Peptidoglycan DD-peptidases have distinct activities that impact fitness of Acinetobacter baumannii

ABSTRACT

The gram-negative cell envelope is a critical interface between the bacterium and its environment, serving as a selective barrier for nutrient uptake and defense against harmful agents. It also facilitates environmental sensing and adaptive responses. Structurally, it comprises the outer membrane, inner membrane, and periplasmic space, which contains the peptidoglycan layer—a conserved polymer that maintains cell shape and withstands internal turgor pressure. Peptidoglycan consists of glycan strands connected by short peptides, forming a mesh-like structure. In gram-negative bacteria, most peptidoglycan subunits contain tetrapeptides, generated by DD-carboxypeptidases (DD-CPases) that cleave the terminal D-alanine from pentapeptides. Although gram-negative bacteria encode multiple DD-CPases, their precise roles in maintaining cell shape and structural integrity remain poorly understood. The nosocomial pathogen Acinetobacter baumannii encodes three putative DD-CPases. To investigate their functions, we generated single and double mutants in dacC, dacD, and pbpG, which encode homologs of Escherichia coli DD-CPases PBP5 and PBP6a, PBP6b, and the endopeptidase PBP7, respectively. We assessed the mutants for changes in cell morphology, growth dynamics, and pH-dependent stress tolerance. Additionally, we analyzed their peptidoglycan composition to determine the biochemical consequences of enzyme inactivation. Each mutant showed distinct alterations in coccobacillary morphology and growth. Peptidoglycan analysis showed DD-CPase activity, with PBP6b also exhibiting endopeptidase activity. Together, our results demonstrate that each peptidoglycan-modifying enzyme contributes uniquely to cell growth, morphology, and pH tolerance, underscoring their non-redundant functions.

IMPORTANCE

DD-peptidases, including carboxypeptidases and endopeptidases, are crucial for maintaining cell envelope homeostasis, with distinct roles for each enzyme in cell wall biogenesis and structural integrity. The enzymatic characterization presented in this study not only advances our understanding of fundamental Acinetobacter baumannii biology but also highlights these enzymatic activities as targets for the development of innovative therapeutic strategies to combat infections caused by this multidrug-resistant microbe.

INTRODUCTION

The gram-negative cell envelope is structurally organized into three distinct layers, each critical for maintaining cellular integrity and environmental fitness. The hallmark of gram-negative bacteria is the asymmetric outer membrane (OM), which consists of an inner leaflet of glycerophospholipids and an outer (surface-exposed) leaflet enriched with lipopolysaccharide (LPS) or lipooligosaccharide. The lipid asymmetry forms a robust permeability barrier that restricts the entry of toxic compounds (1). Beneath the OM lies the cytoplasmic (inner) membrane, a symmetric bilayer composed of glycerophospholipids. Between the two membranes is the periplasmic space, which contains a thin peptidoglycan (PG) layer (2). The PG is the primary determinant of cell shape and mechanical strength, protecting the cell from osmotic lysis (3). However, recent studies suggest that the LPS-rich OM also contributes significantly to the mechanical stability and shape maintenance during osmotic changes (4, 5). Despite these insights, the molecular mechanisms that coordinate the biogenesis and integration of the OM and PG layers remain poorly understood. Emerging evidence points to direct interactions between PG maturation and the β-barrel assembly machinery (BAM) complex, suggesting a physical and functional linkage connecting the envelope layers (6). Disrupting these connections may represent a promising strategy to compromise envelope integrity and sensitize bacteria to environmental and antibiotic stress.

Penicillin-binding proteins (PBPs) are a diverse group of enzymes essential for the biosynthesis, maturation, and remodeling of the PG layer. These enzymes catalyze key reactions such as formation or breaking of peptide bonds, which are critical for forming and modifying peptide cross-links within the PG meshwork. PBPs are broadly categorized into high molecular weight (HMW) and low molecular weight (LMW) groups, each with distinct enzymatic functions that contribute to PG assembly and homeostasis (7). HMW PBPs are PG synthases and are further divided into two classes: Class A PBPs, which possess both glycosyltransferase and transpeptidase activities, and Class B PBPs, which function as transpeptidases (7, 8). The former enzymes are responsible for polymerizing glycan strands and cross-linking peptides, thereby constructing the PG scaffold. LMW PBPs are generally non-essential under standard growth conditions but play important roles in PG maturation, remodeling, and turnover (9–13). They function mainly as DD-endopeptidases (DD-EPases) and/or DD-carboxypeptidases (DD-CPases), cleaving amide bonds within stem peptides to regulate PG architecture and facilitate cell wall plasticity during growth and environmental adaptation (7, 14, 15).

DD-CPases are critical enzymes involved in the maturation and remodeling of the PG layer. These enzymes hydrolyze the peptide bond between the fourth and fifth amino acids (both usually D-Ala) of pentapeptides, thereby releasing a terminal D-alanine and generating tetrapeptides (16). This PG maturation occurs within minutes of the incorporation of new PG strands in Escherichia coli (17). Tetrapeptides, in turn, serve as substrates for various downstream cell envelope processes. In Acinetobacter baumannii, for instance, tetrapeptides are utilized by the LD-transpeptidase (LD-TPase) LdtJ to form alternative 3-3 crosslinks and add D-amino acids to position 4 of tetrapeptides in PG (18, 19). Beyond their role in PG architecture, tetrapeptides have been shown to bind to the BAM complex and inhibit the assembly of OM proteins, dampening OM protein assembly near mature PG (e.g., the old poles) during the cell cycle (6). Moreover, tetrapeptide formation has been implicated in establishing physical and functional connections between the PG layer and OM proteins, contributing to the overall stability of the cell envelope via LD-TPase-mediated attachment of the OM-anchored lipoprotein Lpp in E. coli. DD-CPases are broadly conserved across gram-negative bacteria (20), with orthologs identified in Enterobacterales, Pseudomonas aeruginosa, and A. baumannii.

