PBP2a in β-Lactam–Resistant Laboratory Mutants and Clinical Isolates: Disruption Versus Reduced Penicillin Affinity

Bacterial strains, plasmids, and growth conditions

Streptococcus pneumoniae R6 is a nonencapsulated derivative of the Rockefeller University strain R36A. ⁴³ The cefotaxime-resistant mutants, all derived from the R6 strain, are independent spontaneous mutants described previously ⁴² ; R6 derivatives used in the present study containing various combinations of PBP genes are listed in Table 1. The streptomycin-resistant R6 derivative containing the rpsL41 allele was needed to replace wild-type pbp2a by other pbp2a alleles using the Janus gene replacement procedure. ⁴⁴ R6ciaR::aad9 has been described. ⁴⁵

Clinical isolates of S. pneumoniae have been described: strains 8249 and 29044²²; strains from Poland ⁴⁶ ; and Romanian isolates were obtained from Marina Pana, Bucharest. ⁴⁷ Pneumococci were cultivated without aeration in C-medium ⁴⁸ supplemented with 0.2% yeast extract (Difco) at 30°C or on D-agar plates ⁴⁹ supplemented with 3% defibrinated sheep blood. Growth in liquid culture was monitored by nephelometry and is given in nephelometry units (N). Escherichia coli DH5a [Φ80dlacZΔM15 Δ(lacZYA-argF) recA1 endA1 hsdR17 supE44 thi-1 gyrA96 phoA relA1] was used for propagation of plasmids. Escherichia coli BL21 [B: F⁻ dcm ompT hsdS (rB⁻ mB⁻) gal] was used for propagation of the pGEX-6P-1 derivative expressing GST-PBP2a*. E. coli strains were grown in LB medium or on LB agar. ⁵⁰ Growth of E. coli was followed by measuring of the optical density at 600 nm using a spectrophotometer.

Transformation experiments

Transformation of S. pneumoniae was performed according to published procedures ⁵¹ ; for transformation of C303 and C306 as recipient, cells were grown in the presence of 20 ng/ml of competence stimulating peptide to overcome the competence deficiency due to ciaH mutations in these strains. Cells were incubated for 30 minutes with DNA at 30°C, followed by a 2-hour phenotypic expression period at 30°C. Antibiotic resistance genes used for chromosomal integrations in S. pneumoniae were selected with 200 μg/ml kanamycin (Kan), 200 μg/ml streptomycin (Str), or 20 μg/ml spectinomycin (Spc). In case of altered pbp2a, at least five transformants from each experiment were analyzed for PBP profiles and cefotaxime MICs, and at least two transformants were selected for DNA sequencing of pbp2a; correct DNA sequences of pbp2x and ciaH were also confirmed by sequencing of PCR products using oligonucleotides listed in Table 2. E. coli was transformed using chemically competent cells, ⁵⁰ and transformants were selected in the presence of 100 μg/ml ampicillin, 15 μg/ml tetracycline for pGEX0-, or 500 μg/ml erythromycin for pJDC9-derivatives.

DNA manipulations and oligonucleotides

All DNA manipulations were performed using standard methods. ⁵⁰ Chromosomal DNA was isolated from S. pneumoniae as described ⁵² ; in case for S. mitis strains, lysozyme (25 mg/ml) plus cellosyl (0.5 mg/ml) was added during the lysis procedure as described. ⁵³ Plasmid purification from E. coli was done according to the manufacturer's protocol using the QIAprep Spin Miniprep Kit (Qiagen).

Restriction enzymes and T4 DNA ligase were purchased from New England Biolabs and Roche Diagnostics and used as described by the manufacturers. PCR products and DNA recovered after restriction endonuclease digestions were purified using a JetQuick^® PCR Purification Spin Kit (Genomed). DNA fragments from agarose gels were purified using NucleoSpin^® Extract II (Macherey-Nagel). PCRs from DNA were performed using GoldStar Red Taq polymerase (Eurogentec) or high fidelity iProof™ High Fidelity DNA polymerase (Bio-Rad), and in case of S. pneumoniae cells with DreamTaq DNA polymerase (Thermo Fisher Scientific) according to the manufacturer's instructions. The oligonucleotides used in this study are listed in Table 2 and were obtained from MWG-Biotech and Eurofins MWG Operon. The inserts of recombinant plasmids were sequenced to confirm the insertion and subsequently used to transform pneumococcal strains. Plasmid integration into pneumococcal chromosome was verified by PCR amplification and sequencing.

