carbapenemase of Aeromonas hydrophila - Journal of Antimicrobial

Journal of Antimicrobial Chemotherapy (1996) 37, 33-44

The 'hidden' carbapenemase of Aeromonas hydrophila

M. V. Hayes, C. J. Thomson and S. G. B. Amyes*

Department of Medical Microbiology, The Medical School, University of Edinburgh,Teviot Place, Edinburgh EH8 9AG, UK

It has been presumed that there are just two /Mactamases in the motile Aeromonasspecies, a carbapenemase and a cephalosporinase, based on the premise that all/?-lactamases can be detected by hydrolysis of the chromogenic cephalosporin,nitrocefin. However, when it was recently found that a non-motile species ofAeromonas that causes furunculosis in salmon, contained three /?-lactamases, oneof which was a carbapenemase which could not be detected with nitrocefin, it washypothesised that genetic exchange could occur between fish pathogens and humanpathogens resulting in the transfer of the carbapenemase-encoding gene. This couldhave a potentially serious impact on intensive therapy units where carbapenems areemployed. The purpose of this study was to determine whether the human pathogenAeromonas hydrophila demonstrated the same /?-lactamase profile. After anion andcation exchange chromatography had been employed to separate the /J-lactamasesof a clinical strain of A. hydrophila, three different /Mactamases were found, one ofwhich is a carbapenemase which does not hydrolyse nitrocefin. It is, therefore,probable that many strains of Aeromonas spp. contain a similar array of /Mactamaseswhich include a carbapenemase that cannot be detected with nitrocefin. Similarcarbapenemases may well remain hidden in other species of bacteria unlessappropriate techniques to detect the enzymes are employed .

Introduction

It has been previously reported that the motile Aeromonas species possess two/Mactamases, a carbapenemase with a high pi of 8.0 and a cephalosporinase with a lowpi of 7.0, both of which are detectable with nitrocefin (Shannon, King & Phillips, 1986;Bakken et al., 1988; Iaconis & Sanders, 1990). Massidda, Rossolini & Satta (1991) havecloned the cphA gene of Aeromonas hydrophila which codes for a carbapenemase thatslowly hydrolyses penicillins and cephalosporins including nitrocefin. Recently, three/Mactamases have been found in the non-motile species, Aeromonas salmonicida subsp.achromogenes which causes furunculosis in fish, (Hayes, Thomson & Amyes, I994a,b).Two of the enzymes, a penicillinase and a cephalosporinase, hydrolysed nitrocefinwhereas the third one, a carbapenemase, did not. (Hayes, Thomson & Amyes, I994a,b).These were the first studies to identify three separate /Mactamases from Aeromonasspp. It was postulated that the genes encoding these /Mactamases might be transferableto one of the virulent motile aeromonads associated with human infections such as

'Corresponding author.

330305-7453/96/010033 + 12 $12.00/0 £, 1996 The British Society for Antimicrobial Chemotherapy

Dow

nloaded from https://academ

ic.oup.com/jac/article/37/1/33/751319 by guest on 30 D

ecember 2021

34 M. V. Hayes et al.

gastroenteritis, wound infections, bacteraemia and pneumonia (Picard & Goulet, 1987;Pazzaglia et al., 1990) that affect immunocompetent as well as immunocompromisedhosts (Gracey, Burke & Robinson, 1982; George et al., 1985; Sawle et al., 1986;). Basedon MICs, carbapenems would seem the most reliable drugs for treating seriousinfections due to these bacteria but it is conceivable that treatment of thesecarbapenemase-harbouring strains with a carbapenem or another /Mactam antibioticmight result in an outbreak of carbapenem-resistant Aeromonas spp. in certain settings.

The purpose of this study was to separate completely the /Mactamases of the humanpathogen A. hydrophila, to ascertain if they are similar in profile to those found in thefish pathogen A. salmonicida subsp. achromogenes, and to demonstrate whether clinicalisolates do, in fact, contain /Mactamases that cannot be detected with nitrocefin.