E. coli encodes at least seven DD-CPases, including PBP4 (also has DD-EPase activity), PBP4b, PBP5, PBP6a, PBP6b, AmpC, and AmpH (21–25), as summarized in Table S1. In addition, class A PBPs can also have considerable DD-CPase activity (26). PBP5, encoded by dacA, is the most abundant DD-CPase in E. coli (27). Deletion of dacA along with at least two other DD-CPase genes results in minor morphological abnormalities, such as kinks, bends, or branches (28). However, significant morphological defects only arise when all DD-CPase genes are deleted, in addition to the deletion of mrcA (encoding PBP1A), highlighting their overlapping yet distinct functions in PG maintenance. When E. coli grows under low-pH conditions, PBP6b takes over as the most important DD-CPase from PBP5 (29). Under conditions of severe outer membrane biogenesis stress, the seemingly “minor” PBP6b, but not PBP5, becomes essential (30). The weak beta-lactamase AmpC and the beta-lactamase AmpH also contribute to shape maintenance (25). Whether the other DD-CPases have specialized functions is not known.

In contrast, A. baumannii encodes only three predicted DD-CPase genes. dacC encodes a protein with high homology to both E. coli PBP5 (42% identity) and PBP6a (41% identity), denoted herein as PBP5/6a; dacD encodes PBP6b; and pbpG encodes PBP7, with the latter having both DD-EPase and DD-CPase activity (31), as summarized in Table S1. While A. baumannii PBP7 has both DD-EPase and DD-CPase activity (31), E. coli PBP7 has only DD-endopeptidase activity and has a role in cell division (32, 33). In A. baumannii, PBP7 has been shown to be important for bacterial survival during antibiotic treatment, although the precise mechanism underlying the increased susceptibility upon its inactivation remains unclear (31). The functional redundancy in E. coli contrasts with the streamlined DD-CPase system in A. baumannii, making it a powerful model for dissecting the specific contributions of individual DD-CPases to cell envelope integrity and bacterial fitness.

In this study, we generated targeted mutations in genes encoding the predicted DD-CPases, including dacC and dacD, and the predicted DD-CPase/DD-EPase pbpG, to investigate their individual and combined roles. We systematically assessed the morphological and physiological consequences of single and double mutations by analyzing changes in cell shape, growth kinetics, and sensitivity to environmental stress. Additionally, we analyzed the PG composition in the single mutants and biochemically defined the enzymatic activities of PBP5/6a and PBP6b. Our findings demonstrate that each DD-CPase contributes uniquely to A. baumannii growth, morphology, and overall fitness. These enzymes function in distinct, non-redundant pathways to support PG biosynthesis and homeostasis across different phases of the cell cycle.

RESULTS

Morphological defects associated with deletion of dacC, dacD, or pbpG

Given that A. baumannii encodes only three to four predicted DD-CPases, relative to at least seven in E. coli (29), we hypothesized that single-gene deletions in A. baumannii may result in distinct morphological phenotypes, offering fundamental insights into the specific roles of these enzymes.

Morphological defects were evident in each A. baumannii DD-CPase mutant (Fig. 1). The ΔdacC mutant exhibited a population of short, spherical cells relative to wild-type (WT) strain ATCC 17978 (Fig. 1A). Quantitative measurements confirmed a significant reduction in cell length and width (Fig. 1B; Fig. S1A). Complementation of PBP5/6a from a non-native locus rescued the cell morphology (Fig. 1A and B; Fig. S1A and B).

Fig 1.

Microscopy of A. baumannii mutants in logarithmic growth phase. (A) Phase-contrast (left), fluorescence (middle), and merged (right) images of WT strain ATCC 17978, ΔdacC, ΔdacC/pPBP5/6a, ΔdacD, and ΔdacD/pPBP6b. Scale bar: 5 μm. (B) Quantification of cell length (pole to pole) for each cell population (n ≥ 300), measured using ImageJ software with the MicrobeJ plugin. Each dot represents a single cell. Error bars indicate standard deviation. Statistical significance was determined using one-way ANOVA (***P < 0.001, ****P < 0.0001, ns = not significant). (C) Fluorescence intensity quantifications for ΔdacC and ΔdacD. a.u., arbitrary units. (D) Microscopy images, as described in panel A, showing strains ATCC 17978 ΔpbpG, ΔpbpG/pPBP7, AB5075 WT, and AB5075 ΔpbpG. (E) Quantification of coccobacilli and tetrad formations in strain ATCC 17978 WT and ΔpbpG using ImageJ software (n ≥ 300). Tetrads are reported as a percentage of the total population. Data were collected from three independent experiments; one representative image and data set are shown. (F) Quantification of cell length for strain ATCC 17978 ΔpbpG, as described in panel B. (G) Fluorescence intensity quantifications for strain ATCC 17978 ΔpbpG. (H) Quantification of cell length for strain AB5075 WT and ΔpbpG, as described in panel B. (I) Fluorescence intensity quantifications for strain AB5075 WT and ΔpbpG.