Amplification and sequencing of PBP2a genes

PBP2a genes from different S. pneumoniae strains and mutants were amplified using the oligonucleotide pairs 2a1U plus 2adown and pbp2afarup plus pbp2afardown2. PBP2a genes from clinical strains were amplified with 2a2U or 2a1U plus 2a2268R. Streptococcus mitis B6 pbp2a was amplified with 2a1U_B6 and B6_2aR. Amplification of chromosomal DNA with PCR was carried out in a biomed Thermocycler for 30 cycles of denaturation at 96°C for 30 seconds, annealing at 52°C for a 1 minute and extension at 72°C for 1 minute, followed by a 3-minute delay period at 72°C after the last cycle. The reaction mixture (100 μl) contained 10 pmol of each oligonucleotide primer and 2.5 U of Taq polymerase (PerkinElmer, Norwalk, CT). PCR amplified DNA was directly sequenced using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (PerkinElmer). Automatic cycle sequencing was conducted with an ABI 377 sequencer.

Construction of mutants and plasmids

Insertion-duplication mutagenesis of pbp2a was performed using an internal 616-bp fragment of pbp2a (nucleotide 984–1,600) cloned into the vector pJDC9. ⁵⁴ First, a pbp2a fragment was PCR amplified from Streptococcus pneumoniae R6 with the primer pair FU4 + FL4 and cloned directly into the pGEM^®-T Easy Vector (Promega) suitable for cloning of PCR products. After restriction with EcoR1, the insert was cloned into pJDC9 resulting in the plasmid pIM2a. Plasmids were transformed into S. pneumoniae and transformants were selected with 1 μg/ml erythromycin. Insertion and orientation of the pJDC9 constructs were examined using the primers 2a_5U located within the pbp2a fragment, in combination with primers 1201S and 1211S upstream and downstream of the integrated region. The successful disruption of pbp2a in the S. pneumoniae transformants was verified by PBP profile analysis and DNA sequencing.

Different pbp2a variants were introduced into Streptococcus pneumoniae R6 and derivatives by the Janus gene replacement procedure. ⁴⁴ Three PCR fragments were amplified containing the kan-rpsL region (primers janFOR and janREV), ⁴⁴ the 5′-region of pbp2a (primers 2a1U and 2aJANr_link), and the 3′-pbp2a region (primers 2a2268r and 2aJANf_link). Fragments were purified and combined by overlapping PCR. ⁵⁵ PCR products were transformed into R6rpsL41 and transformants selected with 200 μg/ml kanamycin. The resulting strains expressing a Str^S and Kan^R phenotype (rpsL41 pbp2a:: kanR rpsL) were used as recipients for the various pbp2a alleles by selecting streptomycin-resistant transformants obtained with PCR amplified chromosomal DNA. Presence of the correct pbp2a allele was verified by DNA sequencing. The presence of rpsL41 does not influence CiaRH-mediated regulation. ⁵⁶

To construct a strain producing PBP2a with specific mutations, site-directed mutagenesis was performed by an overlapping PCR approach. A PCR fragment carrying the mutation A₁₂₃₁>G was obtained with the primers 2a1U, 2aT>Ar, 2aT>Af, and 2a2268r to obtain R6pbp2aA⁴¹¹. DNA of this strain was then used to obtain a PCR fragment with primers 2a1U, 2aQ>Lr, 2aQ>Lf, and 2a2268r to obtain R6 pbp2aA⁴¹¹pbp2aLeu⁴³¹. PCR fragments were cloned into the pGEM-T Easy Vector before transfer into Streptococcus pneumoniae R6 through the Janus replacement procedure as previously described. R6 containing the PBP2x mutation Ala⁵⁵⁰ was obtained by transformation with pbp2x^C403 and selection with cefotaxime. The correct mutations were verified by DNA sequencing.

