Role for a Lytic Polysaccharide Monooxygenase in Cell Wall Remodeling in Streptomyces coelicolor

Abstract

Peptidoglycan is a major constituent of the bacterial cell wall and an important determinant for providing protection to cells. In addition to peptidoglycan, many bacteria synthesize other glycans that become part of the cell wall. Streptomycetes grow apically, where they synthesize a glycan that is exposed at the outer surface, but how it gets there is unknown. Here, we show that deposition of the apical glycan at the cell surface of Streptomyces coelicolor depends on two key enzymes, the glucanase CslZ and the lytic polysaccharide monooxygenase LpmP. Activity of these enzymes allows localized remodeling and degradation of the peptidoglycan, and we propose that this facilitates passage of the glycan. The absence of both enzymes not only prevents morphological development but also sensitizes strains to lysozyme. Given that lytic polysaccharide monooxygenases are commonly found in microbes, this newly identified biological role in cell wall remodeling may be widespread. IMPORTANCE Lytic polysaccharide monooxygenases are used in industry for the efficient degradation of recalcitrant polysaccharide substrates. Only recently, we have begun to appreciate some of their important biological roles. In this article, we provide evidence that these enzymes are involved in remodeling peptidoglycan, which is a conserved component of the bacterial cell wall. Given that lytic polysaccharide monooxygenases are commonly found in microbes, this newly identified biological role in cell wall remodeling may be widespread.

Keywords: LPMO; apical growth; cell wall; cell wall biosynthesis; cellulose; glycan; morphology; peptidoglycan.

FIG 1

Comparative analysis of glycoside hydrolase family 6 proteins. (A) MultiGeneBlast output showing gene clusters of filamentous actinobacteria, which are homologous to the cslA-glxA-cslZ gene cluster of S. coelicolor involved in synthesis of a cellulose-like polymer. Clusters have a minimal identity of 30% and minimal sequence coverage of 25% to the S. coelicolor gene cluster. (B) Phylogenetic tree of members of the GH6 family, including CslZ (S. coelicolor), XpCel6A (Xylanimicrobium pachnodae), CelAB (Teredinibacter turnerae T7901), CbhA (Cellulomonas fimi ATCC 484), XylK2 (Cellulosimicrobium sp. HY-13), CbhII (Streptomyces sp. M23), TfCel6B (Thermobifida fusca YX), CenA (Mycobacterium tuberculosis H37Rv), EGI (Neisseria sicca SB), TfCel6A (Thermobifida fusca YX), McenA (Micromonospora cellulolyticum), and TbCelA (Thermobispora bispora), which were selected based on the availability of experimental data on their substrates. (C) Alignment of the catalytic centers of CslZ and other GH6s hydrolases, including TfCel6A, TfCel6B, CelAB, and Cel6H. The conserved residues in the catalytic centers are gray-colored and the key catalytic residue Asp is labeled with a red arrowhead. The full-length alignments of the GH6 domains are available in Fig. S2.

FIG 2

The absence of lpmP and cslZ affects mycelial morphology in S. coelicolor. Pellet morphology of strains lacking or overexpressing genes involved in glycan biosynthesis and degradation. Pellets were imaged after 48 h of growth in TSBS. The double mutant strain lacking lpmP and cslZ (ΔcslZ/ΔlpmP) is no longer able to form pellets and is phenotypically similar to the cslA mutant (ΔcslA). Reintroduction of both genes expressed from the constitutive gapAp promoter (plasmid hpXZ4) in the ΔcslZ/ΔlpmP double mutant restored wild-type pellet morphology. Pellets of the complemented single mutants expressing cslZ (plasmid hpXZ2) or lpmP (plasmid pXZ3) under the control of the constitutive gapAp promoter have a denser appearance compared to the wild-type strain. Pellets of the strain containing the empty pSET152 plasmid (pM145) were comparable to those of the wild type. Scale bar represents 100 μm.