In contrast, the ΔdacD mutant displayed an elongated cell morphology with some heterogeneity (Fig. 1A and B; Fig. S1A), and increased D-amino acid modification of PG (Fig. 1C). Cells were stained with a fluorescent derivative of D-alanine (HADA) (34), which is incorporated into PG by PBPs and LD-TPases (35–38), and enables D-amino acid modification measurements. Increased modification with D-alanine may arise from increased tetrapeptide pools, which are substrates for the LD-TPase LdtJ (18), or from loss of PBP6b-mediated D-amino acid trimming from pentapeptides. While the precise mechanism remains unclear, these findings suggest a regulatory role for PBP6b in D-amino acid incorporation into position 5 of the peptide. Growth rate was similar to wild type (Fig. S1C), and complementation fully restored the wild-type morphology and fluorescence pattern (Fig. 1A through C; Fig. S1A).

To assess strain-specific effects, we analyzed DD-CPase mutants in a second A. baumannii background, the well-characterized clinical isolate AB5075 (39), using transposon mutants from the Manoil library (40). Similar shape and growth defects were observed in the AB5075 ΔdacC and ΔdacD mutants (Fig. S2), supporting the same effects for enzymatic inactivation across different strains, with the exception that the ΔdacD mutant did not show an elongated cell morphology in strain AB5075. Length, width, fluorescence intensity, and growth curve calculations were similar to trends described in strain ATCC 17978 (Fig. S2A through E).

Our previous work demonstrated that PBP7 (encoded by pbpG) exhibits DD-CPase and DD-EPase activity and is important for antibiotic survival (31). In strain ATCC 17978, ΔpbpG mutants formed spherical cells that clustered into tetrads, suggesting altered septum positioning and/or impaired cell separation (Fig. 1F and G). These cells also showed reduced length (Fig. 1H), width (Fig. S3A), and slightly decreased fluorescence upon treatment with HADA (Fig. 1I). Complementation restored wild-type morphology and growth (Fig. 1F, H, and I; Fig. S3A and B). This phenotype may reflect a combined lack of DD-CPase and DD-EPase activities.

In strain AB5075, ΔpbpG mutants also formed spherical cells but did not form tetrads (Fig. 1F). Instead, they formed chains of rounder cells (Fig. 1J and Fig. S3C), indicating a defect in cell elongation and separation. Cells also showed increased HADA fluorescence (Fig. 1K), suggesting increased PG D-amino acid modification and/or reduced removal. The strain-specific differences suggest that while PBP7 plays a consistent enzymatic role affecting morphology, additional functions may vary in different A. baumannii genetic backgrounds.

Morphological and growth defects associated with combined deletions of dacC, dacD, and pbpG

While single DD-CPase mutants in A. baumannii exhibit distinct morphological and growth phenotypes, previous studies in E. coli have shown that only multiple deletions (28), particularly in the ΔdacA background, are required to elicit significant morphological defects, suggesting functional redundancy among DD-CPases. To determine whether paired deletions in A. baumannii would exacerbate phenotypic defects, we engineered the ΔdacC ΔdacD, ΔdacC ΔpbpG, and ΔdacD ΔpbpG double mutants in strain ATCC 17978 (Fig. 2). Surprisingly, the double mutants did not exhibit exacerbated phenotypes. Instead, each double mutant was characterized by a combination of the respective single-mutant characteristics. In the ΔdacC ΔdacD mutant (Fig. 2A), cells appeared spherical, like ΔdacC cells, but with increased morphological heterogeneity (Fig. 2B). Interestingly, the additional deletion of dacD rescued the slower growth observed in the ΔdacC single mutant (Fig. S3D). HADA incorporation was significantly higher relative to wild type (Fig. 2C), resembling the ΔdacD phenotype (Fig. 1C). These findings suggest that PBP5/6a and PBP6b have distinct functions: PBP5/6a in cell elongation, whereas PBP6b is involved in D‑amino acid incorporation into PG. The increase in D-amino acid modification may compensate for the loss of PBP5/6a, thereby restoring normal growth.

Fig 2.

Microscopy of A. baumannii strain 17978 double mutants. (A) Phase-contrast (left), fluorescence (middle), and merged (right) images of WT, ΔdacC ΔdacD, ΔdacC ΔpbpG, and ΔdacD ΔpbpG. Scale bar: 5 μm. (B) Quantification of cell length (pole to pole) for each cell population (n ≥ 300) was calculated using ImageJ software with the MicrobeJ plugin. Each dot represents one cell. Error bars represent standard deviation. Statistical significance was determined using one-way ANOVA (****P < 0.0001, ns = not significant). (C) Quantification of fluorescence intensity. (D) Quantification of coccobacilli and tetrad cell formations using ImageJ software (n ≥ 300). Tetrads are reported as a percentage of the total population. Data were collected from three independent experiments; one representative image and data set are shown.

In the ΔdacC ΔpbpG double mutants, the cells were spherical (Fig. 2A and B), consistent with elongation defects observed in both single mutants (Fig. 1A). However, the phenotype was not more severe, implying that PBP5/6a and PBP7 may have overlapping roles in cell elongation. HADA fluorescence intensity remained similar to wild type (Fig. 2C). Notably, tetrad cell clusters, indicative of altered septum positioning and/or cell separation defects, were formed (Fig. 2A), mirroring the ΔpbpG phenotype (Fig. 1F and K). A growth defect was also observed, and after 12 h, cell growth appeared compromised (Fig. S3D), suggesting a synthetic fitness defect in the double mutant.

The ΔdacD ΔpbpG mutant displayed short, spherical cells similar to the ΔpbpG single mutant cells (Fig. 2A and B). However, unlike in other ΔpbpG mutants in the ATCC 17978 background, these cells did not consistently form tetrads. While some tetrads were present, most cells existed as single short units or chains (Fig. 2D), indicating that ΔdacD disruption partially rescues the defect associated with pbpG deletion. This mutant also showed improved growth (Fig. S2D) and elevated HADA fluorescence intensity, consistent with increased D-amino acid incorporation or less removal (Fig. 2C). Together, these results suggest that A. baumannii DD-CPases have both distinct and overlapping roles in PG biosynthesis and remodeling and envelope homeostasis. The absence of additive morphological defects in DD-CPase double mutants highlights the complexity of their functions and suggests compensatory mechanisms that buffer against loss of individual enzymes.