Replacement of pbp2a with the kanamycin cassette aphIII

PCR fragments were amplified and primer pairs shown in brackets representing sequences upstream (fragment 1; pbp2a up f CS and PM 364) and downstream (fragment 2; pbp2a down r1 and PM 363) of pbp2a. The kanamycin resistance cassette aphIII was obtained by cleavage of plasmid pUC19 aphIII with EcoRI and XbaI. The three PCR fragments were ligated after restriction of fragment 1 with EcoRI and fragment 2 with XabI. Transformation into the recipient S. pneumoniae strains and selection with kanamycin resulted in replacement of pbp2a with aphIII. Correct integration of the aphIII was verified by PCR of S. pneumoniae cells and the primer pair pbp2a up ext CS/pbp2a down r CS.

Detection of PBPs

Preparation of samples, PBP labeling with BocillinFL, and separation of proteins by sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis (PAGE) were carried out as described previously. ⁵⁷ In case of S. mitis, cells were lysed by adding 0.5 mg/ml cellosyl and 0.8 mg/ml lysozyme to the lysis buffer. ⁵³ BocillinFL-PBP complexes were visualized by fluorography detection with a FluorImager 595 fluorescence scanner (Molecular Dynamics) at 488 nm.

After SDS-PAGE, the proteins were transferred to a PVDF membrane (Roche Diagnostics) and probed with affinity purified polyclonal anti-PBP2a antibody followed by incubation with alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin G (Sigma-Aldrich) and staining with 4-nitrobluetetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate (Roche Diagnostics). The presence of equal amounts of protein in the samples was verified on separate SDS-polyacrylamide gels by Coomassie blue staining.

Anti-PBP2a antibodies

A soluble PBP2a GST-tagged (glutathione S-transferase) derivative was constructed and coupled to GSH-Sepharose. In brief, a PCR product of pbp2a between T^234–T²¹⁹⁵ (residues 78–731) was cloned into the vector pGEX-6P-1-tet ⁵⁸ after restriction with BamHI and XhoI and ligation. The resulting pbp2a* was overproduced in Escherichia coli BL21 and purified as described by chromatography on GSH-Sepharose. ⁵⁸ Rabbit anti-PBP2a antisera were obtained from Seqlab GmbH and affinity-purified using PBP2a* coupled to 1 ml HiTrap NHS-activated Sepharose (GE Healthcare). After washing with 20 ml sodium phosphate buffer (10 mM, pH 7.2), PBP2a-specific antibodies were eluted with glycine (100 mM, pH 2.2). Fractions (3 ml) were collected directly in 250 μl Tris buffer (1.5 M, pH 8.8) and used at a dilution of 1:2,000.

PBP2a in high level cefotaxime–resistant S. pneumoniae

Cefotaxime-resistant mutants isolated from the laboratory strain Streptococcus pneumoniae R6 have been described in which PBP2a was not detectable after labeling with β-lactams and protein separation on SDS-polyacrylamide gels, neither with radioactive β-lactam nor with anti-β-lactam antibiotics where high penicillin concentrations for labeling of PBPs can be applied to detect PBP-β-lactam complexes on Western blots.^42,59 This phenotype was observed in three independently isolated mutants with increasing concentrations of cefotaxime: C403and C406 obtained after four selection steps and C604 obtained after six selection steps. All three mutants contained mutations in PBP2x and CiaH and C604 one mutation in PBP3 as well (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Cefotaxime-resistant lineages of laboratory mutants. The cefotaxime-resistant mutant lineages and the occurrence of PBP2a mutations during the selection procedure are shown. Mutations identified in PBP2x, ⁵² PBP3, ³⁵ and CiaH ⁷³ are shown. The first digit of the mutant indicates the selection step, and the last digit specifies the mutant lineage.

To identify PBP2a mutations, pbp2a was amplified by PCR from chromosomal DNA of the three mutants for DNA sequencing. Surprisingly, the PCR fragment obtained with C406 was larger than that compared to the PCR products of the other two mutants and the parental strain R6, indicating some type of insertion in pbp2a^C406 (Fig. 2A).

OPEN IN VIEWER

FIG. 2.