Materials and methods

Bacterial strains

A. hydrophila G19 and A. hydrophila G75 were kindly donated by Dr J. Hood of theGlasgow Royal Infirmary. A. hydrophila C19897 was isolated at the City Hospital,Edinburgh from the faeces of a patient with diarrhoea. Aeromonas caviae R4007, A.hydrophila R6407 and A. caviae R8856 were isolated at the Royal Infirmary, Edinburghfrom stool specimens of patients with diarrhoea and A. hydrophila 5B160 and A.hydrophila 9B273 were isolated from blood cultures. A. salmonicida subsp.achromogenes ASA111 was isolated by Dr A. Barnes at the Scottish Office Agricultureand Fisheries Department Marine Laboratory, Aberdeen from a salmon suffering fromfurunculosis. A. hydrophila NCTC 8049 was included to represent the species. Stockcultures were stored in tryptone soya broth (TSB) (Oxoid Unipath, Basingstoke, UK)supplemented with 10% glycerol.

Antibiotics and chemical agents

All antibiotics and chemical agents were obtained from the manufacturers; amoxycillin(Bencard, Hertfordshire, UK), carbenicillin, clavulanic acid and BRL42715(SmithKline Beecham, Brockham Park, UK), nitrocefin and cephaloridine (GlaxoGroup Research, Greenford, UK), cefotaxime (Hoechst-Roussel Pharmaceuticals,Hounslow, UK), imipenem (Merck Sharp & Dohme, Hoddesdon, UK), ampicillin,cefoxitin, EDTA, zinc sulphate, 2-[N-morpholine]ethanesulfonic acid (MES),piperazine-N,N'-bis[2-ethanesulfonic acid] (PIPES) and Coomassie brilliant blue R250(Sigma Chemical Co. Poole, Dorset, UK) and silver nitrate (Pharmacia LKBBiotechnology, Uppsala, Sweden).

Antibiotic susceptibility tests

The MIC of each antibiotic was determined by the agar dilution method of the NationalCommittee for Clinical and Laboratory Standards (NCCLS, 1993) except thatIso-Sensitest agar (Oxoid) was used instead of Mueller-Hinton medium. Antibioticswere distributed in two-fold dilutions ranging from 0.02 to 1024 mg/L. Ampicillin wastested alone and in combination with clavulanic acid. The inoculum was prepared bygrowing the bacteria in TSB for 24 h at 30°C, and then diluting in single strength Davis

Dow



ecember 2021

Carbapenemase in A. hydrophUa 35

and Mingioli media (Davis & Mingioli, 1950) so that 1 /iL delivered 104cfu per spot.Plates were incubated at 30°C for 24 h. The MIC was taken as the lowest concentrationto inhibit visible growth.

fl-Lactamase studies

The bacteria were grown in 1 L of TSB at 37°C for 24 h and /Mactamases were inducedby exposing the culture to 1/4 of the MIC of ampicillin for a further 4 h. The cells wereharvested by centrifugation at 6000 g for 15 min, discarding the supernatant andwashing the pellet in 50 mM phosphate buffer (pH 7.0) at 1/20 of the original volumeand centrifuging as before. The pellet was resuspended in 1 mL phosphate buffer,sonicated as described by Reid & Amyes (1986) and centrifuged at 31,650 g for 1 h toremove cell debris. The /Mactamase preparations were stored at — 20°C until requiredfor use. The crude preparation was applied to a polyacrylamide gel containingampholines (Sigma Chemical Co.) with a pH range of 3.5-10. The gel was stained byoverlaying it with filter paper soaked in 500 mg/L nitrocefin for 5-30 min. (Matthewset a!., 1975). The isoelectric points of the proteins were determined by comparing theirpositions on the gel with those of standard /Mactamases of known pis.

