Production of BRO β-Lactamases and Resistance to Complement in European Moraxella catarrhalis Isolates

Abstract

Of the 419 Moraxella catarrhalis isolates collected during the 1997-1999 European SENTRY surveillance study, 385 (92%) were β-lactamase positive. Twenty-two (5.7%) produced BRO-2 β-lactamase. Twenty-one new mutations were found in the putative promoter region of the bro genes. Nineteen percent of all isolates tested were complement sensitive. Resistance to β-lactams is not linked to the phylogenetic lineages associated with susceptibility to complement.

Moraxella catarrhalis was long considered a harmless commensal of the respiratory tract and uniformly susceptible to penicillins. In recent years, however, its pathogenic potential has come to light (9, 20, 27). The production of a new β-lactamase enzyme, designated BRO (from Branhamella and Moraxella), by this bacterium was an event that occurred virtually simultaneously around the world (9, 13, 19, 20). Two distinct BRO β-lactamase enzymes, namely, BRO-1 and BRO-2, have since been found in strains of M. catarrhalis. These enzymes are identical in substrate and inhibition profile but differ by a single amino acid (2-5). The increased level of resistance conferred by BRO-1 compared to BRO-2 appears to be related to the relative amount of each enzyme produced by the strains (5). This difference in production may be the result of differences in the promoter strength of the two genes. Bootsma et al. (5) reported a 21-bp deletion in the promoter region of the β-lactamase gene from a BRO-2-producing strain. In addition, Richter et al. (22) suggested that additional mutations or deletions or insertions in the promoter sequence might explain the variations seen in antibiotic susceptibilities within the populations of BRO-1- and BRO-2-producing M. catarrhalis isolates.

The first goal of this study, therefore, was to characterize the BRO β-lactamases and their putative promoter regions in the European M. catarrhalis isolates collected between 1997 and 1999 during the European SENTRY surveillance study. The relative percentages of the two enzyme types were also determined.

Besides protecting the bacteria against β-lactam antibiotics, β-lactamase can also have indirect pathogenic effects. For example, it can block the antibiotic treatment of concomitant infections with more dangerous pathogens (e.g., pneumococci), as experimentally confirmed by Hol et al. (14-16). This, together with the discovery of M. catarrhalis as a pathogen in its own right, has aroused scientific interest in other possible virulence factors. One of these virulence factors is the ability of M. catarrhalis to resist the action of complement (9, 12, 17, 20, 24-26). Based on molecular typing data, i.e., pulsed-field gel electrophoresis and PCR-restriction fragment length polymorphism analysis, it has been suggested that complement-resistant M. catarrhalis isolates form a genetically distinct lineage with a different pathogenic potential within the species (3, 26, 28). The second aim of this study, therefore, was to determine the percentage of complement-resistant M. catarrhalis isolates in the recent SENTRY surveillance study and to analyze their association with BRO-1 and BRO-2 enzymes.

A total of 419 M. catarrhalis strains were isolated between 1997 and 1999 in 24 European university hospitals in 13 countries as part of the European SENTRY study. The selected isolates originated from specimens from patients with either community-acquired respiratory tract infections or nosocomial pneumonia, as reported previously for U.S. isolates (7). Only one isolate per patient was permitted. The identity of the organism was confirmed by conventional criteria (18).

To detect β-lactamase activity, we used a commercially available chromogenic cephalosporin disk β-lactamase assay containing nitrocefin as the substrate (Cefinase disks; BBL Microbiology Systems, Cockeysville, Md.). All of the β-lactamase-positive isolates were selected for β-lactamase extraction and isoelectric focusing (IEF) of the enzymes as described by Richter et al. (22). IEF of β-lactamase enzymes was performed with commercially prepared polyacrylamide gels (Ampholine PAGplate; Pharmacia Biotech, Uppsala, Sweden) with a pH range of 5.5 to 8.5 (22).

PCR amplification and sequencing of the putative promoter region and of a part of the bro gene were performed as described by Bootsma and colleagues (5). Two independent amplicons were generated and at least one strand of the second PCR product was sequenced to confirm that no mismatches were created upon amplification. Primers (forward, 5′CTTGGCGATGTCTACACC-3′; reverse, 5′AAGTTTGGCATTGACACG-3′) were used for 25 cycles of amplification (30 s at 95°C, 1 min at 55°C, and 1 min at 72°C) with a ThermoCycler (PerkinElmer, Weiterstadt, Germany).