PBP2a in cefotaxime-resistant laboratory mutants. (A). Agarose gel electrophoresis of PCR products obtained with oligonucleotide primers FU5 and FL3 from three mutants C403, C406, and C604 with apparent PBP2a mutations; the mutant C303 containing only PBP2x and CiaH mutations and the parental strain Streptococcus pneumoniae R6 are included for comparison. The size in kb of the PCR products is shown on the right and of the ladder marker on the left. (B). The domain structure of the 731 amino acid PBP2a consists of a 51 amino acid cytoplasmic tail, a hydrophobic membrane anchor region between amino acid 52–77 (gray) is followed by the transglycosylase (GT) domain; amino acid 264–301 represents the predicted junction between GT and transpeptidase (TP) domain (gray); and the predicted TP domain ⁶⁴ is shown in black. The conserved active site motifs of the TP domain are indicated. Mutations observed in the three mutants C403, C604, and C406 are shown below. The codon representing stops in pbp2a resulting in PBP2a fragments is indicated; black boxes represent distinct peptides due to tandem duplication of the regions as indicated below the bars. In C406, the stop codon (black arrow) is located within the duplicated region (dashed).

DNA sequencing revealed that it contained a 119 nucleotide tandem duplication within pbp2a (nucleotide 1,569–1,687), resulting in a change of amino acid composition after residue 564 and a truncated predicted peptide of 566 amino acid compared to the 731 amino acid PBP2a of strain R6 (Fig. 2B). Similarly, pbp2a of C403 contained a tandem duplication of eight nucleotides (84–91) resulting in an altered peptide sequence after amino acid 30 and a stop codon after 42 amino acid, whereas C604 contained a single mutation in codon 215 TGG encoding Trp²¹⁵ which was changed to TAG, resulting in a stop codon after amino acid residue 214 (Fig. 2B).

Thus, none of the three mutants apparently produced a functional PBP2a. Tandem duplications have been observed in a macrolide-resistant S. pneumoniae, which contained an 18-bp tandem repeat in the gene encoding the ribosomal protein L22. ⁶⁰ Interestingly, such duplications have been recognized in phase variants of serotype 3, 8, and 37 S. pneumoniae, and the frequent isolation of capsular revertants indicated a precise excision mechanism. ⁶¹ No instability of the PBP2a mutants was observed.

Cefotaxime resistance mediated by the lack of PBP2a

Initial transformation experiments with chromosomal DNA of the mutants C403, C604, and C406, using C303 as recipient, resulted in a large number of transformants after cefotaxime selection with MICs similar to C403. Selected transformants all showed a lack of the PBP2a band after labeling with BocillinFL and SDS-PAGE on fluorograms (not shown). However, in this and subsequent experiments additional mutations in ciaH were detected in several transformants, potentially contributing to cefotaxime resistance, clearly indicating that cefotaxime selection is not the method of choice to determine the impact of pbp2a on the resistance phenotype. In particular strongly activating ciaH alleles, including ciaH^C103, are genetically unstable, ⁵⁶ and therefore, ciaH was sequenced in all constructs to verify the presence of the ciaH wild-type allele.

To avoid cefotaxime selection for testing whether a disrupted pbp2a indeed confers cefotaxime resistance, first pbp2a^C403 was introduced into strain C303 through the Janus cassette, a bicistronic cassette conferring a Kan^R Str^S phenotype, which enables the replacement of genes through negative selection in a two-step procedure. ⁴⁴ Insertion of the Janus cassette in the first step results in the disruption of the gene, and thus, this technique can only be applied for nonessential genes, such as pbp2a,^12,13 or nonessential regions adjacent to essential genes. Streptomycin-resistant recipients are required for this procedure mediated by the rpsL41 allele, which does not influence cefotaxime resistance. ⁵⁶ We chose pbp2a^C403 because of the clear difference in MIC values for cefotaxime between C303, which contains only mutations in PBP2x and CiaH, and C403, which shows affinity changes in PBP2a as well (0.3 μg/ml versus 1.1–1.2 μg/ml; Table 1).