Ion exchange chromatography

The cell-free extract from A. hydrophUa G19 was divided into two aliquots, of whichone was dialysed at 4°C in 50 mM Tris buffer (pH 8.5) and the other in 50 mM MESbuffer (pH 6.5) each for 4 h. Twenty-five milligrams of the Tris-dialysed protein wasloaded onto a Mono-Q anion exchange column (Pharmacia LKB Biotechnology,Uppsala, Sweden) which had been previously equilibrated with Tris buffer (pH 8.5). Alinear concentration gradient of nil to 1 M NaCl in Tris buffer (pH 8.5) was applied tothe column to elute the bound proteins in 1 mL fractions, each of which was maintainedat 4°C and assayed for /Mactamase activity against nitrocefin and imipenem. The sameprocedure as above was followed to elute the proteins using the same quantity ofMES-dialysed protein which was applied to a Mono-S cation exchange column(Pharmacia LKB Biotechnology, Uppsala, Sweden) that had been equilibrated with50 mM MES buffer (pH 6.5). Assuming the metallo-enzyme had a pi of c. 8.0, thecell-free extract was loaded onto a cation exchange column equilibrated at pH 6.5.Fractions containing /Mactamase activity were purified further by reloading the fractiononto the same column from which it had been eluted.

Determination of the molecular mass

The sizes of the /Mactamases were determined by sodium dodecyl sulphate-polyacryl-amide gel electrophoresis (SDS-PAGE) as described by Laemmli (1970) employing thePharmacia PhastSystem (Pharmacia LKB Biotechnology, Uppsala, Sweden). Theproteins were boiled in SDS (final concentration of 2.5% w/v), /?-mercaptoethanol (finalconcentration of 5% v/v), 10 mM Tris/HCl (pH 8.0) and 1 mM EDTA for 5 min afterwhich bromophenol blue was added to yield a final concentration of 0.01% v/v. Both/Mactamases and standards were run on a polyacrylamide gel with a continuous gradientof 10-15% and stained with either Coomassie brilliant blue R250 or silver nitrate.

Dow



ecember 2021


Substrate and inhibitor profiles

/?-Lactamase activity was measured with a Perkin Elmer X2 spectrophotometer at 37°Cin 50 mM PIPES buffer (pH 7.0) and with freshly prepared solutions of antibiotics (Reid& Am yes, 1986). The enzymes were assayed with decreasing concentrations ofampicillin, imipenem, nitrocefin, cephaloridine and cefoxitin to determine the values ofV^ and Km by the method of Lineweaver & Burk (1934). The method of Waddell (1956)was employed to determine the concentration of protein in the /Mactamasepreparations. The concentrations of clavulanic acid, BRL42715, EDTA and zincsulphate required to reduce enzyme activity by 50% (IC50) were determined bypre-incubating the enzyme with various concentrations of each inhibitor for 5 min at37°C before adding the antibiotic. The rate of hydrolysis of the reporter substrate wasdetermined for each concentration of inhibitor until the IC50 could be determined.

Results

Susceptibility tests

All strains were resistant to ^128 mg/L ampicillin (Table I) and seven were resistantto > 64 mg/L carbenicillin. MICs of ampicillin were lowered considerably to 8 or16 mg/L by the addition of 4 or 8 mg/L clavulanic acid. This reduction in the MIC wasnot solely a result of the clavulanic acid blocking the active site of the /Mactamase, asthe MIC of clavulanic acid on its own was 4-16 mg/L therefore demonstrating its abilityto bind directly to the penicillin binding proteins and act as an antibiotic. All strainswere susceptible to ^0.25 mg/L imipenem, < 1 mg/L cefotaxime and ^ 8 mg/L cefoxitin.

A. hydrophila G19 was resistant to ampicillin (MIC 512 mg/L), carbenicillin (MIC128 mg/L) and cephaloridine (MIC 128 mg/L) and susceptible to imipenem (MIC0.125 mg/L), cefoxitin (MIC 4 mg/L) and cefotaxime (MIC 0.03 mg/L). The MIC ofampicillin was lowered to 8 mg/L in the presence of 4 mg/L clavulanate which wasprobably due to the antibacterial activity of the inhibitor since the MIC was 4 mg/L.

