis. After all, even after very long storage of spores, germination still occurs at the spore “poles,” suggesting that physical marks to the PGs, such as rare modifications, may occur. A previous study showed that mutation of the gene dacA that encodes d-alanyl–d-alanine carboxypeptidase disrupts spore maturation and germination, where one could influence the other. This indicates that either pentapeptides inhibit spore maturation or that a large amount of tetrapeptides is important (64). Indeed, we report a large amount of tetrapeptides in spores, 48% of the monomers. A necessity for tetrapeptides might be linked to the formation of 3-3 cross-links, which require tetrapeptides, rather than pentapeptides, as a substrate. Indeed, spores carry 3-3 cross-links in 35% of the dimers, which probably strongly contribute to structural stability. Interestingly, a relatively large amount of MurN-Tri (3.5%) was identified in the spore PGs, while this molecule was almost completely absent in bldD mutants, which are arrested in the vegetative growth phase. A small amount of MurN-Tri (0.4% to 1.2%) was found in whiG mutants, which do develop aerial hyphae but do not sporulate. It will be interesting to see what the biological significance is of the overrepresentation of MurN-Tri in aerial and spore PGs. This underlines the importance of analyzing the cell walls of different culture types, as it reveals novel features that may play key roles in development.

Conclusions.

We have provided a detailed analysis of the peptidoglycans of Streptomyces mycelia and spores and developed a reliable and fast method to compare larger numbers of samples. Our data show significant changes over time, among which are changes in the amino acid chains, hydrolysis of dimers, and the accumulation of the rare MurN-Tri specifically, in the spores. The cell wall likely plays a major role in the development of streptomycetes, with implications for germination and the switch to development and antibiotic production (via PCD-released cell wall components). The dynamic process that controls the remodeling of the cell wall during tip growth is poorly understood, but we anticipate that the local cell wall structure at sites of growth and branching may well be different from that in older (nongrowing) hyphae. This is consistent with the changes we observed over time between the younger and older mycelia. A detailed localization of cell wall modifying enzymes and of specific cell wall modifications, in both time and space, should provide further insights into the role of the cell wall in the control of growth and development of streptomycetes.

MATERIALS AND METHODS

Bacterial strains and culturing conditions.

Streptomyces coelicolor A3(2) M145 (41), bldD mutant J774 (cysA15 pheA1 mthB2 bldD53 NF SCP2* [19]), and whiG mutant J2400 (whiG::hyg [65]) were obtained from the John Innes Centre strain collection. All media and methods for handling Streptomyces are described in the Streptomyces laboratory manual (41). Spores were collected from soy flour-mannitol (SFM) agar plates. Liquid cultures were grown shaking at 30°C in a flask with a spring, using normal minimal medium with phosphate (NMM+) supplemented with 1% (wt/vol) mannitol as the sole carbon source; polyethylene glycol (PEG) was omitted to avoid interference with the MS identification. Cultures were inoculated with spores at a density of 10⁶ CFU/ml. A growth curve was constructed from dry weight measurements by freeze-drying the washed biomass obtained from 10 ml of culture broth (three biological replicates). To facilitate the harvest of mycelium from agar plates, they were grown on cellophane slips, after which the biomass was scraped off the cellophane. Spores were collected from SFM agar plates by adding 0.01% (wt/vol) SDS to facilitate spore release from the aerial mycelium, scraping them off with a cotton ball, and drawing the solution through a syringe. Spores were filtered with a cotton filter to separate spores from residual mycelium.

PG extraction.

Cells were lyophilized for biomass measurement; 10 mg biomass was directly used for PG isolation. PG was isolated according to reference 32, using 2-ml screw-cap tubes for the entire isolation. Biomass was first boiled in 0.25% SDS in 0.1 M Tris-HCl (pH 6.8), thoroughly washed, sonicated, and treated with DNase, RNase, and trypsin, and proteins were inactivated by boiling in and washing with water. Wall teichoic acids were removed with 1 M HCl. PGs were digested with mutanolysin and lysozyme (66). Muropeptides were reduced with sodium borohydride, and the pH was adjusted to 3.5 to 4.5 with phosphoric acid.

To validate the method, we compared it to the method described previously (33). For this, S. coelicolor mycelia were grown in 1 liter NMM+ medium for 24 h. After washing the mycelia, the pellets were resuspended in boiling 5% (wt/vol) SDS and stirred vigorously for 20 min. Instead of sonicating the cells, they were disrupted using glass beads, followed by the removal of the teichoic acids with an HF treatment at 4°C as described.

LC-MS analysis of monomers.

The LC-MS setup consisted of a Waters Acquity ultra-performance LC (UPLC) system (Waters, Milford, MA) and an LTQ Orbitrap XL hybrid ion trap Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA) equipped with an Ion Max electrospray source.

Chromatographic separation was performed on an Acquity UPLC HSS T3 C₁₈ column (1.8 μm, 100 Å, 2.1 mm by 100 mm). Mobile phase A consisted of 99.9% H₂O and 0.1% formic acid, and mobile phase B consisted of 95% acetonitrile, 4.9% H₂O, and 0.1% formic acid. All solvents used were of LC–MS-grade or better. The flow rate was set to 0.5 ml/min. The binary gradient program consisted of 1 min 98% A, 12 min from 98% A to 85% A, and 2 min from 85% A to 0% A. The column was then flushed for 3 min with 100% B, the gradient was then set to 98% A, and the column was equilibrated for 8 min. The column temperature was set to 30°C and the injection volume used was 5 μl. The temperature of the autosampler tray was set to 8°C. Samples were run in triplicates.

MS/MS was performed both on the full chromatogram by data-dependent MS/MS and on specific peaks by selecting the mass of interest. Data-dependent acquisition was performed on the most intense detected peaks, the activation type was collision-induced dissociation (CID). Selected MS/MS was performed when the resolution of a data-dependent acquisition lacked decisive information. MS/MS experiments in the ion trap were carried out with relative collision energy of 35%, the trapping of product ions was carried out with a false-discovery rate (q value) of 0.25, and the product ions were analyzed in the ion trap. Data were collected in the positive electrospray ionization (ESI) mode with a scan range of m/z 500 to 3,000 in high-range mode. The resolution was set to 15.000 (at m/z 400).

Data analysis.

Chromatograms were evaluated using the free software package MZmine (http://mzmine.sourceforge.net/ [67]) to detect peaks, deconvolute the data, and align the peaks. Only peaks with a mass corresponding to that of a muropeptide were saved, other data were discarded. The online tool MetaboAnalyst (68) was used to normalize the data by the sum of the total peak areas and then normalize the data by log transformation. The normalized peak areas were exported, and a final table which shows peak areas as the percentage of the whole was produced in Microsoft Excel.

Muropeptide identification.

The basic structure of the peptidoglycan of S. coelicolor has been published previously (30). Combinations of modifications were predicted and the masses were calculated using ChemDraw Professional (PerkinElmer). When a major peak had an unexpected mass, MS/MS helped resolve the structure. MS/MS was used to identify differences in cross-linking and to confirm predicted structures.