Transformants were readily obtained, and sequencing proved the existence of the truncated pbp2a^C403; consequently a PBP2a band could not be detected in their PBP profiles on fluorograms or by Western blotting using anti-PBP2a antibodies (Fig. 3, left side). Surprisingly, the MIC values of transformants obtained did not reach the MIC of the donor strain C403 by far (Table 1), similar to transformants obtained by insertion-duplication using the vector pJDC9 and which did not contain additional mutations in ciaH (not shown). The effect of a pbp2a^null genotype was investigated in C303, as well as in C306, using allelic replacement of pbp2a with the gene aphIII conferring kanamycin resistance. Again, these transformants did not contain PBP2a (e.g., see Fig. 3, right side) and showed only a slight increase in resistance which did not reach the level of the next step mutant in which this mutation had been originally identified (Table 1).

OPEN IN VIEWER

FIG. 3.

PBP2a in cefotaxime-resistant laboratory mutants and transformants of the R6 strain. Cell lysates were labeled with BocillinFL (top), and PBPs were detected after SDS-PAGE and fluorography with the FluorImager. The PBP profile of the parental strain Streptococcus pneumoniae R6 is included for comparison. The positions of the PBPs are marked on the left side. Bottom: Western blot developed with anti-PBP2a antibodies. The arrow shows the position of PBP2a present in strains R6 and C306. PBP, penicillin-binding protein; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis.

No attempts were made to use C504pbp3^C604 as recipient in such experiments. It should be noted that relative differences in cefotaxime MICs between the various constructs (although as low as 0.05 μg/ml) were consistent in all experiments, including one final test that included all mutant constructs described in this study. However, differences below 0.05 μg/ml cefotaxime cannot be excluded.

The results strongly suggest that in addition to pbp2a mutations, other still unknown genes are affected in C403 and C406. Nevertheless, a slight effect on cefotaxime resistance in a genetic background with a mutated pbp2x was mediated by the absence of an intact PBP2a gene.

PBP2a in clinical isolates

PBP2a genes were examined in 19 PRSP of the Kaiserslautern clinical isolate collection. Isolates with very high β-lactam resistance levels were chosen, including isolates from Poland and Romania, countries where high rates of PRSP have been recorded and where affinity changes of PBP2a have been observed as described in strains 8249 (South Africa) and 29044 (Czech Republic). ²² In addition, another two pbp2a described in previous publications were used. These included pbp2a from a high level cefotaxime–resistant clinical strain from Japan SP00084. ³⁸ Smith et al. ³⁹ described three clonally related isolates with identical mosaic pbp2x, pbp2b, and pbp1a, whereas two (isolate 1 and isolate 2) contained the same altered pbp2a and expressed higher MIC values compared to isolate 3. Isolate 3 could be transformed to higher β-lactam resistance levels with pbp2a of isolate 1.

Moreover, pbp2a of the high-level β-lactam–resistant Streptococcus mitis strains B6 ⁶² and DD28 ⁶³ was used for comparison, as were pbp2a of the penicillin-sensitive Streptococcus pneumoniae R6 and Streptococcus mitis strains NCTC12261, NCTC10712, and M3 as references.

The full-length pbp2a sequence encoding the 731 amino acid PBP2a was retrieved from genome data if available and was obtained for pbp2a of strain Streptococcus pneumoniae 29044. Sequences covering at least codons 258–651 were obtained in all other cases, which include most of the TP domain located between codons 301–678 as defined by Di Guilmi et al. ⁶⁴ It should be noted that Streptococcus mitis B6 has an extension of two amino acids at the C-terminus due to the integration of an ISSmi1 element at the very end of pbp2a. ⁶⁵

Properties of the strains and accession numbers of the genomes, respectively, PBP2a genes are listed in Table 3.

PBP2a gene sequences covering nucleotide positions 903–1,973 were used for a phylogenetic analysis with the MEGA6 program. Since mosaic genes were included in this analysis, the tree does not reveal the phylogenetic relationship, but rather was used to provide a visual aid to assess the sequence similarities. As shown in Figure 4, 11 sequences clustered together with pbp2a of the penicillin-sensitive reference strain Streptococcus pneumoniae R6 (R6-cluster), whereas all others were more dispersed. All pbp2a outside the R6 cluster diverged from pbp2a^R6 by between 3.3% (SP00084) and 6.4% (Hu11) over the entire sequence.

