AmpD_I Is Involved in Expression of the Chromosomal L1 and L2 β-Lactamases of Stenotrophomonas maltophilia

Abstract

Two ampD homologues, ampD_I and ampD_II, of Stenotrophomonas maltophilia have been cloned and analyzed. Comparative genomic analysis revealed that the genomic context of the ampD_II genes is quite different, whereas that of the ampD_I genes is more conserved in S. maltophilia strains. The ampD system of S. maltophilia is distinct from that of the Enterobacteriaceae and Pseudomonas aeruginosa in three respects. (i) AmpD_I of S. maltophilia is not encoded in an ampDE operon, in contrast to what happens in the Enterobacteriaceae and P. aeruginosa. (ii) The AmpD systems of the Enterobacteriaceae and P. aeruginosa are generally involved in the regulation of ampR-linked ampC gene expression, while AmpD_I of S. maltophilia is responsible for the regulation of two intrinsic β-lactamase genes, of which the L2 gene, but not the L1 gene, is linked to ampR. (iii) S. maltophilia exhibits a one-step L1 and L2 gene derepression model involving ampD_I, distinct from the two- or three-step derepression of the Enterobacteriaceae and P. aeruginosa. Moreover, the ampD_I and ampD_II genes are constitutively expressed and not regulated by the inducer and AmpR protein, and the expression of ampD_II is weaker than that of ampD_I. Finally, AmpD_II is not associated with the derepression of β-lactamases, and its role in S. maltophilia remains unclear.

Stenotrophomonas maltophilia, a gram-negative rod, shows increasing prevalence in clinical infections (2, 4). S. maltophilia produces two inducible β-lactamases (the L1 and L2 enzymes) which together confer resistance to all β-lactams (4). L1, a molecular class B and functional group 3 β-lactamase, is a Zn²⁺-dependent metalloenzyme with a broad substrate profile including penicillins, cephalosporins, and carbapenems (25). L2 displays a hydrolytic ability toward penicillins, cephalosporins, and monobactams (26). A LysR-type transcriptional regulator gene, ampR, is contiguous to the L2 gene but divergently transcribed (10). No putative transcriptional regulator gene has been discovered in the proximity of the L1 gene. AmpR has been shown to be a key regulator for L1 and L2 β-lactamase induction in response to β-lactam challenge (24). Recently, the regulatory characteristics of AmpR have been further elucidated (17).

ampR-ampC systems, homologous to the ampR-L2 gene module, are widely distributed among several species of Enterobacteriaceae and in Pseudomonas aeruginosa (18, 21). A similar induction mechanism involving AmpR has been proposed for the ampR-ampC systems and the ampR-L2 gene module (17). The induction of chromosomal ampC is linked with bacterial cell wall recycling, and several regulatory genes, such as ampR, ampG, and ampD, are involved in the induction course (6). The degraded cell wall product GlcNAc-anhMurNAc-peptide (N-acetylglucosaminyl-1,6-anhydro-N-acetylmuramyl-peptide), including GlcNAc-anhMurNAc-tripeptide, GlcNAc-anhMurNAc-tetrapeptide, and GlcNAc-anhMurNAc-pentapeptide, is transported into the cytosol by the AmpG protein and further cleaved into 1,6-anhydromuramic acid and peptide by AmpD. The peptide is reused by the enzymes of the cell wall-recycling pathway, resulting in the formation of UDP-N-acetylmuramic acid-pentapeptide. AmpR, a transcriptional regulator of ampC expression, is regulated allosterically by two cell wall components, anhMurNAc-peptide and UDP-N-acetylmuramic acid-pentapeptide. The binding between AmpR and anhMurNAc-peptide activates AmpC gene expression. In contrast, AmpR binding with UDP-N-acetylmuramic acid-pentapeptide represses the expression of AmpC.

