Beta-lactam antibiotics: from antibiosis to resistance …szabojudit.web.elte.hu/antibiotikum...

Review Article

Beta-lactam antibiotics: from antibiosis to resistance

and bacteriology

KOK-FAI KONG1, LISA SCHNEPER2 and KALAI MATHEE2

1Department of Biological Sciences, College of Arts and Sciences; and 2Department of Molecular Microbiologyand Infectious Diseases, College of Medicine, Florida International University, Miami, FL, USA

Kong K-F, Schneper L, Mathee K. Beta-lactam antibiotics: from antibiosis to resistance and bacteriol-ogy. APMIS 2010; 118: 1–36.

This review focuses on the era of antibiosis that led to a better understanding of bacterial morphology,in particular the cell wall component peptidoglycan. This is an effort to take readers on a tour de forcefrom the concept of antibiosis, to the serendipity of antibiotics, evolution of beta-lactam development,and the molecular biology of antibiotic resistance. These areas of research have culminated in a deeperunderstanding of microbiology, particularly in the area of bacterial cell wall synthesis and recycling. Inspite of this knowledge, which has enabled design of new even more effective therapeutics to combatbacterial infection and has provided new research tools, antibiotic resistance remains a worldwidehealth care problem.

Key words: Penicillin; beta-lactam; penicillin-binding protein; Alexander Fleming; beta-lactamases;amp genes.

Kalai Mathee, Department of Molecular Microbiology and Infectious Diseases, Florida InternationalUniversity, Modesto A. Maidique Campus, Miami, FL 33199, USA. e-mail: [email protected]

Present address: Kok-Fai Kong, Department of Cell Biology, La Jolla Institute for Allergy and Immu-nology, La Jolla, CA, USA.

The concept of using chemicals to alleviatediseases, particularly infectious diseases, datesback to Ancient Egypt, Babylon, the Far Eastand the Incas (1). These early examples ofremedies include applying molds to open cutsand wounds, and eating radishes, leeks, garlicand onions that are now known to be antibacte-rial.The possibility of one organism interfering

with the growth of another has been thesubject of intense investigation since the earlydays of microbiology. In 1871, Sir JohnBurdon-Sanderson observed that mediaexposed to air rapidly became turbid withbacteria, but if a Penicillium mold happened

to grow on the surface of the broth, lessturbidity ensued (2). Others, including WilliamRoberts, John Tyndall, and Joseph Lister,reported similar observations with Penicillium(3–5). Although perhaps only Roberts andLister realized that their findings could be dueto the action of an antibacterial compound,these observations are noteworthy nonetheless(6). One of the earliest references to the thera-peutic potential of the inhibition of bacterialgrowth by other microorganisms was in Pasteurand Joubert’s description of the inhibition ofanthrax by air-borne organisms (7). Furtherevidence that Lister understood the clinical rele-vance of the antagonism is his reported treat-ment of infected wounds with Penicillium in1884 (6).Received 11 October 2009. Accepted 26 October 2009

APMIS 118: 1–36 � 2009 The Authors

Journal Compilation � 2009 APMIS

DOI 10.1111/j.1600-0463.2009.02563.x

1

PRE-PENICILLIN ERA

Although many observed microbial antago-nism, the concept that the phenomenon wascaused by the production of a compound by onemicroorganism that could kill another was notimmediately apparent. For example, in a laterwork, Tyndall surmised that the bacterialgrowth inhibition observed in the presence ofPenicillium was due to limiting oxygen condi-tions (8). Initial work by Corneil and Babes,and substantiated by experiments by Garre,firmly established that microbial antagonismwas caused by the action of a diffusible sub-stance produced by one organism on another (9,10). By the end of the 19th century, not onlywas this phenomenon well accepted, but it wasalso given a name – antibiosis (11).A few researches showing the contradictory

effects between Penicillium and bacteria predatethe now famous studies of Alexander Flemingand received little attention until after the clini-cal importance of penicillin was established.While completing his doctoral degree in 1897,Ernest Duschesne found antagonism betweenPenicillium and Escherichia coli in vitro. He thenfound that co-injection of a normally lethal doseof Salmonella Typhi, the causative agent oftyphoid fever, with Penicillium glaucum pre-vented the animals from contracting typhoid(12).Another notable pioneer was Gratia, a

Belgian scientist who did ground-breakingresearch on the phenomenon of bacteriolysisand its application to the defeat of bacterialpathogens (13). In classic experiments per-formed in 1925–1926, Gratia and his assistants

exposed a 2% water agar plate containing deadStaphylococcus to the laboratory air. A cultureof white actinomycete grew on the plate sur-rounded by a clear zone of dissolved bacteria.This airborne contaminant was next demon-strated to attack killed cultures of Pseudomonasaeruginosa, Mycobacterium tuberculosis andE. coli. Gratia called the lytic agent ‘mycoly-sate’. The mycolysate was non-toxic whenintroduced into guinea pigs and rabbits. Inaddition, he used the mycolysate on humansubjects in 1930 and obtained considerablesuccess (13).Fleming therefore, was not the first to observe

antibiosis between the fungus and bacteria in avessel, but without controversy, he was the firstto study the substance he termed ‘penicillin’.

BETA-LACTAM DISCOVERY AND

DEVELOPMENT FOR CLINICAL USE

Discovery of penicillin

The story began in 1921. Alexander Fleming(Fig. 1), a Scottish bacteriologist, sufferingfrom a common cold, inoculated agar plateswith his own nasal secretion to determine thechange in his nasal bacterial flora (14). Nocolonies appeared for several days. Thissuggested the presence of a diffusible sub-stance in the nasal secretion that affected theability of bacteria to grow at all. The sub-stance, a protein Fleming named lysozyme,was found to be a potent enzyme capable ofdissolving the cell wall and causing lysis inmany Gram-positive bacteria. The discovery

Sir Alexander Fleming(1881–1955)

Sir Howard W.Florey(1898–1968)

Ernst Boris Chain(1906–1979)

Fig. 1. Antibiotic hall of fame. Courtesy of Nobel e-Museum (http://www.nobel.se/medicine/laureates/1945/index.html).

KONG et al.

2 � 2009 The Authors Journal Compilation � 2009 APMIS

of lysozyme paved the path for the laterserendipitous discovery of penicillin.In the summer and early fall of 1928, Fleming

was studying the relationship between colonymorphology of Staphylococcus and theirvirulence (14). Before leaving for vacation, heinoculated culture plates with Staphylococcuscolonies and stacked the plates on the corner ofhis laboratory bench. When he returned, hefound several cultures contaminated withmolds. He discarded the contaminated platesinto a Lysol basin. He worked with many cul-tures that day, and a few culture plates restedabove the level of the liquid antiseptic, escapingthe disinfectant. The next day, in the process ofdescribing his experiments to a colleague, Flem-ing dug up some previously discarded cultureplates (14). Upon re-examination, one of theplates contained a contaminating mold whosepresence seemed to be influencing the morphol-ogy of the surrounding Staphylococcus colonies:colonies in proximity to the mold were transpar-ent and seemed to be undergoing lysis. Thisobservation was reminiscent of what Fleminghad previously seen with lysozyme, and led himto realize that the mold contained a microbialantagonistic property.Fleming carefully sub-cultured and preserved

the Penicillium notatum, which he originallydescribed as Penicillium rubrum. He describedthe growth property of the mold, elaborated anextraction procedure, and designed a rapidantimicrobial assay (15). Fleming, togetherwith his assistants Dr. Stuart Craddock andMr. Frederick Ridley, set out to purify the lyticagent, which he dubbed ‘penicillin’, releasedinto the broth by the mold (16). According toCraddock’s notebook, Craddock and Ridleyfound that activity could be recovered frombroth and concentrated by evaporation as longas the pH was kept at 6.9 or below and that theactivity could be extracted with ether, acetone,or alcohol (16). Serial dilutions of these extractsinhibited growth of Staphylococcus, Streptococ-cus pyrogenes and Pneumococcus cultures.However, the active ingredient did not harmE. coli, Haemophilus influenzae, SalmonellaTyphi, P. aeruginosa, Bacillus proteus or Vibriocholerae and thus was found to be useful in cul-turing these organisms. The extract was alsoapplied directly to animal tissues, specificallywounds and human eyes, and found to be

neither toxic nor irritating (15). These legend-ary experiments formed the foundation for thesubsequent purification of penicillin and aneasy antimicrobial testing technique which isbeing widely used today. However, Fleming didnot extend his work to clinical study because hewas not able to purify enough penicillin for theexperiments. So, while the discovery was madein 1928, the use of penicillin as a therapeuticagent to treat infections did not happen untilthe 1940s.

Bringing penicillin to bedside

In 1939, 12 years after Fleming’s ground-break-ing discovery, Howard Walter Florey (Fig. 1),Head of the Sir William Dunn School of Pathol-ogy, Oxford University, England, obtained a$25 000 research grant for a period of five yearsfrom the Rockefeller Foundation to investigatethe chemical and biological properties of anti-bacterial substances produced by molds andother bacteria (17). He recruited Ernst Chain todevelop the biochemical methodologies (Fig. 1).Among the molds, they selected Fleming’sstrain of P. notatum for further study because ofits potent antimicrobial activity reported byFleming (16). Florey, Chain and eight co-work-ers became legendary for the progress made onFleming’s penicillin in a surprisingly short per-iod. In August 1940, they reported successfuluse of penicillin to cure infections in mice, rats,and cats (18). A year later, the first clinical useof penicillin in 10 human subjects suffering fromStaphylococcus aureus infection was describedextensively (19).With the initial success, the Oxford group was

prepared to enter the next phase, a large-scaleclinical trial involving 100 patients. However,these experiments required an enormousamount of penicillin, which was impossible tomanufacture in a small laboratory. Hence, Flo-rey and Heatley, another group member,crossed the Atlantic Ocean to the United Statesof America in June 1941 to seek aid in mass pro-ducing penicillin (20). Through several contacts,they were introduced to the new large-scalefermentation laboratory, the Northern RegionalResearch Laboratory of the Department ofAgriculture at Peoria, Illinois (20). In Peoria,they were able to isolate a novel Penicilliumstrain, Penicillium chrysogenum, that produced

BETA-LACTAM RESISTANCE

� 2009 The Authors Journal Compilation � 2009 APMIS 3

substantially more penicillin than previouslyused strains. This work began to stimulate theinterest of several pharmaceutical firms, amongothers, Charles Pfizer & Co., E.R. Squibb &Sons, and Merck & Co. as well as the govern-ment (17, 21). The strong monetary investmentfrom these firms enabled the mass production ofpenicillin. This led to multiple large-scale clini-cal trials in England and in the United States inlate 1942 that continued into 1943 (22–27). By1946, a year after the Nobel Prize for Medicinewas awarded to Fleming, Florey, and Chain,penicillin was finally available in the openmarket.

Penicillin as b-lactam

The work at the Northern Regional ResearchLaboratory also revealed that different strains,culture conditions, and media result in theproduction of different penicillin compounds.Penicillin from the United States strain wassuccessfully crystallized and analyzed at E.R.Squibb & Sons (17) and Florey’s group subse-quently crystallized the compound from theEnglish strain (28). Florey and colleagues demon-strated that their strain used for the English clini-cal trials was primarily 2-pentenylpenicillin(PenicillinForI)while theUnitedStatespenicillin(Penicillin G or Penicillin II) was mainly benzyl-penicillin (29). It was later shown that both formsofpenicillincontainedb-lactamrings (Fig. 2).

b-Lactam derivatives

With the diverse chemical structure of b-lactammolecules and their respective antibacterialpotency, chemists were eager to look beyondwhat nature had supplied, and how sciencecould contribute and substitute. Early work

using different fermentation conditions was notvery successful (30). The b-lactam nucleus,6-aminopenicillanic acid (6-APA; Fig. 2),proved to be the key in penicillin synthesis andmodification. J.C. Sheehan produced PenicillinV by acylation of chemically synthesized 6-APA(31). This important piece of work opened upthe floodgate where novel b-lactam agents couldbe mass-produced by adding unusual sidechains to 6-APA. Thereafter, semi-syntheticb-lactam compounds have been developed con-tinuously and systematically. For example,methicillin, approved for clinical use in the Uni-ted States in 1960, was the first semi-syntheticpenicillin introduced to resist hydrolysis by pen-icillinase (32). More significantly, a semi-syn-thetic compound effective against P. aeruginosa,carbenicillin, a penicillin analog with a carboxylgroup in place of amino group of ampicillin,was introduced in 1967 (33).Though the generation of semi-synthetic com-

pounds presented a great opportunity, naturalsources continued to be explored. Abraham andNewton isolated cephalosporin C from a strainof Cephalosporium acremonium (34). This com-pound generated an entirely new family of b-lac-tam antibiotics because, instead of 6-APA, itpossesses a nucleus of 7-aminocephalosporanicacid (7-ACA; Fig. 2). Using 7-ACA as the pre-cursor, several generations of cephalosporinswith potent broad-spectrum activity have beensynthesized. Many related naturally occurringcompounds with important antibiotic activitywere isolated from bacteria. These includecarbapenem from Streptomyces spp. (35), andmonobactam from Pseudomonas acidophila(36), Chromobacterium violaceum (37) andAgrobacterium radiobacter (38). Penem andmonobactam contained different b-lactamnuclei from 6-APA and 7-ACA. Derivatives of

HA B C

H

S SCH3

CH3

COOH

N NO O

O

O

H2N H2N

7-Aminopenicillanic acid 7-Aminocephalosporanic acidCOOH

CH3

D-Alanyl-D-alanine

H2N H CH3

HN CH3

HOCOOH

Fig. 2. Molecular structure of (A) 7-aminopenicillanic acid, (B) 7-aminocephalosporanic acid and (C) D-alanyl-D-alanine.

