PBP4 Mediates β-Lactam Resistance by Altered Function

ABSTRACT

Penicillin binding protein 4 (PBP4) can provide high-level β-lactam resistance in Staphylococcus aureus. A series of missense and promoter mutations associated with pbp4 were detected in strains that displayed high-level resistance. We show here that the missense mutations facilitate the β-lactam resistance mediated by PBP4 and the promoter mutations lead to overexpression of pbp4. Our results also suggest a cooperative interplay among PBPs for β-lactam resistance.

TEXT

Antibiotic resistance in Staphylococcus aureus is an important cause of concern for the health care system worldwide. β-Lactam antibiotics are a prominent class of antibiotics used to treat infections caused by the bacteria. Resistance to traditional β-lactam antibiotics, such as penicillin, methicillin, or their derivatives, is widespread among S. aureus and is primarily mediated by penicillin binding protein 2a (PBP2a) (1), which is encoded by mecA or mecC (2, 3).

We previously reported that S. aureus strains lacking mecA can develop high-level β-lactam resistance on passage (4–6). Appearance of this mode of resistance in different strains of S. aureus (COLnex and SF8300ex, i.e., COLn and SF8300 strains lacking mecA) and in different β-lactam drugs (ceftaroline, ceftobiprole, and nafcillin) suggested that the underlying resistance mechanism is a general one. Genome sequences of the strains obtained from passage in ceftaroline and ceftobiprole indicated high frequencies of missense and promoter mutations in pbp4 among these strains (6). Although the role of the pbp4 mutations remains unclear, wild-type strains lacking pbp4 and mecA were unable to develop high-level resistance (5), suggesting that PBP4 is essential for this mode of resistance. PBP4 is an uncanonical, low-molecular-weight penicillin binding protein of S. aureus whose mechanism of action is poorly characterized.

To determine whether the basis of high β-lactam resistance of the previously generated nafcillin-passaged strains (COLnex and SF8300ex resistant to nafcillin [CRN and SRN, respectively] [Table 1]) follows the same underlying principles of ceftaroline and ceftobiprole resistance, their genomes were sequenced. First, three colonies each from CRN- and SRN-passaged strains were chosen. All three strains displayed high-level resistance to nafcillin and ceftaroline (Table 1), suggesting a common mechanism of action. One clone each from CRN and SRN were randomly chosen, and their genomes were sequenced using the method described in Text S1 in the supplemental material.

TABLE 1.

Wild-type, passaged, and mutant strains used in this study

Strain	Driver for selectiona	MIC (μg/ml) tob:			Mutations in:
		CPT	NAF	VAN	PBP1	PBP2	PBP3	pbp4 promoter	PBP4
COLnexc	NA	<0.25 S	0.5 S	2 S
CRB	BPR	64 R	128 R	0.5 S				36-bp duplication at 290 bp upstream of pbp4 start codone	E183A; F241R
CRB Δpbp4		0.25 S	0.25 S					36-bp duplication at 290 bp upstream of pbp4 start codone
CmTc	CPT	>64 R	>256 R	0.5 S		D156N		A → C at −399 bp upstream of pbp4 start codonf	T201A; F241L
CmTc Δpbp4		0.25 S	0.25 S			D156N		A → C at −399 bp upstream of pbp4 start codonf
CRN	NAF	64 R	256 R	1 S				191-bp deletion at 135 bp upstream of pbp4 start codonf	R200L; F241L
CRN Δpbp4		0.25 S	0.5 S					191-bp deletion at 135 bp upstream of pbp4 start codonf
SF8300exd	NA	0.25 S	0.5 S	1 S
SRB	BPR	4 S	8 S	1 S	H499R; E567K	Y437C; V445L; Q453R; M559I	W228Xg		E183V; F241R
SRB Δpbp4		0.125 S	0.25 S		H499R; E567K	Y437C; V445L; Q453R; M559I	W228Xg
SRT	CPT	>64 R	64 R	1 S				“A” deletion at −378 bp upstream and 11-bp deletion at 300 bp upstream of pbp4 start codonf	N138K; H270L
SRT Δpbp4		0.125 S	0.25 S					“A” deletion at −378 bp upstream and 11-bp deletion at 300 bp upstream of pbp4 start codonf
SRN	NAF	32 R	128 R	1 S		Δ105 bp	T619R
SRN Δpbp4		0.25 S	0.5 S			Δ105 bp	T619R

^aBPR, ceftobiprole; CPT, ceftaroline; NAF, nafcillin.