AmpD, a cytosolic anhydro-N-acetylmuramyl-l-alanine amidase, acts as a key enzyme to balance the concentration of GlcNAc-anhMurNAc-peptide and UDP-N-acetylmuramic acid-pentapeptide in the cytosol. The role of the AmpD protein in the Enterobacteriaceae, including Enterobacter cloacae and Citrobacter freundii, has been well studied (8, 22). In addition, three ampD genes (ampD, ampDh2, and ampDh3) are present in P. aeruginosa, and a new mechanism of stepwise upregulation has been proposed (13). Moreover, a recent study has demonstrated that other regulatory pathways are involved in the derepressed phenotype of P. aeruginosa, in addition to the three known ampD genes (27). Recently, an AmpD-like lipoprotein, AmiD, was identified and characterized in Escherichia coli (30). Apparently, multiple anhydro-N-acetylmuramyl-l-alanine amidase-like genes can be commonly seen in a microorganism. Multiple AmpD homologues may coordinately regulate the expression of the ampC gene in the Enterobacteriaceae and P. aeruginosa, resulting in a two-step or three-step derepression phenotype (13, 19).

The genome sequences of two S. maltophilia strains, K279a (2; http://www.sanger.ac.uk/Projects/S_maltophilia) and R553-1 (http://www.genome.gov/page.cfm?pageID=10506376), have been determined. A genome-wide search of the sequence databases showed that two AmpD homologues are found in S. maltophilia K279a (Smlt0154 and Smlt1562) and R553-1 (Smal0115 and Smal1320). The exact functions of these AmpD-like proteins in the induction of the L2 β-lactamase (and perhaps of L1) in S. maltophilia are still only partially understood. This study aimed to gain a better understanding of the functions of the AmpD homologues in S. maltophilia.

MATERIALS AND METHODS

Bacterial strains, plasmids, and primers.

The bacterial strains, plasmids, and primers used in this study are listed in Table 1. S. maltophilia KJ is a clinical isolate producing the L1 and L2 β-lactamases, as identified by isoelectric focusing electrophoresis in our previous study (10). S. maltophilia KJ displays a low basal L2 lactamase activity, and expression of L1 and L2 is highly inducible. Moreover, the inducibility of L1 and L2 is AmpR dependent (17).

TABLE 1.

Bacterial strains, plasmids, and primers used in this study

Strain, plasmid, or primer	Genotype, sequence, or properties	Reference
S. maltophilia
KJ	Wild type; a clinical isolate from Taiwan	9
KJΔDI	S. maltophilia KJ ampDI isogenic mutant; deletion of 103-bp internal DNA fragment of ampDI gene	This study
KJΔDII	S. maltophilia KJ ampDII isogenic mutant; deletion of 149-bp internal DNA fragment of ampDII gene	This study
KJΔDIΔDII	S. maltophilia KJ double mutant of ampDI and ampDII genes; deletion of 103-bp and 149-bp internal DNA fragments of ampDI and ampDII genes, respectively	This study
KJΔR	S. maltophilia KJ ampR isogenic mutant; deletion of 468-bp internal DNA fragment of ampR gene	This study
E. coli
DH5α	λ− ϕ80dlacZ&Dgr;M15 &Dgr;(lacZYA-argF)U169 recA1 endA1 hsdR17(rK− mK−) supE44 thi-1 gyrA relA1	Invitrogen
S17-1	λ pir+ mating strain
Plasmids
pEX18Tc	sacB oriT Tcr	6
pAmpDI	pEX18Tc vector with an 820-bp DNA fragment of S. maltophilia KJ containing the intact ampDI gene and the 50-bp upstream region; Tcr	This study
pAmpDII	pEX18Tc vector with a 2,172-bp DNA fragment of S. maltophilia KJ containing the intact ampDII gene and the 858-bp upstream region; Tcr	This study
pAmpDIIDI	pEX18Tc vector with a 2,502-bp DNA fragment containing the intact ampDI and ampDII genes; Tcr	This study
pΔDI	pEX18Tc vector with a 1,918-bp DNA fragment of S. maltophilia KJ containing the ampDI gene with an internal 103-bp deletion, the 841-bp upstream region, and 349-bp downstream region of ampDI; Tcr	This study
pΔDII	pEX18Tc vector with a 2,089-bp DNA fragment of S. maltophilia KJ containing the ampDII gene with an internal 149-bp deletion, the 889-bp upstream region, and 523-bp downstream region of ampDI; Tcr	This study
Primersa
AmpDI-F	5′-CGAAGCTTCGACAAGGAAAGGGAAGGCAG-3′	This study
AmpDI-R	5′-CAAGATCTGCACCCACCAACAGCGGCAG-3′	This study
AmpDII-F	5′-CACTTCCACTGTCCTCGTTC-3′	This study
AmpDII-R	5′-CCCTTGCCCTTCAGTTCC-3	This study
DIN-F	5′-TGAAGCTTCCAATGGTGGCAGTGG-3′	This study
DIN-R	5′-GGTCTAGAAGTGGCAGGCGGTCTTC-3′	This study
DIC-F	5′-GGTCTAGACAACAGCGGGCATTTCTAC-3′	This study
DIC-R	5′-CTGAATTCCGCACGCATCTACGCCGAC-3′	This study
16rDNAQ-F	5′-GACCTTGCGCGATTGAATG-3′	This study
16rDNAQ-R	5′-CGGATCGTCGCCTTGGT-3′	This study
L1Q-F	5′-ACCCCTGGCAGATCGGCAC-3′	This study
L1Q-R	5′-CAGCAGCACCGCCGTTTC-3′	This study
L2Q-F	5′-AACGCACCCACCGATGCC-3′	This study
L2Q-R	5′-CGCCTGTCCAGCAATGCC-3′	This study
AmpDIQ-F	5′-CTACGAAGACCGCCTGCC-3′	This study
AmpDIQ-R	5′-GAAATGCCCGCTGTTGCC-3′	This study
AmpDIIQ-F	5′-CCACCACACCGAGCAGAAG-3′	This study
AmpDIIQ-R	5′-ATCTGCGCCGCACTGAAC-3′	This study

