Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Biochemistry. Author manuscript; available in PMC 2011 Sep 21.
Published in final edited form as:
PMCID: PMC2939205

Molecular and structural analysis of mosaic variants of penicillin-binding protein 2 conferring decreased susceptibility to expanded-spectrum cephalosporins in Neisseria gonorrhoeae: role of epistatic mutations


Mutations in penicillin-binding protein 2 (PBP 2) encoded by mosaic penA alleles are critical for intermediate resistance to the expanded-spectrum cephalosporins ceftriaxone and cefixime in Neisseria gonorrhoeae. Three of the ~60 mutations present in mosaic alleles of penA, G545S, I312M, and V316T, have been reported to be responsible for increased resistance, especially to cefixime (Takahata et al. 2006. Antimicrob Agents Chemother 50:3638-45). However, we observed that the minimum inhibitory concentrations (MICs) of penicillin, ceftriaxone, and cefixime for a wild type strain (FA19) containing a penA gene with these three mutations increased only 1.5-, 1.5-, and 3.5-fold, respectively. In contrast, when these three mutations in a mosaic penA allele (penA35) were reverted back to wild type and the gene transformed into FA19, the MICs of the three antibiotics were reduced to near wild type levels. Thus, these three mutations display epistasis, in that their capacity to increase resistance to β-lactam antibiotics is dependent on the presence of other mutations in the mosaic alleles. We also identified an additional mutation, N512Y, that contributes to decreased susceptibility to expanded-spectrum cephalosporins. Finally, we investigated the effects of a mutation (A501V) currently found only in non-mosaic penA alleles on decreased susceptibility to ceftriaxone and cefixime, under the expectation that this mutation may arise in mosaic alleles. Transfer of the mosaic penA35 allele containing an A501V mutation into FA6140, a chromosomally mediated penicillin-resistant isolate, increased the MICs of ceftriaxone (0.4 μg/ml) and cefixime (1.2μg/ml) to levels above their respective breakpoints. The proposed structural mechanisms of these mutations are discussed in light of the recently published structure of PBP 2.

Neisseria gonorrhoeae is the etiologic agent of the sexually transmitted infection gonorrhea. In 2007, there were over 350,000 infections reported in the United States (1). Gonococcal infections are often asymptomatic, especially in females, and if left untreated can cause pelvic inflammatory disease and disseminated infections, as well as contribute to the spread of the disease. Because of the lack of a vaccine, antibiotics are the primary treatment for gonococcal infections.

Traditionally, gonorrhea was treated with penicillin G, but in 1986 this antibiotic was discontinued in the USA due to the onset of resistance. For the same reason, tetracyclines were also withdrawn. More recently, high-level resistance to fluoroquinolones led to the withdrawal of these antibiotics from the recommended list in the USA, leaving only expanded-spectrum cephalosporins such as ceftriaxone and cefixime as reliable treatments for gonococcal infections (2). Importantly, strains with decreased susceptibility to these antibiotics (also called cephI strains) are becoming widespread in the community (3, 4), making it crucial that we understand the mechanisms that lead to cephI resistance so that new antibiotics and new approaches be found for treating gonococcal infections before we are left with no effective antibiotics (4). This need is highlighted by recent reports of treatment failures using oral expanded-spectrum cephalosporins in Japan and Hong Kong (5, 6), as well as two confirmed treatment failures of pharyngeal gonococcal infections with ceftriaxone (7), although the latter two cases probably reflect the well-recognized difficulties in eradicating pharyngeal gonorrhea (4).

The most common mechanism of penicillin resistance in N. gonorrhoeae, termed chromosomally mediated resistance, involves at least 5 resistance determinants (8, 9). These determinants, which are mutated forms of normal genes/loci, can be transferred from a high-level resistant donor to a susceptible strain by homologous recombination and selection, with transfer occurring in a defined order (8, 10, 11). The first step is transfer of penA, which encodes altered forms of penicillin-binding protein 2 (PBP 2), the lethal target of penicillin G in N. gonorrhoeae (12). The second determinant transferred is mtrR, which leads to increased efflux pump expression and activity (13). The third determinant transferred is penB, which encodes altered forms of the major outer membrane porin, PorB1B (14, 15). The final steps, which result in high-level penicillin resistance, include acquisition of ponA, encoding an altered form of PBP 1, and a resistance determinant whose identity is unknown (8). The sum effect of these determinants is to increase the minimum inhibitory concentration (MIC) of penicillin by 400-fold (from 0.01 μg/ml to 4 μg/ml).

The major difference between penicillin-resistant and cephI strains is the presence of mosaic penA alleles that encode PBP 2 variants containing up to 60 amino acid changes compared to PBP 2 from wild type strains. We and others have proposed that the rapid emergence of cephI strains has occurred by horizontal transfer of mosaic penA genes, which were generated by recombination events between N. gonorrhoeae and commensal Neisseria species, into chromosomally mediated penicillin-resistant strains already harboring the necessary determinants to increase resistance to intermediate levels (16, 17).

