New trends in enterobacterial beta-lactamases
P Nordmann, Service de Microbiologie, Faculté
de Médecine Paris-Ouest, Garches, France; T Naas, Abteilung Mikrobiologie, Biozentrum der Universität,
Basel, Switzerland; and R Labia, Laboratoire de Biochimie, Museum d'Histoire Naturelle, Paris, France
In the last four years, numerous beta-lactamases have been described
in enterobacterial species. These enzymes included carbapenemases, plasmid-mediated cephalosporinases, inhibitor-resistant
penicillinases and novel extended spectrum beta-lactamases. Their discovery may change the potential benefit of
novel beta-lactam for clinical therapy.
Carbapenem-hydrolysing beta-lactamases or carbapenemases
Imipenem (the only commercially-available carbapenem) and meropenem are broad-spectrum beta-lactams. Their activities
are not significantly inhibited by the commonly found enterobacterial beta-lactamases. AmpC cephalosporinases from
Enterobacter sp., Serratia sp., Citrobacter freundii, TEM and SHV extended-spectrum derivatives commonly found
in Escherichia coli, Klebsiella pneumoniae and Proteus mirabilis and class A beta-lactamases found in P. rettgeri
and P. penneri , K. oxytoca and C. diversus cannot alone lead to a clinically significant decrease in carbapenem
However, acquired carbapenem resistance has been reported in E. cloacae, E. aerogenes and P. rettgeri. It results
from a sharp decrease in outer-membrane permeability to carbapenems combined with an overproduction of AmpC cephalosporinase
(1-4). These strains have mainly been isolated from patients treated with moxalactam (a cephamycin), others resulted
from in-vitro moxalactam selection. Carbapenems do not seem to select easely for such strains. These strains are
not widespread; they represent less than 1% of clinical isolates in France(unpublished data).
This type of carbapenem-resistance mechanism may not be limited to overproduction of cephalosporinase. We have
recently analyzed one clinical strain in which decrease in outer membrane permeability (or lipopolysaccharide alteration)
combined with a an overproduction of broad spectrum beta-lactamase (TEM or SHV derivatives) lead to a moderated
level of carbapenem resistance (unpublished data).
More interesting, a few clinical enterobacterial strains have been isolated which produce biochemically-characterized
carbapenemases: two Serratia marcescens isolates, S6 and S8 in England found in 1982 and published in 1990 (5);Enterobacter
cloacae, strain NOR-1 in France found in 1990 and published in 1993 (6) and very recently published in 1994 (7)
a third Serratia marcescens, strain TN9106 found in 1993 in Japan. S. marcescens S6 and S8 produce a similar pI
9.7 beta-lactamase. This enzyme has been purified; it hydrolyzes imipenem but to a lesser extend meropenem, weakly
extended-spectrum beta-lactams such as cefotaxime and ceftazidime but significantly aztreonam. As many penicillinases,
it also strongly hydrolyzes benzylpenicillin, ticarcillin, mezlocillin and cephalothin. Moxalactam and cefoxitin
are not hydrolyzed by this enzyme. Its activity is weakly inhibited by clavulanic acid or tazobactam but not by
Its gene is chromosomally located. Recently, this gene has been cloned and expressed in E. coli. Sequence analysis
revealed that this novel enzyme named Sme-1 is a class A beta-lactamase which shares close to 50% homology with
the commonly found enterobacterial class A beta-lactamases such as TEM or SHV derivatives (P. Nordmann, unpublished
data). S. marcescens S6 and S8 may be special enterobacterial strains which naturally possess a carbapenemase gene
no evidence for a transposon sequence or inverted repeats was found on either side of the beta-lactamase gene.
At the 1990 ICAAC meeting (9), other S. marcescens strains were reported to produce carbapenemase. However, no
detailed biochemical analysis of these enzymes was made, so that we are not able to compare them with Sme-1 enzyme.
The second enterobacterial carbapenemase reported has been described from E. cloacae isolated from a polywounded
patient treated with one dose (500 mg) of imipenem (6). This enzyme named NmcA has been biochemically characterized.
It is a pI 6.9 beta-lactamase of 30 kDa. It hydrolyzes strongly benzylpenicillin, ticarcillin, imipenem more than
meropenem and aztreonam. Extended-spectrum beta-lactams such as cefotaxime, ceftriaxone, ceftazidime and cephamycins
(cefoxitin and moxalactam) are not significantly hydrolyzed. Its activity is partially inhibited by clavulanic
acid or tazobactam but not by EDTA. Therefore, its biochemical properties are very similar to those of Sme-1. As
described for AmpC cephalosporinase in E. cloacae, the synthesis of NmcA carbapenemase is inducible by imipenem
or cefoxitin. Therefore in E. cloacae NOR-1, cefoxitin and imipenem induce both carbapenemase and AmpC cephalosporinase.