Muropeptide analysis in the DD-CPase mutants reveals altered PG structures

To investigate how PBP5/6a and PBP6b contribute to A. baumannii PG structure, we analyzed the muropeptide compositions of the wild type and ΔdacC, and ΔdacD mutants in exponential and stationary phases (Fig. 3; Table S2), as previously done with the ΔpbpG mutant (31). PG was isolated and digested with a muramidase, and the resulting muropeptides were reduced with sodium borohydride and analyzed by high-performance liquid chromatography (HPLC), as previously described (18, 31, 41). Each mutant exhibited distinct muropeptide profiles relative to wild type. Wild-type strain ATCC 17978 showed the typical profile of a tetrapeptide-rich PG with only minor differences between growth phases (Fig. 3A and D). In contrast, ΔdacC mutants displayed the presence of more pentapeptide-containing muropeptides during both exponential and stationary phases (Fig. 3B and E), indicating that PBP5/6a is required for trimming D-alanine residues from pentapeptides—a hallmark of DD-CPase activity (16).

Fig 3.

Muropeptide analysis in A. baumannii strain 17978 ΔdacC and ΔdacD mutants. Peptidoglycan was isolated from (A) wild-type, (B) ΔdacC, and (C) ΔdacD strains during logarithmic (left) or stationary (right) growth phases and analyzed by HPLC. The ΔdacC mutant exhibited enrichment of pentapeptides, while the ΔdacD mutant showed increased levels of D-modified muropeptides relative to wild type. Red arrows indicate peaks corresponding to specific muropeptide structures. (D, E, and F) Structural illustrations of the labeled muropeptides shown in panels A–C.

ΔdacD mutants also showed pentapeptide enrichment during exponential growth, consistent with DD-CPase activity. However, in both growth phases, these mutants exhibited elevated levels of D-amino acid-modified tetrapeptides (Fig. 3C and F). This increase correlates with the enhanced fluorescence observed in ΔdacD cells labeled HADA (Fig. 1A and C), suggesting that PBP6b may regulate D-amino acids incorporation into PG.

Peptides with D-amino acids (e.g., HADA) at position 4 and 5 are formed by exchange reactions with the terminal D-Ala on tetrapeptides and pentapeptides, respectively. PBPs exchange D-amino acids at position 5, resulting in tetrapeptide-D-amino acid (tetrapeptides with an additional amino acid at position 5), while LD-TPases exchange D-amino acids at position 4, resulting in tripeptide-D-amino acids. DD-CPases are only expected to remove D-amino acids from position 5. Hence, D-amino acid-modified tetrapeptides could be enriched due to loss of PBP6b-mediated trimming of D-amino acids from pentapeptides. Notably, pentapeptide enrichment was only observed during exponential growth, suggesting that PBP6b functions as a DD-CPase primarily during active cell wall synthesis while promoting D-amino acid removal during stationary phase.

Together, these results support distinct roles for PBP5/6a and PBP6b in PG remodeling: PBP5/6a primarily trims pentapeptides to tetrapeptides and regulates cell elongation, while PBP6b modulates D-amino acid incorporation, in the stationary phase, both potentially influencing PG cross-linking and envelope stability.

PBP5/6a and PBP6b are DD-CPases that catalyze tetrapeptide formation, and PBP6b is also an EPase

To validate the proposed enzymatic functions of PBP5/6a and PBP6b supported by PG analysis, we performed activity assays using the purified recombinant enzymes (Fig. S4). Native PBP5/6a and PBP6b, along with their respective catalytically inactive mutants (PBP5/6a_S121A and PBP6b_S122A), in which the catalytic active serine was substituted with alanine, were expressed and purified. Each enzyme was incubated with muropeptides from E. coli CS703-1—enriched in pentapeptides—as previously described (31), and reactions were conducted at both pH 7.5 and pH 5.0 to assess pH-dependent activity.

PBP5/6a exhibited robust DD-CPase activity, as evidenced by a marked reduction in pentapeptide peaks and a corresponding increase in tetrapeptides at both pH values (Fig. 4A). No activity was observed in the reactions containing the PBP5/6a_S121A active site mutant or in the no-enzyme controls. Additionally, PBP5/6a reduced TetraPenta levels while increasing TetraTetra muropeptides, further supporting its DD-CPase activity.

Fig 4.

PBP5/6 and PBP6b exhibit DD-carboxypeptidase activity, and PBP6b also functions as an endopeptidase. (A) Recombinant PBP5/6 and its active site mutant PBP5/6_S121A, and (B) PBP6b and its active site mutant PBP6b_S122A, were incubated with peptidoglycan isolated from E. coli D456 (CS703-1), which contains Tetra, Penta, TetraTetra, and TetraPenta as the major components. Both PBP5/6 and PBP6b exhibited DD-carboxypeptidase activity, cleaving pentapeptides, whereas the active-site mutants showed no activity. PBP6b also demonstrated endopeptidase activity at both neutral (pH 7.5) and acidic pH (pH 5.0) conditions. Enzymes were used at a final concentration of 10 μM. Muropeptides are labeled in red.

PBP6b was also active against pentapeptides and tetrapentapeptides, converting these to Tetra and TetraTetra muropeptides, respectively, at pH 7.5 (Fig. 4B). Interestingly, under acidic conditions at pH 5.0, more so than at pH 7.5, PBP6b showed robust DD-EPase activity. This dual activity was dependent on the catalytic serine residue, as the PBP6b_S122A active site mutant and no-enzyme controls showed no such changes.