Takahata et al. have reported that three mutations in mosaic variants of PBP 2, G545S, I312M, and V316T, are responsible for decreased susceptibility to cefixime (18). However, when we incorporated these mutations into a wild type penA gene and transformed the gene into FA19, resistance to penicillin, ceftriaxone, and cefixime increased only marginally (≤ 3-fold), putting the original conclusion in doubt (see Results). Therefore, we set out to identify important regions of PBP 2, and amino acids within them, that confer resistance to penicillin and expanded-spectrum cephalosporins. Our data indicate that more complex and subtle mechanisms are at play; that is, mosaic PBP 2 variants display epistasis, in which the three residues identified by Takahata et al. (18)are important in decreasing the rate of acylation, but only in the presence of other residues that have little to no apparent effect on their own. We also find that if an A501V mutation, which has been observed recently in non-mosaic penA alleles (1921), were to emerge in mosaic penA alleles, the MICs of ceftriaxone and ceftriaxone would increase to above the established break points of resistance.


Strains and Plasmids

The strains and plasmids used in this study are shown in Table 1. FA19 is a penicillin- and cephalosporin-susceptible strain that served as the recipient strain for most of the studies described herein (22). In the experiments examining the effect of the A501V mutation, FA6140, a penicillin-resistant, cephalosporin-susceptible strain (23) that contains all of the known resistance determinants (i.e., penA, mtrR, penB, and ponA)as well as the unknown determinant, also served as a recipient strain. pUC18us-penA and pUC18us-penA35 contain the wild type penA gene from FA19 and the mosaic penA gene from the cephI strain 35/02 (24), respectively, along with 300 bp of downstream sequence and an uptake sequence to facilitate homologous recombination. To construct the different chimeric PBPs, silent restriction sites were introduced into the coding sequences of pUC18us-penA and pUC18us-penA35 by the QuikChange method (Stratagene, Carlsbad, CA). These sites allowed us to swap out individual cassettes from penA35 with the corresponding cassettes from penA to create the “−mod” constructs (Fig. 1, Supplemental Fig. 1). Other point mutations were introduced into the appropriate constructs by overlap extension PCR (25). All constructs were verified by sequencing before proceeding with transformation experiments.

Fig. 1
Chimeric penA genes used in this study
Table 1
Strains and plasmids used in this study.


Transformation of the chimeric and mutant constructs into FA19 and FA6140 was accomplished as described previously (8). Transformants were selected on GCB plates containing various amounts of antibiotics just above their respective MICs. To verify correct recombination, transformants were passaged on GCB plates, and the next day, colonies were boiled in 30 μl water, spun briefly to pellet cell debris, and the supernatants were used as templates in PCR with the appropriate penA primers. PCR products were sequenced by the UNC sequencing facility or by Eton Bioscience Inc. (Research Triangle Park, NC).

MIC measurements

The MICs for penicillin, ceftriaxone, and cefixime were determined exactly as described previously (16). The antibiotics were tested in ~1.5-fold increments to increase the accuracy of the MIC determination. At least two (and often up to 4) colonies from each transformation were tested, each verified by PCR amplification and sequencing as described above. At least three independent MIC experiments were carried out, and the MICs reported represent the averages of all experiments. Error bars indicate the standard deviation of the determinations.

Purification of PBP 2 variants

PBP 2, PBP 235/02, and PBP 235/02-A501V were purified as described previously (26). Briefly, the genes encoding each variant were cloned into derivative of pMAL-C2 (New England Biolabs, Beverly, MA), pMAL-C2KV, which fuses His6-maltose-binding protein and an intervening tobacco etch virus (TEV) protease site to amino acid 44 of PBP 2, and the proteins were expressed in E. coli. The fusion proteins were purified on a Ni2+-NTA column (GE Healthcare, Piscataway, NJ), cleaved with His6-tagged tobacco etch virus (TEV) protease, and the digests were re-chromatographed over a Ni2+-NTA column. The purified proteins were eluted in buffer containing 15 mM imidazole, while uncleaved protein, His6-TEV andHis6-maltose-binding protein were eluted with 250 mM imidazole. The proteins were dialyzed to remove imidazole, concentrated to ~ 6 mg/ml, and frozen at −80°C.