Moreover, detailed analysis of MICs of beta-lactams for E. cloacae NOR-1 and E. coli JM109 harboring a recombinant
plasmid which contains the NmcA gene indicates that E. cloacae NOR-1 may also have decreased outer-membrane permeability
for carbapenems. The localization of the NmcA gene is chromosomal as described for Sme-1.
NmcA sequence analysis revealed that it is weakly related to TEM, SHV or CARB beta-lactamases although being a
class A beta-lactamase (10). It is strongly related to Sme-1 with which it shares 70% amino-acid identity and 80%
amino-acid homology. However, NmcA is not a point mutant derived from Sme-1 as described for TEM or SHV extended-spectrum
beta-lactamases which are point-mutant derivatives of TEM-1/TEM-2, or SHV-1 restricted-spectrum beta-lactamases.
Sme-1 and NmcA may represent a new class A beta-lactamase subgroup which is related to class A beta-lactamases
described in Yersinia enterocolitica, Klebsiella oxytoca and Citrobacter diversus. Detailed analysis of conserved
amino-acids at specific positions indicates that the above quoted beta-lactamases, NmcA and Sme-1 are all chromosomally
located as opposed to CARB, TEM and SHV derivatives which may be either chromosomal, or plasmid and or transposon-located.
This result may indicate that carbapenemase class A beta-lactamase have low probability to spread among enterobacterial
isolates which is usually correlated with a plasmid location.
Analysis of upstream DNA sequences of nmcA revealed presence of another open reading frame which encodes a regulator
NmcR (10). This 33-kDa protein belongs to the LysR-type regulator family. This regulator family includes AmpR regulators
which modulate AmpC cephalosporinase synthesis as described in E. cloacae, C. freundii, Y. enterocolitica and P.
aeruginosa. AmpR acts as negative regulator in the absence of inducers and as a positive regulator in the presence
of beta-lactam inducer such as cefoxitin and imipenem. On the opposite, NmcR acts as a strong positive regulator
in the absence of inducer and as a weak positive regulator in the presence of cefoxitin or imipenem. This system
of class A beta-lactamase regulation is the first described in enterobacterial species and differs from the one
of AmpC cephalosporinase.
Epidemiological analysis of NmcA clinical distribution has not found this gene among 500 gram-negative clinical
isolates (P. Nordmann, unpublished data). Therefore, it is unlikely that this gene is widespread or unexpressed
among gram-negative bacteria.
Very recently, a novel carbapenem-resistant S. marcescens has been reported in Japan (17). Strain TN109 encodes
a chromosomally-located carbapenemase (IMP-1) which hydrolyzes imipenem and meropenem, ampicillin, ceftazidime,
cefotaxime, moxalactam but not aztreonam. Its activity is not inhibited by clavulanic acid but by EDTA. Sequence
analysis indicated that, as opposed to the previously described enterobacterial carbapenemases, it is not a class
A beta-lactamase but a class B metallo-enzyme. This enzyme shows homology with the sequenced metallo-carbapenemases
reported from Aeromonas hydrophila, Bacillus cereus and Bacteroides fragilis (7, 11). This indicates that in S.
marcescens TN109 the chromosomal location of bla IMP-1 could have resulted from an intergeneric DNA transfer. Such
finding is very uncommon for class B metallo-enzymes which were so far considered as stable chromosomally-located
enzymes as opposed to class A beta-lactamases. Clonal dissemination of such strain may lead to therapeutic problems
such as extended-spectrum cephalosporin and carbapenem hydrolysis. However, it is interesting to note that imipenem
and extended-spectrum beta-lactams such as ceftazidime have been introduced into the north american and european
markets at about the same period, around 1985. Almost ten years later, clinical strains resistant to extended-spectrum
cephalosporins have been extensively described. In France, 10 to 20% of Klebsiella pneumoniae isolated in university
hospitals possess an extended-spectrum beta-lactamase of TEM or SHV series (12). In comparison, carbapenem-resistance
remains much more limited among Enterobacteriaceae.