Together, these results confirm that PBP5/6a and PBP6b catalyze formation of tetrapeptides from pentapeptides, consistent with their role as DD-CPases, as anticipated by sequence homology and PG analysis. Moreover, PBP6b exhibits pH-dependent DD-EPase activity, potentially contributing to PG remodeling under acidic conditions, as previously observed for E. coli PBP6b, which only showed DD-CPase activity (29).

PBP5/6a and PBP7 play unique roles under pH stress

To determine whether A. baumannii DD-CPases contribute to environmental stress tolerance, we assessed the growth and cell morphology defects of ΔdacC, ΔdacD, ΔpbpG mutants under acidic or alkaline conditions (Fig. 5). The ΔdacC mutant exhibited growth defects in acidic (pH 5.0) and alkaline conditions (pH 9.2), showing reduced growth in both relative to neutral pH (pH 7.5) (Fig. 5A). These findings suggest that PBP5/6a is required for maintaining cell envelope integrity across a broad pH range and that its function cannot be compensated by other DD-CPases.

Fig 5.

Colony growth and cell morphology of A. baumannii strain ATCC 17978 mutants under acidic and alkaline stress conditions. (A) Colony spot assays of WT and mutant strains serially diluted 10-fold starting at OD₆₀₀ 0.05, plated on Luria-Bertani (LB) agar adjusted to the indicated pH values. (B) Phase-contrast (left), fluorescence (middle), and merged images of WT, ΔdacC, ΔdacC/pPBP5/6, ΔdacD, and ΔdacD/pPBP6b cells grown at pH 5.0. (C) Same as panel B, but cells grown at pH 9.2. Scale bar: 5 μm. (D) Quantification of cell length (pole to pole) at pH 5.0 of cell populations (n ≥ 300), measured using ImageJ with the MicrobeJ plugin. Each dot represents a single cell. Error bars indicate standard deviation. Statistical significance was determined using one-way ANOVA (*P < 0.05, **P < 0.01, ****P < 0.0001, ns = not significant). (E) Quantification of fluorescence intensity at pH 5.0. (F) Same as panel D, but for cells grown at pH 9.0. (G) Same as panel E, but for cells grown at pH 9.0. (H) Phase-contrast (left), fluorescence (middle), and merged (right) images of WT, ΔpbpG, and ΔpbpG/pPBP7 cells grown at pH 5.0. (I) Same as panel H, but cells grown at pH 9.0. Each experiment was independently replicated three times, and one representative data set was reported.

In contrast, the ΔdacD mutant did not show obvious growth defects under any tested condition (Fig. 5A). Interestingly, this mutant displayed enhanced tolerance to alkaline and acidic pH relative to wild type, suggesting that PBP6b inactivation confers a selective advantage in stress environments, potentially by altering PG remodeling or D-amino acid incorporation (Fig. 3C and 4B).

The ΔpbpG mutant was highly sensitive to both acidic and alkaline conditions, with complete growth inhibition at pH 9.2 (Fig. 5A). This indicates that PBP7 is essential for survival under pH stress and, like PBP5/6a, its function is not redundant with other DD-CPases.

DD-carboxypeptidase mutants exhibit unique morphological responses under pH and osmotic stress

To further investigate the role of DD-CPases under environmental stress, we examined the morphology of ΔdacC, ΔdacD, and ΔpbpG cells using microscopy under acidic (Fig. 5B, D, and H, and Fig. S5A) or alkaline (Fig. 5C, F, G, and I, and Fig. S5B) conditions. In alkaline conditions, all strains—including wild type—exhibited enrichment of spherical cells (Fig. S5C, D, F, H, I, and K) and reduced HADA fluorescence intensity (Fig. S5E, G, and J) compared to neutral pH. These observations suggest that alkaline stress may affect cell elongation and D-amino acid modification of PG, likely through mechanisms independent of PBP5/6a and PBP6b. In contrast, acidic conditions led to shorter, brighter cells in the ATCC 17978 wild-type strain, consistent with reduced D-amino acid removal (Fig. S5E, G, and J). Mutants displayed similar length phenotypes at acidic pH relative to those observed at a neutral pH: ΔdacC cells were shorter, and ΔdacD cells were longer (Fig. 5D). However, a notable difference was the occasional appearance of chaining phenotypes in ΔdacD cells under both pH extremes (Fig. 5B through D), suggesting a potential role for PBP6b in cell separation during stress. HADA fluorescence intensity analysis revealed a striking shift at pH 5.0: ΔdacC cells showed increased fluorescence intensity relative to wild type, while ΔdacD cells showed reduced fluorescence (Fig. 5E). This suggests that PBP5/6a may remove incorporated D-amino acids under acidic conditions, a role previously attributed to PBP6b at neutral pH.

In the ΔpbpG mutants, tetrad-like cell clusters persisted under both acidic and alkaline conditions (Fig. 5H and I), but the nature of these clusters varied. Under acidic conditions, tetrads appeared as a single group of clustered cells, indicating a severe defect. In alkaline conditions, multiple tetrads aggregated into larger clusters, further compounding the phenotype. These morphological abnormalities likely contribute to the complete growth inhibition at pH 9.2 (Fig. 5A). The distinct responses under each stress condition suggest that PBP7 may have additional, condition-specific roles in regulating cell wall synthesis and envelope remodeling.