k2/KS measurements of the rate of acylation by β-lactam antibiotics

The reaction of β-lactam antibiotics with PBP 2 is denoted by the following equation: E+SKsESk2ESk3E+P, where E•S is the non-covalent enzyme-antibiotic complex, E-S’ is the acyl-enzyme complex, and P the hydrolyzed antibiotic. k2/Ks constants, which are a direct measure of the ability of an antibiotic to inhibit a PBP (27), were calculated from first order rates of acylation of purified, soluble PBP 2 variants by [14C]penicillin G (Moravek, Brea, CA) as previously described (26, 28). Graphs of PBP 2-[14C]penicillin G complex formation versus time were obtained by incubating 27 μg of protein with 1.0 μM [14C]penicillin G, and aliquots of ~4 μg were removed at 15 sec intervals, precipitated with 5% trichloroacetic acid, filtered over Whatman GC-A filters, and the filters were submitted to scintillation counting. The concentration of [14C]penicillin G was increased to 25 and 50 μM for determination of k2/KS values with PBP 235/02 and PBP 235/02-A501V, respectively. The k2/Ks values of non-radioactive cephalosporin antibiotics were obtained in competition experiments with [14C]penicillin G using the following equation: (k2/Ks)ceph=(k2/Ks)penG([penG][ceph]0.5), where [penG] is the concentration of [14C]penicillin G used in the reaction and [ceph]0.5 is the concentration of cephalosporin antibiotic that inhibits the binding of [14C]penicillin G by 50% (27).


Analysis of mosaic PBP 2 mutations I312M, V316T, and G545S in cephI resistance following incorporation into a wild-type penA background

Takahata et al. reported that three amino acid mutations found in mosaic PBP 2 variants, I312M, V316T and G545S, are responsible for decreased susceptibility to cefixime in N. gonorrhoeae (18). To confirm these data, we introduced a single mutation (G545S), a double mutation (I312M/V316T), or all three mutations (G545S/I312M/V316T) into the wild type penA gene from FA19, and used these plasmids to transform FA19 to increased penicillin G or cefixime resistance. No transformants could be isolated with either the single or double mutants, presumably because they did not confer an increase in resistance. However, we were able to select for transformants harboring the triple mutant, but the MICs of penicillin, ceftriaxone, and cefixime for these transformants increased by only 1.5-, 1.5-, and 3.5-fold, respectively, compared to 12-, 20-, and 100-fold increases, respectively, with the full penA35 gene (16). Thus, it appears that together these three mutations confer only minimal increases in resistance to these antibiotics when incorporated into the penA gene from FA19.

Construction and analysis of PBP 2 chimeras

These data indicated that the mechanisms leading to emergence of cephI strains are more complex than originally reported, and thus we constructed chimeras between wild type and mosaic penA genes to identify the regions and amino acids that are important in decreasing the rate of acylation by the these antibiotics. As shown in Fig. 1, six regions in penA35 bounded by silent restriction sites, termed “modules” and denoted mod0 through mod5, were replaced by their corresponding modules from penA, and vice versa, and these constructs were used to transform FA19 to increased cefixime or penicillin resistance. Each module contained between 3 and 15 amino acid changes, with mod0 comprising the entire N-terminal domain of PBP 2 and mods 1–5 covering the C-terminal transpeptidase/β-lactam-binding domain (Figs. 1, S1). No transformants could be selected when individual penA35 modules were transferred into wild type penA in an attempt to show gain-of-function, indicating that no single region contained changes that were capable of conferring increased resistance to the antibiotics.

In contrast, we were able to select for transformants of FA19 with penA35 constructs in which individual mosaic penA35 modules were replaced with their wild type counterpart in penA (Fig. 2). In all cases, the penA35/penA chimeras resulted in lower levels of resistance relative to penA35, and except for a few instances, the MICs of the three antibiotics (compared to penA35) decreased similarly for each chimera. For clarity, the different chimeric constructs are defined in terms of the fold decrease in MIC compared to that conferred by penA35.

Fig. 2
MICs of penicillin, ceftriaxone, and cefixime for FA19 harboring 3X mutant and chimeric −mod penA genes

Replacing modules 0 or 2 of penA35 with wild type modules had little to no effect on the MICs of the three antibiotics, while replacing modules 1, 3 or 4 of penA35 with the corresponding wild type module decreased MICs from 2- to 3-fold compared to penA35 (Fig. 2). Replacing modules 1 or 4 of penA35 resulted in a greater decrease in the MICs than replacement of module 3, whereas replacement of modules 3 and 4 together decreased the MICs of the three antibiotics slightly more than replacement of the individual modules. Despite repeated attempts, we could not select transformants in which module 5 of penA35 was replaced by wild type, most likely because this chimera did not increase the MIC of FA19 above that conferred by wild type penA. These data indicate that mutations within module 5 are the most critical for decreased susceptibility, whereas mutations in modules1 and 4 are important but less so than those in module 5.

Role of G545S and I312M/V316T mutations in cephI resistance

Importantly, modules 1 and 5 contain the I312M/V316T and G545S mutations, respectively, initially described by Takahata et al. (18). To examine the role of the G545S mutation in resistance, we incorporated this mutation back into the −mod5 chimera (i.e. −mod5 + G545S). Unlike the parent −mod5 construct, which did not confer resistance over wild type levels, the resulting transformants were only slightly less resistant to penicillin and ceftriaxone compared to those containing the unmodified penA35 gene (Fig. 3), demonstrating the importance of this mutation. Resistance to cefixime, however, was nearly 3-fold lower, indicating that the G545S mutation is less important for cefixime resistance and that one or more additional mutations within module 5 appear to be required to reach the level of penA35. Overall, these data indicated that G545S was the most important mutation within mod5 for increasing resistance, but the degree to which the mutation is responsible for resistance depended on the antibiotic being examined.