OPEN IN VIEWER

FIG. 4.

Phylogenetic tree of PBP2a genes. The tree was generated from pbp2a between codons 301 and 658 covering the TP domain. Bootstrap values (percentages) are based on 1,000 replications; the bar refers to genetic divergence as calculated by the MEGA6 software. ⁷⁴ White circles: Streptococcus pneumoniae R6 related pbp2a (<1% nucleotide divergence); black circles: sequences differing from pbp2a of Streptococcus pneumoniae R6 by > 4.5% nucleotide; penicillin-susceptible S. mitis: light blue; penicillin-resistant S. mitis: dark blue. The mosaic structure is indicated on the left with the domain structure of PBP2a shown on top. The gray area represents the TP domain used for the phylogenetic tree. The colors mark pbp2a sequences diverging by <1%; gray: unrelated sequences differing by > 4.5%.

A closer look revealed sequence blocks almost identical to Streptococcus mitis B6 sequences in 11 of these sequences corresponding to a mosaic structure of pbp2a, whereas pbp2a of strain 8249 was distinct (Fig. 4). Five of the serotype 23F strains from Poland, all members of the clone Poland^23F-16 or single locus variants thereof, ⁴⁶ carried the same mosaic variant; pbp2a of strain 29044 was highly similar to that of the PRSP clone CSR^14–10. This strongly indicates that occasionally pbp2a in PRSP has acquired sequence blocks from S. mitis by genetic transformation and, thus, represents a mosaic gene in particular clones. All other S. pneumoniae pbp2a were highly similar (<1% divergence) and differed from S. mitis pbp2a between 5.5% and 6.6%.

Distinct alleles occurred in S. mitis even in penicillin susceptible strains, similar to as described for other PBP genes.^27,28 The PBP2a gene of Streptococcus mitis NCTC10712 differed from pbp2a of Streptococcus mitis NCTC12261 by only 0.9%, but from pbp2a of the other S. mitis by 5.7–5.9%. BLAST searches of unique pbp2a sequences revealed no match above 96% in case of Streptococcus pneumoniae 8249 and Streptococcus mitis DD28; Streptococcus mitis M3 pbp2a was highly similar to that of Streptococcus mitis SK578 (JPFY01000008). No sequence similarities with Streptococcus oralis Uo5 pbp2a, which differs from pbp2a of Streptococcus pneumoniae R6 by almost 20% and does not play a role in this high level β-lactam–resistant strain, ⁶⁶ were detected in any of the pbp2a genes analyzed in this study.

Mutations in PBP2a of PRSP

So far, mutations relevant for penicillin resistance have only been identified within the TP domains of PBPs. Structural information for the 731 amino acid PBP2a-protein is not available. A PBP2a fragment spanning amino acid 301–678, obtained by limited proteolysis with trypsin of a purified PBP2a derivative, was still able to bind penicillin. ⁶⁴ This fragment contains the three active site motifs and, thus, includes the TP domain. Therefore, we concentrated our search for mutations potentially related to β-lactam resistance on this region in PBP2a of clinical isolates.

The deduced amino acid sequences revealed a change of Thr⁴¹¹Ala, next to the active site Ser⁴¹⁰, in all deduced PBP2a with B6-related sequence blocks, as well as in PBP2a of Streptococcus pneumoniae 8249 and Streptococcus mitis DD28 (Fig. 5). This mutation has been highlighted in strain SP00084 with high-level cefotaxime resistance. ³⁸ Ala⁴¹¹ is also present in PBP2a of isolate 1, ³⁹ and occurs in PBP2a of two PRSP strains 1983 and 55938 (as deduced from the available sequence reads), and contributes to β-lactam resistance as shown by genomic sequencing of Streptococcus pneumoniae R6 transformants obtained by serial transformations. ³⁷

OPEN IN VIEWER

FIG. 5.