^bVAN, vancomycin; R, resistant; S, susceptible.

^cCOL strain with mecA excised, parent of CRB, CmTc, and CRN.

^dSF8300 strain with mecA excised, parent of SRT, SRB, and SRN.

^eSee reference 13.

^fSee reference 6.

^gX, truncation of protein.

Both CRN and SRN showed mutations targeting PBPs similar to those observed before in ceftaroline- and ceftobiprole-resistant strains. CRN had R200L and F241L PBP4 missense mutations near the active site of PBP4 (5) and a 191-bp deletion starting at −135 bp upstream of the pbp4 start codon. Notably, SRN lacked pbp4 mutations and had a 105-bp deletion at the C-terminal end of the pbp2 gene and a T619R missense mutation in PBP3 (Table 1). The 105-bp deletion in pbp2 did not affect the transpeptidase (TPase) or glycosyltransferase (GTase) domains of PBP2 but affected a region that shares no similarity with any known domains through BLAST searches.

To identify accessory gene mutations that might be commonly present among the passaged strains, genome sequences of all six passaged strains were compared (Table 1). This revealed a total of six genes to be mutated at a frequency of at least twice among the six passaged strains (see Table S1 in the supplemental material). Five of these candidates, i.e., GdpP, FmtA, RpoB, Stp1, and ClpX, have already been implicated in β-lactam resistance (7–11), but their precise role in resistance is currently unknown. The sixth candidate, TcaA, was shown to be upregulated on treatment with cell wall-active antibiotics (12). Notably, all six passaged strains had mutations in gdpP (Table S1).

PBP4 drives β-lactam resistance among passaged strains. We previously showed that pbp4 plays an important role in the resistance among passaged strains, as deletion of pbp4 in CRB, SRB, and SRT rendered them completely susceptible to β-lactams (5, 13). All three strains had mutations in pbp4 (Table 1). To determine pbp4's role in CRN and SRN, in-frame deletion of pbp4 was carried out as previously described (5). Deletion of pbp4 in CRN made it completely susceptible to β-lactams (ceftaroline and nafcillin; (MICs, ≤0.5 μg/ml)). This result was expected, but strikingly, pbp4 deletion in SRN, although it has no pbp4 mutation, also made it completely susceptible to β-lactam drugs (ceftaroline and nafcillin; MICs, ≤0.5 μg/ml) (Table 1). These results indicated that pbp4 played a central role in resistance not only in the CRN strain but also in the SRN strain. The roles of pbp2 and pbp3 mutations in SRN are unknown, but they may play a supportive role in the resistance process, underscoring the complex interplay among PBPs in S. aureus. Deletion of pbp4 in CmTc, which also has pbp4 mutations, likewise turned it into a completely susceptible strain (Table 1).

PBPs perform the penultimate steps of bacterial cell wall synthesis through their transpeptidase (TPase) and glycosyltransferase (GTase) domains (14). They are also the exquisite targets of the β-lactam class of antibiotics (1). S. aureus has five PBPs, of which PBP4 is considered uncanonical because it possesses only the TPase domain and is roughly half the size of the other PBPs. TPase activity mediates the cross-linking of bacterial peptidoglycan by the formation of a pentaglycine crossbridge between two adjacent PG molecules, whereas GTase activity mediates formation of a glycosidic bond between peptidoglycans (15–17). Thus, in principle, a monofunctional PBP (such as PBP4) has to work in concert with a bifunctional PBP (such as PBP2, the only known bifunctional PBP in S. aureus) for effective cell wall synthesis. PBP2 was previously implicated to function in concert with PBP4, although a direct interaction between them has not been shown experimentally (18). Thus, PBP2 missense mutations that were detected in CmTc and SRB apart from SRN (Table 1) probably also play a yet-to-be-determined role in resistance.