KONG et al.


these compounds, either naturally occurring orsemi-synthetic, further expanded the spectrumof b-lactam agents.

b-LACTAMMECHANISM OF ACTION

Interest quickly developed, after the introduc-tion of penicillin, as to the mechanism of peni-cillin action on bacteria. Investigations of themode of action of b-lactam antibiotics had pro-gressed through several seemingly independentroutes. These routes included (i) physiologicalanalysis of the effects of b-lactam on growth,viability, shape, division, and integrity of bacte-rial cells; (ii) biochemical analysis of the cell wallstructure and the enzymes involved in itsbiosynthesis; (iii) biophysical analysis of thecellular components that tightly bind penicillinand attempts at correlating this binding with itsphysiological effects; and (iv) genetic analysis ofmutants with altered response to the b-lactamantibiotics. These lines of research ultimatelyconverged and became coherent. Readers areencouraged to seek other recent reviews formore comprehensive overviews of specific topicsof interest (39–41).

Physiological analysis

Gardner observed that at low concentrations ofpenicillin bacteria formed filaments (42). Thisprovided an early indication that penicillininterferes with the maintenance of cell shape. Ina subsequent study, using different concentra-tions of penicillin, Duguid showed that penicil-lin interfered with cell division and maintenanceof the cell structures (43). These morphologicalchanges suggested penicillin inhibited thesynthesis of an unknown outer supporting cellwall structure (the nature of the supporting wallwas not known at that time).Research into the bacterial cell wall, particu-

larly the murein, was revived when the linkbetween b-lactam antibiotic and cell wall wasestablished. The earliest documentation of thepresence of a bacterial cell wall was traced asearly as the age of Antonie van Leeuwenhoek.Under his microscope, he observed that the bac-teria were bonded by some sort of structure andexpressed his hope of resolving the question ofwhat held them together (44). This turned out to

be a nontrivial problem. Hans Christian Gram,a Danish bacteriologist, devised a stainingprocedure that allowed him to observe bacteriaunder the light microscope (45). Since then, histechnique has been widely adopted and used toclassify bacteria into two categories: the Gram-positive and the Gram-negative (45). Althougha variety of distinguishing features or chemicaldifferences in eubacteria have been suggestedas being important for differentiating Gram-positive vs Gram-negative organisms (45), it isapparent that the murein content of the cell wallis the primary differentiating architecture (46).The breathtaking first thin-section images ofBacillus cereus vividly demonstrated the bacte-rial surfaces and architecture of the cell wall,making it apparent, for the first time, that thecell wall of Gram-positive bacteria was predom-inately single-layered (47). On the other hand,the cell wall of Gram negative bacteria wascontroversial until visualization by the freeze-substitution cryotechnique on E. coli. Theresulting Gram-negative images were soremarkable that microbiologists consequentlyre-evaluated their previous analysis (48). Theterm ‘periplasmic’ was coined when Kellen-berger’s group from Switzerland producedtangible microscopic images of a rich, denseperiplasmic space of E. coli’s cell envelope (49).

Biochemical analysis

Bacterial cell wall biology was a relative new-comer to the scene of biochemical research inthe 1950s. The first biochemical clue to the siteof penicillin action was provided by Park andJohnson (50) and Park (51) who demonstratedthat several novel uridine peptides accumulatedin the cytoplasm of penicillin-treated S. aureus.Park and Strominger subsequently observedthat the sugar and amino acid compositions ofthose uridine peptides were similar to those ofthe bacterial cell wall, suggesting that they werecell wall precursors which accumulated as aresult of penicillin inhibition of cell wall bio-synthesis (52). Originally, molecules comprisedof approximately equal proportions of proteinand sugar components were called peptido-glycans and the peptidoglycan making up thebacterial cell envelope was referred to asmurein. Over time, these terms became usedinterchangeably as they are in this review.



Murein sacculus refers exclusively to the rigidmurein polymer that is present in the cellenvelope.Ironically, prior to 1951, virtually nothing

was known about the chemical composition ofbacterial envelopes beyond the fact that suchwell-known structural polysaccharides of plantcell walls as cellulose and chitin were absent. In1951, Salton and Horne described a large-scalemethod for cell wall isolation (53). Thistechnique was based on the fact that cell wallsare physically quite resistant to mechanicalstress. Electron microscopic illustrations accom-panying every step of purification showed thatthere was no gross physical contamination from

other cellular compounds (53). Qualitativeexaminations of the purified cell walls from alarge number of Gram-positive eubacteriademonstrated the presence of N-acetylglucos-amine (GlcNAc) and N-acetylmuramic acid(MurNAc) in all these cellular extracts (54, 55).In addition, the amino acids glycine, D-glutamicacid, D- and L-alanine and L-lysine weredetected in each cell wall preparation (Table 1)(56). Four amino acids were present in the formof a tetrapeptide whose sequence was deter-mined to be L-alanyl-D-glutamyl-L-lysyl-D-alanine (Fig. 3). Glycine was present as aninterpeptide cross-bridge. These data correlatedwell with the first general hypothesis about the

Table 1. Chemical composition of Gram-positive and Gram-negative bacterial murein

Gram-positive Gram-negative

Sugar Glucosamine GlucosamineAcetylmuramic Acetylmuramic

Amino acid D-Glutamic acid D-Glutamic acidD- and L-Alanine D- and L-AlanineL-Lysine or L-diaminopimelic acid glycine L-Diaminopimelic acid

A B

Fig. 3. Schematic representation of the (A) Staphylococcus aureus and (B) Escherichia coli murein monomers.(A) In S. aureus, N-acetylglucosamine and N-acetylmuramic acid form the backbone of the murein. It is attachedto pentapeptides containing L-alanine, D-glutamic acid, L-lysine, and D-alanyl-D-alanine. The two pentapeptidesare linked with a penta-glycine cross-bridge. (B) In Escherichia coli, the murein backbone is composed of thesame sugar moieties as S. aureus, N-acetylglucosamine and N-acetylmuramic acid. The backbone is attached topentapeptides containing L-alanine, D-glutamic acid, L-diaminopimelic acid, and D-alanyl-D-alanine. Cross-link-ing occurs between diaminopimelic acid and D-alanine. The D-alanines in red boxes are usually liberated upontranspeptidation.

KONG et al.


basal component of cell wall structure, namelymucopeptides, consisting of a hexosaminewhich is either GlcNAc or MurNAc and a pep-tide component made up of alanine, glutamicacid, glycine, and either diaminopimelic acid orlysine (Fig. 3) (55, 57). In fact, this compoundwas exceedingly similar to the nucleotides Parkobserved accumulating in S. aureus as a resultof penicillin attack resulting in the convergenceof these two lines of investigation (52).Though early investigation into Gram-

positive organisms successfully elucidated thechemical structure of bacterial cell walls, itbecame apparent that this was not the case inGram-negative bacteria when the purifiedGram-negative cell wall was subjected to chemi-cal examination. Specifically, the Gram-negativecell wall preparation contained more lipid andless amino sugar than the Gram-positive cellwall (58). The detection of a lipid component inthe Gram-negative’s cell wall was not surprisingbecause the presence of lipopolysaccharide wasknown from the 18th century as the substancethat was associated with fever and diseases (59).Furthermore, a full range of amino acids,presumably corresponding to normal proteinstructures, were identified (58). However, theobservations that in E. coli, lysis occurred bytreatment with lysozyme, and that penicillinproduced osmotic fragility, suggested the pres-ence of peptidoglycan (60).Although the basic biochemical compositions

were known, the architecture of the Gram-nega-tive bacterial cell wall remained elusive. Elegantwork by Weidel and Primosigh indicated thatphenol could remove the lipopolysaccharidelayer (80% of the whole) from E. coli, leavingbehind a layer consisting of sugars, lipids, andamino acids. This layer retained the shape of theintact wall and was therefore assumed to beresponsible for its rigidity (61). Subsequently,modified purification procedures allowed themto show that the ‘wall’ remaining after phenoltreatment contained glucosamine, muramicacid, and a preponderance of alanine, glutamicacid, and diaminopimelic acid (62). Quantitativedissection revealed that various E. coli strainsgave essentially the same molar proportions asS. aureus: 1.0 glutamic acid, 1.5 alanine, 0.2lysine, 0.65 diaminopimelic acid, 0.8 glucosa-mine, and 0.5 muramic acid (63). Essentially,the same mucopeptide composition is found in

the cell walls of a wide range of Gram-negativespecies. The absence of glycine in the Gram-negative cell wall strongly indicated the absenceof interpeptide cross-bridges (Fig. 3).

Murein biosynthesis

Murein is a macromolecule of enormous size.It completely envelopes and maintains thestructure of bacteria. The realization of themode of b-lactam agents on the murein layer isan essential prelude to the study of its biosyn-thesis. Ten years were devoted to elucidating theenzymatic reactions involved in the biosynthesisof the complex murein molecules. Key enzymesinvolved in the murein synthesis and cell wallrecycling and the E. coli and P. aeruginosagenes encoding them are listed in Table 2.Biosynthesis of the cell wall is comprised of

three distinct stages occurring at three distinctsites in bacterial cells (64). The first site is in thecytoplasm. This chain of reactions is initiatedwith the committed step in which a three-carbonphosphoenolpyruvic acid (PEP) is attachedto UDP-N-acetylglucosamine (UDP-GlcNAc),leading to the formation of UDP-N-acetylmu-ramic acid (UDP-MurNAc; Fig. 4). This isfollowed by an enzymatic addition of aminoacids in a stepwise manner, and finally to theproduction of UDP-MurNAc pentapeptide, thecompound first isolated by Park and Johnson(Fig. 4) (50). The second stage of the bacterialmurein synthesis occurs at the cytoplasmicmembrane of the cell. At this site, the intra-cellular precursors, UDP-GlcNAc and UDP-MurNAc pentapeptide, are joined at the sugarends to form a linear murein with protrudingshort peptide chains at every N-acetylmuramicacid sugar (Fig. 4). In addition, depending onthe organism, the murein peptides are modified.For the genus Staphylococcus, the a-carboxylgroup of D-glutamic acid in the pentapeptide isamidated to yield D-isoglutamic acid (65).The final transpeptidation step does not

require any input of energy. In vivo investigationof crude protein preparations of E. coli showedthat there are, in fact, two concomitant trans-peptidation reactions (66, 67). Cross-linkingbetween two murein strands is introduced bytranspeptidase, followed by a second reactioncatalyzed by a D-alanine carboxypeptidase inwhich the terminal D-alanine residue of the



Table 2. List of genes involved in cell wall biogenesis

Proteins Genes inEscherichia coli

Genes inPseudomonasaeruginosa

Functional annotation

Penicillin-binding proteinsPBP1a ponA ⁄mrcA PA5045 Membrane carboxypeptidase;

catalyzes the polymerization ofthe glycan subunit and cross-linkingof muropeptides

PBP1b ponB ⁄mrcB PA4700 Membrane carboxypeptidasePBP1c pbpC PA0378 TransglycosylasePBP2 pbpA ⁄mrdA PA4003 Chain length regulation; murein

elongation initiating enzymePBP3 ftsI ⁄ pbpB PA4418 Transpeptidation of muropeptides

exclusively for septal mureinsynthesis; septal murein-synthesizingenzyme; important for cell division

PBP3a None PA2272 TranspeptidasePBP4 dacB PA3047 D-alanyl-D-alanine endopeptidasePBP5 dacA PA3999? D-alanyl-D-alanine carboxypeptidasePBP6 dacC D-alanyl-D-alanine carboxypeptidase;

stabilization of the murein at thestationary phase

PBP6b dacD D-alanyl-D-alanine carboxypeptidasePBP7 ⁄ 8 pbpG PA0869 D-alanyl-D-alanine carboxypeptidase

Other proteins involved in bacterial cell shapeEnvA envA ⁄ asmB ⁄ lpxC PA4406 UDP-3-O-acyl-N-acetylglucosamine

deacetylaseRodA rodA ⁄mrdB PA4002 Rod shape-determining proteinRodX rodX NoneRodY ⁄MreB ⁄EnvB rodY ⁄mreB ⁄ envB PA4481 Rod shape-determining proteinMreC mreC PA4480 Rod shape-determining proteinMreD mreD PA4479 Rod shape-determining protein

Other proteins involved in bacterial cell wall metabolismMpaA mpaA None Gamma-D-glutamyl-meso-

diaminopimelate hydrolaseMtgA mtgA (mgt) PA0378 Glycan strand polymerizationMpl mpl PA4020 UDP-N-acetylmuramate : L-alanyl-

gamma-D-glutamyl-meso-diaminopimelate ligase

SltY sltY PA3020? Lytic transglycosylaseMltA mltA PA1222 Lytic transglycosylaseMltB mltB PA4444 ⁄

mltB1 ⁄ sltBMembrane-bound lytic murein

transglycosylase BPA4001 ⁄

sltB1 ⁄mltB2Lytic transglycosylase

MltC mltC PA3020? Lytic transglycosylaseMltD mltD PA1812 Lytic transglycosylaseMltE mltE ⁄ emtA PA3764 Lytic transglycosylaseAmiA amiA PA5538 N-acetylmuramoyl-L-alanine

amidase; cell separationAmiB amiB PA4947 Cell separation

N-acetylmuramoyl-L-alanineamidase

AmiC amiC PA4947? ⁄PA5538?