Fig. 3
The G545S and I312M/V316T mutations are critical for conferring resistance to penicillin and expanded-spectrum cephalosporins only when present in the penA35 background

To examine the role of the I312M and V316T mutations in resistance, we replaced module 1 within the −mod5 + G545S chimera with the wild type sequence of penA (−m1,5 + G545S). The MICs of all three antibiotics for the resulting transformants decreased by ~2-fold compared to −mod5 + G545S (Fig. 3). Incorporation of the I312M/V316T double mutation back into this construct (i.e. −m1,5 + I312M/V316T/G545S) restored the MICs of ceftriaxone and cefixime in these transformants to the same level as −mod5 + G545S, whereas the MIC of penicillin did not change. These data suggest that all of the decrease in the MICs of ceftriaxone and cefixime caused by replacement of module 1 of penA35 with wild-type sequence was due to the I312M/V316T mutations, while the drop in penicillin MIC was due to other residue(s) within the module.

Reversion of G545S and I312M/V316T mutations in mosaic penA35

Reversion of the G545S mutation in penA35 back to wild type decreased the MICs of penicillin, ceftriaxone, and cefixime for the resulting transformants 2-, ~6- and 8-fold, respectively, compared to penA35, whereas reversion of the I312M/V316T mutations to wild type decreased MICs of penicillin, ceftriaxone, and cefixime by 2-, 4- and 4-fold, respectively. When all three mutations were reverted to wild type, the MICs of penicillin, ceftriaxone, and cefixime were decreased by 2.5-, 16-, and 25-fold, respectively, relative to penA35. Importantly, the MICs of ceftriaxone and cefixime for the triple reversion mutant were nearly identical to that of FA19. These data are consistent with the idea that the three mutations are important in resistance, but only in the context of some or all of the other 55 mutations found in mosaic penA alleles.

Role of module 4 amino acids in antibiotic resistance

Replacement of module 4 from penA35 with wild type sequence resulted in a ~3-fold decrease in the MIC of the three antibiotics when the chimera was transformed into FA19 (Fig. 2). This result was intriguing, since module 4 has only three amino acid changes, F504L, A510V, and N512Y, the first two of which are also found in the non-mosaic penA4 allele from FA6140 (26), a high-level, penicillin-resistant but cephalosporin-susceptible strain (8, 26). These three mutations are clustered on the β3-β4 loop of PBP 2, which is just C-terminal to the KTG active site motif. We therefore reverted the three changes individually in penA35 back to the wild type sequence, and determined the MICs of penicillin, ceftriaxone, and cefixime of the resulting transformants.

Surprisingly, the individual mutations had different effects on the MIC depending on the antibiotic being examined. For the expanded-spectrum cephalosporins, the most important reversion was Y512N, which was responsible for most if not all of the decrease in the MICs observed in strains containing penA35 −mod4. For penicillin, the Y512N mutation had a minor effect on resistance; instead, both the L504F and V510A reversions decreased resistance to the same level as the −mod4 construct (Fig. 4). These data highlight the importance of the β3-β4 loop for the reactivity of β-lactam antibiotics toward PBP 2 and demonstrate the differential effects of these mutations on the different antibiotics.

Fig. 4
MICs of penicillin, ceftriaxone, and cefixime for FA19 harboring penA35 containing reversion mutations from module 4

Effects of the A501V mutation in PBP 235/02

We also examined the effects of the A501V mutation in both mosaic and non-mosaic penA genes by incorporating this mutation into the penA4 and penA35 genes and transforming these genes into FA19 and FA6140 (Fig. 5). When transformed into FA19, the penA4-A501V and penA35-A501V mutant alleles decreased the MIC of penicillin by 20% and 50%, respectively, compared to levels conferred by the parental alleles (Fig. 5a). In contrast, penA4-A501V and penA35-A501V both increased ceftriaxone and cefixime MICs between 2-fold and 4-fold above those conferred by penA4 and penA35, respectively. In general, the effects of the A501V mutation were greater in the penA35 background than in the penA4 background.