Isoelectric focusing

Apart from A. hydrophila G75 and A. hydrophila C19897, the majority of strainsappeared to possess two /Mactamases when the crude extracts were visualised on anisoelectric focusing gel which had been stained with nitrocefin. One of the /Mactamasesof A. hydrophila G19 had a pi of 7.0 and the other a pi of 8.0 (Figure 1). The otherstrains had /Mactamases with similar isoelectric points.

Separation of the fi-lactamases

Each of the strains was initially separated on the Mono-Q anion exchange column whichhad been equilibrated at pH 8.5. Examination of the /Mactamase activity in the fractionscollected indicated that seven of the nine strains possessed both a carbapenemase whichcould hydrolyse ampicillin and nitrocefin and a cephalosporinase. The carbapenemaseactivity of each strain was also found to be inhibited by EDTA whereas thecephalosporinase and penicillinase activities in this fraction were not sensitive to EDTA(Figure 1). It was assumed that there were three /Mactamases present in each strain soonly A. hydrophila G19 was chosen for further examination.

Dow



ecember 2021

Table I. Minimum inhibitory concentrations (mg/L) for A. hydrophila strains

Strain

NCTC8049G19G755B16O9B273R4007R6407R8856C19897

Ampicillin

> 1024512128

1024256

1024512512512

Amoxycillin/clavulanate

16:88:48:4

16:816:816:816:816:816:8

Clavulanicacid

848

168

16888

Carbenicillin

> 10246464

2561024

643232

128

Cephaloridine

64128

264166464

12816

Cefoxitin

840.125214281

Cefotaxime

<0.03<0.03<0.03<0.03

1<0.03

0.030.060.03

Imipenem

0.250.1250.250.250.1250.1250.1250.060.25

Carb;

0

"8

g5*

•§

a:5"

Dow



ecember 2021

133

Xto

aB

g.B

B D G H I K M N O Q R

Figure I. Isoelectric focusing gel of /f-lactamases of A hydroplnla G19 after staining with nitrocefin. Lane A. A hvdroplula NCTC 8049; B. A hydrophilaNCTC 8049 after 10 min incubation with 0 I M EDTA, C. A. hydrophila G19; D, A hydrophila G19 after 10 min incubation with 0.1M EDTA; E. A.hvdrophila G75; F, A hydrophila G75 after 10 min incubation with O.I M EDTA; G, A. hydrophila 5BI60; H, A. hydrophila 5B160 after 10 min incubationwith 0.1 M EDTA; I. A. hydrophila 9B273: J. A. hydrophila 9B273 after 10 min incubation with 0 1 M EDTA; K, A caviae R4007; L. A. caviae R4O07after 10 min incubation with 0.1 M EDTA. M. A hydrophila R6407. N. A hydrophila R6407 after 10 min incubation with 0.1 M E D T A ; O . / ( caviae R8856;P. A. caviae R8856after 10 min incubation with O.i M EDTA. Q. A hydrophilaC19897. R. A. hydrophilaC19897 after 10 mm incubation with0.1 M EDTA;S, A salmomcida subspp achromogenes ASAll l . T. A salmomcula subspp achromogenes ASAlll after 10 min incubation with 0 1 M EDTA