The M. catarrhalis strains were subsequently tested for their sensitivity to serum-mediated lysis in a microtiter bactericidal assay that used 50% pooled human serum incubated at 37°C (3, 24). Using the sensitivity results, the strains could be subdivided into three categories: resistant (>50% survival of the bacteria after a 3-h incubation in 50% pooled human serum), sensitive (<3% survival of the bacteria after a 1-h incubation in 50% pooled human serum), and intermediate. The samples were observed at several different times of incubation up to the 3-h point.

MICs of amoxicillin, penicillin, cefepime, and cefixime were determined by a broth microdilution procedure according to National Committee for Clinical Laboratory Standards guidelines (21).

Of the 419 M. catarrhalis isolates collected, 385 (92%) were β-lactamase positive. The incidence of these β-lactamase-producing isolates varied from 83 to 98% in the 13 participating countries, and no increase was observed during the study period. All of the β-lactamase-producing isolates were screened for the presence of BRO-1 and BRO-2 β-lactamase by IEF (22) and for the bro1 and bro2 genes by PCR and sequencing methods (5). IEF and sequencing displayed corresponding results for all β-lactamase-producing isolates.

BRO-2 was detected in 22 (5.7%) of the 385 isolates. The remaining 94.3% of the isolates had a BRO-1 β-lactamase. The 22 BRO-2-positive isolates came from 10 of the 24 participating hospitals, located in Belgium, France, Germany, Greece, Italy, Poland, Spain, Switzerland, and the United Kingdom. β-Lactamase production was detected at similar levels in hospital and community isolates, regardless of the specimen source or the age of the subject. BRO-1-producing isolates were detected at all participating centers. As expected, all β-lactamase-negative isolates gave negative results in the IEF as well as in PCR screening for bro genes.

The incidence of BRO β-lactamase production has continued to increase around the world during the last 2 decades, and today more than 90% of clinical isolates produce these enzymes (1, 6-8, 10, 11, 23). Studies have also shown that 90 to 95% of the β-lactamase-producing strains generate the BRO-1 enzyme, while less than 10% produce the BRO-2 enzyme (6, 9, 18, 20). With these results in mind, our data suggest that the proportion of strains producing each enzyme has remained fairly constant in Europe over the last 10 years (9-20).

Although the genes coding for BRO-1 and BRO-2 display only minor differences in their coding sequences, a more pronounced sequence diversity has been observed in the upstream region of the two genes (5). The 21-bp deletion in the BRO-2 promoter region, for example, is the one most likely to have a notable effect on promoter activity, affecting both the potential −10 and −35 sequences of BRO-1. However, additional, as-yet-unknown mutations or deletions and/or insertions in the regulatory sequence might also explain the variations seen in the antibiotic susceptibilities of the BRO-1- and BRO-2-producing clinical isolates (22). Because of this, we sequenced the putative promoter region and a part of the gene of all of the β-lactamase-positive isolates (5). We found the 21-bp deletion in all of the sequences of the bro2-containing isolates but no additional deletions and/or insertions. We were also able to confirm the sequence reported by Bootsma et al. (5) for 16 of the 22 BRO-2 β-lactamase-producing isolates. The new additional nucleotide changes we found in the putative regulatory region of the six remaining isolates did not affect either the −10 and −35 sequences or the putative ribosome-binding site. These mutations were as follows: position 1583, A→C; 1586, G→A; 1589, G→A; 1590, T→C; 1592, G→C; 1595, T→C; 1598, C→T; 1601, G→A; 1604, T→C; 1610, G→A; 1613, A→G; 1616, C→T; 1619, G→A; 1625, A→G; 1627, A→C; 1736, A→T; 1752, T→G; 1753, A→G; 1754, A→G; and 1763, A→G. None of the BRO-1-producing isolates we studied exhibited a deletion or an insertion in the regulatory sequence; however, there were some new mutations. We were able to confirm the regulatory sequence reported by Bootsma et al. (5) in 93 of the 363 bro1-containing isolates. The mutations in the remaining isolates were as follows: position 1583, C→A; 1586, A→G; 1590, C→T; 1592, C→G; 1595, C→T; 1598, T→C; 1601, A→G; 1604, T→C; 1610, G→A; 1613, A→G; 1619, G→A; 1622, G→A; 1625, A→G; and 1627, A→C.