Cell separation

MepA mepA None LD-endopeptidase

KONG et al.


pentapeptide in the other strand is removed(68–70). Among the enzymatic reactions, onlythe transpeptidase and D-alanine carboxypepti-dase are sensitive to b-lactam antibiotics (66).However, the inhibition of the D-alaninecarboxypeptidase enzyme occurs at a muchlower minimum inhibitory concentration (MIC)than that for the organism (66, 67). Thus, it is

proposed that the inhibition of this activity isnot lethal. In sharp contrast, the concentrationrequired to inhibit growth was virtually similarto the concentration required to halt the activityof transpeptidase, the enzyme carrying out theactual cross-linking reaction, suggesting thatthis is the crucial step inhibited by b-lactamantibiotics (66, 67, 71).

Table 2. (Continued)

Proteins Genes inEscherichia coli

Genes inPseudomonasaeruginosa

Functional annotation

NagZ nagZ PA3005 b-N-acetyl-D-glucosaminidaseLdcA ldcA PA5198 LD-carboxypeptidaseEnvC envC PA5133 EndopeptidaseFlgJ flgJ PA1085 Muramidase (flagellum-specific)

Amp and Pox proteinsAmpC ampC PA4110 AmpC beta-lactamaseAmpR None PA4109 Transcriptional regulator AmpRAmpD ampD PA4522 1,6-anhydro-N-acetylmuramyl-L-

alanine amidasePA5485 (ampDh2) N-acetyl glucosamidasePA0807 (ampDh3) N-acetyl glucosamidase

AmpE ampE PA4521 Hypothetical membrane proteinAmpG ampG PA4218 Transport of degraded muropeptides:

GlcNAc-anhMurNAc peptidesAmpF None PA4219 Hypothetical membrane proteinAmpP None PA4393 PermeaseAmpO None PA4219 Hypothetical membrane proteinPoxA None PA5513 Hypothetical hydrolasePoxB None PA5514 Class D beta-lactamase

Mur &Mra proteinsMurA murA PA4450 UDP-N-acetylglucosamine enolpyruvyl

transferaseMurB murB PA2977 UDP-N-acetylpyruvoylglucosamine

reductaseMurC murC PA4411 UDP-N-acetylmuramate-alanine ligaseMurD murD PA4414 UDP-N-acetylmuramoylalanine-D-glutamate

ligaseMurE murE PA4417 UDP-N-acetylmuramoylalanyl-D-glutamate-2,

6-diaminopimelate ligaseMurF murF PA4416 UDP-N-acetylmuramoylalanyl-D-glutamyl-2,

6-diaminopimelate-D-alanyl-D-alanyl ligaseMurG murG PA4412 UDP-N-acetylglucosamine-N-acetylmuramyl-

(pentapeptide) pyrophosphoryl-undecaprenolN-acetylglucosamine transferase

MurI murI PA4662 Glutamate racemaseMraY mraY PA4415 Phospho-N-acetylmuramoyl-pentapeptide-

transferaseMraW mraW PA4420 Peptidoglycan biosynthetic process with methyl

transferase activityMraZ mraZ PA4421 Role in cell wall biosynthesisMraR mraR ⁄ ftsL None Cell division

Information compiled from the EcoCyc, PseudoCyc, and Pseudomonas databases (236–239).



Penicillin is a cyclic dipeptide of two aminoacids, L-cysteine and D-valine. The conformationof one edge of the penicillin molecule is nearlyidentical to the conformation of the backboneof the D-alanyl-D-alanine (Fig. 2). A crucialfeature in this case is the fact that the highlyreactive CO–N bond in the b-lactam ring of thepenicillin molecule lies in exactly the same posi-tion as the CO–N bond in D-alanyl-D-alaninewhich is the target of transpeptidation (Fig. 2).Owing to the striking structural similarity, it

was thus hypothesized that penicillin acts as asubstrate analog, binding to the substrate-anchoring site normally occupied by D-alanyl-D-alanine (69).

Biophysical analysis

Transpeptidases have been characterized aspenicillin-sensitive proteins as their activity isspecifically prevented by the covalent bindingof b-lactam antibiotics to their active site (72).The inference that transpeptidase is a penicillin-binding protein (PBP) contributes yet anotherline of independent research to our overallunderstanding of the penicillin action.As early as 1949, ingenious work had

employed radioactive penicillin to locate the siteof its action on bacterial cells (73–76). Thesestudies clearly indicated that penicillin bound adefinite trace target, termed penicillin-bindingcomponent (PBC) (74). It appeared that theentire penicillin molecule reacted with PBC.Once bound, radioactive penicillin was unableto be removed by neutral buffers, acids, ordetergent, but treatment with dilute alkalireleased it as penicilloic acid (74, 76, 77). Inaddition, the PBC was destroyed at high tem-perature and treatment with strong solventsindicative of protein components (78). Takentogether, this suggested penicillin is covalentlybound to PBC. When the PBC–penicillincomplex was subjected to sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE), the PBC resolved into several proteinsranging in molecular weight from 40 to 90 kDa(79). These proteins were named PBPs and num-bered according to their descending molecularweights. The number of PBPs, their molecularweight, their concentration (molecule per cell)and their sensitivity to b-lactam antibioticsvaried widely between bacterial species (e.g.Table 2) (80). Some of the low-molecular-weightPBPs, PBP5, PBP6, and PBP7, were foundabundantly in bacilli but not in cocci (80).Gram-positive cocci, for example the two cous-ins, S. aureus and Staphylococcus epidermidis,express only four penicillin-binding enzymes,whereas Bacillus subtilis and B. cereus harborseven to eight PBPs (80, 81). In E. coli, PBP1-6were identified using radioactive benzyl-penicil-lin (79). Similar numbers of PBPs were alsofound in P. aeruginosa, Enterobacter cloacae,

Fig. 4. The biosynthetic pathway of bacterial murein.The pathway initiates with committing steps, generat-ing UDP-acetylmuramic acid (UDP-MurNAc).L-Alanine, D-glutamic acid, lysine or diaminopimelicacid, and D-alanyl-D-alanine are enzymatically addedto UDP-MurNAc in a stepwise manner. The forma-tion of UDP-acetylmuramyl-pentapeptide (UDP-MurNAc-pentapeptide) occurs in the cytoplasm. TheUDP-MurNAc-pentapeptide is then transferred tothe cytoplasmic membrane where two lipid carriers(Lipid I and Lipid II) are involved in the joining ofUDP-MurNAc-pentapeptide and UDP-acetylglucos-amine (UDP-GlcNAc) to form a linear murein withprotruding short peptide chains. Finally, the linearmurein chains are cross-linked with each other and topre-existing murein. A few chemical compounds areknown to block this pathway. b-Lactams areknown to inhibit the cross-linking, whereas bacitracininhibits the recycling of undecaprenyl. Moenomycinand vancomycin inhibit the function of Lipid II thatcarries the disaccharide peptide monomer units.

KONG et al.


Salmonella enterica serovar Typhimurium(S. Typhimurium), and Serratia marcescens (80).

Genetic analysis

Functional and morphological correlations ofthe PBPs strongly relied on their bindingspecificity of b-lactam antibiotics and isola-tion of temperature-sensitive mutants. AsPBPs were named based on their SDS-PAGEmigration, PBP nomenclature is organism-dependent. For example, E. coli PBP2 is mostsimilar to S. aureus PBP3. The numberingsystem used below is based upon E. coliPBPs. Much has been elucidated about thestructure and biochemical properties of PBPs.Readers are referred to recent reviews focus-ing on PBPs (82–85).

PBP1 – Cephaloridine, a cephalosporin agentwhich blocks cell elongation at sub-inhibitoryconcentration, has the highest affinity for E. coliPBP1, suggesting that PBP1 might be involvedin cell elongation (86). The high-molecular-weight PBP1 band on SDS-PAGE was actuallycomposed of two polypeptides (79). Spratt firstobserved the resolution of these two closelyrelated proteins into two fragments, but thiswas inconclusive as the separation was inconsis-tent (79). Subsequent modification of the SDS-PAGE technique allowed resolution of PBP1into two distinct components, PBP1a andPBP1b (87).The gene coding for PBP1a was mapped

to ponA or mrcA using N-methyl-N-nitro-N-nitrosoguanidine chemical treatment (87,88). Strains containing mutations in ponA ⁄mrcA appeared to grow normally, suggestingthat that PBP1a might have a subtle role incell growth and morphology (87, 88). How-ever, PBP1a mutants underwent a slow rateof b-lactam-induced lysis, suggesting involve-ment in murein metabolism (87). Subsequentstudies showed that PBP1a catalyzed thepolymerization of the glycan subunit andcross-linking of muropeptides (88).PBP1b could be further resolved into at least

three bands (88); however, chemical mutagene-sis of a single gene, ponB or mrcB, resulted inthe loss of all three fragments. The three couldbe recovered simultaneously by P1 transductionof the wild-type gene located at 3.3 min of the

E. coli chromosome, suggesting that PBP1bcould exist in three isoforms (88). By examiningthe in vitro incorporation of radioactive UDP-MurNAc-pentapeptide into the murein, mem-brane fractions isolated from ponB ⁄mrcBmutants had significantly decreased mureinbiosynthetic activity, indicating that PBP1bcatalyzed the majority of this activity in vitro(88). Subsequently, PBP1b was functionallyindicated to be responsible for transglycosylaseand DD-transpeptidase enzymatic reactions(89–91).PBP1a and PBP1b differ markedly in their

affinities for several b-lactams. For example,cephalothin and cephaloridine both bind PBP1astrongly, while PBP1b has high affinity only forcephaloridine (87). In agreement with this,E. coli cells lacking PBP1b are hypersensitiveto cephalothin, owing to the binding of thisb-lactam antibiotic to the PBP1a (88). Althoughindividual loss of PBP1a or PBP1b did notconfer thermosensitive growth, the loss of bothproteins was lethal, indicative of redundancy inthe system (90). Inhibition of both PBP1a andPBP1b was necessary to prevent cell elongationand thereby induce lysis (90).PBP1c, a third PBP related to PBP1a and

PBP1b, has also been identified in E. coli (92).Unlike PBP1a and PBP1b which function asboth transglycosidases and transpeptidases,PBP1c appears to only function as a transgly-cosylase. Overexpression of PBP1c, which isencoded by pbp1c located at 56.9 min on theE. coli chromosome, cannot suppress the effectof mutating both PBP1a and PBP1b. Mutationin PBP1c does affect cell murein synthesis (92).

PBP2 – PBP2 was the first PBP to be dissectedfunctionally due to the ability of this enzyme tospecifically bind to mecillinam (79). Covalentbinding of mecillinam to PBP2 caused alterationin E. coli shape from rod into ovoid (93). Thismorphologic alteration allowed the isolation ofmutants lacking PBP2 after an intensive chemi-cal mutagenesis (94). These thermolabilemutants grew slowly as rod-shaped cells at30 �C and became round cells upon shifting thetemperature to 42 �C (94). Hence, PBP2 wasthought to be the major protein involved inmaintenance of cell shape. Absence of PBP2also caused inhibition of cell division anddelayed muropeptide synthesis (94, 95).



The actual functions of PBP2 attracted a lotof attention. In vitro, upon the presence ofmecillinam, the carboxypeptidase, transpepti-dase, and endopeptidase activities were unaf-fected, suggesting that PBP2 might catalyze anovel reaction (93, 96). However, it was alsopostulated that PBP2 mediated a topologicallyrestricted transpeptidase reaction, perhapsduring a specific phase of the cell cycle. Severallines of evidence agreed with this postulation:

• Inactivation of PBP2 resulted in a doublingof the E. coli cell wall anhydromuramic acidcontent, thus implicating PBP in cell wallglycan chain length regulation (97).

• In the presence of cefsulodin and cefmetaz-ole, cephalosporins that bind to all PBPs butPBP2, the murein synthesized acquired adegree of cross-linkage even higher than thecontrol cells, demonstrating that PBP2, orother proteins, might have a transpeptidaseactivity in vivo able to produce a highlycross-linked peptidoglycan (95).