^aUnderlining indicates the restriction sites introduced for cloning.

Cloning and sequencing of the ampD_I and ampD_II genes.

In order to PCR amplify the two ampD genes of S. maltophilia KJ, two sets of primers derived from the K279a genome sequence (2), AmpDI-F/AmpDI-R for ampD_I and AmpDII-F/AmpDII-R for ampD_II, were used (Table 1). The AmpD homologues of Smlt1562 and Smlt0154 from S. maltophilia KJ were designated AmpD_I and AmpD_II, respectively. All the PCR amplifications were performed using the following program: 5 min at 94°C and 30 cycles of 1 min at 94°C, 1 min at 60°C, and 1 min at 72°C, followed by a final extension step of 10 min at 72°C. The amplified products contained the putative promoter region and structural ampD genes, with amplicons of 820 bp for the ampD_I gene and 2,172 bp for the ampD_II gene. The PCR amplicons were treated with the restriction enzymes HindIII and XbaI and then cloned into the plasmid pEX18Tc (7) treated with the same enzymes. Both strands of the cloned DNA were sequenced. The resultant plasmids were named pAmpDI and pAmpDII. A 1.4-kb ampD_II gene fragment obtained from pAmpDII was inserted into the HindIII-treated pAmpDI, resulting in the recombinant plasmid pAmpDIIDI.

Construction of ampD_I, ampD_II, and ampD_ID_II mutants.

A gene replacement strategy was used for the construction of the ampD_I, ampD_II, and ampD_ID_II mutants. A 944-bp PCR amplicon containing the upstream and N-terminal regions of the ampD_I gene was obtained by PCR using the primer sets DIN-F/DIN-R (Table 1) and cloned into the yT&A vector (Yeastern Biotech Co.) (pTDIN). Similarly, the recombinant plasmid pTDIC, which contained the 793-bp C-terminal and downstream regions of the ampD_I gene, was obtained using the PCR primer sets DIC-F/DIC-R (Table 1). pTDIN was digested with HindIII and XbaI and cloned into pEX18Tc, and then the 793-bp XbaI-EcoRI fragment cut from pTDIC was further ligated. The resultant plasmid, pΔDI, with an internal 103-bp deletion of the ampD_I gene, was used for the construction of mutant KJΔDI. A similar strategy was employed for constructing the recombinant plasmid pΔDII, which had an internal 149-bp deletion of the ampD_II gene. The resultant plasmids, pΔDI and pΔDII, were transformed into E. coli S17-1 and further introduced into S. maltophilia KJ by conjugation (10). Transconjugants were selected on LB plates containing norfloxacin (2.5 μg/ml) and tetracycline (30 μg/ml). For mutant selection, the transconjugants were further transferred onto 10% sucrose-containing plates. The authenticity of the mutants was checked by colony PCR amplification (16) and sequencing. The double mutant was constructed sequentially from the single mutant by the same procedure.

