Altered PBP4 and GdpP functions synergistically mediate MRSA-like high-level, broad-spectrum β-lactam resistance in Staphylococcus aureus

Abstract

Infections caused by Staphylococcus aureus are a leading cause of mortality worldwide. S. aureus infections caused by methicillin-resistant Staphylococcus aureus (MRSA) are particularly difficult to treat due to their resistance to next-generation β-lactams (NGBs) such as methicillin, nafcillin, and oxacillin. Resistance to NGBs, which is alternatively known as broad-spectrum β-lactam resistance, is classically mediated by PBP2a, a penicillin-binding protein encoded by mecA (or mecC) in MRSA. Thus, presence of mec genes among S. aureus spp. serves as the predictor of resistance to NGBs and facilitates determination of the proper therapeutic strategy for a staphylococcal infection. Although far less appreciated, mecA-deficient S. aureus strains can also exhibit NGB resistance. These strains, which are collectively termed as methicillin-resistant lacking mec (MRLM), are currently being identified in increasing numbers among natural resistant isolates of S. aureus. The mechanism/s through which MRLMs produce resistance to NGBs remains unknown. In this study, we demonstrate that mutations that alter PBP4 and GdpP functions, which are often present among MRLMs, can synergistically mediate resistance to NGBs. Furthermore, our results unravel that this novel mechanism potentially enables MRLMs to produce resistance toward NGBs at levels comparable to those of MRSAs. Our study provides a fresh new perspective about alternative mechanisms of NGB resistance, challenging our current overall understanding of high-level, broad-spectrum β-lactam resistance in S. aureus. It thus suggests reconsideration of the current approach toward diagnosis and treatment of β-lactam-resistant S. aureus infections.

Importance: In Staphylococcus aureus, high-level, broad-spectrum resistance to β-lactams such as methicillin, also referred to as methicillin resistance, is largely attributed to mecA. This study demonstrates that S. aureus strains that lack mecA but contain mutations that functionally alter PBP4 and GdpP can also mediate high-level, broad-spectrum resistance to β-lactams. Resistance brought about by the synergistic action of functionally altered PBP4 and GdpP was phenotypically comparable to that displayed by mecA, as seen by increased bacterial survival in the presence of β-lactams. An analysis of mutations detected in naturally isolated strains of S. aureus revealed that a significant proportion of them had similar pbp4 and GGDEF domain protein containing phosphodiesterase (gdpP) mutations, making this study clinically significant. This study not only identifies important players of non-classical mechanisms of β-lactam resistance but also indicates reconsideration of current clinical diagnosis and treatment protocols of S. aureus infections.

Keywords: gdpP; methicillin-resistant lacking mec (MRLM); pbp4; β-lactam resistance.

Fig 1

GdpP and PBP4 mutations that were identified in MRLM isolates. (A) Venn diagram describing the occurrence of pbp4 and/or gdpP mutations in MRLM strains described in previous studies. (B) Venn diagram describing the occurrence of pbp4 regulatory site associated and/or missense mutations in MRLM strains described previously (C) Schematic diagram of pbp4 and abcA genes separated by a 420-bp intergenic, regulatory site region for wild type (Wt) and the CRB mutant. (D) Illustration and distribution of gdpP-associated mutations (insertion or deletion (INDELs), missense, and nonsense) detected in MRLM isolates. gdpP-associated mutations were further classified based on location of mutation with respect to the DHH/DHHA1 catalytic domain. (E) Schematic representation of different domains in the gdpP gene. The phosphodiesterase catalytic activity of GdpP is mediated by its DHH/DHHA1 domains.

Fig 2

Phenotypic characterization of the strains used in this study. (A) List of strains used for obtaining the results for panels B–E. The regulatory site mutation (36-bp duplication, 36 bp upstream of the start codon) is represented as Ppbp4*. Missense mutations (E¹⁸³A and F²⁴¹R) are represented as pbp4**. (B) Immunoblotting with membrane fraction to detect PBP2a (α-PBP2a, 76.1 kDa, upper panel) was present in only lane 1 (W) and Sortase-A (α-SrtA, lower panel) as a loading control that was detected for all samples. “M” represents protein molecular weight marker. (C) Nitrocefin disk test indicated that only W had β-lactamase activity, as seen by the color change of the disk from yellow to red. The color of the disks streaked with other strains remained unchanged. (D) Bocillin assay with membrane fraction to detect PBPs 1–4. Strains with pbp4 regulatory site mutations (lanes 4 and 5) showed enhanced PBP4 expression compared to the isogenic Wtex strain (lane 2). Strains with the pbp4 S⁷⁵A mutation (lanes 6 and 7) did not show any PBP4 band. ΔgdpP (lanes 5 and 7) did not have an effect on the expression of PBPs. PBP expression levels were visualized by the Typhoon 9410 imager (Amersham/GE Healthcare). (E) Measurement of intracellular levels of CDA for the studied strains. Strains with the ∆gdpP mutation show high levels of CDA compared to their isogenic pairs. P = 0.0003 for Wtex versus Wtex ΔgdpP; P < 0.0001 for Wtex Ppbp4* pbp4** versus Ppbp4* pbp4** ΔgdpP, and Wtex Ppbp4* pbp4** S⁷⁵A versus Wtex Ppbp4* pbp4** S⁷⁵A ΔgdpP. *** represents P ≤ 0.001 and **** represents P ≤ 0.0001.