• A mutation in ponB or mrcB, encodingPBP1b, resulted in the over-production ofPBP1a and PBP2. This was consistent withthe suggestion that PBP2 might act to ensurethat new initiation sites for elongation wereintroduced with the correct rod orientation(86).

These data argued that in the absence ofactive PBP2, novel glycan chains might be uti-lized preferentially by the septum-formingenzyme system resulting in growth as roundcells (98).Because the strains lacking PBP2 were

isolated upon intensive chemical treatments,mapping the gene encoding PBP2 was non-trivial. Several genes were found to be involvedin round-shaped morphology and mecillinamresistance. In fact, the earliest report that associ-ated the appearance of spherical E. coli withantibiotic resistance, interestingly, pre-dated thediscovery of PBPs (99). Normark, investigatingthe effect of ultraviolet light tolerance, observedround-shaped E. coli under the electron micro-scope when the strains harbored a mutation inenvB (99). This envB gene was subsequentlymapped to the 64 min of the chromosome (99).Half a decade later, a second gene, rodA,responsible for coccal-shaped cells, was mapped

to 15 min between the purE and pyrC loci (100).E. coli strains with mutation in rodA were alsofound to be tolerant towards ultraviolet irra-diation (100). Moreover, intensive chemicalmutagenesis analysis mapped two other genes atmin 14 (rodX) and 70 (rodY) that were involvedin the maintenance of the rod shape (101). How-ever, as the round-shaped morphology could bedue to the mutation of any of these genes, thesereports did not correlate the morphologicalchange with the loss of the PBP2 protein. Even-tually, the gene encoding PBP2, with the con-comitant loss of the PBP2 protein, was mappedto a region at 14.4 min (88, 102). Using a seriesof defective lambda transducing phage, theregion was mapped to a 12-kb DNA containingthe lip-dacA-rodA-pbpA-leuS region (103). Thiscluster of genes was involved in cell shape deter-mination and murein synthesis (103). Analysisof the proteins produced from lambda transduc-ing phage carrying a well-defined DNA frag-ment showed that pbpA encoded PBP2, whereasdacA coded for PBP5 (103). Although mutationin the rodA gene resulted in spherical cells, lossof RodA did not affect PBP2 activity (103).RodA was later found to act synergistically withPbpA as transglycosylase and transpeptidase(104). Elegant genetic analysis of pbpA demon-strated that pbpA spherical mutants could beisolated that retained penicillin binding activity(98). Further analysis showed that deletion ofpbpA is lethal, but may be suppressed by over-expression of ftsZ, either directly or by increas-ing the intracellular ppGpp pool (105, 106).Together, this suggests that PBP2 plays anessential role in cell division.

PBP3 – E. coli strains expressing temperature-sensitive PBP3 mutated proteins were isolatedfrom nitrosoguanine mutagenesis (79). At thepermissive temperature of 30 �C, these mutantsappeared slightly longer than the parental strain(102). On shifting to the restrictive temperatureof 42 �C, cell division ceased. But, the continu-ous increase in cell density and the incorpora-tion of [3H]-thymidine suggested that cellgrowth and DNA replication were not affectedin the absence of PBP3 at restrictive tempera-ture. As a result of continuous replication with-out division, long filamentous polynucleatedE. coli was observed (102). This observationstrongly argued that PBP3 was an essential cell

KONG et al.


division protein. Supporting evidence for its rolein cell division came from the use of PBP3-specific b-lactam antibiotics, furazlocillin andpiperacillin (107). PBP3 is recruited to the septalring where it functions as a transpeptidase tocross-link the peptidoglycan at the divisionseptum (108, 109). Due to its function in celldivision, PBP3 is also known as FtsI.Biochemically, at the permissive temperature,

both [14C]-diaminopimelic acid incorporationinto murein of intact cells and [14C] N-acetyl-glucosamine incorporation into murein werereduced in septating PBP3 mutants, as com-pared to the wild-type while no inhibition wasdetected in filaments growing at the nonpermis-sive temperature (108). The authors suggestedthat PBP3 is involved in transpeptidation ofmuropeptides exclusively for septal murein syn-thesis but not for the elongating murein produc-tion (107, 108).

Low-molecular-weight PBPs – The low-molecu-lar-weight PBPs 4, 5, and 6 were shown asthe major DD-carboxypeptidases (67). Theseenzymes displace the terminal D-alanine residuefrom one of the two peptide chains to formcross-linked murein (67). It was argued thatDD-carboxypeptidase regulated the extent ofcross-linking in murein by removing the terminalD-alanine from a proportion of the pentapeptideside chains of nascent peptidoglycan, thuspreventing their involvement in cross-links (67).This DD-carboxypeptidase activity was dividedinto two enzymatic reactions, namely IA and IB,each catalyzed by different PBPs (67, 110).

• Enzyme IA was shown to catalyze DD-car-boxypeptidase and transpeptidase reactions,bind to [14C]-penicillin G, and confer a weakpenicillinase activity, but was devoid ofendopeptidase activity (110).

• Enzyme IB was demonstrated to catalyzeDD-carboxypeptidase and endopeptidasereactions, confer weak penicillinase activity,but not bind [14C]-penicillin G and had poortranspeptidase activity (110).

PBP4 – E. coli strains lacking the highly penicil-lin-sensitive activities of DD-carboxypeptidaseIB and DD-endopeptidase showed a concom-itant loss of PBP4 (111, 112). This mutation was

mapped to the dacB gene located at 68 min onthe E. coli map (112). These mutants grew nor-mally under a wide range of growth conditions,suggesting that PBP4 was not essential for thegrowth of E. coli under laboratory conditions(88, 111). However, overexpression of PBP4showed an increase in DD-carboxypeptidaseand DD-endopeptidase, which did not alter thetranspeptidation reaction (113). PBP4 wasdemonstrated as the only PBP of E. coli thatpossessed DD-endopeptidase activity. Thisenzyme was able to cleave the D-alanyl-c-meso-2,6-diaminopimelyl peptide bond in the cross-links of murein (114).

PBP5 – Purified DD-alanine carboxypeptidaseIA activity in E. coli was attributed to two poly-peptides (115). Both proteins bound strongly tobenyzlpenicillin and were identified as PBP5and PBP6 (79). These membrane-bound pro-teins were shown to catalyze a transpeptidasereaction and had a weak penicillinase activity(110). Loss of PBP5 proteins could be comple-mented with P1 phage that carried the 14 minregion of the chromosome, indicating that thePBP5 gene was located within the lip-leuSregion (116). Using lambda transducing phagecarrying well-defined DNA fragments, the geneencoding PBP5 was mapped to dacA (103).Mutation in dacA abolished the DD-alaninecarboxypeptidase IA activity and rendered thecells sensitive to penicillin, though they exhib-ited normal morphology (117). A dacA dacBdouble mutant demonstrated a negligible carbo-xypeptidase activity, suggesting that PBP4 andPBP5 contributed at least 90% of the DD-carboxypeptidase activity (117).

PBP6 – The dacC-encoded PBP6 protein shares62% sequence identity with PBP5 (118). Bothpurified PBP5 and PBP6 were able to catalyzeidentical reactions but PBP5 had a three-to four-fold higher specific activity towarduncross-linked peptidoglycan than PBP6 (119).Deletion of dacC had no effect on cell morphol-ogy and growth rate (120). However, strainslacking PBP6 showed a very slight increase inantibiotic sensitivity (120). Comparison of theexpression of PBP5 and PBP6 at differentphases of growth exposed the temporallydifferential functions of PBP5 and PBP6. Theconcentration of PBP5 remained fairly constant



throughout all phases, whereas the amount ofPBP6 increased two- to ten-fold in stationaryphase as compared to exponential phases (121).This growth-phase dependent expression ofPBP6 strongly suggested that it might beinvolved in stabilizing the murein at stationaryphase.Another low-molecular-weight PBP that is

similar to PBP5 and PBP6 was identified inE. coli, PBP6b (122). PBP6b has DD-carboxy-peptidase activity and binds [3H] benzylpeni-cillin. The gene encoding PBP6b, dacD, islocated at min 44 on the E. coli chromosomeand was identified using the Kohara lambdaphage library (122). Like PBP5 and PBP6,PBP6b is non-essential for viability (122).

PBP7 ⁄8 – PBP7 and PBP8 were characterizedmore recently than most of the other PBPs inpart due to their irreproducible presence in bac-terial preparations. PBP8 is an OmpT proteo-lytic product of PBP7 (123). PBP8 has beenshown to be a D-alanyl-DAP-endopeptidase(124) and increased expression is related toincreased cephaloridine and ceftazidime resis-tance (125). Both are soluble periplasmicproteins that are peripherally associated with themembrane. Use of the Kohara lambda phageminiset library narrowed the genomic locationof the PBP7-encoding gene to 47.8 and 48 minon the E. coli chromosome (126). Subsequentsubcloning of one open reading frame (ORF),yohB now renamed pbpG, and insertional muta-genesis confirmed that pbpG encoded PBP7.

HIDDEN BLACKBOX: ANTIBIOTIC

RESISTANCE

The introduction of penicillin from laboratoryto bedside heralded a new era in the clinicalsetting. Physicians had a landslide victory intreating many dominating lethal infectiousdiseases such as meningitis, tuberculosis, pneu-monia, and so forth. The development of otherclasses of antibiotics, such as erythromycin,tetracyclines, and chloramphenicol, strength-ened the foundation of infection chemotherapy.Antimicrobial therapy provided physicians withthe ability to prevent some infections, to cureothers, and to curtail the transmission of certaindiseases. At the beginning of the 20th century,

infectious diseases were the leading cause ofdeath worldwide. The success of vaccinationand the discovery of antibiotics changed thistrend. Between 1900 and 1980, in the UnitedStates, mortality from infectious diseases fellfrom 797 to 36 per 100 000 individuals (127). In1969, amid the success of vaccination and thedevelopment of antibiotics, the SurgeonGeneral told the United States Congress that itwas time to ‘close the book on infectious dis-eases’ (128). The availability and success of anti-biotics resulted in confidence that technologyand modern medicine would be victoriousagainst infectious diseases. The emergence ofoutbreaks and epidemics of antimicrobial resis-tant infections resulted in a re-evaluation of thisoptimism.At the very beginning, resistance to antibacte-

rial agents was recognized as a potential prob-lem. Resistance to penicillin was actually thefocus of Alexander Fleming’s 1929 ground-breaking masterpiece where he exploited thesubstance’s ability to eliminate Staphylococcusand Streptococcus contamination and thus iso-lated pure cultures of E. coli, S. enterica serovarTyphi and H. influenzae (15). Yet the potentialof penicillin as a therapeutic agent overshad-owed the resistance issue.All was not lost as this report intrigued some

scientists, particularly Abraham and Chain.They prepared to answer the question of whysome bacteria are more resistant than others.In 1940, they found penicillinase enzyme inthe culture fluid of a penicillin-resistant Gram-negative rod that had contaminated Penicilliumcultures and in an extract of E. coli, but detectedno such activity in a penicillin-sensitive strainof S. aureus (18). The purified enzyme fromStaphylococcus was able to inactivate penicillineffectively (129). Thus, it was known from thebeginning of the human-pathogen battle thatbacteria have the ability to counteract anti-biotics. Furthermore, Abraham et al. also dem-onstrated that cultures of staphylococci couldbe made resistant by continuous subculture inthe presence of the antibiotic in vitro (19).Within a year of introduction, four penicillin-

resistant staphylococci strains were isolatedfrom patients during the course of treatment oflocalized infections (130). This was followedby a number of reports showing the emergenceof penicillin-resistant staphylococci (131–133).

KONG et al.