Fig. 5
MICs of penicillin, ceftriaxone, and cefixime for FA19 and FA6140 harboring the penA4 (from FA6140) or penA35 (from 35/02) allele with or without an A501V mutation

Consistent with our results in FA19, when the penA35-A501V gene was transformed into FA6140, the MIC of penicillin for FA6140 penA35-A501V was nearly half of that for FA6140 penA35 (Fig. 5b), whereas the MICs of both ceftriaxone and cefixime increased by over 2-fold. Importantly, the MICs of ceftriaxone and cefixime for the resulting strains, 0.4 and 1.2 μg/ml, respectively, are well above the breakpoints for “resistance” (>0.25 μg/ml for both antibiotics). These data suggest that emergence of this mutation in mosaic penA genes, which to date has not occurred, could render both ceftriaxone and cefixime ineffective for treating gonococcal infections.

k2/Ks constants of wild type PBP 2, PBP 235/02, and PBP 235/02-A501V

To compliment our MIC data, we determined the acylation rate constants of penicillin, ceftriaxone, and cefixime for purified wild type PBP 2, PBP 235/02, and PBP 235/02-A501V. For wild type PBP 2, [14C]penicillin G had a k2/KS of 76,000 M−1s−1, whereas ceftriaxone and cefixime displayed k2/KS constants that were markedly higher (1.7 × 106 and 1.5 × 106 M−1s−1, respectively; Table 2), consistent with their low MICs for wild type gonococcal strains. The acylation rates of penicillin G, ceftriaxone, and cefixime for PBP 235/02 decreased 150-to 200-fold compared to wild type, underscoring the marked effect of the multiple substitutions in mosaic PBP 2 on the kinetics of β-lactam binding. The acylation rate of penicillin for PBP 235/02-A501V increased ~2.7-fold compared to PBP 235/02, while the k2/KS of cefixime decreased by 2.3-fold (Table 2). Surprisingly, the rate for ceftriaxone increased 1.8-fold, opposite of what we expected based on MIC experiments. Except for this latter result, these data are consistent with the changes observed in strains transformed with the mosaic penA allele.

Table 2
k2/KS acylation rates for wild type PBP 2, PBP 235/02, and PBP 235/02-A501V


Takahata et al. reported that three amino acid mutations found in mosaic PBP 2 variants, I312M, V316T and G545S, are responsible for decreased susceptibility to cefixime in N. gonorrhoeae (18). In that study, the authors determined the MICs of a range of expanded-spectrum cephalosporins for transformants of FA1090 harboring penAFA1090 alleles containing either the G545S mutation alone or G545S together with either the I312M or V316T mutation (a strain containing all three mutations could not be isolated). The MICs of ceftriaxone and cefixime increased 2-fold for the G545S transformant and 4-fold and 8-fold, respectively, for the G545S/I312M or G545S/V316T double transformants. From these data, the authors concluded that these three mutations were responsible for the decreased MICs conferredby mosaic penA alleles.

However, when we incorporated the three mutations into the wild type penA allele from FA19 and transformed it into FA19, the MICs of ceftriaxone and cefixime increased by only 1.5- and 3.5-fold, respectively, well below the levels obtained with the full mosaic penA35 gene (Fig. 2). An important difference between our work and the study by Takahata et al. is the choice of the parental penA allele. Whereas we used penA from FA19, which we consider to be a true “wild type” allele (22), Takahata et al. (18) used penA from FA1090, which contains an Asp345a insertion that decreases the k2/KS acylation rate constant of penicillin G by 6-fold (26). Presumably, the increased levels of resistance observed by Takahata et al. were due to the Asp-345a insertion adding to or amplifying the effects of the mutations. Since the Asp-345a insertion is not present in mosaic penA alleles, the penA gene from FA1090 is not an optimal background to examine the effects of these mutations.

While the three mutations (I312M, V316T, and G545S) do not increase resistance markedly when incorporated into a wild type penA gene, they have a striking effect on resistance when reverted back to the wild type in the mosaic penA35 background, indicating that these mutations require at least some of the other amino acid changes found in 35/02 to decrease susceptibility. This phenomenon, referred to as epistasis, has been elegantly described by Thornton and colleagues (29, 30), who identified several “permissive” mutations in steroid hormone receptors that, while having no functional importance on their own, stabilized function-switching mutations and facilitated the evolution of new steroid hormone-binding activity. Weinreich et al., studying the evolution of hydrolytic activity against cefotaxime in TEM β-lactamase, have described a similar epistatic phenomenon (31). These authors showed that only a few pathways to full resistance are permissible, since some mutations do not increase cefotaxime resistance in certain allelic backgrounds. In an example involving PBPs, Hedge and Spratt (32) defined the steps for acquisition of resistance of E. coli to a range of cephalosporins through serial mutagenesis of PBP 3 (the functional equivalent of PBP 2 in N. gonorrhoeae). The authors were able to obtain high-level resistance to cephalosporins in four steps. Importantly, two of the steps had little effect on resistance, but increased protein stability and thermostability of PBP 3 and allowed for subsequent isolation of resistance-conferring mutations (32).