Dow



ecember 2021

Carbapenemase in A. hydrophila 39

The hydrolysing activity of nitrocefin peaked in the fourth fraction after the start ofthe salt gradient (approximately 0.035 M NaCl) and carbapenemase activity peaked inthe seventh and eighth fractions after being subjected to a salt gradient of 0.13 M NaCl(Figure 2). The carbapenemase demonstrated weak nitrocefin hydrolysis that wasnot inhibited by 0.1 M EDTA, unlike the hydrolysis of imipenem. These fractions werethen reloaded onto the Mono-S column and the carbapenemase activity was againdetected but now there was no nitrocefin hydrolysis. Similarly, the cephalosporinaseactivity which was detected in the fourth fraction after the start of the salt gradientpossessed some carbapenem-hydrolysing activity but this was not found once thefraction had been purified further. Nitrocefin and imipenem hydrolysis had also beenobserved in the initial fractions which contained the proteins that had not bound to thecolumn. This activity would have been due to the /Mactamase with a pi of 7.0 whichcannot bind to the column when it is equilibrated at pH 6.5 and also a result ofoverloading the column. When Tris-dialysed protein was loaded onto the anionexchange column which was equilibrated at pH 8.5, neither the carbapenemase nor the/Mactamase with a pi of 8.0 bound to the column, whereas the /Mactamase with a piof 7.0 did and was eluted by 0.13 M NaCl (Figure 3). The separation of the /Mactamaseswhich could hydrolyse nitrocefin was confirmed by isoelectric focusing (data notshown).

The purified enzymes were assayed for their activity on nitrocefin, imipenem,ampicillin, cephaloridine and cefoxitin. The carbapenemase ACP {Aeromonascarbapenemase) only hydrolysed imipenem and showed no hydrolysis of nitrocefin(Table II). The enzyme was highly resistant to inhibition by clavulanic acid andBRL42715 (Table II) and was sensitive to EDTA but, unusually, it was not inhibitedby low concentrations of zinc.

The /Mactamase with a pi of 8.0 was a penicillinase, APE {Aeromonas penicillinase),although, unlike the other enzymes, its activity on ampicillin did not follow first orderkinetics (Table II). Nitrocefin was the only cephalosporin hydrolysed and the enzymewas extremely sensitive to inhibition by clavulanic acid and BRL42715 (Table II). The/Mactamase isolated on the Mono-Q column was a cephalosporinase, ACE {Aeromonas

o.i

1 2 30

5 fi 7 >>• < ! ' i 1 1 1 2 1 3 M 1 5

Fraction numberFigure 2. Separation of /J-lactamases from A. hydrophila G19 by cation exchange chromatography. D.

Specific activity with imipenem; O, specific activity with nitrocephin; , sodium chloride concentration.

Dow



ecember 2021


0.15 -

CO

0.05 -

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Fraction number

Figure 3. Separation of /Mactamases from A. hzydrophila G19 by anion exchange chromatography. • ,Specific activity with lmipenem; O» specific activity with nitrocephin; , sodium chloride concentration.

cephalosporinase) with a pi of 7.0. The enzyme hydrolysed nitrocefin faster than

cephalondine and cefoxitin (Table II) and possessed a similar IC50 for BRL42715 to that

of APE, but it was not as sensitive to inhibition by clavulanic acid (Table II).

Table II. Substrate profile of three /}-lactamases isolated from A. hydrophila G19

0-lactamase

ACPimpenemnitrocefinampicillincephaloridinecefoxitin

APEimipenemnitrocefinampicillincephaloridinecefoxitin

ACEimipenemnitrocefinampicillincephaloridinecefoxitin

(/jmoles/min/mL)

0.8————

—0.062NF——

—1.25—

NMNM

130————

—5.1

NF——

—16

—NMNM

0.006————

—0.012NF——

—0.075

—NMNM

Relative K j

100208

100

6410

'Vm,, relative to nitrocefin.—, No detectable hydrolysisNF, The hydrolysis of ampicillin did not follow first-order kinetics and thus the determination of

V*u and A« were impossible.NM, Not measurable because the rate of hydrolysis was too low to measure accurately and thus

the determination of k m l and K* were impossible.