This phenomenon did not raise any immediatealarm as the initial increase in penicillin resis-tance could be compensated for by with anelevated dosage of the drug, as the antibioticwas relatively safe and became more widelyavailable and inexpensive (134). By 1947, just5 years after its introduction, the majority ofhospital isolates of S. aureus were resistant topenicillin G and six other common antibacterialagents, including streptomycin, tetracycline,and erythromycin (135). This prevalent inter-hospital spreading S. aureus strain 80 ⁄81complex was the first encounter with multidrug-resistant pathogens (135). Barber in Englandwas the first to call attention to the problem ofthe increasing occurrence of penicillin-resistantstaphylococci in hospitals (136–139). World-wide alarm regarding S. aureus resistancequickly followed as reports emerged from manyhospitals and laboratories in the United States(135), Canada, Australia, Norway, Sweden,Denmark, France, India, and Chile (134).In the 1960s, analysis of the organisms associ-

ated with nosocomial infections shifted awayfrom Gram-positive species to Gram-negativespecies such as E. coli, Klebsiella pneumoniaeand Proteus mirabilis, a group of bacteria ofwhich scientists and bedside physicians had littleunderstanding (140). According to Finland’slongitudinal survey of the panorama of infec-tions, the introduction of antimicrobial agentswas accompanied by a negative rather thanpositive overall impact on the patients (140). Infact, the potential therapeutic effect was nulli-fied by a shift toward more drug-resistantorganisms and serious infections by morevirulent organisms. In addition, organisms chal-lenged with antibiotics increased their number,and enhanced their pathogenicity and invasive-ness (141).At the same time, the medical renaissance

increased the life expectancy, but not the lifequality, of the human race with sophisticatedtreatments, including intensive care for terminalpatients, artificial instrument insertions andchemotherapy of cancer patients. As a result,there was a sudden surge of patients who wereimmuno-compromised or required frequentmedical attention. These events directly aug-mented the opportunity for infections by agroup of pathogens such as P. aeruginosa,S. epidermidis, K. pneumoniae and S. marcescens

(140). Unlike S. aureus, E. coli and SalmonellaTyphi, this group of bacteria are successful envi-ronmental organisms, evolutionarily equippedwith an armory to counteract naturally occur-ring growth inhibitors, including naturallyoccurring antibiotics, antiseptics, and heavymetals (140).The failure of penicillin to be the magic bullet

and the shift in infection pattern forced investi-gators to seek alternative routes. The age of‘drug rejuvenation’ began with the artificial syn-thesis of 6-APA (31). This discovery allowed therapid introduction of b-lactam antibiotics bychanging the chemical structures in vitro. For ashort period of time, drug company chemistsmanaged to keep ahead in the race against anti-biotic resistance by making slight changes in thestructures of their antibiotics. The synthesis ofsemi-synthetic methicillin, which was b-lactam-ase stable, temporarily counterbalanced thefailure of penicillin (142). Meanwhile, theproduction of ampicillin, another semi-syntheticb-lactam antibiotic, inhibited a wide range ofGram-negative organisms, including E. coli,H. influenzae, Salmonella Typhi, and Shigella.As a result of the explosive antibiotic develop-ment, the market was flooded with more thanone hundred antibacterial agents (143). Think-ing that they had won a total victory, manypharmaceutical firms started to reduce theirefforts to develop new antibiotics in the 1980s,instead focusing on drugs for chronic illnessessuch as heart disease, cancer, and diabetes, theleading causes of mortality and morbidity indeveloped countries (128). Thus far, antibioticshave been used with unrestrained passion farsurpassing the needs of management and infec-tion control. This is owing to an unorthodoxbelief that antibiotics play a life-saving role inall debilitating diseases (144). Over the years,antibiotic resistance issues frequently comeunder the spotlight in the media.As the gruesome antibiotic resistance alarm

became inescapable, international conferences,workshops, and other actions were organized topersuade funding agencies to take notice andbring the issue to the forefront (145). In 1990,Gail Cassell, President of the American Societyof Microbiology (ASM) and a member of theadvisory council of the National Institutes ofAllergy and Infectious Diseases (NIAID),warned, ‘we are at a very critical crossroads in



terms of readiness to deal with infectious dis-eases and antibiotic resistance from a fundingstandpoint’ (145). Even though better monitor-ing or a surveillance program could help preventand manage outbreaks of antibiotic-resistantdisease, most scientists argued that basicresearch on how bacteria defy antibiotics wasequally pressing and seriously lacking.

BETA-LACTAMASES

b-Lactam was the first antibiotic to be described(28). Consequently, resistance to b-lactam anti-biotics was the first to be understood. The mosteffective way for bacteria to counteract thesechemicals has been by producing b-lactamases,enzymes that inactivate the drugs by hydro-lyzing the b-lactam ring (18). Based onthe sequence analysis, b-lactamases and thePBPs are believed to diverge from a commonancestor. All PBPs possess b-lactam-catalyzingcapability to a smaller extent. For instance,PBP5 was demonstrated to have the highestb-lactamase activity. At pH 7.0 and 30 �C,the half-life of penicilloyl-PBP5 was averagedat 10 min (119). In addition, the monofunction-al penicillin-binding DD-peptidases and penicil-lin-hydrolyzing serine beta-lactamases retainthe same tertiary folding, three-motif aminoacid (Lys-Thr-Gly) sequence signature, serine-assisted catalytic mechanism, and active-sitetopology (Fig. 5) (72).Thus far, more than 300 enzymes have been

documented (146). Because of the diversity ofenzymatic characteristics of the many b-lacta-mases discovered, multiple attempts have beenmade to categorize them since the late 1960s(147–150). These classifications involve twomajor approaches: the first and older one isbased on the biochemical and functional charac-teristics of the enzyme, whereas the secondapproach is based on the molecular structure ofthe enzyme. In the former classification scheme,several criteria are used, including the spectrumof antimicrobial substrate profile, enzyme inhi-bition profile, hydrolysis rate (Vmax), bindingaffinity (Km), isoelectric focusing (pI), proteinmolecular weight, and amino acid composition(149, 150). The molecular classification of b-lac-tamases is based on the nucleotide and aminoacid sequences of these enzymes. To date, four

classes are recognized (A–D), correlating withthe functional classification. Classes A, C, andD act by a serine-based mechanism, whereasclass B or metallo b-lactamases need zinc fortheir action (147, 148).

Ambler class A – Ambler molecular class A wasinitially described in Gram-positive plasmids.However, plasmid-, transposon- and chromo-somally-mediated b-lactamases (VHS, PER,TEM and SHV) in Gram-negative bacteria havealso been reported (151). This class of enzymes,dubbed penicillinases, exhibits the highestdegree of sequence variability and kineticproperties (152). Plasmid-encoded Ambler classA b-lactamases identified in P. aeruginosa areactive against carbenicillin hence referred to asCARB or Pseudomonas-specific b-lactamases

A

B

Fig. 5. Structure of the active site of DD-carboxy-peptidase and b-lactamase. The X-ray crystal struc-ture of (A) E. coli PBP5 (240) and (B) Bacilluslicheniformis BS3 b-lactamase (241). The conservedSerXXLys, Ser[Tyr]XAsn, and Lys[His]Thr[Ser]Glymotifs (242) are yellow, orange, and green, respec-tively. The catalytic serine of the SerXXLys motif isred.

KONG et al.


(PSE) (153, 154). In addition, VHS-, PER- andTEM-derived b-lactamases have been reportedin P. aeruginosa (155).

Ambler class B – The class B enzymes contain asmall number of Zn2+ metallo-b-lactamases,whereby their activities could be inhibited byEDTA (156). IMP-1 is the first metallo-b-lac-tamase described inP. aeruginosa (157). Its gene,blaIMP, appears to be dispersed among P. aeru-ginosa and other Gram-negative rods in Japan(158). An integron-borne metallo-b-lactamasegene, blaVIM, which was originally described inP. aeruginosa isolated in Italy and is prevalentlyfound in Greece, gives rise to the resistance ofmeropenem and imipenem (159, 160).

Ambler class C – The Ambler class C enzymesare active against cephalosporins, hence dubbedcephalosporinases (161). They are chromosome-encoded and synthesized bymost Gram-negativebacteria. The known sequences of these enzymesare highly conserved (162). The P. aeruginosaclass C cephalosporin-hydrolyzing chromo-somal b-lactamase, encoded by ampC (PA4110),has been cloned and sequenced (163).

Ambler class D – The enzymes belonging to theclass D group of b-lactamases are called oxacil-linases as they are able to degrade isoxazolylb-lactams such as oxacillin and methicillin(164), while clavulanic acid serves as a goodinhibitor for these enzymes (165). Due to thestructural similarity between class A and class Denzymes, Couture et al. suggested the use of the26 conserved amino acid residues as the class Dstandard numbering scheme (DBL numeration)(151). Plasmid- and transposon-mediated oxa-cillin-hydrolyzing b-lactamases in P. aeruginosaare common and complex. The characterizationof a P. aeruginosa ampC mutant led to the dis-covery of a novel chromosomally encoded classD b-lactamase, encoded by poxB [PA5514;(166)]. This class D b-lactamase was indepen-dently reported by Girlich et al., and theenzyme was named OXA-50 (167). However,sequence analysis argues that PoxB is distinctfrom OXA enzymes (166). A nosocomialoutbreak with P. aeruginosa extended-spectrumof class D b-lactamases was reported (168) andmore episodes are expected to arise in the nearfuture.

THE CHROMOSOMAL b-LACTAMASE

INDUCTIONMECHANISM

There are two kinds of chromosomal-encodedb-lactamases: constitutive and inducible. Theformer is present at a predictable level underall circumstances, whereas the latter is repressedin the absence of antibiotics and induced inthe presence of b-lactam agents (169). The ana-lysis of chromosomal-mediated b-lactamasesrevealed that in bacterial species that possessa specific transcriptional regulatory system,synthesis of b-lactamase is inducible, whereas instrains without this system, synthesis is usuallyconstitutive (170, 171).The Gram-negative organisms possess chro-

mosomally determined b-lactamases and thetypes of b-lactamase produced are often speciesspecific (171). The inducible expression ofb-lactamases of Gram-negative bacteria suggeststhat this mode of bacterial resistance does notinvolve a single-protein one-step mechanism.Indeed, genetic studies performed on E. cloacae,Citrobacter freundii and E. coli revealed thatleast five genes found in three loci directly partic-ipate in b-lactam resistance. They are ampC,ampR, ampD, ampE, and ampG. Many humanpathogens harbor AmpR and AmpC homo-logs, namely, Burkholderia cenocepacia (172),Yersinia enterocolitica (173), C. freundii (174),E. cloacae (175), Morganella morganii (176),S. marcescens (177), Laribacter hongkongensis(178), andOchrobactrum anthropi (179).

ampC

The chromosomally encoded b-lactamases havebeen a subject of intense study since the 1980s(161). Most of these enzymes are cephalosporin-ases that are capable of hydrolyzing third-gener-ation cephalosporins with low Vmax value andhigh Km values (161). They are classified as thetype I b-lactamases in the Richmond and Sykesscheme, and class C enzymes in Ambler’s classi-fication (147, 150). These are relatively largeenzymes (30–42 kDa), usually with alkaline iso-electric points (150).The chromosomal b-lactam hydrolyzing

gene, ampC, of E. coli K-12 was the first to becloned and sequenced (180). This 1536-bp geneencodes a protein of 377 amino acids. Thefirst 19 amino acids form a signal peptide,



suggesting that this protein is translated as aprecursor and transported into the periplasm asa mature 39.6 kDa enzyme (180). The AmpCb-lactamase with substrate specificity for ceph-alosporins showed no significant sequencehomologies with b-lactamases of the Ambler’sclass A penicillinase or with the D-alanine carb-oxypeptidases (161).The expression of AmpC in E. coli K-12 was

constitutively low in the absence of b-lactamantibiotics (181). However, the amount ofAmpC protein increased proportionately withgrowth rate and nutrient availability (181). Thecheckpoint for the limited expression of thisenzyme was due to the presence of an attenuatorlocated between the promoter and the structuralgene (182). This attenuator formed a stem-loopstructure followed by a string of Us similar to aq-independent transcriptional terminator. Usingan in vitro transcription system, Jaurin et al.determined that less than one-fourth of the tran-scripts initiated at the promoter escape attenua-tion (182). This mechanism suggested the lowexpression level of E. coli b-lactamase.Unlike the ampC of E. coli K-12, the expres-

sion of E. cloacae ampC was repressed in theabsence of b-lactam antibiotics, and induced inthe presence of b-lactam agents (183, 184).Intriguingly, the E. cloacae AmpC did nothydrolyze the third-generation cephalosporins(185). However, the authors concluded thatb-lactamase expression was solely responsiblefor resistance toward cefoperazone, cefotaxime,ceftriaxone, ceftizoxime, ceftazidime, andmoxalactam, because the cloning of the ampCgene into E. coli conferred an identical resistantpattern as E. cloacae (185).The b-lactamase gene of C. freundii was first

cloned and transferred to E. coli as a 1.5-kbDNA fragment (186). The presence of this frag-ment in E. coli conferred a similar resistanceprofile to third-generation cephalosporins as theC. freundii AmpC. Subsequently, the ampCof C. freundii strain OS60 was sequenced anddemonstrated to encode a 380 amino acid poly-peptide with a 19 residue signal peptide (187).The mature protein had a molecular weight of39.8 kDa. It shared 77% amino acid identitywith E. coliK-12 AmpC (187).The ampC of P. aeruginosa PAO1 was cloned

and sequenced in the 1990s (163). Sequenceanalysis of the cloned 1.3 kb P. aeruginosa

chromosomal DNA showed the presence of anampC gene with a high sequence identity toE. cloacae and C. freundii AmpC proteins (163).When the putative ampC gene was transformedinto a non-b-lactamase-producing P. aeruginosastrain, there was detectable chromogenicactivity (163). Western blotting using anti-P. aeruginosa AmpC antibody indicated thatP. aeruginosa AmpC had an approximatemolecular weight of 35.5 kDa (188). However,using crude protein extracts, the isoelectricpoint (pI) of P. aeruginosa b-lactamases wasshown to be heterogenous (189). Multi-steppurification of the b-lactamase enzymes fol-lowed by protein sequencing suggests thatC-terminal truncations of AmpC might causethe variations in isoelectric point (190). How-ever, Gates et al. proposed that multiple b-lac-tamase genes might be present in P. aeruginosa(191).Similar to C. freundii and E. cloacae, the

expression of P. aeruginosa chromosomal b-lac-tamases could be induced in the presence ofb-lactam antibiotics (191). Examination of alarge number of P. aeruginosa clinical strainsrevealed four patterns of b-lactamase expres-sion: low-basal, inducible; moderate-basal,inducible; moderate-basal, constitutive; andhigh-basal, constitutive (189). As the existenceof PoxB was not known, and the biochemicalassays used were not sensitive enough to distin-guish these two b-lactamases, it was presumedthat AmpC was overexpressed in the hyperpro-ducing clinical isolates (189). A high constitutiveexpression of b-lactamase seen in P. aeruginosastrains with ampR mutation suggests that theclinical P. aeruginosa isolates that are hyperpro-ducing b-lactamase may harbor poxB and amutation in ampR (166). A Southern blotanalysis of a few of these clinical P. aeruginosaisolates with ampR- and ampC-specific probesshowed no gross changes in the ampR-ampCintergenic DNA (192). This led the authorsto rule out any changes in ampR and ampCgenes, although this technique is not sensitiveenough to detect minor deletions or point muta-tions (192). In fact, one of these P. aeruginosab-lactamase hyperproducers was shown to havea point mutation in ampR (193). It is not clear ifthis clinical isolate harbors the poxB gene,though it has been demonstrated thatpoxB is fairly ubiquitous and is found in