Since the 21-bp deletion was found in all of the sequences of the bro2-containing isolates but in none of the bro1-containing genes, a PCR detection of this difference in M. catarrhalis isolates might be enough for a reasonable discrimination between BRO-1- and BRO-2-producing isolates for epidemiological purposes.

Although BRO-1-producing isolates are more resistant to the β-lactams tested than BRO-2 strains (Table 1), probably due to differences in enzyme production (9, 20), the above-mentioned additional new mutations in the putative regulatory region of the bro genes did not have an observable effect on MICs. Hence, the already-reported 21-bp deletion seems to be the most important difference in the regulatory region. However, MICs are crude measurements and may not reflect subtle changes in promoter strength. Furthermore, it could not be excluded that other putative promoter sequences exist upstream from the bro genes and that alterations in these sequences may result in different MICs.

TABLE 1.

Susceptibility of M. catarrhalis to four antimicrobial agents

Antimicrobial agent and phenotype (no. of isolates)	MIC (μg/ml)a
	50%	90%	Range
Amoxicillin
β-Lactamase negative (34)	≤0.06	0.25	≤0.06-0.5
β-Lactamase positive; BRO-1 (363)	2	8	≤0.06->8
β-Lactamase positive; BRO-2 (22)	0.25	8	≤0.06->8
Penicillin
β-Lactamase negative	≤0.03	0.25	≤0.03-0.5
β-Lactamase positive; BRO-1	4	>4	≤0.03->4
β-Lactamase positive; BRO-2	1	>4	0.25->4
Cefepime
β-Lactamase negative	0.12	0.5	≤0.06-2
β-Lactamase positive; BRO-1	0.5	2	≤0.06-4
β-Lactamase positive; BRO-2	0.25	1	0.12-4
Cefixime
β-Lactamase negative	≤0.03	0.03	≤0.03-0.5
β-Lactamase positive; BRO-1	0.12	0.5	≤0.03-2
β-Lactamase positive; BRO-2	0.06	0.25	≤0.03-0.5

^a50% and 90%, MICs at which 50 and 90% of isolates were inhibited, respectively.

Complement resistance is an important virulence factor in M. catarrhalis. Studies comparing serum resistance and disease induction by M. catarrhalis have indicated that isolates from healthy persons are more likely to be sensitive to complement-mediated lysis (43 to 58%) than clinical isolates (10%) (9, 20, 28). Recently, Bootsma et al. (3) and Verduin et al. (26) suggested that complement-resistant adherent strains form a subpopulation that diverged from a common ancestor more recently than the complement-sensitive nonadherent strains. Nineteen percent of the isolates we tested were complement sensitive. The distributions of β-lactamase-negative and β-lactamase-positive strains and of BRO-1- and BRO-2-producing strains were comparable in the two complement populations: β-lactamase production was observed in 94% of the complement-sensitive isolates and in 91% of the complement-resistant isolates. The same was found for the BRO-1- and BRO-2-producing strains: the bro2 gene was found in 6% of the complement-sensitive strains and in 5% of the complement-resistant strains. These data support the theory that transfer of the bro gene is horizontal. We also found that there was no difference in antibiotic susceptibilities between complement-sensitive and complement-resistant isolates. Therefore, if complement-resistant M. catarrhalis isolates really form a genetically distinct lineage within the species, then the resistance to antibiotics is not linked to phylogenetic lineages.

In summary, of the 419 M. catarrhalis isolates collected during the 1997-1999 European SENTRY surveillance study, 385 (92%) were β-lactamase positive. Twenty-two (5.7%) of the 385 isolates produced BRO-2 β-lactamase. The additional new mutations in the putative regulatory region did not have an observable effect on the MIC distributions among the BRO-1- and BRO-2-producing isolates. Nineteen percent of the tested isolates were complement sensitive. The distributions of β-lactamase-negative and β-lactamase-positive strains and of BRO-1- and BRO-2-producing strains were comparable in the two complement populations.