To suggest how the three mutations described in this study might impact the rate of acylation, we examined our recent crystal structure of PBP 2from N. gonorrhoeae, solved in the apo form. G545S, which is the most important mutation for resistance in mosaic penA alleles, is present at the start of the α11 helix, one of the two helices that pack on top of the 5 anti-parallel β strands typical of penicillin-interacting proteins. The main chain amides of G545 and G546 are within hydrogen-bonding distance to the side chain hydroxyls of Thr498 and Thr500, respectively, located within the KTG(T) active-site motif (Fig. 6A). By analogy with other PBPs such as E. coli PBP 5, the main chain amide of Thr500 is predicted to stabilize the oxyanion hole for the transition state, and so one effect of the G545S mutation might be to lower acylation by compromising the geometry of the transition state/tetrahedral intermediate. Alternatively, because the hydroxyl side chains of the equivalent residues to Thr498 and Thr500 in S. pneumoniae PBP 2x (Ser548 and Thr550) interact with the β-lactam carboxylate in the covalent complex, alteration of these contacts may be another mechanism that lowers acylation. Ile312 and Val316 are located on the opposite side of the helix as Ser310 and Lys313 of the SxxK motif and pack into a hydrophobic pocket (Fig. 6B). Mutation to larger (I312M) or more hydrophilic (V316T) side chains might disrupt these interactions such that the position of the SxxK motif is altered, leading to decreases in acylation rates with β-lactam antibiotics. A similar argument has been put forth for the M339F mutation in S. pneumoniae PBP 2X (33).

Fig. 6
Structure of PBP 2 showing the location of important mutations and the active site sequence motifs

In addition to the three mutations discussed above, we also showed that reversion of N512Y in penA35 decreases resistance to ceftriaxone and cefixime by 2-fold without affecting penicillin resistance. This result is consistent with its emergence in mosaic penA alleles and its relative absence in non-mosaic penA alleles found in penicillin-resistant strains(18, 19, 21). Asn512 is relatively distant from the active site (Fig. 6B) and the exact impact of the N512Y mutation on acylation is difficult to predict, but it is located on the β3-β4 loop that contains mutations that are known to be important for resistance to penicillin in non-mosaic forms of PBP 2 (26). One possibility is that such a mutation perturbs the architecture of the KTG motif of β3.

Our work has also highlighted the impending danger of the A501V mutation arising in mosaic penA alleles, because if it happens this would increase the MICs of ceftriaxone and cefixime above their breakpoints for resistance. The alteration of A501 in PBP 2 appears to be a gonococcal-specific alteration, as it has not yet been observed in commensal Neisseria species, which suggests that it arose through mutation in response to selective pressure with expanded-spectrum cephalosporins instead of by transformation. Indeed, Takahata et al. (18)reported the isolation of a spontaneous A501V mutation during transformation experiments. Thus, it may be a matter of when and not if mutations in Ala501 arise in mosaic penA genes. Ala501 resides on the β3-β4 loop where mutation of the methyl side chain to the bulkier side chain of valine could clash with the R1 substituent of the cephalosporin (Fig. 6B). Consistent with this hypothesis, an A501T mutation, which introduces a similar branched side chain, affects resistance in a similar manner as A501V (MU, unpublished observations).

The k2/KS acylation rate constants of the three antibiotics for wild type PBP 2, PBP 235/02 and PBP 235/02-A501V were determined to assess directly the effects of the PBP mutations on reactivity with β-lactam antibiotics. The k2/KS rate constants for acylation are largely, but not entirely, consistent with MIC values. For example, acylation rates are dramatically impaired in PBP 235/02 compared to wild type, with 150-to 200-fold decreases in acylation rate for all three antibiotics. However, since the MICs increase 10-, 20- and 100-fold for penicillin, ceftriaxone, and cefixime, respectively, when penA35 is transformed into FA19, it is clear that other intrinsic factors, e.g. diffusion through wild type porin channels, likely play an important role in defining the MIC. One anomaly in the acylation data is the ~2-fold increase in k2/KS of ceftriaxone for the PBP 235/02-A501V variant, which is inconsistent with the ~2-fold increase in the MIC for transformants harboring this mutant.

In conclusion, the emergence of decreased susceptibility to expanded-spectrum cephalosporins through remodeling of the active site of PBP 2 is more complicated than originally envisioned, with important resistance-conferring mutations showing epistasis. The acquisition of further mutations, such as A501V, in mosaic penA alleles is likely to increase resistance to levels that render expanded-spectrum cephalosporins ineffective in treating gonococcal infections. The need to identify new antibiotics against this organism is therefore of prime importance in the treatment of STIs.

Supplementary Material

Fig. 1S

Abbreviations and Textual Footnotes

penicillin-binding protein
minimum inhibitory concentration
cephalosporin-intermediate resistance
nitriloacetic acid
tobacco etch virus
sexually transmitted infection


This work was supported by grants AI36901 (to R.A.N.) and GM66861 (to C.D.) from the National Institutes of Health and a grant from the Research Committee of Örebro County, the Örebro University Hospital Research Foundation, Örebro, Sweden (to M.U.).

Supporting Information Available

Alignment of wild type PBP 2 with PBP 235/02 showing the location of mutations in each module. This material is available free of charge via the Internet at http://pubs.acs.org.