KONG et al.


chromosomes of both clinical and environmen-tal P. aeruginosa isolates (166).

ampR

Unlike E. coli, the AmpC b-lactamases ofC. freundii, E. cloacae, and P. aeruginosa showdifferential expression in the absence andpresence of b-lactam antibiotics, suggestingthat the regulatory systems governing theampC gene expression are very different fromthat of E. coli. Bergstrom et al. showed that inC. freundii, the ampC gene was in close proxim-ity to the fumarate reductase operon (frd) as inE. coli (194). However, whereas the frd operonand ampC overlapped in E. coli, they were sepa-rated by a 1100-bp region in C. freundii (194).This region appeared to contain the regulatoryelement required for the inducibility of AmpC(194). Indeed, when this 1100-bp region wascloned independently from the ampC gene, itrepressed the expression of ampC, suggestingthat it encoded a trans regulatory element (174).Using a minicell system, the small 1-kb regionwas also demonstrated to express a 31-kDa pro-tein, known as AmpR (174). Notably, as com-pared to the absence of ampR, the presence ofC. freundii ampR resulted in a more than two-fold increase in C. freundii AmpC b-lactamaseactivity in the presence of b-lactam agents inE. coli (174). This indicated that AmpR mighthave a dual role in b-lactamase expression, as arepressor in the absence of an inducer, and as anactivator in the presence of the inducer (174).Similar to C. freundii, in E. cloacae, the frd

operon and ampC gene do not overlap (194).Using a minicell system, Honore et al. andLindberg et al. demonstrated that the regionbetween the frd operon and ampC gene encodedfor a polypeptide of 31.5 kDa (175, 195). Con-sistent with the nomenclature in C. freundii, thisgene was called ampR. Under non-inducedconditions, E. cloacae ampC expression wasconstitutively low in E. coli, even in the presenceof ampR (195). This observation suggested thatunlike C. freundii AmpR, E. cloacae AmpRmight not act as a repressor of ampC in theabsence of an inducer. The repressive functionof C. freundii AmpR was further confirmedusing C. freundii ampC in an E. coli host, inwhich the synthesis of C. freundii AmpCb-lactamase was 11-fold lower in the presence of

the homologous ampR gene than when comple-menting with ampR from E. cloacae (195).However, the inductive role of C. freundii

and the E. cloacae ampR gene product wasindistinguishable because the E. cloacae AmpCb-lactamase was elevated approximately 42-foldand 40-fold, respectively, when the E. cloacaeor C. freundii ampR was present in trans (195).So, though it is highly homologous, C. freundiiand E. cloacae AmpR exhibited differentproperties in the regulation of the cognateampC gene.The ampR gene encodes a DNA-binding pro-

tein that belongs to the LysR family of regula-tory proteins (196). The LysR family consistsof small sized, autoregulatory transcriptionalregulators. They have a helix-turn-helix motifat the N-terminus and substrate-bindingdomain(s) at the C-terminus. The function ofthe DNA-binding motif depends upon bindingof the substrate-binding domain to small mole-cules as co-inducers or co-repressors (196).Mutational studies and amino acid sequencesimilarities identify: (i) a DNA-binding domainemploying a helix-turn-helix motif (residues1–65), (ii) domains involved in co-inducer recog-nition and ⁄or response (residues 100–173 and196–206), (iii) a domain required for both DNAbinding and co-inducer response (residues 227–253) (196).Gel-mobility-shift assay is a common way to

examine the interaction of a DNA-bindingmolecule with a DNA fragment. Crude cellularextracts of E. coli harboring the C. freundiiampR in trans were demonstrated to retard theradioactively labeled ampR-ampC intercistronicregion (197). No significant difference in mobil-ity shift was detected when extracts from thewild-type E. coli were used. This retardation ingel mobility clearly indicated that the AmpR isa DNA-binding protein (197). It interacts spe-cifically with the ampR-ampC intercistronicregion and thus regulates the expression ofampC (197).In order to determine the DNA binding

region more precisely, a more refined technique,DNase I footprinting, is often used. In DNase Ifootprinting using cellular extracts from E. coliexpressing C. freundii ampR, a DNA segment of38 bp was indicated to bind to AmpR (197).This 38-bp region was located immediatelyupstream of the ampC promoter. The addition



of b-lactam antibiotics did not change thefootprinting properties of AmpR (197). How-ever, a slight change in the footprinting ofC. freundii AmpR with muramyl pentapeptidehas been observed and is discussed in detailbelow (198).The 38-bp AmpR binding site overlaps the

AmpR promoter (197). This led to the postula-tion that AmpR binding represses its ownexpression, an auto-regulatory mechanismfrequently seen with other LysR transcriptionalfactors (196). However, the auto-regulatorymechanism of AmpR was not consistent. In anin vivo system using an E. coli heterologoushost, the PampR-lacZ activity was repressedthreefold in the presence of C. freundii ampR(174). However, this mode of regulation was lostin a minicell system with E. cloacae AmpR(195). The latter concurred with studies ofC. freundii and P. aeruginosa ampR: the amountof C. freundii AmpR expressed in E. coli mini-cells remains constant in the absence and pres-ence of inducer (174), constitutive ampRexpression is low with a short half-life mRNA(175), and the ampR transcription (197, 199)and the level of AmpR were unaffected in theabsence and presence of antibiotics (197).The sequence of the 5¢-end of P. aeruginosa

ampR was initially reported along with theampC gene and, subsequently, the wholesequence was described (163, 200). Based on thesequence analysis, the P. aeruginosa ampR geneproduct was predicted to be a polypeptide of323 amino acids with a helix-turn-helix motif atthe N-terminus (200). By aligning sequenceswith those of E. cloacae and C. freundii, Lodgeet al. proposed that the translation initiationcodon of the P. aeruginosa ampR was a rareTTG codon, rather than the usual ATG startcodon (163). Overall, this rare TTG codonaccounts for only 1% of the known initiatorsequences (201). The derived amino acidsequence showed a high degree of identity withE. cloacae AmpR (62.5%) and C. freundiiAmpR (58.3%) (200). The binding motif ofC. freundii AmpR was also present in theampR-ampC intercistronic region of P. aerugin-osa (200). In a gel mobility-shift assay, crude cellextracts of E. coli containing the P. aeruginosaampR were shown to bind to the region �50 bpupstream of ampR, but not to the promoterregion of ampC (200).

Recent work on P. aeruginosa AmpR showsthat it is a global regulator because the straincarrying the ampR mutation produced higherlevels of pyocyanin and LasA protease, andlower levels of LasB elastase (199). The produc-tion of LasA protease has been correlated withthe LasI-LasR quoromone system (202, 203). Inthe P. aeruginosa ampR mutant, the increase inLasA levels positively correlated with the lasI ⁄Rquorum sensing gene expression (199). Thereduction in the LasB activity was positivelycorrelated with the rhlI ⁄R quorum sensing geneexpression. Thus, AmpR plays a dual role, posi-tively regulating ampC, lasB, and rhlR expres-sion, and negatively regulating poxB, lasA, lasI,and lasR expression (199).

ampD and ampE

When E. coli K-12 harboring C. freundii ampR-ampC was grown with high concentration ofcefotaxime, chromosomal mutants producinghigh level of b-lactamase constitutively arose ata considerable rate (10)6 to 10)7) (174). Analy-ses of these mutants showed the presence of anintact functional ampC-ampR intergenic region,and the phenotype could not be complementedusing a C. freundii frd-ampC region, suggestingthe presence of other regulatory genes (174). Inaddition, when the frd-ampC region of a b-lac-tamase-overproducing E. cloacae strain wastransformed into E. coli, an inducible expres-sion of AmpC b-lactamase was observed, rulingout any loss of integrity of the frd-ampC region(204). These led to the hypothesis that the con-stitutive b-lactamase overproduction was due toloss of function in a gene required for inducibil-ity that resided outside of ampR-ampC in thechromosome of E. coli (205). In addition, theconstitutive overproduction of b-lactamase waslost in the absence of ampR, suggesting that thefunction of the gene was ampR dependent (174).The mapping of this novel mutation was a

tedious task. Using a series of Tn10 transposonmutants and Hfr strains, the mutation inE. coli was localized to the nadC-lpd region(205). An E. coli strain with a deletion atnadC-aroP region also showed constitutiveproduction of C. freundii AmpC b-lactamase,suggesting that a gene necessary for inducibilitywas located within this region (205). By supple-menting the nadC-aroP region in trans, the

KONG et al.


inducible b-lactamase phenotype could berestored from E. coli strains that overproducedthe enzyme (205). Using an approach similar tothat of Lindberg et al. (205), Honore et al. iden-tified that a loss of E. coli nadC-aroP region alsocaused an overproduction of E. cloacae AmpC(206).Dideoxy sequencing of the nadC-aroP region

of E. coli revealed the presence of two ORFs(206, 207). The upstream ORF was designatedampD, overlapping the downstream ampE. A S1nuclease protection assay confirmed that thetwo ORFs constituted an operon (206). Primerextension analysis performed on ampDE tran-script showed that the 5¢-end of the ampDmRNA was localized to an adenosine residue 29nucleotides upstream from the translationalstart codon (206). This knowledge of thetranscript start site enabled the promoter to bepinpointed. The promoter of the ampDE operondid not possess a canonical -35 region, recog-nized by the r70 RNA polymerase holoenzyme.The -10 region (5¢-TACTCT-3¢) resembled theconsensus sequence and was preceded by a TGmotif at -12 (206). The TG motif has beenreported to greatly strengthen promoters lack-ing the consensus -35 element (208, 209).The E. coli ampD gene product is 183 amino

acids long with an apparent molecular weight of25 kDa. The E. coli AmpE, with 284 aminoacids, has its putative initiation codon overlap-ping the ampD stop codon. E. coli AmpE has anapparent molecular weight of 28 kDa (207).Using a minicell fractionation technique, AmpDwas demonstrated to localize in the cytoplasm.This was in agreement with its hydropathy pro-file that is indicative of a hydrophilic protein(207).Fractionation of minicell harboring E. coli

ampE showed that AmpE was located in theinner membrane fraction. The AmpE hydro-phobicity plot suggested the presence of fourtransmembrane regions: residues 44–60, 68–84,195–211, and 268–284 (207). The primarysequence of E. coli AmpE possessed two of thethree motifs found in b-lactam binding proteins(210). These are the Ser-Arg-Asn correspondingto the Ser-X-Asn motif, and the Ser-Met-Ala-Lys corresponding to the Ser-X-X-Lys motif(206). However, E. coli AmpE failed to bind toany b-lactam antibiotics (207). In addition,AmpE could be an ATP-binding protein

because the Walker’s ATP-binding consensus isfound at residues 74–87 and 184–198 (207). Thefunction of AmpE in E. coli is not clear;however, unpublished data suggest that it playsa role in D-alanyl-D-alanine murein recycling(40).Phenotypically, the loss of E. coli ampD

caused a constitutive overproduction of AmpCb-lactamase that could be further induced bythe addition of an inducer, suggesting thatampD acts as a negative regulator of ampC. Thiswas in contrast to the loss of E. coli ampD andampE that caused a constitutive high level ofb-lactamase production, and the level could notbe increased any further, even with the additionof an inducer (207). Moreover, the complemen-tation of E. coli ampE in an ampDE mutantbackground restored only 50% of the E. cloacaeAmpC b-lactamase induction (206). These datastrongly suggest that ampE also played a role inthe inducibility of ampC. However, intriguingly,the inducible phenotype of the ampDE mutantcould be complemented with ampD alone (207).This could have been due to overproduction ofAmpD from a multicopy plasmid.At this point, it is worth noting that all the