Until recently, AmpC cephalosporinases have been recognized as naturally-present and chromosomally-located in various
enterobacterial species. However, plasmid-mediated cephalosporinases such as MIR-1 and CMY-2 have been reported
in the U.S.A. and in Europe (13, 14). These enzymes hydrolyze strongly ampicillin, ticarcillin and weakly ticarcillin,
cephamycins, extended-spectrum cephalosporins but not carbapenems. Their expression is constitutive as opposed
to the inducible expression of the chromosomally-encoded cephalosporinases. MIR-1 and CMY-2 amino-acid homology
are hugh with AmpC cephalosporinase from E. cloacae and from C. freundii. respectively. Similarly, CMY-1, BIL-1
and LAT-1 as well as an enzyme recently described in K. pneumoniae possess cephalosporinase properties (14, 15,
16, 17). However, their amino-acid sequence has yet not been reported.
Dissemination of these enzymes among clinical enterobacterial strains especially those not possessing a chromosomally-located
cephalosporinase (E. coli, P. mirabilis, K. pneumoniae) may lead to resistance to extended-spectrum cephalosporins.
As opposed to extended-spectrum beta-lactamases of TEM and SHV series, the enzyme activities are not inhibited
by clavulanic acid, tazobactam or sulbactam. These inhibitors are therefore not able to counteract the activity
of these enzymes, especially when prescribed for treating urinary tract infections.
TEM and SHV series beta-lactamases are widely disseminated in enterobacterial species. TEM-1/TEM-2 and SHV-1 hydrolyzed
significantly ampicillin, amoxycillin and ticarcillin. However, their effect could be, until recently, counteracted
by clavulanic acid or sulbactam addition. These antibiotic combinations were therefore widely prescribed for treating
urinary tract infection in non-hospitalized patients.
However, several inhibitor-resistant penicillinases (TRI) have been reported recently in E. coli isolates in France,
Great-Britain, Spain and Greece (18, 19, 20, 21, 22). Although differing from the parent beta-lactamases by a few
amino-acid changes, the reasons for which these TEM derivatives are resistant to inhibitors are still unknown.
At the last 1993 ICAAC meeting, in vitro inhibitor-resistant penicillinases have been obtained in E. coli. Some
authors have reported that such enzymes may be present in as high as 20% of E. coli isolates in France (Paris and
Clermont-Ferrand, unpublished data). However, it is difficult to evaluate the dissemination of TRI enzymes since
they may give similar patterns of antibiotic resistance as TEM-1/-2 or SHV-1 beta-lactamase overexpression. Extended-spectrum
beta-lactamases of TEM and SHV series resistant to inhibitors have not been described so far. Inhibitor-resistant
beta-lactamases have not yet been reported in H. influenzae or in Staphylococcus aureus. The outcome of class A
inhibitor-resistant penicillinases may abolish the beneficial effect of inhibitors widely prescribed in urinary
tract, respiratory tract or skin infections.
Novel extended-spectrum beta-lactamases (ESBLA)
Since 1983, extended-spectrum beta-lactamases of TEM and SHV series have been extensively described in enterobacterial
strains (23). A recent study indicates that they are especially well distributed in K. pneumoniae (10-20%). A tremendous
amount of plasmid-mediated enzymes has been reported often differing by single amino-acid changes. All the enterobacterial
species may possess such enzymes. They have recently been described in Proteus mirabilis where the plasmidic TEM-3
gene was found to be specific to this species (25, 26). Spontaneous mutations occurring in TEM/SHV restricted spectrum
beta-lactamases harbored on low-frequency mating plasmids may be another source of dissemination in some enterobacterial
ESBLA hydrolyze significantly, but to various extent, cefotaxime, ceftriaxone and ceftazidime and usually aztreonam
at a low level. Recently TEM-22 has been reported in K. pneumoniae hydrolyzing significantly aztreonam has not
yet been published (27). Description of MEN-1, an ESBLA from E. coli (28) and a novel ESBLA from K. oxytoca (8)
show that these ESBLA are not limited to TEM and SHV derivatives. As for SHV derivatives, MEN-1 may have originated
from Klebsiella spp. but is poorly related to TEM and SHV derivatives. Both TEM-1 and SHV-1 have been reported
as transposon-located. However, clear evidence of ESBLA transposon location is lacking but seems credible for TEM
Since 1990, novel and powerful beta-lactamases among enterobacterial beta-lactamases were described, which indicates
that the enzymes for resistance to all commercially available beta-lactams now exist.
However, their diffusion is still limited. It is likely that the small incidence of these enzymes may result from
an increase in hygiene measures and alternance in prescription of beta-lactam therapy at a same hospital site.
As is well known, treatment for adequate periods of time may also limit emergence of novel mechanisms of resistance.