Dow



ecember 2021

Carbapenemase in A. hydrophila 41

BTable III. Inhibitor profile of three /Mactamases isolated from A. hydrophila G19

/Mactamase

A C PAPE'ACE*TEM*

Pi

~8.28.07.05.4

Clavulanic acid

2.1 x2.0 x59 x4.3 x

io-3

10-.oio-3

10"'

IC» (M)BRL42715

2.5 x lO"6

3.0 x 10""3.7 x 10""3.3 x 10-'

EDTA

1.8 x IO-4

Zinc

1.5

sulphate

x IO-5

'Reporter substrate was imipenem'Reporter substrate was nitrocefin.—, Not measured.TEM-1 0-lactamase punned from E coli J62-2 (Rl)

Molecular mass determinations. The sizes of ACP, APE and ACE were determinedby SDS-PAGE to be 31,000, 43,250 and 20,000 Da respectively.

Discussion

Initial studies indicated that the majority of A. hydrophila possessed two different/Mactamases which could be separated on an anion exchange chromatography columnequilibrated at pH 8.5. One of the /Mactamases (pi of approximately 7.0) was acephalosporinase whereas the other (pi of approximately 8.0) could hydrolyseimipenem, nitrocefin and ampicillin. These results were similar to those found byBakken et al., (1988), Iaconis & Sanders (1990) and Shannon et al. (1986). Thecarbapenemase was assumed to be a metallo-jS-lactamase belonging to Ambler class B(Ambler, 1980) and it was therefore expected that its activity would be inhibited byEDTA. However, the crude /Mactamase preparations were still visible on an isoelectricfocusing gel stained with nitrocefin despite pre-incubation with the metal ion chelator.After running the extract through a cation exchange column equilibrated at pH 6.5,three peaks were observed. The early fractions contained proteins that had not boundto the column and these hydrolysed both nitrocefin and imipenem. The fourth fractioneluted after the start of the salt gradient (0.035 M NaCl) showed a peak innitrocefin-hydrolysing activity and also exhibited some carbapenemase activity whichfurther purification showed to have been the result of contamination. This /Mactamasewas shown to be a penicillinase with a pi of 8.0. The carbapenemase activity peakedin the seventh and eighth fractions after the start of the salt gradient (0.13 M NaCl).Again, these fractions demonstrated some nitrocefin-hydrolysing activity initially butthis was also found to be the result of contamination once the /Mactamase had beenpurified further. The fraction with maximal enzyme activity was collected from theanion exchange column and contained the /Mactamase with a pi of 6.0 which was foundto be a cephalosporinase.

Thus, A. hydrophila has three different /Mactamases as has been noted by Segatoreet al. (1993) although they reported that the carbapenemase exhibited penicillinase andcephalosporinase activity and slow hydrolysis of nitrocefin. However, as we have shown,the cephalosporinase activity may have been the result of incomplete purification of thecarbapenemase. Moreover, attempts to amplify the gene encoding ACP by thepolymerase chain reaction using primers from the sequence of cphA have failed(unpublished observations) suggesting that these enzymes are not the same. In 1992,

Dow



ecember 2021


seven of the nine isolates possessed carbapenem hydrolysing activity. However, threeyears later, only four retained their carbapenemase activity suggesting that the geneencoding the enzyme had been lost or repressed as a result of subculturing or storingthe strains on antibiotic-free medium.