Fig 3

pbp4 mutations and deletion of gdpP synergistically mediate high-level β-lactam resistance in S. aureus. Population analysis with (A) nafcillin and (B) oxacillin. The triple mutant, Wtex Ppbp4* pbp4** ΔgdpP (blue squares), had increased survival compared to Wtex (pink squares), Wtex ΔgdpP (beige squares), and Wtex Ppbp4* pbp4** (green squares). Resistance displayed by the triple mutant was comparable to that displayed by WT (red squares). Functional mutation of pbp4 due to introduction of S⁷⁵A [Wtex Ppbp4* pbp4** S⁷⁵A (black squares) and Wtex Ppbp4* pbp4** S⁷⁵A ΔgdpP (gray squares)] resulted in NGB susceptibility.

Fig 4

Complementation of gdpP restored β-lactam susceptibility of the NGB-resistant triple mutant. Population analysis of complemented strains carried out with (A) nafcillin and (B) oxacillin. Complementation of the triple mutant with a functional GdpP [Wtex Ppbp4* pbp4** ΔgdpP (gdpP), empty blue squares] resulted is a loss of high-level resistance in a profile similar to Wtex Ppbp4* pbp4** (E) (filled green squares). Complementation of the triple mutant with an empty vector [Wtex Ppbp4* pbp4** ΔgdpP (E), filled blue squares] preserved the resistant phenotype associated with the mutations. (C) Measurement of intracellular levels of CDA for complemented strains. Complementation of the triple mutant with a functional GdpP [Wtex Ppbp4* pbp4** ΔgdpP (gdpP), empty blue squares] led to decreased levels of CDA within cells that were similar to Wtex Ppbp4* pbp4** (E) (filled green squares), whereas complementation of the triple mutant with an empty vector (Wtex Ppbp4* pbp4** ΔgdpP (E), filled blue squares) maintained high levels of CDA.*** represents P ≤ 0.001.

Fig 5

Alteration of GdpP does not affect the composition of the cell wall. Muropeptide purification and analysis profiles of (A) WT, Wtex, and Wtex Ppbp4* pbp4**; (B) Wtex and Wtex ΔgdpP; and (C) Wtex Ppbp4* pbp4** and Wtex Ppbp4* pbp4** ΔgdpP. pbp4-associated mutations led to increased cell wall cross-linking (increased levels of peak 12 and peak 18+, i.e., the “hump”) compared to Wtex. This increase was independent of ΔgdpP. Similarly, deletion of gdpP did not contribute to any significant alterations in the cell wall, irrespective of the presence of pbp4-associated mutations.

Fig 6

Deletion of gdpP led to NGB tolerance in S. aureus. TD test analysis to estimate tolerance levels in S. aureus strains. (A and B) TD test with nafcillin. (C and D) TD test with oxacillin. CFUs in the 50% inhibition zone of the strains containing gdpP deletion, namely Wtex ΔgdpP and Wtex Ppbp4* pbp4** ΔgdpP are higher compared to their isogenic strains with a functional GdpP, namely Wtex and Wtex Ppbp4* pbp4**. All data are from three independent experiments and presented as mean ± SD with one-way analysis of variance. **P ≤ 0.01, ***P ≤ 0.001. ns, not significant.

Fig 7

Synergistic action of PBP4 and GdpP alterations resulted in NGB therapy failure. Percent survival of C. elegans when infected with bacteria over a period of 72 h in the presence of (A) 4-mg/L, (B) 8-mg/L, or (C) 16-mg/L nafcillin. Worms infected with Wtex had 100% survival, whereas infection with Wtex Ppbp4* pbp4** ΔgdpP had decreased survival, similar to that seen by W. (D) The triple mutant, Wtex Ppbp4* pbp4** ΔgdpP, had attenuated virulence compared to Wt and Wtex the the absence of nafcillin. Worms infected with OP50, the E. coli control, showed 100% worm survival. Enumeration of bacterial load within the gut of C. elegans in the presence of (E) 4-mg/L, (F) 8-mg/L, or (G) 16-mg/L nafcillin. Compared to Wtex, which was was not detected, W and Wtex Ppbp4* pbp4** ΔgdpP had significant colonization within the gut. (H) Gut bacterial enumeration when infection was carried out without nafcillin indicated that W, Wtex, and Wtex Ppbp4* pbp4** ΔgdpP had similar colonization abilities. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001; ns denotes P value of >0.05.

Fig 8

Growth assay in presence of ceftaroline. (A–F) Treatment of W, Wtex, and Wtex Ppbp4* pbp4** ΔgdpP in increasing concentrations of ceftraroline (0–1 mg/L) demonstrated that W and Wtex were susceptible to ceftaroline. However, Wtex Ppbp4* pbp4** ΔgdpP was not susceptible to the drug and survived even in the presence of the highest concentration of the drug.