extensive molecular studies were performedusing E. coli ampD and ampE, which comple-mented the constitutive expression of C. freundiiand E. cloacae b-lactamases (205–207). Thusfar, no homologous system has been developedto understand the mechanism. Hence, little isknown about the C. freundii and E. cloacaeampDE genes, except for the presence of ampDEhomologs in these organisms (205, 211).Sequencing of the wild-type ampD gene ofC. freundii and E. cloacae revealed that they are564-bp, corresponding to a gene product of 187amino acids (212). Sequence comparisonrevealed that they share 74–83% similarity toE. coli ampD at the nucleotide level (212). Thislevel of homology resembles that found for theirrespective ampC and ampR genes, reflecting thatthese molecules evolved simultaneously (162,175, 212). More importantly, the three AmpDproteins are highly similar except for their car-boxyl termini with a loss of four amino acids inE. coli AmpD. As compared to E. coli, E. cloa-cae and C. freundii AmpD carry four additionalamino acids yielding a Ser-X-X-Lys motif typi-cally found in b-lactamases and PBPs (210).However, AmpD is unlikely to belong to this



family of proteins because other signaturemotifs of b-lactam binding proteins are notpresent (212). To determine the importance ofthe carboxyl terminus of E. cloacae AmpD, astop codon was introduced, giving rise to a trun-cated ampD gene product lacking the 16 car-boxyl-terminal residues (212). Though thetruncated AmpD protein is stable, the truncatedprotein failed to complement the E. coli ampDmutation, indicating that the carboxyl terminusof C. freundii and E. cloacae AmpD was crucialfor its function (212). Meanwhile the identity ofC. freundii and E. cloacae ampE remains unre-ported; however, our analysis of the sequencedCitrobacter rodentium genome revealed the pres-ence of an AmpE homolog.Mutation of the ampD gene caused an uncon-

trolled expression of AmpC b-lactamase, lead-ing to the hypothesis that AmpD was mostlikely a negative regulator of ampC transcrip-tion. However, the primary protein structure ofAmpD had no characteristic helix-turn-helixmotif or Zn-binding finger, two well-knownDNA-binding motifs (206). So, it was postu-lated that AmpD might not affect the transcrip-tion of ampC directly, but it might act throughthe transcription factor AmpR as the constitu-tive expression required ampR (174). In addi-tion, it had been speculated that AmpD playeda role in murein metabolism as E. coli strainswith defective ampD lost up to 40% of their cellwall per generation (213, 214). The function ofAmpD was elucidated when a highly purifiedfunctional AmpD protein was incubated withE. coli sacculi (215). The presence of AmpDdecreased the relative amount of all anhydro-muropeptides, suggesting that AmpD wasinvolved in the metabolism of murein (215).Using a more refined isolate of muropeptides,AmpD was shown to degrade anhydro-N-acetyl-muramyl-L-Ala-D-Glu-DAP (anhMurNAc-trip-eptide) over time, giving rise to anhMurNAcand tripeptide (216). Hence, AmpD proteincleaved the amide bond between the carboxylgroup of anhydromuramic acid and the a-aminogroup of L-Ala-D-Glu-DAP (216). AmpDappeared to be specifically cleaving the 1,6anhydro bond on the muramic acid asMurNAc-peptides, and the cytoplasmic precur-sors of murein, UDP-MurNAc-peptides, werenot cleaved by AmpD (216). This unique 1,6 an-hydro bond was the result of an intramolecular

transglycosylation reaction that occurred whenthe peptidoglycan was degraded by the periplas-mic E. coli lytic transglycosylases (217). Mureinprecursors did not contain the 1,6 anhydrobond, allowing AmpD to distinguish the precur-sors from the murein-degraded products. Thisfunctional specificity of AmpD was similar tothe membrane-bound N-acetyl-anhydromura-myl-L-alanine amidase of E. coli (218). How-ever, instead of interacting with the cytoplasmicmembrane, AmpD is a cytosolic N-acetyl-anhydromuramyl-L-alanine amidase (216).Besides C. freundii, E. coli, and E. cloacae,

the ampDE region had also been determined inP. aeruginosa. The P. aeruginosa ampDE regionwas detected using a degenerate probe against aphage genomic library of PAO1 (219). Thisstrategy located ampD gene to a 2.6-kb genomicregion. Sequencing of this 2.6-kb DNA frag-ment revealed two ORFs (219). The P. aerugin-osa ampD gene was 567 bp long, correspondingto a gene product of 188 amino acids with a pre-dicted molecular mass of 21 kDa. The deducedP. aeruginosa AmpD protein exhibited 65%,63%, and 62% similarity to the E. coli,C. freundii, and E. cloacae AmpD proteins,respectively (219). However, our analysis showsthat although the carboxyl terminus of C. freun-dii, E. cloacae, and P. aeruginosa AmpD pro-teins are highly conserved, the Ser-X-X-Lysmotif found at the carboxyl terminus ofC. freundii and E. cloacae was not found inP. aeruginosa AmpD.Similar to the gene organization in E. coli,

the P. aeruginosa ampE start codon over-lapped the translational termination codon ofampD. P. aeruginosa ampE encoded a 278amino acid polypeptide with predicted molec-ular weight of 31 kDa (219). The predictedP. aeruginosa AmpE gene product had thecharacteristic of a cytoplasmic membrane pro-tein with five predicted transmembranedomains. However, it showed a relatively lowdegree of homology (33%) to the counterpartin E. coli (219).In spite of the sequence, virtually nothing is

known about the molecular biology of P. aeru-ginosa ampDE. Similar to E. coli, P. aeruginosaampDE was believed to form an operon andtranscript simultaneously. Unfortunately, with-out accompanying primer extension analysis orany techniques to determine the 5¢-end of the

KONG et al.


ampDE transcript, Langaee et al. claimed thatthe putative -10 and -35 promoter sequenceswere 14- and 36-bp, respectively, upstream of aputative Shine-Dalgarno sequence of ampD(219). However, without any conventional tech-niques, the authentic promoter of P. aeruginosaampDE remains elusive. In addition, our analy-sis shows the absence of any strong q-indepen-dent terminator at the 3¢-end of ampE,suggesting that the termination of the transcriptmay be q-dependent.To determine the function of P. aeruginosa

ampD and ampE gene product, two E. colistrains, ampD and ampDE mutants, harboringE. cloacae ampR-ampC genes in trans were usedfor complementation analysis (219). The pre-sence of P. aeruginosa ampD alone and ampDErestored the low basal and b-lactam antibiotic-dependent inducible levels of E. cloacae AmpCb-lactamase in the E. coli ampD and ampDEmutant strains. However, the b-lactamaseinduction was reduced two-fold and 1.5-fold inthe presence of P. aeruginosa ampE, as com-pared to the absence of ampE and E. coli ampE,respectively (219).In a subsequent analysis of P. aeruginosa

ampDE, a polar gentamycin cassette wasinserted within the ampD gene, generating anisogenic P. aeruginosa ampDE mutant, notwith-standing the author’s claim that it was anampD-defective strain (220). The ampDE muta-tion of P. aeruginosa led to six- and 32-foldincreases in the MIC of ampicillin and cefotax-ime, respectively, as well as a 37-fold increase ofthe basal non-induced b-lactamase level (220).Unlike E. coli where the loss of both ampD andampE caused a non-inducible high level of b-lac-tamase synthesis (207), b-lactamase productionof a P. aeruginosa ampDE mutant was induced22-fold in the presence of an inducer (220). Thisdiscrepancy suggested that P. aeruginosa ampDand ⁄or ampE might play a different role thanthat of E. coli. However, no complementationanalysis was performed to confirm their finding(220). Some of these discrepancies could beexplained by the recent realization that P. aeru-ginosa encodes three ampD homologs, ampD,ampDh1, and ampDh2, and these are responsiblefor step-wise upregulation of b-lactamaseexpression (221). Although single or doublemutants of the ampD homologs do not affectP. aeruginosa fitness in vitro, deletion of all three

genes does (222). This cost of fitness is due toupregulation of ampC because ampC deletionrestores fitness (222). Interestingly, in a mousemodel, ampD ampDh2 or ampD ampDh3 doublemutants as well as the triple mutant affect fit-ness. This effect of the triple mutant is ampCindependent. Of the three, ampDh3 plays thegreatest role in virulence in a mouse model sys-tem; an ampD ampDh2 double mutant also hasreduced virulence (222). This study also lookedat the ampD homologue sequences of b-lactam-ase wild-type and overproducing pairs of clini-cal isolates from 10 different subjects (222).Although six of the overproducing strains con-tained mutations in ampD, loss of functionmutations were not identified in ampDh2 orampDh3, consistent with a high biological fit-ness cost of the double mutants (222). Thesefindings also suggest that the four groups of b-lactamase producers identified by Campbellet al. (192) may harbor mutations in the variousampD genes.

ampG

The ampG gene was discovered when an E. cloa-cae strain 55M-L was reported to produce con-stitutive low levels of b-lactamase (223). Thisintriguing strain was derived from an ampDparental strain, E. cloacae 55M, which synthe-sized a high basal level of b-lactamase constitu-tively (223). Complementation analyses withwild-type ampR-ampC and ampD failed torecover the parental phenotype (224). Further-more, the 55M-L ampR-ampC and ampDregions retained the activities of the parentalstrain, suggesting that a novel locus, known asampG, which was unlinked to ampR-ampC orampD, was responsible for the low-basalb-lactamase level (224). Based on the assump-tion that low-level constitutive expression ofb-lactamase in E. cloacae was due to a novelmutation in ampG, it was proposed that a wild-type ampG allele of the gene should be dominantover the mutated allele (224). If this was correct,the introduction of the wild-type ampG alleleshould change E. cloacae 55M-L into a resistantstrain as it was mutated in the ampD gene (224).A random genomic library was constructedusing a wild-type E. cloacae chromosome andtransformed into E. cloacae strain 55M-L.Selection of the resistant transformants showed



that a 2.9-kb fragment was able to transformE. cloacae 55M-L into a b-lactam-resistantstrain that produced b-lactamase at levelscomparable to the parental strain, 55M (224).The presence of this 2.9-kb fragment alongwith the ampD gene in E. cloacae M55re-established the inducibility of the AmpCb-lactamase (224). As a mutation in ampG ledto an effect that was dominant over the pheno-types of ampR and ampD, it was proposed thatAmpG might play a role upstream of AmpRand AmpD in the induction of AmpC b-lactam-ase (224).The presence of an ampG locus in E. coli

was confirmed through Southern blotting usingthe E. cloacae 2.9-kb ampG fragment (225). A3.2-kb fragment containing ampG from E. coliwas able to complement E. cloacae 55M-L(225). An ampG knockout mutant was gener-ated in E. coli using allelic replacement viahomologous recombination (225). The resultingampG mutant was unable to induce C. freundiiAmpC b-lactamase in the presence of an indu-cer, confirming that the induction of C. freundiiAmpC required a functional ampG gene (225).Sequencing of the E. coli 3.2-kb fragment

showed the presence of two ORFs flanked byan upstream bloA gene and a downstreamcyoA gene, which encoded a morphogene thataffected carboxypeptidase activity, and the cyto-chrome o ubiquinol oxidase, respectively (225).The deduced AmpG protein sequence yielded a491 amino acid polypeptide with a predictedmolecular mass of 53.2 kDa. However, minicellscarrying E. coli ampG could not detect thepresence of AmpG in the cytoplasmic fraction(225). Bioinformatic analysis and fusion analy-sis suggested the existence of 10 membrane-spanning regions in the ampG gene product(225, 226).Upstream of ampG, there was an ORF of

579-bp in the same orientation. This precedingORF might encode a putative lipoprotein as itcontains a strong lipoprotein signal of Leu-X-Gly-Cys (225). A polar insertional inactivationof this ORF showed a non-inducible phenotypeof E. coli harboring C. freundii ampR-ampC,similar to an ampG mutant (225). But, the b-lac-tamase inducibility was regained when thestrain was complemented with ampG alone ororf-ampG, suggesting that the lipoprotein playedno role in b-lactamase induction (225).