Undoubtfully, Enterobacteriaceae which are the main source of nosocomial infections possess efficient adaptative
properties due to their genetic plasticity. Recently, cefepime and cefpirome (extended-spectrum cephalosporins)
have been introduced on the antibiotic market. Their potentiality to counteract the in vivo effects of ESBLA still
remains to be shown.
In view of the difficulties to discover new beta-lactams, it is questionable whether, within the next few years,
widespread distribution of powerful beta-lactamases may lead to major problems for treating nosocomial infections.
- Lee E. H., Nicolas M. H. , Kitzis M. D., Pialoux G.,
Collatz E., Gutmann L. Antimicrob Agents Chemother 1991;35:1093-1098.
- Raimondi A., Traverso A., Nikaido H. Antimicrob Agents
- Tzouvelekis L.S., Tzelepi E., Mentis A.F., Vatopoulos
A.C., Tsakris A. FEMS Microbiol Lett 1992;95:195-200.
- De Champs C., Henquell C., Guelon D., Sirot D., Gazuy
N., Sirot J. J Clin Microbiol 1993; 31:123-127.
- Yang Y., Wu P., Livermore D.M. Antimicrob. Agents Chemother
- Nordmann P., Mariotte S., Naas T., Labia R., Nicolas
M.H. Antimicrob Agents Chemother 1993;37:939-946.
- Osane E., Arakawa Y., Wachasotayankun R., Ohta M.,
Horii T., Ito H., Yoshimura F., Kato N. Antimicrob Agents Chemother 1994; 38:71-78.
- Wu S. W., Dornbusch K., Norgren M., Kronvall G. J Antimicrob
- Medeiros A.A., Hare R.S. Progr Abstr Intersci Conf
Antimicrob Agents Chemother 1986, abst 116.
- Nordmann P., Naas T. Progr Abstr Intersci Conf Antimicrob
Agents Chemother 1993, abst 586.
- Livermore D.M. Carbapenemases. J. Antimicrob Chemother
- Sirot D.L., Goldstein F.W., Soussy C.J., Courtieu A.L.,
Husson M.O.,Lemozy J., Meyran M., Morel C., Perez R., Quentin-Noury C., Reverdy M.E., Scheftel J.M. Rosenbaum M.,
Rezvani Y. Antimicrob Agents Chemother 1992;36:1677-1681.
- Papanicolaou G. A., Medeiros A. A., Jacoby G. A. Antimicrob
Agents Chemother 1990;34:2200-2209.
- Bauernfeind A., Mangold P., Schweighart S., Miller
G.H., Shaw K.J., Giamarellov H., Dornbush K. Progr Abstr Intersci Conf Antimicrob Agents Chemother 1992, abst 1268.
- Payne D. J. , Woodford N. , Amyes S. G. B. J Antimicrob
- Tzouvelekis L.S., Tzelepi E., Mentis A.F., Tsakris
A. J Antimicrob Chemother 1993; 31: 645-654.
- Horii T., Arakawa Y., Ohta M., Ichiyama S., Wacharotayankun
R., Kato N. Antimicrob Agents Chemother 1993; 37:984-990.
- Vedel G., Belaaouaj L., Gilly L. Labia R., Philippon
A., Nevot P., Paul G. J Antimicrob Chemother 1992;30;449-462.
- Thomson C.J., Amyes S.G.B. J Antimicrob Chemother 1992;
- Thomson C.J., Amyes S.G.B. FEMS Microbiol Lett 1992;
- Blazquez J., Baquero M. R., Canton R., Alos I., Baquero
F. Antimicrob Agents Chemother 1993; 37: 2059-2063.
- Tzouvelekis L.S., Tzelepi E., Mentis A.F., Tsakris
A. J Antimicrob Chemother 1993;31:645-654.
- Jacoby G.A., Medeiros A.A. Antimicrob Agents Chemother
- Mariotte S., Nordmann P., Nicolas M.H. J Antimicrob
Chemother, in press.
- Espinasse F., Mariotte S., Labia R., Nicolas M.H. Progr
Abstr Intersci Conf Antimicrob Agents Chemother 1992, abst 1272.
- Rossi M.A., Gutkind G., Quinteros M., Marino E., Couto
E., Tokumoto M., Woloj M., Miller G., Medeiros A. Progr Abstr Intersci Conf Antimicrob Agents Chemother 1990, abst
- Arlet G., Rouveau M., Fournier G., Lagrange P. H.,
Philippon A. Antimicrob Agents Chemother 1993;37:2020-2023.
- Barthélémy M., Péduzzi J. , Bernard
H., Tancrède C., Labia R. Biochem Biophys Acta 1992;1122:15-22.