1. CDC; Services, U. S. D. o. H. a. H. Sexually Transmitted Disease Surveillance 2007 Supplement, Gonococcal Isolate Surveillance Project (GISP) Annual Report 2007. Atlanta, Georgia: 2009.
2. CDC. Update to CDC’s sexually transmitted diseases treatment guidelines, 2006: fluoroquinolones no longer recommended for treatment of gonococcal infections. MMWR Morb Mortal Wkly Rep. 2007;56:332–336. [PubMed]
3. Ameyama S, Onodera S, Takahata M, Minami S, Maki N, Endo K, Goto H, Suzuki H, Oishi Y. Mosaic-like structure of penicillin-binding protein 2 Gene (penA) in clinical isolates of Neisseria gonorrhoeae with reduced susceptibility to cefixime. Antimicrob Agents Chemother. 2002;46:3744–3749. [PMC free article] [PubMed]
4. Tapsall JW, Ndowa F, Lewis DA, Unemo M. Meeting the public health challenge of multidrug- and extensively drug-resistant Neisseria gonorrhoeae. Expert Rev Anti Infect Ther. 2009;7:821–834. [PubMed]
5. Lo JY, Ho KM, Leung AO, Tiu FS, Tsang GK, Lo AC, Tapsall JW. Ceftibuten resistance and treatment failure of Neisseria gonorrhoeae infection. Antimicrob Agents Chemother. 2008;52:3564–3567. [PMC free article] [PubMed]
6. Yokoi S, Deguchi T, Ozawa T, Yasuda M, Ito S, Kubota Y, Tamaki M, Maeda S. Threat to cefixime treatment for gonorrhea. Emerg Infect Dis. 2007;13:1275–1277. [PMC free article] [PubMed]
7. Tapsall J, Read P, Carmody C, Bourne C, Ray S, Limnios A, Sloots T, Whiley D. Two cases of failed ceftriaxone treatment in pharyngeal gonorrhoea verified by molecular microbiological methods. J Med Microbiol. 2009;58:683–687. [PubMed]
8. Ropp PA, Hu M, Olesky M, Nicholas RA. Mutations in ponA, the gene encoding penicillin-binding protein 1, and a novel locus, penC, are required for high-level chromosomally mediated penicillin resistance in Neisseria gonorrhoeae. Antimicrob Agents Chemother. 2002;46:769–777. [PMC free article] [PubMed]
9. Shafer WM, Folster JP, Nicholas RA. Molecular Mechanisms of Antibiotic Resistance Expressed by the Pathogenic Neisseria. In: Genco CA, Wetzler L, editors. Neisseria: Molecular Mechanisms of Pathogenesis. Caister Academic Press; Norfolk, UK: 2010. pp. 245–270.
10. Veal WL, Nicholas RA, Shafer WM. Overexpression of the MtrC-MtrD-MtrE efflux pump due to an mtrR mutation is required for chromosomally mediated penicillin resistance in Neisseria gonorrhoeae. J Bacteriol. 2002;184:5619–5624. [PMC free article] [PubMed]
11. Faruki H, Sparling PF. Genetics of resistance in a non-β-lactamase-producing gonococcus with relatively high-level penicillin resistance. Antimicrob Agents Chemother. 1986;30:856–860. [PMC free article] [PubMed]
12. Spratt BG. Hybrid penicillin-binding proteins in penicillin-resistant strains of Neisseria gonorrhoeae. Nature. 1988;332:173–176. [PubMed]
13. Hagman KE, Shafer WM. Transcriptional control of the mtr efflux system of Neisseria gonorrhoeae. J Bacteriol. 1995;177:4162–4165. [PMC free article] [PubMed]
14. Gill MJ, Simjee S, Al-Hattawi K, Robertson BD, Easmon CS, Ison CA. Gonococcal resistance to beta-lactams and tetracycline involves mutation in loop 3 of the porin encoded at the penB locus. Antimicrob Agents Chemother. 1998;42:2799–2803. [PMC free article] [PubMed]
15. Olesky M, Hobbs M, Nicholas RA. Identification and analysis of amino acid mutations in porin IB that mediate intermediate-level resistance to penicillin and tetracycline in Neisseria gonorrhoeae. Antimicrob Agents Chemother. 2002;46:2811–2820. [PMC free article] [PubMed]
16. Zhao S, Duncan M, Tomberg J, Davies C, Unemo M, Nicholas RA. Genetics of chromosomally mediated intermediate resistance to ceftriaxone and cefixime in Neisseria gonorrhoeae. Antimicrob Agents Chemother. 2009;53:3744–3751. [PMC free article] [PubMed]
17. Ohnishi M, Watanabe Y, Ono E, Takahashi C, Oya H, Kuroki T, Shimuta K, Okazaki N, Nakayama S, Watanabe H. Spread of a chromosomal cefixime-resistant penA gene among different Neisseria gonorrhoeae lineages. Antimicrob Agents Chemother. 2010;54:1060–1067. [PMC free article] [PubMed]