Deduced amino acid sequences of PBP2a in penicillin-susceptible and -resistant Streptococcus pneumoniae and Streptococcus mitis. PBP2a of Streptococcus pneumoniae R6 was used as a reference. The positions of the amino acids are indicated vertically in the top three rows. Only positions with altered residues are shown; residues identical to those in PBP2a of strain R6 are indicated by dots. The mutations A⁴¹¹ and L⁴³¹ common to mosaic Streptococcus pneumoniae PBP2a (boxed) are highlighted in pink; light blue indicates mutations that occur only in a subset of sequences containing mosaic blocks (see text for details). S. mitis sequences are shaded gray.

Most PBP2a with Ala⁴¹¹ also contained the changes Gln⁴³¹Leu and Val⁴⁴⁴Ala, except for PBP2a of strain 8249 which has Val⁴⁴⁴. Both mutations are absent in Streptococcus mitis DD28 PBP2a. Four PBP2a (strains Hu11, isolate 1, SP00084, and k-598) contained the mutation Glu⁶⁰¹Asp. All these changes were present in Streptococcus mitis B6 PBP2a and absent in other PBP2a variants of S. pneumoniae with different levels of β-lactam resistance. Ile⁵⁸⁶, Ser⁴²⁵, Ile³⁵⁴, and Ser⁶⁵⁶, present in the nonmosaic PBP2a of our study, were also mentioned in a study by Chesnel et al. ³⁴ It is interesting that the authors observed sequence differences in one pbp2a variant consistent with a recombination event, but this sequence type was found in a susceptible strain.

The presence of PBP2a mutations did not appear to be related to the resistance level of the strains. For example, the three Romanian strains with benzylpenicillin MICs of 4–8 μg/ml contained only amino acid changes that were present in susceptible S. pneumoniae such as strains R6 (Gly⁴²⁵Ser) and 657 (Val⁵⁸⁶Ile Pro⁶⁵⁶Ser); Ser⁵⁴⁷Asn is also present in strain Hun663 (clone Hungary^19A-6) where PBP2a does not seem to play a role in penicillin resistance, ⁶⁷ and in strain 670 (Spain^6B-2). This is not surprising given the fact that other PBPs and non-PBP components as described above feed into the phenotypic expression of penicillin resistance. It is likely that murM contributes to high-level resistance as shown for PRSP isolates from Romania ⁶⁸ and for members of the Hungary^19A-6 clone^67,69 where the ciaH232 allele also plays a role.^56,70

Ala⁴¹¹ mediates low affinity to BocillinFL

Since the mutations Ala⁴¹¹ and Leu⁴³¹ were prevalent among mosaic PBP2a from PRSP, mutated pbp2a was constructed by site-directed mutagenesis that contained the single Ala⁴¹¹ mutation or both, Ala⁴¹¹ plus Leu⁴³¹. They were introduced into strain R6pbp2x²³⁴⁹ by the Janus procedure to investigate the role of these mutations. S. pneumoniae 2349 is a member of the PRSP clone Spain^23F-1.^36,41 The mutation Ala⁴¹¹ resulted in slow binding and low labeling intensity of BocillinFL, and binding reached a plateau of ∼30% compared to the value obtained for PBP2a^R6 after 16 minutes (Fig. 6). The introduction of Leu⁴³¹ led to an altered electrophoretic mobility so that PBP2x and PBP2a^A411L431 were difficult to resolve on SDS-gels, and therefore, the affinity change could not be analyzed quantitatively in this case.

OPEN IN VIEWER

FIG. 6.

Binding of PBP2a with mutations Ala⁴¹¹ and Leu⁴³¹ to BocillinFL. (A) Cell lysates of R6pbp2a^A411 and R6pbp2a^A411L431 were incubated with 4.8 μM BocillinFL at 15°C to obtain better resolution of binding activities. The parental strain R6 was included for comparison. Samples were taken at the time indicated and placed into a boiling water bath upon addition of 20 μl sampling buffer. After SDS-PAGE, PBP2a-BocillinFL complexes of strain R6 and R6pbp2a^A411 were visualized with a FluorImager. (B) Fluorescence intensities of PBP-BocillinFL complexes for R6 and R6pbp2a^A411 shown in (A) were recorded with the Scion Image software.