Complementation studies.

The plasmids pAmpDI, pAmpDII, and pAmpDIIDI were introduced into strain KJΔDIΔDII by conjugation. The transconjugants were selected on LB plates containing 2.5 μg/ml norfloxacin and 30 μg/ml tetracycline and further confirmed by colony PCR (16) using AmpDIQ-F/AmpDIQ-R and AmpDIIQ-F/AmpDIIQ-R as the primers.

Preparation of β-lactamase extracts.

Cells were grown to mid-log phase, and an inducer at a specific concentration was added for the production of β-lactamase. After further incubation for 2 h, the cells were harvested by centrifugation and washed with 10 mM sodium phosphate buffer at pH 7.0. Following resuspension of the pellet in 2 ml of the same buffer, the cells were disrupted by sonication to yield the cell extract, which was ready for isoelectric focusing and β-lactamase activity assays.

Determination of β-lactamase activity.

Specific β-lactamase activity was determined spectrophotometrically on crude sonic cell extracts (17). The differential L1 and L2 β-lactamase activities were determined by the modified nitrocefin-EDTA method (9). The specific activity (U/mg) was expressed as nanomoles of nitrocefin hydrolyzed per minute per milligram of protein, using an extinction coefficient (Δɛ) of 20,500 M⁻¹ cm⁻¹ for nitrocefin at 486 nm, as suggested by the manufacturer (Oxoid, United Kingdom). The protein concentration was determined using the Bio-Rad protein assay reagent, with bovine serum albumin as a standard.

QRT-PCR.

Total RNA from cultures grown to log phase with and without inducers was isolated by using the PureLink Total RNA Purification System (Invitrogen, Carlsbad, CA) and treated with 1 unit of RNase-free DNase I (Invitrogen, Carlsbad, CA) for 15 min to eliminate DNA contamination. cDNA was generated from equal amounts of RNAs of the assayed strains by using an MMLV Reverse Transcriptase 1st Strand cDNA Synthesis Kit (Epicentre Biotechnologies, Taiwan). Quantitative real-time PCR (QRT-PCR) was performed with Smart Quant Green Master Mix (Protech Technology Enterprise Co., Ltd.) using the programmed ABI Prism 7000 Sequence Detection System (Applied Biosystems) (11). The primers used for QRT-PCR are listed in Table 1. The level of gene expression was calculated using the comparative ΔΔC_T method (20). The expression of assayed genes was normalized to the endogenous 16S rRNA gene for variation in RNA quantity and quality. Each experiment was performed three times.

Antimicrobial susceptibility test.

MICs were determined in triplicate by a standard twofold serial agar dilution method according to the guidelines of the Clinical Laboratory Standards Institute (1). All antibiotics were purchased from Sigma. The MICs of cefepime, imipenem, and meropenem were quantified using Etest strips (AB Biodisk, Solna, Sweden) according to the manufacturer's instructions.

Nucleotide sequence accession numbers.

The nucleotide sequences of S. maltophilia KJ ampD_I and ampD_II have been deposited in GenBank under accession no. EJ447465 and EJ447466.

RESULTS AND DISCUSSION

The ampD_I gene and its genomic context organization are conserved in different S. maltophilia strains.

Sequence analysis of the 820-bp PCR amplicon from strain KJ revealed a putative ampD_I gene. The predicted 188-amino-acid AmpD_I protein exhibited 98, 94, 67, 32, 33, 33, and 31% identity, respectively, to Smlt1562 of S. maltophilia K279a (accession no. NC_010943), Smal1320 of S. maltophilia R551-3 (accession no. NC_011071), XCC1539 of Xanthomonas campestris pv. campestris ATCC 33913 (accession no. NC_003902), PA4522 (AmpD) of P. aeruginosa PAO1 (accession no. NC_002516), AmpD of E. cloacae 14 (accession no. Z14003), AmpD of C. freundii OS60 (accession no. Z14002), and AmpD of E. coli K-12 (accession no. AAC73221). The intraspecies diversity in the AmpD_I protein was at least 6%, as revealed by comparison among S. maltophilia strains KJ, K279a, and R551-3. This diversity value is consistent with those of E. cloacae (22) and C. freundii (29). A highly conserved “common core region” of the AmpD protein and seven strictly conserved residues revealed by Jacobs et al. (12) were also identified in the AmpD_I protein of S. maltophilia KJ. Furthermore, the AmpD_I protein of S. maltophilia KJ conserved the four residues essential for C. freundii AmpD activity proposed by Genereux et al. (5).