18. Takahata S, Senju N, Osaki Y, Yoshida T, Ida T. Amino acid substitutions in mosaic penicillin-binding protein 2 associated with reduced susceptibility to cefixime in clinical isolates of Neisseria gonorrhoeae. Antimicrob Agents Chemother. 2006;50:3638–3645. [PMC free article] [PubMed]
19. Osaka K, Takakura T, Narukawa K, Takahata M, Endo K, Kiyota H, Onodera S. Analysis of amino acid sequences of penicillin-binding protein 2 in clinical isolates of Neisseria gonorrhoeae with reduced susceptibility to cefixime and ceftriaxone. J Infect Chemother. 2008;14:195–203. [PubMed]
20. Whiley DM, Limnios EA, Ray S, Sloots TP, Tapsall JW. Diversity of penA alterations and subtypes in Neisseria gonorrhoeae strains from Sydney, Australia, that are less susceptible to ceftriaxone. Antimicrob Agents Chemother. 2007;51:3111–3116. [PMC free article] [PubMed]
21. Lee SG, Lee H, Jeong SH, Yong D, Chung GT, Lee YS, Chong Y, Lee K. Various penA mutations together with mtrR, porB and ponA mutations in Neisseria gonorrhoeae isolates with reduced susceptibility to cefixime or ceftriaxone. J Antimicrob Chemother. 2010;65:669–675. [PMC free article] [PubMed]
22. Maness MJ, Sparling PF. Multiple antibiotic resistance due to a single mutation in Neisseria gonorrhoeae. J Infect Dis. 1973;128:321–330. [PubMed]
23. Danielsson D, Faruki H, Dyer D, Sparling PF. Recombination near the antibiotic resistance locus penB results in antigenic variation of gonococcal outer membrane protein I. Infect Immun. 1986;52:529–533. [PMC free article] [PubMed]
24. Lindberg R, Fredlund H, Nicholas RA, Unemo M. Neisseria gonorrhoeae isolates with reduced susceptibility to cefixime and ceftriaxone: association with genetic polymorphisms in penA, mtrR, porB1b, and ponA. Antimicrob Agents Chemother. 2007;51:2117–2122. [PMC free article] [PubMed]
25. Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR. Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene. 1989;77:51–59. [PubMed]
26. Powell AJ, Tomberg J, Deacon AM, Nicholas RA, Davies C. Crystal structures of penicillin-binding protein 2 from penicillin-susceptible and -resistant strains of Neisseria gonorrhoeae reveal an unexpectedly subtle mechanism for antibiotic resistance. J Biol Chem. 2009;284:1202–1212. [PMC free article] [PubMed]
27. Frere JM, Nguyen-Disteche M, Coyette J, Joris B. The chemistry of β-lactams. Chapman & Hall; Glasgow: 1992. Mode of action: Interaction with the penicillin binding proteins; pp. 148–196.
28. Stefanova ME, Tomberg J, Olesky M, Holtje JV, Gutheil WG, Nicholas RA. Neisseria gonorrhoeae penicillin-binding protein 3 exhibits exceptionally high carboxypeptidase and beta-lactam binding activities. Biochemistry. 2003;42:14614–14625. [PubMed]
29. Bridgham JT, Ortlund EA, Thornton JW. An epistatic ratchet constrains the direction of glucocorticoid receptor evolution. Nature. 2009;461:515–519. [PubMed]
30. Ortlund EA, Bridgham JT, Redinbo MR, Thornton JW. Crystal structure of an ancient protein: evolution by conformational epistasis. Science. 2007;317:1544–1548. [PMC free article] [PubMed]
31. Weinreich DM, Delaney NF, Depristo MA, Hartl DL. Darwinian evolution can follow only very few mutational paths to fitter proteins. Science. 2006;312:111–114. [PubMed]
32. Hedge PJ, Spratt BG. Resistance to β-lactam antibiotics by re-modelling the active site of an E. coli penicillin-binding protein. Nature. 1985;318:478–480. [PubMed]
33. Chesnel L, Pernot L, Lemaire D, Champelovier D, Croize J, Dideberg O, Vernet T, Zapun A. The structural modifications induced by the M339F substitution in PBP2x from Streptococcus pneumoniae further decreases the susceptibility to beta-lactams of resistant strains. J Biol Chem. 2003;278:44448–44456. [PubMed]
34. Gordon E, Mouz N, Duee E, Dideberg O. The crystal structure of the penicillin-binding protein 2x from Streptococcus pneumoniae and its acyl-enzyme form: implication in drug resistance. J Mol Biol. 2000;299:477–485. [PubMed]
35. Collaborative Computation Project, N. The CCP4 suite: programs for protein crystallography. Acta Crystallographica. 1994;D50:760–763. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...