When introduced into R6pbp2x²³⁴⁹, only minor changes in cefotaxime susceptibility were conferred by Ala⁴¹¹, or by both mutations Ala⁴¹¹ and Leu⁴³¹, increasing the MIC from 0.3 (R6pbp2x²³⁴⁹) to 0.35 μg/ml only (Table 1). Since the affinity of PBP2a for penicillins is lower than for cephalosporins ⁴⁰ it would be interesting to analyze the affinity for cefotaxime in the PBP2a mutant constructs with purified protein derivatives.

The full-length pbp2a^B6, which contains the mutations Ala⁴¹¹ and Leu⁴³¹, is known to confer cefotaxime resistance after transformation in Streptococcus pneumoniae R6 containing the PBP2x gene of Streptococcus mitis B6. ³⁶ When introduced into the R6pbp2x²³⁴⁹ background, pbp2a^B6 also mediated an over twofold increase of cefotaxime MIC from 0.3 to 0.65 μg/ml (Table 1), strongly suggesting that other mutations in PBP2a^B6 are important for resistance. No difference in cefotaxime MIC was apparent between R6 and R6pbp2a^B6, that is, in the absence of a mosaic pbp2x, at least within the range of 0.05 μg/ml used for MIC determination (Table 1). It should be noted that R6pbp2a^B6 grew similar to the parental R6 strain and did not lyse rapidly as the strain R6pbp2a^C403 (not shown).

Clinical isolates versus laboratory mutants: a conclusion

Once more, distinct modes of penicillin resistance development became apparent in cefotaxime-resistant laboratory mutants versus clinical isolates in terms of PBP2a mutations. Clinical isolates have the opportunity to capture genes from other species resulting in the well-known mosaic gene structures of pbp2x, pbp2b, and pbp1a and, as described here, also of pbp2a in particular PRSP clones. Moreover, clinical isolates had sufficient time to evolve compensatory mutations that help to alleviate potential fitness loss associated with some mutations that increase the β-lactam resistance level.

The finding that mutants selected in the laboratory with cefotaxime contained truncated versions of pbp2a rather than mutations that lead to amino acid changes in the protein as seen in PRSP is in agreement with fundamental differences between these two settings. This has been pointed out before in context with other β-lactam resistance determinants: mutations selected in PBP2x and PBP2b are distinct from those identified in clinical isolates in most cases ³¹ ; ciaH mutations are frequent in laboratory mutants but rare in PRSP and less activating^56,70; and finally, alterations in pbp1a which are a prerequisite for high resistance levels in PRSP have not been observed in laboratory mutants. PBP1a mutations in PRSP are likely to have altered enzymatic activity as suggested previously. ⁷⁰

Since double mutants in PBP2a/PBP1a are not viable, the absence of an enzymatically functional PBP2a in the three cefotaxime-resistant mutants might prevent mutations in PBP1a. In this context it should be pointed out that R6 derivatives may express phenotypes distinct from its progenitor D39 due to the presence of ≥ 80 mutations in R6 compared to D39, ⁷¹ at least with respect to several genes implicated in cell wall biosynthesis as reported recently. ⁷² Therefore, Streptococcus pneumoniae R6 might not represent the ideal genetic background to evaluate the impact on β-lactam resistance of pbp2a mutations, and this should be considered in future experiments.

The exact function of PBP2a remains largely unknown. ¹⁷ Interestingly, it has been shown that PBP2a resides in complexes with the cytoplasmic protein GpsB, a major regulator of peptidoglycan biosynthesis and other cell division proteins, and that GpsB is most likely required for PBP2a activation. ⁷² The authors also mention a recognition sequence for GpsB present in the 51aa positively charged, cytoplasmic amino domain of PBP2a. We do not know whether PBP2a fragments are expressed in the cefotaxime-resistant mutants with premature stop codons in pbp2a, but it is suggestive that they all have the potential to encode the cytoplasmic tail at least up to Gln ⁴¹ , which includes the highly positively charged region R ³⁰ RSRSDRKK. Thus it is possible that the mutations in the laboratory mutants maintain the ability for GpsB-PBP2a interaction to ensure proper PG synthesis.

Further studies will be required to identify the still unknown mutations selected in the laboratory by genome sequencing, to unravel the impact of individual PBP2a mutations on resistance in context with mutations in other genes, and to reveal the function of PBP2a in peptidoglycan biosynthesis and cell division.