The AmpD_I protein is likely a cytoplasmic protein, according to the SignalP3.0 Server prediction (http://www.cbs.dtu.dk/services/SignalP/).

Two whole genomic sequences of S. maltophilia strains K279a and R551-3 are available in the public database. The genomic organizations surrounding ampD_I in strains K279a and R551-3 were compared. Highly conserved genetic contexts were found to surround the ampD_I gene. A 528-bp open reading frame (ORF) encoding a hypothetical protein was located upstream of the ampD_I gene, while a putative sulfurtransferase-encoding gene was observed downstream. However, no ampE homologue was present in the neighborhood of the ampD_I gene. This structure is different from the genomic organization of the ampD-ampE operon of the Enterobacteriaceae and P. aeruginosa (8, 15).

The ampD_II gene and its genomic context are more divergent in different S. maltophilia strains.

The ampD_II gene from strain KJ was located on a 2,172-kb PCR amplicon. Sequencing of the amplicon revealed a complete 747-bp ORF and two truncated ORFs. The complete ORF encoded AmpD_II, which showed 99, 90, 67, 38, 33, and 40% homology, respectively, to Smlt0154 of S. maltophilia K279a (accession no. CAQ_43763), Smal0115 of S. maltophilia R551-3 (accession no. ACF_49820), XCC3805 of X. campestris pv. campestris ATCC 33913 (accession no. AAM_43051), PA5485 (AmpDh2) (accession no. NP_254172) and PA0807 (AmpDh3) (accession no. NP_249498) of P. aeruginosa PAO1, and AmiD of E. coli (accession no. NP_415388). AmpD_I and AmpD_II of S. maltophilia KJ displayed 29% identity. As predicted by the LipoP 1.0 server (14; http://www.cbs.dtu.dk/services/LipoP/), a putative signal peptidase II cleavage site was detected between the 18th Ala and the 19th Cys residues, indicating that AmpD_II could be a lipoprotein. According to the “+2 rule” of lipoprotein sorting (28), AmpD_II is likely to be transferred to the outer membrane.

Interestingly, the genetic contexts of ampD_II in S. maltophilia strains K279a and R551-3 were different. In K279a, ampD_II was flanked by a putative ammonia transporter gene (Smlt0153) and a putative murein-degrading transglycosylase gene (Smlt0156), while in R551-3, there was an approximately 17.5-kb DNA fragment inserted between the putative ammonia transporter (Smal 0111) and ampD_II (Smal0115) genes, as well as an extra 4-kb DNA fragment inserted between ampD_II and the putative murein-degrading transglycosylase (Smal0119) gene. The ampD_II genomic organization of strain KJ is more similar to that of strain K279a.

AmpD_I, but not AmpD_II, is involved in regulation of β-lactamase expression.

To investigate the biological functions of AmpD_I and AmpD_II, the isogenic mutants KJΔDI, KJΔDII, and KJΔDIΔDII were constructed. It has been found that inactivation of ampD of the Enterobacteriaceae and P. aeruginosa generally leads to a derepression phenotype (12, 19). Therefore, induction assays and β-lactamase activity measurements were performed with the wild-type strain KJ and the isogenic mutants KJΔDI, KJΔDII, and KJΔDIDII. Cefoxitin (50 μg/ml) was used as the inducer, and the nitrocefin-EDTA method (9) was used to differentially determine the L1 and L2 β-lactamase activities. Strain KJΔDII displayed an induction pattern (low basal level and high induced activity) similar to that of the wild-type strain KJ, whereas strains KJΔDI and KJΔDIΔDII had a fully derepressed phenotype (constitutive hyperexpression) with an equivalent fully derepressed β-lactamase activity (Table 2). Furthermore, the mutants KJΔDI and KJΔDIΔDII expressed both the L1 and L2 β-lactamases in the basal condition (Table 2), indicating that AmpD_I is involved in the expression of both the L1 and L2 genes.

TABLE 2.