FIG 3

Deposition of the β-(1-4)-glycan at hyphal tips is abolished in the absence of LpmP and CslZ. Calcofluor white (CFW) staining was used to detect β-(1-4) glycans in S. coelicolor strains lacking genes involved in glycan biosynthesis and degradation. As expected, tip staining (arrowheads) is evident in the wild-type strain and control strain (pM145) and absent in the ΔcslA mutant (see insets). Tip staining is reduced in the ΔcslZ and the ΔlpmP single mutants but is absent in the ΔlpmP/ΔcslZ double mutant. Reintroduction of both genes expressed from the constitutive gapAp promoter (plasmid hpXZ4) in the ΔcslZ/ΔlpmP double mutant restored tip staining. The complemented single mutants expressing cslZ (plasmid hpXZ2) or lpmP (plasmid pXZ3) under the control of the constitutive gapAp show an increased staining compared to the wild-type (see also Fig. S4). Scale bars represent 100 μm (main images) and 20 μm (insets).

FIG 4

The absence of the CslA-produced polymer causes lysozyme sensitivity in S. coelicolor. (A) Growth of the wild-type strain, the ΔcslA mutant, and the ΔlpmP/ΔcslZ double mutant on plates with or without lysozyme (0.25 mg mL⁻¹). No growth is observed for the ΔcslA mutant and the ΔlpmP/ΔcslZ double mutant on plates containing lysozyme. (B) Quantitative assessment of the relative number of CFU obtained following growth in the presence and absence of lysozyme. Percentages were determined by dividing the number of colonies on plates with 0.25 mg mL⁻¹ lysozyme by the number of colonies on plates without lysozyme. The values represent the average of triplicate experiments. The error bars indicate the standard errors of the mean (P < 0.01).

FIG 5

LpmP facilitates hydrolysis of peptidoglycan by lysozyme and CslZ. (A) SDS-PAGE gel showing purified LpmP (18.4 kDa) and CslZ (31.9 kDa) heterologously produced in E. coli. (B) In vitro binding assays of LpmP and CslZ to PG, cellulose, and chitin. CslZ or copper-loaded LpmP were incubated with PG from S. coelicolor, microcrystalline cellulose, or α-chitin for 3 h at room temperature. The supernatants, containing the unbound proteins (NB), were collected by centrifugation. The pelleted insoluble polysaccharides were washed, after which the bound (B) proteins were extracted with 4% SDS. The unbound (NB) and bound (B) proteins were analyzed using a 15% SDS-PAGE gel, revealing that LpmP binds weakly to chitin and strongly to PG. No binding was observed for CslZ. (C) LpmP facilitates hydrolysis of PG by CslZ. PG from S. coelicolor was incubated with CslZ (5 μM), CslZ^D120A (5 μM), LpmP (1 μM), apo-LpmP (1 μM), or combinations thereof. The difference in absorbance (ΔA₆₀₀) was used as a measure for the degradation of PG. (D) Quantitative assessment of PG hydrolysis by lysozyme in the presence and absence of LpmP. PG from S. coelicolor was incubated with lysozyme (2.5 μM), LpmP (1 μM), apo-LpmP (1 μM), or combinations thereof. The difference in absorbance (ΔA₆₀₀) was used as a measure for the degradation of PG. (E) Degradation of S. coelicolor PG by lysozyme (2.5 μM) with and without prior treatment with LpmP. Data points were fitted with a linear regression plot. Error bars represent the standard error of the mean of triplicate experiments.

FIG 6

Proposed model for assembly and deposition of the apical glycan produced by CslA in Streptomyces. CslA utilizes UDP sugars to synthesize a glycan, which is possibly modified by the activity of the copper-containing enzyme GlxA. LpmP binds to PG and introduces random cleavages, allowing further degradation by CslZ to create a passage that allows exposure of the glycan at the cell surface. The polymer is then integrated in the cell wall, presumably via interactions involving teichoic acids (6).