In addition to pH stress, we assessed whether DD-CPase mutants exhibit altered tolerance to osmotic stress. To test osmotic stress tolerance, we performed colony spot assays under hypo- and hyperosmotic conditions using LB agar with various NaCl concentrations (Fig. S6). Across all tested conditions, including Miller LB (171 mM NaCl), low-salt (0 mM NaCl), and high-salt environments (+250 mM NaCl), survival patterns were similar to wild type at under low salt conditions, except for ΔpbpG, which showed slightly increased susceptibility at high salt. These findings suggest that PBP7 may contribute to adaptation under hypoosmotic stress.

DD-carboxypeptidases are not redundant in Acinetobacter baumannii

To further investigate the potential redundancy among DD-CPases, we performed complementation experiments by expressing other DD-CPases in single-mutant backgrounds (Fig. S7). In ΔdacC mutants complemented with pPBP6b or pPBP7, cells remained short and round, similar to the ΔdacC phenotype, indicating that neither enzyme restored elongation (Fig. S7A and B). Likewise, the ΔdacD mutant complemented with either pPBP5/6a or pPBP7 did not regain wild-type morphology. PBP5/6a expression produced a heterogeneous population of oblong and elongated cells, while PBP7 expression maintained elongated cells. Fluorescence intensity also remained high, confirming that dacD disruption could not be compensated for by other DD-CPases. In ΔpbpG mutants complemented with PBP5/6a and PBP6b, tetrad formation was eliminated, but the cells were still short and round, similar to the original ΔpbpG phenotype. These findings suggest that while other DD-CPases cannot compensate for elongation defects caused by pbpG loss, they may partially restore cell separation.

DISCUSSION

PG homeostasis—through coordinated synthesis, remodeling, and recycling—is essential for bacterial survival, morphology, and adaptation to environmental stress. Among the enzymes involved in PG remodeling and turnover, DD-CPases remain to be further studied in gram-negative bacteria other than E. coli. These enzymes trim terminal D-alanine residues from pentapeptides, generating tetrapeptides that influence PG cross-linking, cell shape, and stress resilience. In E. coli, DD-CPases are functionally redundant, with multiple enzymes compensating for one another depending on environmental conditions (29).

In this study, we characterized the roles of the putative DD-CPases PBP5/6a and PBP6b, and the DD-CPase/EPase PBP7 in A. baumannii, revealing different roles than their homologs in E. coli. Our results demonstrate that each DD-CPase in A. baumannii performs a unique, non-redundant function in PG homeostasis, cell shape maintenance, and environmental stress adaptation.

The ΔdacC mutant exhibited short spherical cells, implicating PBP5/6a in cell elongation. ΔdacD mutants, while slightly longer than wild type, displayed increased D-amino acid incorporation, suggesting that PBP6b may remove D-amino acids added by PBP-mediated exchange. ΔpbpG mutants exhibited the most severe defects, forming tetrad-like clusters, consistent with an important role for PBP7 in cell morphology and separation. These observed phenotypes were consistent also in the clinical isolate AB5075, although with some strain-specific morphological variation in the ΔdacD and, particularly, the ΔpbpG mutants.

Double mutant analysis further supported the non-redundant nature of these enzymes. Combinatorial deletions (ΔdacC ΔdacD, ΔdacC ΔpbpG, ΔdacD ΔpbpG) resulted in additive or compensatory phenotypes, but not in synthetic lethality. Notably, ΔdacD ΔpbpG mutants showed partial restoration of the ΔpbpG phenotype, suggesting that PBP6b may compensate for the morphological defects associated with PBP7 loss.

Muropeptide profiling revealed that ΔdacC mutants accumulated pentapeptides, confirming PBP5/6a’s role in pentapeptide trimming. ΔdacD mutants showed enrichment of D-arginine- and D-lysine-acid-modified muropeptides, suggesting that PBP6b may specifically cleave these modified substrates or indirectly regulate their incorporation. These findings highlight the specialized substrate preferences and regulatory roles of individual DD-CPases in A. baumannii.

Enzymatic assays confirmed that both PBP5/6a and PBP6b possess DD-CPase activity. Additionally, PBP6b exhibited DD-EPase activity that increased at acidic pH, similar to the previously reported DD-EPase activity for PBP7 (31). This reduced specificity—combining DD-CPase and DD-EPase activities—may explain the lack of redundancy among A. baumannii DD-CPases, in contrast to the environment-specific redundancy observed in E. coli, aside from PBP4’s DD-CPase/EPase activity (29).

Environmental stress assays further emphasized the unique roles of these enzymes. ΔdacC and ΔpbpG mutants exhibited growth defects across all tested pH conditions, while ΔdacD mutants showed enhanced tolerance to alkaline stress. Morphological analysis under pH stress revealed that ΔpbpG mutants experienced exacerbated shape defects, with tetrads forming as single unseparated cells under acidic conditions and as aggregated clusters under alkaline conditions. These observations suggest that PBP7 may have additional, condition-specific roles in cell wall remodeling.

In conclusion, our study reveals that A. baumannii DD-peptidases are not functionally redundant but instead perform distinct, crucial roles in PG remodeling, cell shape maintenance, and stress adaptation. The multifunctionality of these enzymes—combining DD-CPase and DD-EPase activities—may reflect an evolutionary adaptation to optimize PG regulation in this clinically relevant pathogen. This streamlined adaptation could be advantageous for survival in hospital environments where this pathogen is frequently found and subjected to antibiotic pressures and desiccation. A simplified set of multifunctional enzymes could reduce metabolic burden and increase regulatory efficiency. A. baumannii presents a unique system of multifunctional DD-CPases. Only one other group of the ESKAPE pathogens presents consolidated DD-peptidases with the gram-positive Staphylococcus aureus containing one dual-activity DD-CPase/DD-EPase (PBP4) (7). Future studies should explore the broader regulatory networks and potential therapeutic vulnerabilities associated with these unique enzymes.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Strains and plasmids used in this study are listed in Table S3. Acinetobacter baumannii strains were cultured aerobically from frozen stock on LB agar or in LB broth at 37°C. Unless otherwise specified, antibiotics were used at the following concentrations: 25 mg/L kanamycin and 10 mg/L tetracycline.