Basal and induced β-lactamase activities of S. maltophilia strains

Strain	β-Lactamase activity (Ua/mg)
	Basal			Inducedb
	(L1 + L2)c	L1d	L2e	(L1+L2)c	L1d	L2e
KJ	20	0	20	1,958	605	1,437
KJΔDI	4,478	1,155	3,323	4,379	1,064	3,315
KJΔDII	16	1	15	2,105	624	1,481
KJΔDIΔDII	4,431	896	3,535	4,478	1,006	3,472
KJΔDIΔDII (pKJAmpDI)	20	1	19	1,186	282	904
KJΔDIΔDII (pKJAmpDII)	4,706	1,272	3,434	4,539	1,475	3,064
KJΔDIΔDII (pKJAmpDIIDI)	14	2	12	743	366	377

^aOne unit of β-lactamase activity is defined as 1 nanomole of nitrocefin hydrolyzed per minute. The results are means of three independent determinations. Standard derivations were within 10% of the means in all cases.

^bWith 50 μg/ml cefoxitin as the inducer.

^cThe total β-lactamase activity (L1 and L2) was determined by the nitrocefin method.

^dThe L1 β-lactamase activity was calculated by subtracting the L2 β-lactamase activity from the total β-lactamase activity.

^eThe L2 β-lactamase activity was determined by the nitrocefin-EDTA method.

Isoelectric focusing analysis of β-lactamases in the cell extracts revealed two identical bands, corresponding to L1 and L2, in the induced strain KJ and in the uninduced strain KJΔDI. This finding confirmed that both the L1 and L2 genes are coderepressed in the AmpD_I-null mutant.

The RNA transcripts of the L1 and L2 genes from strains KJ and KJΔDI were quantified by QRT-PCR. Under the uninduced condition, the transcript ratio of KJΔDI to KJ was 89 for the L1 gene and 184 for the L2 gene. However, this ratio was 3 for the L1 gene and 21 for the L2 gene under the induced condition. Consequently, the AmpD_I protein has a stronger repressor effect on L1 and L2 β-lactamase expression regardless of the presence of β-lactams.

Complementation assays were performed by introducing pAmpDI, pAmpDII, and pAmpDIIDI into strain KJΔDIΔDII, and the basal and induced β-lactamase activities of these transconjugants were quantified. Strains KJΔDIΔDII(pAmpDI) and KJΔDIΔDII(pAmpDIIDI) exhibited the wild-type β-lactamase expression pattern, i.e., low basal level and inducibility, as observed in the wild-type strain KJ, while complementation with pAmpDII did not restore the inducible phenotype in KJΔDIΔDII (Table 2).

These results confirmed that AmpD_I is a functional anhydro-N-acetylmuramyl-l-alanine amidase in S. maltophilia and is associated with the regulation of L1 and L2 gene expression. Although strains KJ, KJΔDIΔDII(pAmpDI), and KJΔDIΔDII(pAmpDIIDI) displayed β-lactamase inducibility (Table 2), the induced β-lactamase activities were notably different, since KJΔDIΔDII(pAmpDI) showed an induced β-lactamase activity lower than that of the wild-type strain, KJ (Table 2). This difference could be related to the copy number of plasmid pAmpDI. More AmpD_I proteins in KJΔDIΔDII(pAmpDI) could result in a stronger repressor effect.

As for AmpD_II, results obtained with KJΔDII suggested that AmpD_II has no significant effect on the expression of the L1 and L2 genes, and this hypothesis was confirmed by data from the complementation assays. However, it is noteworthy that the induced β-lactamase activity of strain KJΔDIΔDII(pAmpDIIDI) was lower than that of strain KJΔDIΔDII(pAmpDI) (743 versus 1,186 [Table 2]), suggesting that the AmpD_II protein could exhibit a synergistic effect on the function of the AmpD_I protein. The synergistic effect of AmpD_II on the AmpD_I function was not detectable with mutants, which could be due to lower expression of the ampD_II gene when carried in the chromosomal context than when carried on the multicopy plasmid.

The effect of AmpD inactivation on β-lactam resistance was also determined. All tested β-lactams exhibited the same MICs for the wild-type strain, KJ, and the ampD-associated mutants except cefepime, which was slightly less active against the KJΔDI and KJΔDIΔDII mutants (Table 3). These results suggest that ampD_I inactivation likely does not have clinical relevance.

TABLE 3.