The discovery of a /Mactamase which does not hydrolyse nitrocefin also explains theambiguous results reported by Iaconis & Sanders (1990) who noted that their A2/Mactamase had vastly different IC50S to clavulanic acid when they used differentreporter substrates. They explained this to be a result either of the clavulanic acidonly allowing the smaller imipenem molecule into the active site of the enzyme or tothe A2 /Mactamase having two different binding sites for the substrates. Our resultssuggest that neither explanation is correct but that the anomaly is, in fact, due to thepresence of two different enzymes, of which only one is detectable using nitrocefin.Rasmussen et al. (1994), have recently cloned a cephalosporinase and an oxacillinasefrom a strain of Aeromonas sobria which had previously been reported to possesscarbapenemase activity (Iaconis & Sanders, 1990). Both of these enzymes weredetectable with nitrocefin and had been previously visualised on an isoelectric focusinggel (Iaconis & Sanders, 1990), but the cephalosporinase had only minimalcarbapenemase activity. Rasmussen et al. (1994) proposed that the carbapenemasedetected in the original strain had been lost during storage, which again supportsour proposition that there are three /Mactamases present in many Aeromonasspp. Similarly, a cephalosporinase and a penicillinase were cloned from A. sobria intoEscherichia colt but the carbapenemase which had been detected in the original straincould not be cloned (Walsh et al., 1995). The non-motile species A. salmonicida subsp.achromogenes was also found to possess three /Mactamases one of which onlyhydrolysed imipenem (Hayes et al., 19946) which suggests either genetic transferbetween the different Aeromonas species or that all of the species of the Aeromonas genuspossess a gene for intrinsic resistance which codes for this highly unusualcarbapenemase.

Our data suggest that many strains of Aeromonas spp. possess these three/Mactamases. As many researchers have only detected two /Mactamases in this genus(for instance Bakken et al., 1988; Iaconis & Sanders, 1990; Shannon et al., 1986), wemust question whether the detection of carbapenemases in other genera has been grosslyunderestimated, firstly, by only looking for carbapenemases in bacteria resistant toimipenem and secondly, by assuming that all carbapenemases can be detected bynitrocefin hydrolysis.

We speculate that many environmental strains have the potential to code for acarbapenemase even though they have probably never been exposed to a carbapenemor possibly to any other antibiotic apart from naturally occurring penicillins andcephalosporins. Iaconis and Sanders selected imipenem-resistant mutants of a strain ofA. hydrophila and a strain of A. sobria by exposing them to sub-MICs of imipenem orcefoxitin (Iaconis & Sanders, 1990), so it is conceivable that other strains of Aeromonasspp. may become resistant to imipenem if the appropriate selective pressure is applied.Consequently, it is important for clinicians to realise the potential threat that thesehidden carbapenemases could pose to patients being treated with imipenem for infectiondue to Aeromonas spp. Laboratories should also be aware that the detection of/Mactamases with nitrocefin is not completely reliable and that, unless appropriatetechniques are applied, /Mactamase activity may go undetected in Aeromonas spp aswell as in other species of bacteria.

Dow



ecember 2021

Carbapenetnase in A. hydrophila 43

Acknowledgement

We wish to thank the Biotechnology and Biological Sciences Research Council for thegrant to MVH. We are also extremely grateful to Dr A. Cronshaw and the BiochemistryDepartment for the use of their FPLC equipment.

References

Ambler, R. P. (1980). The structure of /?-lactamases. Philosophical Transactions of the RoyalSociety of London—Series B 289, 321-31.

Bakken, J. S., Sanders, C. C , Clark, R. B. & Hori, M. (1988). 0-Lactam resistance in Aeromonasspp. caused by inducible /Mactamases active against penicillins, cephalosponns, andcarbapenems Antimicrobial Agents and Chemotherapy 32, 1314—9.

Davis, B. D. & Mingioli, E. S. (1950). Mutants of Escherichia coli requiring methionine or vitaminB12. Journal of Bacteriology 60, 17-28.

George, W. L., Nakata, M. M., Thompson, J. & White, M. L. (1985). Aeromonas-rdatcd diarrheain adults. Archives of Internal Medicine 145, 2207-11.

Gracey, M., Burke, V. & Robinson, J. (1982). Aeromonas-associaled gastroenteritis. Lancet ii,1304-6.

Hayes, M. V., Thomson, C. J. & Amyes, S. G. B. (1994a). Carbapenemase in Aeromonassalmonicida. In Abstracts of the 94th Annual Conference of the American Society forMicrobiology, Las Vegas 1994. Abstract A63. p. 13, American Society for Microbiology,Washington, DC.