A. baumannii strain ATCC 17978 has been reported to consist of two distinct variants (42). To confirm the variant used in this study, colony PCR was performed using primers specific to the cardiolipin synthase gene (clsC2), which yielded a positive product, indicating that all experiments were conducted using the A. baumannii 17978 UN variant. A. baumannii strains, including AB5075, are known to exhibit phase-variable colony opacity phenotypes (43). Microscopic examination of AB5075 cultures revealed that most colonies were opaque, with only <1% translucent colonies observed among more than 200 colonies per plate (n = 3 plates per strain). For consistency, only opaque variants were used in all experiments, including those involving the AB5075 dacA/C::tn, pbp6b::tn, and pbpG::tn mutants.

Construction of genetic mutants

The primers used in this study are listed in Table S4 A. baumannii mutants, including ΔdacC, ΔdacD, ΔpbpG, and all double mutants, were generated using the recombination-mediated genetic engineering (recombineering), as previously described (18, 44–46). Briefly, a kanamycin resistance cassette flanked by FLP recombination target sites was PCR-amplified from the pKD4 plasmid using primers containing 125 bp homology arms specific to the target gene. The linear PCR product was electroporated into A. baumannii strain ATCC 17978, harboring the pRECAb plasmid (pAT03). Following recovery in LB broth, transformants were plated on LB agar supplemented with 7.5 mg/L kanamycin. All mutants were verified by PCR.

To remove the recombineering plasmid, mutants were grown on LB agar containing 2 mM nickel(II) chloride (NiCl₂) and replica-plated onto LB agar with either kanamycin or tetracycline, as previously done (41). Colonies that were kanamycin-resistant and tetracycline-sensitive were confirmed by PCR to have lost the plasmid.

For markerless deletions, the kanamycin resistance cassette was excised by transforming the cured mutants with pMMB67EH::FLP (pAT08), which expresses FLP recombinase. Transformants were plated on LB agar containing tetracycline and 2 mM IPTG to induce recombinase expression. Successful excision of the resistance cassette was confirmed by PCR.

For complementation, the coding sequences of dacC (PBP5/6a), dacD (PBP6b), and pbpG (PBP7), including 200 bp of upstream and downstream flanking regions, were amplified from A. baumannii ATCC 17978 genomic DNA. These fragments were cloned into the pABBRknR plasmid at the XhoI and KpnI restriction sites. The resulting constructs (pPBP5/6a, pPBP6b, and pPBP7) were introduced into each single deletion mutant for analysis under the control of their native gene promoter.

Fluorescent HADA staining

Fluorescent D-Amino Acid staining was performed as previously described (18, 31, 47). Overnight cultures were grown with shaking at 37°C in 5 mL of LB broth, supplemented with appropriate antibiotics when necessary. The following day, cultures were diluted 1:100 into fresh LB medium (5 mL total volume) and incubated at 37°C with shaking until OD₆₀₀ reached 0.3–0.5. The cells were harvested, washed to remove residual debris, and resuspended in 1 mL LB. Two microliters of 10 mM HCC-(linezolid-7-nitrobenz-2-oxa-1,3-diazol-4-yl)-amino-d-alanine (HADA) were added to each culture. The mixtures were incubated at 37°C for 20 minutes to allow incorporation of the fluorescent alanine. Following staining, cultures were washed with LB broth and fixed in phosphate-buffered saline (PBS) containing a 1:10 dilution of 16% paraformaldehyde.

Microscopy

Microscopy was performed as previously described (18, 31, 47). Paraformaldehyde-fixed cells were mounted on 1.5% agarose pads and imaged using an inverted Nikon Eclipse Ti-2 wide-field epifluorescence microscope. The system was equipped with a Photometrics Prime 95B camera and a Plan Apo 100× objective lens (NA 1.45). Phase-contrast and fluorescence images were acquired using NIS-Elements software. Blue fluorescence was visualized using a Sola LED light engine with the following filter set: 350/50 nm excitation filter, 460/50 nm emission filter, and a 400 nm dichroic mirror.

Image analysis

Microscopy images were processed as previously described (18, 31, 47). Pseudocoloring of fluorescence images was performed using ImageJ Fiji (48). Quantitative measurements, including length, width, fluorescence intensity, and total cell surface area, were measured using MicrobeJ (18, 31, 47, 49). Data were analyzed and visualized using GraphPad Prism version 10.2.2. Each experiment was independently repeated three times. Quantification data and images in the figures are from one representative biological replicate.

Peptidoglycan isolation

PG was isolated as previously described (18, 31, 41). Biological replicates were cultured in 400 mL of LB broth to either mid-logarithmic or stationary phase. Cells were harvested by centrifugation at 7,000 × g for 20 minutes at 4°C using an Avanti JXN-26 Beckman Coulter centrifuge with a JA-10 rotor. Pellets were resuspended in 6 mL of chilled PBS and lysed by dropwise addition of boiling 8% sodium dodecyl sulfate (SDS). PG was purified following the protocol established by Glauner et al. (50). Sacculi were treated with the Cellosyl muramidase (gift from Hoechst, Frankfurt am Main, Germany) to release muropeptides, which were then reduced with sodium borohydride and separated using HPLC on a 250 × 4.6 mm, 3 μm Prontosil 120-3-C₁₈ AQ reversed-phase column (Bischoff, Leonberg, Germany). Elution was monitored by absorbance at 205 nm. Muropeptide peaks were identified by comparison to published chromatograms (18, 41, 51).