MICs of β-lactam antibiotics for S. maltophilia KJ and its derived ampD-associated mutants

Antibiotic	MIC (μg/ml) of antibiotic for strain:
	KJ	KJΔDI	KJΔDII	KJΔDIΔDII
Penicillins
Piperacillin	1,024	1,024	1,024	1,024
Carbenicillin	1,024	1,024	1,024	1,024
Cephalosporins
Cefuroxime	2,048	2,048	2,048	2,048
Cefoxitin	1,024	1,024	1,024	1,024
Ceftriaxone	256	256	256	256
Cefepime	64	96	64	96
Carbapenems
Imipenem	>32	>32	>32	>32
Meropenem	>32	>32	>32	>32
Monobactam
Aztreonam	>2,048	>2,048	>2,048	>2,048

The ampD_I and ampD_II genes are constitutively expressed.

The expression of the ampD_I and ampD_II genes under different conditions, such as the basal state, induced state, and AmpD_I-null state, was studied using QRT-PCR (Fig. 1). The expression of the ampD_I and ampD_II genes was constitutive and was hardly influenced by the addition of the inducer. Under basal or induced conditions, the expression of ampD_I was higher than that of ampD_II, suggesting that in wild-type S. maltophilia, AmpD_I could play a significant role in the cell wall-recycling process. Since strains KJΔDI and KJ expressed equivalent amounts of ampD_II transcript (Fig. 1), no compensatory expression of the ampD_II gene occurs in the absence of the AmpD_I protein.

FIG. 1.

The relative increase in ampD_I and ampD_II mRNA of the basal and induced S. maltophilia strains KJ, KJΔR, and KJΔDI. Each pair of bars indicates the ratio of the mRNA transcript of a specific gene to that of the strain KJ basal ampD_II, as measured by QRT-PCR. The induced condition is 30 μg/ml cefoxitin for 2 h. The mean value is based on three independent experiments. The error bars indicate the standard deviations.

AmpR is not involved in the expression of ampD_I and ampD_II.

In our previous study, AmpR was shown to act as an important regulator in the induction of the L1 and L2 genes (10). To investigate the role of AmpR in the expression of the ampD_I and ampD_II genes, the ampD_I and ampD_II transcripts of strains KJ and KJΔR were evaluated under the basal and induced conditions. Figure 1 shows that the expression of ampD_I and ampD_II was unaffected by ampR inactivation, suggesting that AmpR is not involved in the expression of the ampD_I and ampD_II genes.

Concluding remarks.

Following the release of whole genomic sequences, several bacteria have been found to have multiple ampD-like alleles, such as ampD and amiD in E. coli (30) and ampD, ampDh2, and ampDh3 in P. aeruginosa (13). S. maltophilia, like the Enterobacteriaceae, has genes for two AmpD homologues, i.e., AmpD_I and AmpD_II, in its genome. However, in the present study, the previously mentioned semiconstitutive β-lactamase hyperproduction phenotypes in E. cloacae (19) and the moderate-level (or high-level) hyperinducible phenotypes in P. aeruginosa (13) were not observed in the ampD-associated mutants of S. maltophilia. Recently, Okazaki and Avison have isolated 12 β-lactamase hyperexpression S. maltophilia mutants, 2 of which exhibit the phenotype of semiconstitutive overexpression (24). The ampD_II genes of those mutants have been sequenced, and mutations in the ampD_II gene were not observed. However, ampD_I sequence results were not reported. Nevertheless, the finding of the semiconstitutive overexpression mutants suggests that, in addition to the regulation involving ampD, alternative mechanisms may contribute to the L1 and L2 gene derepression, which is consistent with the results of Schmidtke and Hanson for P. aeruginosa (27).

For some specific enterobacterial strains, such as E. cloacae NOR-1 and Proteus vulgaris, a second set of ampR class A β-lactamase modules exists, in addition to the ampR-ampC module (3, 23). AmpD of E. cloacae NOR-1 has been shown to be involved in the expression of both of the ampR-linked β-lactamases, NmcA and AmpC (22). In this study, AmpD_I in S. maltophilia was found to play a key role, not only in the expression of the ampR-linked L2 gene, but also in the expression of the non-ampR-linked L1 gene. This, together with our previous findings (17), indicates that the L1 and L2 genes may be under the control of the ampD-ampR-associated regulatory system.