Missense mutations in PBP4 provide β-lactam resistance. A total of six missense mutations surrounding the active site of PBP4 (S75) were detected among passaged strains (5). Whether these mutations provide β-lactam resistance or have an indirect role in resistance, such as facilitating interactions with other PBPs or proteins that mediate cell wall synthesis, is currently unknown. To precisely determine the contribution of pbp4, we cloned the wild-type and mutated pbp4 strains from the COLn and passaged strains (CRB, CmTc, CRN, SRB, and SRT) in a constitutive expression, high-copy-number vector (pTX_Δ) as described before (5). These clones were introduced to a surrogate recipient, COLnex Δpbp4, a wild-type background that lacks pbp4, and the β-lactam resistance of these strains was evaluated. Population analysis of the resultant strains showed that the PBP4 missense mutations conferred significant nafcillin resistance to the recipient compared with that of wild-type PBP4 (Fig. 1).

FIG 1.

PBP4 missense mutations confer β-lactam resistance. pbp4 from wild-type (COLn) and mutant passaged strains were cloned in constitutively expressed vector pTX_Δ. The resultant plasmids were transformed into the surrogate recipient COLnex Δpbp4 strain, and population analysis was carried out in nafcillin. Two-way analysis of variance of the data revealed a significant difference (P < 0.0417) between strains.

Promoter mutations in pbp4 lead to its overexpression. Four of our passaged strains had mutations in the pbp4 promoter region (Table 1). The promoter mutations varied widely from a small insertion and deletion to a large duplication and deletions. The 36-bp duplication that was detected in the pbp4 promoter region of CRB resulted in its overexpression (13). To determine the role of the other pbp4 promoter mutations in CRB expression, we cloned the pbp4 promoters (Ppbp4) from wild-type (COLn) and mutant (CRB, CmTc, CRN, and SRT) passaged strains into a lux reporter plasmid (19). The resultant plasmids were introduced into the COLnex strain through transformation, and reporter activity was measured. All strains with a mutated Ppbp4 had higher lux signals than the wild-type strain, suggesting enhanced pbp4 expression due to the promoter mutations (Fig. 2). The results also suggest that the promoter mutations are responsible for pbp4 overexpression and probably lead to considerable β-lactam resistance, as shown previously in the CRB strain (13).

FIG 2.

pbp4 promoter mutations cause enhanced pbp4 expression. pbp4 promoters from wild-type (COLn) and mutant passaged strains were cloned into lux reporter plasmid pAmilux. Lux signals and bacterial optical density at 600 nm (OD₆₀₀) were measured at 4 h postculture, and data are represented by dividing the lux signal by OD₆₀₀. P values of <0.0001 were revealed by nonparametric t test between wild-type and other strains.

PBP4 is generally expressed in very small amounts, as suggested by transcriptional analysis and bocillin assays performed using bacterial whole-cell lysates (13). Thus, pbp4 expression is generally under tight regulatory control in bacterial cells. Enhanced pbp4 expression due to promoter mutations suggests a lack of regulatory control that leads to enhanced pbp4 expression. The regulators that control pbp4 expression are currently unknown.

The resistant passaged strains displayed increased cell wall thickening and abnormal cell morphology. We previously reported that CRB had highly cross-linked peptidoglycan as a consequence of pbp4 overexpression (13). Because peptidoglycans are the building blocks of the bacterial cell wall, enhanced cross-linking may affect bacterial cell wall morphology. To determine if peptidoglycan cross-linking affects bacterial cell wall structure, we performed transmission electron microscopic (TEM) analysis on the COLnex and CRB strains, as described in Text S1. This revealed cell wall thickening of CRB compared with its parental COLnex strain (Fig. 3). TEM analysis on other passaged strains also showed cell wall thickening and abnormal cell morphologies, such as uneven cell division, roughness of the cell surface, and compromised structural integrity (see Fig. S1 in the supplemental material). Thus, cell wall thickening was a common phenomenon among all of the passaged strains.

FIG 3.

Transmission electron microscopy reveals cell wall thickening of CRB.

Since thickening of the bacterial cell wall has been attributed as one of the primary underlying factors of vancomycin intermediate resistance, we analyzed vancomycin resistance among the passaged strains. MIC to vancomycin was unaltered among passaged strains versus their parental strains (Table 1), suggesting that cell wall thickening per se does not impart vancomycin resistance among these strains.

In summary, our results suggest that PBP4 played a critical role in mediating high-level β-lactam resistance among all of the passaged strains. Resistance mediated through pbp4 promoter and missense mutations contributed to the resistance phenotype. Further studies are required to determine its precise mechanism of action.