Construction of PBP5/6a and PBP6b active-site mutants

To generate catalytically inactive variants of PBP5/6a and PBP6b, gene fragments encoding PBP5/6a_S121A (serine 121 substituted with alanine) and PBP6b_S122A (serine 122 substituted with alanine) were synthesized by Twist Biosciences. These mutations target the conserved serine residue in the active site, essential for enzymatic activity.

Construction of PBP5/6a and PBP6b overexpression strains

The coding sequences for dacC and dacD were amplified from A. baumannii ATCC 17978 genomic DNA, while dacC_S121A and dacD_S122A were amplified from synthesized gene fragment DNA. The amplicons were generated using primers that included a C-terminal His_8×-tag sequence to facilitate downstream purification. The resulting PCR products were cloned into the NdeI and BamHI restriction sites of the pT7-7Kn expression vector (52). The constructs pT7-7Kn::dacC, pT7-7Kn::dacC_S121A, pT7-7Kn::dacD, and pT7-7Kn::dacD_S122A were transformed into E. coli C2987 chemically competent cells and confirmed by Sanger sequencing. Verified plasmids were subsequently transformed into E. coli C2527 (BL21) cells (New England BioLabs, Inc.) for protein expression and purification.

Purification of recombinant PBP5/6a and PBP6b

PBP expression and purification were done as previously described (31). Briefly, E. coli BL21 cells harboring pT7-7Kn::dacC, pT7-7Kn::dacC_S121A, and pT7-7Kn::dacD_S122A were cultured in 1,000 mL of LB broth at 37°C until reaching OD₆₀₀ 0.3–0.5. Due to toxicity associated with overexpression, 4,000 mL of culture was needed for E. coli BL21 cells harboring pT7-7Kn::dacD. 1L was sufficient with the other cultures. All cultures were induced with 1 mM IPTG for 4 hours at 25°C.

Following incubation, cells were harvested, washed with cold 1× PBS, and pelleted. Pellets were frozen overnight at −20°C, then thawed on ice and resuspended in 20 mL of lysis buffer (20 mM Tris, 300 mM NaCl, 10 mM imidazole, pH 8). Cells were lysed by sonication using a Fisher Scientific Model 120 Sonic Dismembrator, with 20 seconds on, followed by 20 seconds off, for a total of 10 cycles at 80% amplitude. Lysates were centrifuged at 15,428 × g for 10 minutes at 4°C. Pellets were resuspended in extraction buffer (25 mM Tris/HCl, 10 mM MgCl₂, 1 M NaCl, 0.02% NaN₃, 2% TX-100, pH 7.5) and incubated overnight at 4°C with gentle rocking. The next day, supernatants were collected by centrifugation and incubated with HisPur Ni-nitrilotriacetic acid resin (Thermo Scientific), pre-equilibrated with extraction buffer for 2 h at 4°C with rotation. The resin mixture was transferred to a 10-mL gravity-flow column and washed sequentially with 20 mL of lysis buffer containing increasing imidazole concentrations (0 mM, 15 mM, and 30 mM). Proteins were eluted in eight fractions using 500 µL of elution buffer (20 mM Tris, 300 mM NaCl, 250 mM imidazole, pH 8), incubated for 5 minutes before gravity elution. Fractions containing protein, identified by SDS-PAGE, were pooled and dialyzed overnight at 4°C in dialysis buffer (10 mM Tris, 50 mM KCl, 0.1 mM EDTA, 5% glycerol, pH 8) using 12 mL dialysis cassettes (Thermo Scientific). Protein purity and identity were confirmed by SDS-PAGE and bocillin-binding assays.

Activity assays

PBP5/6a and PBP6b activity assays were performed in a final reaction volume of 50 μL containing either neutral pH buffer (20 mM Tris/HCl, 50 mM NaCl, 2 mM MgCl₂ pH 7.5) or acidic pH buffer (20 mM sodium acetate, 50 mM NaCl, 2 mM MgCl₂, pH 5.0). Each reaction contained 10 μM of purified PBP5/6a, PBP5/6a_S121A, PBP6b, or PBP6b_S122A. PG isolated from E. coli CS703-1, enriched in pentapeptides, was added as a substrate. Reactions were incubated at 37°C for 16 hours. Samples were boiled for 10 minutes to terminate reactions and incubated with 100 μg/mL cellosyl for a further 16 hours at 37°C. Samples were then boiled for 10 minutes and centrifuged at 16,000 × g for 10 minutes in a microfuge. The resulting muropeptides were reduced with sodium borohydride and acidified to a pH of 4.0–4.5. Control samples contained PG from E. coli CS703-1 without enzyme. Muropeptide analysis was performed as previously described (50).

Spot assays with pH medium

Overnight cultures were grown in LB broth at 37°C and then back-diluted to OD₆₀₀ 0.05. Cultures were serially diluted 1:10 across five dilution steps. From each dilution, 5 µL aliquots were spotted onto LB agar plates adjusted to various pH levels. To prepare pH-adjusted LB plates, 50 mM Tris was added to standard LB agar, and the pH was adjusted to the desired value using either HCl or NaOH for the pH stress assays, while NaCl was adjusted to high (+250 mM NaCl) or low (0 mM NaCl). Plates were incubated at 37°C for 16 h, after which images were captured. All images shown are representative of three independent experiments.

Statistical analysis

Statistical analysis for cell morphology and fluorescence intensity data was performed using one-way ANOVA. A P-value of less than 0.05 was considered statistically significant.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/mbio.00072-26.

Supplemental material. mbio.00072-26-s0001.docx.

Figures S1-S7; Tables S1-S4.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.