Hayes, M. V., Thomson, C. J. & Amyes, S. G. B. (1994ft). Three ^-lactamases isolated fromAeromonas salmonicida, including a carbapenemase not detectable by conventional methods.European Journal of Clinical Microbiology and Infectious Diseases 13, 805-11.

laconis, J. P. & Sanders, C. C. (1990). Purification and characterization of inducible /!-!actamasesin Aeromonas spp. Antimicrobial Agents and Chemotherapy 34, 44—51.

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head ofbactenophage T4. Nature 227, 680-5.

Lineweaver, H. & Burke, D. (1934). The determination of enzyme dissociation constants. Journalof the American Chemical Society 56, 658-66.

Massidda, O., Rossolini, G. M & Satta, G. (1991). The Aeromonas hydrophila cphA gene:molecular heterogeneity among class B metallo-/3-lactamases. Journal of Bacteriology 173,4611-7.

Matthew, M., Harris, A. M., Marshall, M. J. & Ross, G. W. (1975). The use of analyticalisoelectric focusing for detection and identification of /?-lactamases. Journal of GeneralMicrobiology 88, 169-78.

National Committee for Clinical Laboratory Standards. (1993). Methods for DilutionAntimicrobial Susceptibility Tests for Bacteria that Grow Aerobicallv. M7-A3, 3rd Edn.NCCLS, Villanova, PA.

Pazzaglia, G., Escalante, J. R., Sack, R. B., Rocca, C. & Benavides, V. (1990). Transient intestinalcolonisation by multiple phenotypes of Aeromonas species during the first week of life.Journal of Clinical Microbiology 28, 1842-6.

Picard, B. & Goullet, P. (1987). Seasonal prevalence of nosocomial Aeromonas hydrophilainfection related to aeromonas in hospital water. Journal of Hospital Infection 10, 152-5.

Rasmussen, B. A., Keeney, D., Yang, Y. & Bush, K. (1994). Cloning and expression of acloxacillin-hydrolysing enzyme and a cephalosporinase from Aeromonas sobria AER 14Min Escherichia coli: requirement of an Escherichia coli chromosomal mutation for efficientexpression of the class D enzyme. Antimicrobial Agents and Chemotherapy 38, 2078-85.

Reid, A. J. & Amyes, S. G. B. (1986). Plasmid penicillin resistance in Vibrio cholerae:identification of new /J-lactamase SAR-1. Antimicrobial Agents and Chemotherapy 30, 245-7.

Sawle, G. V., Das, B. C , Acland, P. R. & Heath, D. A. (1986). Fatal infection with Aeromonassobria and Pleisiomonas shigelloides. British Medical Journal 292, 525—6.

Segatore, B., Massidda, O., Satta, G., Setacci, D. & Amicosante, G. (1993). High specificity ofcphA -encoded metallo-/?-lactamase from Aeromonas hydrophila AE036 for carbapenems andits contribution to /f-lactam resistance. Antimicrobial Agents and Chemotherapy 37, 1324-8.

Dow



ecember 2021

44 M. V. Hayes et at.

Shannon, K., King, A. & Phillips, I. (1986). /?-Lactamases with high activity against imipenemand Sch 34343 from Aeromonas hvdrophila. Journal of Antimicrobial Chemotherapy 17,45-50.

Waddell, W. J. (1956). A simple ultraviolet spectrophotometric method for the determination ofprotein. Journal of Laboratory and Clinical Medicine 48, 311—4.

Walsh, T. R., Payne, D. J., MacGowan, A. P. & Bennett, P. M. (1995). A clinical isolate ofAeromonas sobria with three chromosomally mediated inducible /J-lactamases: acephalosporinase, a penicilhnase and a third enzyme diplaying carbapenemase activityJournal of Antimicrobial Chemotherapy 35, 271-9.

{Received 22 November 1994; returned 14 December 1994; revised 26 January 1995;accepted 1 August 1995)

Dow



ecember 2021

Documents

carbapenemase of Aeromonas hydrophila - Journal of Antimicrobial