When the E. coli ampG gene was cloned withthe upstream intergenic region, a promoter mustbe inserted upstream in order to complement achromosomal ampG mutation. This stronglysuggested that the ampG promoter was notimmediately upstream (225). Furthermore, theshort 43-bp intracistronic sequence did notcontain any E. coli r70 promoter consensus,suggesting that ampG was co-transcribed withan upstream ORF of 495-bp (225). A northernblotting analysis using an ampG probe showed aband of 2.4 kb corresponding to the transcriptsize covering both the ampG and ORF, confirm-ing the hypothesis that ampG and its upstreamORF constituted an operon (225). Using primerextension analysis, the 5¢-end of the transcriptwas mapped to a G residue, 43-bp upstream ofthe proposed AUG start codon of the upstreamORF (225). With this information in hand,the determined promoter of the ampG operonshowed high homology to the consensussequence for RpoD promoters (225).However, interestingly, when a probe hybrid-

izing to the upstream ORF was used in thenorthern blot, two transcripts of 0.7- and 2.4-kbin size were found (225). The 2.4-kb band corre-sponded to the transcript of both the ampG andthe upstream ORF, whereas the 0.7-kb tran-script coincided with the expected size of theupstream ORF alone (225). This suggests thatalthough the two genes were transcribed in anoperonic manner, expression of E. coli ampGwas post-transcriptionally regulated. Examina-tion of the intergenic region between the ORFand ampG revealed the presence of a shortinverted repeat of 11-bp upstream of the startcodon of ampG (225). This inverted repeat wasbelieved to function as a transcriptional termi-nator. In order to further determine the regula-tion of the ampG gene, two promoter fusions,Porf-lacZ and PampG-lacZ, were constructed(225). It was indicated that the b-galactosidaseactivity of Porf-lacZ was always 1.5-fold higherthan the transcription of PampG-lacZ,confirming the northern blot data that therewas down regulation of ampG transcription(225). However, the ampG transcription was notaffected by AmpD or different growth phases(225).Although AmpG protein appeared to be a

transmembrane protein, the function of AmpGremains elusive. The fact that E. coli contains

KONG et al.


ampG and ampD genes, even though this organ-ism lacks the ampR gene and an inducible ampCgene, provided clues that the primary functionsof AmpG and AmpD proteins were notinvolved in AmpC b-lactamase induction (175,205, 225). The initial clue regarding AmpGfunction came from Jacobs et al., who demon-strated that E. coli strains with a defective ampGgene lost 20–40% of the labeled muropeptidesfrom their cell wall in each generation, whereasthe control strain lost only 4–8% per generation(213). In addition, a large proportion of thelabeled muropeptides were released into themedium (213). These data indicated that E. colistrains with a defective ampG gene were unableto reuse the cell wall components during cellregeneration. The data together with the factthat AmpG was a transmembrane protein sug-gest that AmpG protein acted as a permeasethat allowed the transport of peptidoglycanrecycling intermediates from the bacterial peri-plasm to the cytoplasm (213).To ascertain the specific function of AmpG,

hot water extracts of the DAP-labeled wild-type E. coli and a E. coli strain with defectiveampG were analyzed using high-pressure liquidchromatography (HPLC) (213). It was demon-strated that the ampG mutant failed to retainanhMurNAc-peptides in the cytoplasm. How-ever, the presence of anhMurNAc-peptides, theAmpD substrates, remained peculiar becausethe unprocessed enzymatic products of cellwall hydrolases were GlcNAc-anhMurNAc-peptides (213). Thus, the ampG gene productcould be transporting either the GlcNAc-anhMurNAc-peptides or anhMurNAc-peptidesinto the cytoplasm. To determine whetherGlcNAc-anhMurNAc-peptides were cleaved inthe periplasm or cytoplasm, the culturemedium of the ampG mutant was assayed. Alarge amount of GlcNAc-anhMurNAc-pep-tides was found in the culture medium ofE. coli ampG and not in the wild-type, suggest-ing that AmpG protein was a transporter ofGlcNAc-anhMurNAc-peptides from the peri-plasm to the cytoplasm (213). The GlcNAc-anhMurNAc-peptides were then cleaved in thecytoplasm by an intracellular N-acetyl-glucosa-minidase (227, 228).In a different experiment, the substrate speci-

ficity of AmpG was investigated (229). Usingdifferent GlcNAc-anhMurNAc-peptides of

various lengths, it was shown that the principalrequirement of AmpG was the presence of Glc-NAc-anhMurNAc disaccharide, independent ofthe length of the peptide side chain (229). Muro-peptides lacking either GlcNAc or anhMurNAcwere not imported into the cytosol. In addition,the function of AmpG was sensitive to carbonylcyanide m-chlorophenylhydrazone (CCCP), anH+ ionophore that dissipates the H+ gradientand uncouples electron transport from ATPsynthesis (230), suggesting that AmpG trans-port is dependent on the proton motive force.Although the ampG gene was first detected

in E. cloacae (224), all the molecular andbiochemical work were performed on the E. coliampG gene and its gene product (213, 225).The ampG homolog of E. cloacae remains tobe characterized. Analysis of the sequencedP. aeruginosa PAO1 genome shows the presenceof two copies of ampG, currently designated asampG and ampP (Kong, K.-F. and Mathee, K.,unpublished data). Both the permeases areessential for b-lactamase induction.

MUREIN RECYCLING AND b-LACTAMASE

INDUCTION

The most important issue in the study of b-lac-tamase induction is the coordinate regulationof the presence of b-lactam antibiotics in theperiplasm, and the intracellular activation ofb-lactamase. b-Lactam antibiotics are knownto act on the murein localized in the periplasm;thus, these chemicals do not enter the cyto-plasm to mediate their effects. Yet the presenceof b-lactam agents is able to induce b-lactam-ase, suggesting that an efficient communicationmust exist between the extracellular site ofb-lactam actions and the intracellular site ofb-lactamase transcription. Because of structuralhomology between the b-lactam ring andD-alanyl-D-alanine, Tuomanen et al. proposedthat the pathway that leads to the induction ofb-lactamase expression may share regulatorycomponents with the pathway of cell wallmetabolism (214). If this is true, the authorspostulated that mutants with altered b-lactam-ase induction must have aberrant cell wallmetabolism. To test the hypothesis, the differ-ence in peptidoglycan composition betweenwild-type E. coli and an isogenic mutant with



defective ampDE was determined using HPLC(214). Interestingly, it was shown that themuropeptide composition in ampDE mutantwas profoundly altered as compared to thewild-type. However, this alteration could berecovered by complementing the ampD alonein trans, suggesting that ampE played a subtlerole in this event (214). Thus, it is proposedthat the ampD gene product repressedb-lactamase activity by retaining the concentra-tion of GlcNAc GlcMurNAc-L-Ala-D-Glu-meso-DAP-D-Ala-D-Ala (214). Two importantphenomena converged in this study: mureinrecycling and b-lactamase induction.The current molecular model of b-lactamase

induction is proposed (Fig. 6). According tothis model, during cell growth, specific cell wallhydrolyases degrade the peptidoglycan to yieldmurein peptides with different lengths. Thesemuropeptides (GlcNAc-anhMurNAc-peptides)are transported into the cytoplasm by the per-mease AmpG. They are then directly cleaved byAmpD, an N-acetylmuramyl amidase, into theN-acetylmuramic acids and peptides (Fig. 6).During normal physiological growth, ampCexpression is repressed by AmpR binding to thepromoter region (Fig. 6) (198). Jacobs et al.demonstrated that in the presence of b-lactamantibiotics, there was a relative decrease inthe murein precursors, UDP-MurNAc-L-Ala-D-Glu-meso-DAP-D-Ala-D-Ala (213). It waspostulated that the C. freundii AmpR was main-tained in the repressive form by the murein pre-cursors. This was indeed demonstrated withpurified AmpR in an in vitro transcription assaywhere the ampC transcript was diminished withan increase in UDP-MurNAc-pentapeptide(198). In addition, the negative effect of UDP-MurNAc-pentapeptide in b-lactamase induc-tion was also investigated using an AmpRmutant, AmpRGly102Glu (198). Expression of thisAmpR variant resulted in a high basal-levelconstitutive b-lactamase expression in theabsence of b-lactam agents in wild-type as wellas in an ampG mutant (231). In the transcrip-tion assay, purified AmpRGly102Glu behaved likethe wild-type protein in the activation of ampCtranscription. However, increasing concentra-tions of UDP-MurNAc-pentapeptide failed toreduce the level of AmpRGly102Glu-mediatedampC transcription (198). This result clearlyindicated that an UDP-MurNAc-pentapeptide

binding site was present in AmpR aroundamino acid 102.In contrast, gel mobility-shift assay showed

no difference in the DNA binding property bet-ween the wild-type AmpR and AmpRGly102Glu

(198). A detailed study carried out using DNa-seI footprinting also gave rise to an identicalfootprint (198). However, the presence of theUDP-MurNAc-pentapeptide resulted in asmall but consistent alteration in the pro-tection pattern of wild-type AmpR, but notAmpRGly102Glu (198). Specifically, the base inposition -49 (relative to the +1 site of ampC)was not protected by the wild-type AmpR in thepresence of the UDP-MurNAc-pentapeptide,though the significance of this nucleotide wasnot discussed (198).The identification of a repressing ligand

of AmpR was not sufficient to explain the induc-tion of ampC transcription because theincreased pool of the UDP-MurNAc-pentapep-tide in an ampD mutant exhibited a high basallevel of b-lactamase expression constitutively(213). It was then proposed that a differentligand may be responsible for the co-inductionof ampC transcription. In this case, the anhMur-NAc-tripeptide served as a potential candidateas it dramatically accumulated in an ampDmutant (213). In an in vitro transcription experi-ment, the anhMurNAc-tripeptide consistentlyexerted the negative effect of the UDP-Mur-NAc-pentapeptide on ampC transcription (198).Thus, in the presence of b-lactam antibiotics, asa result of increased murein breakdown, highconcentration of the anhMurNAc-tripeptidedisplaced the repressing UDP-MurNAc-penta-peptide from its AmpR-binding site (Fig. 6)(198). However, it was not known if thefootprint of AmpR changed with the binding tothe anhMurNAc-tripeptide.The delicate coordination of murein break-

down and b-lactamase expression was elegantlydemonstrated using a complicated heterologoussystem (Fig. 6) (198). Most of the experimentswere performed using an E. coli backgroundwhich carried the C. freundii or E. cloacaeampR-ampC system. E. coli does not have aninducible AmpC b-lactamase, even in the pres-ence of ampD and ampG. The E. coli ampC isregulated through an attenuator present at the5¢-end of the ampC transcript (182). b-Lactamresistance in E. coli clinical isolates is brought

KONG et al.


about by various mechanisms, including tan-dem amplification of the ampC gene caused byrecombination between short homologies (232),mutations to the promoter region that up-regu-lated ampC expression (233), insertion of an ISelement at the promoter that created a strongr70 promoter (234), and horizontal chromo-somal DNA transfer from Shigella (235).Thus far, no data have demonstrated the impor-tance of E. coli ampD and ampG in the tran-scriptional up-regulation of its cognate ampC.

Moreover, C. freundii and E. cloacae ampD andampG have not been sufficiently demonstratedto involve AmpC b-lactamase induction or cellwall metabolism in a homologous system. TheC. freundii ampR gene product is believed tobind to different murein breakdown productsand murein precursors present in E. coli. Theidentity and the physiological concentration ofthese molecules in C. freundii, E. cloacae,P. aeruginosa or any other organisms remain tobe defined.

Cephalosporin

ampC

OM

PG

PP

IMC

M

Muramyl-peptide??

AmpR

NagZ

??

??

AmpD

Disaccharide-pentapeptide & Disaccharide- tripeptide

UDP- Disaccharide- pentapeptide

AmpC

Am

pE

Am

pE

Am

pG

Am

pG

Fig. 6. Mechanism of b-lactamase induction in Enterobacteriaceae. In the absence of b-lactam antibiotics, themurein is constantly being remodeled for cell elongation and cell division. The breakdown products, muramylpeptides, are released into the periplasm and transported to the cytoplasm via AmpG. These muramyl peptidesare then processed by AmpD for the murein re-cycling. In this instance, the ampC gene is kept at the basal levelby a suppressed AmpR transcriptional regulator. In the presence of b-lactam antibiotics, the homeostasis is dis-rupted because these antibiotics target the murein resulting in an excessive breakdown of the murein and an accu-mulation of the muramyl peptides both in the periplasm and cytoplasm. However, the AmpD protein is thelimiting factor. An accumulation of the muramyl peptides renders the activation of AmpR, which then inducesthe production of AmpC b-lactamase. It is not clear which of the muramyl products directly interact with AmpR.AmpE has been implicated in the transport of free pentapeptide to the periplasm (40). The transport of the mura-myl peptides into the periplasm may involve AmpE or some yet to be unidentified protein. CM, IM, PP, PG, andOM refer to cytoplasm, inner membrane, periplasm, peptidoglycan, and outer membrane, respectively.



CLOSING REMARKS

It is conceivable that the recognition ofantibiotics marks one of the most importantmilestones in human history. The use ofantibiotics has brought immeasurable benefitsto medicine, physiology, and biology. Over thedecades, our understanding of antibiotics andresistance to them has facilitated a deeperunderstanding of the microbial physiology andpathogenesis. In addition, this has provided uswith tools to facilitate the search for new knowl-edge. For example, the ampicillin resistancegene is routinely used as a selectable marker inrecombinant DNA technology. However, asa community, we should also proceed withcaution because the misuse and abuse of thesesubstances will lead to unwanted consequences,including the return of multidrug-resistantbacteria.

ABBREVIATIONS

6-APA, 6-aminopenicillanic acid; 7-ACA,7-aminocephalosporanic acid; anhMurNAc,anhydro-N-acetylmuramyl; ASM, AmericanSociety of Microbiology; ATP, adenosine tri-phosphate; CARB, Carbenicillin b-lactamase;CCCP, carbonyl cyanide m-chlorophenylhyd-razone; DAP, diaminopimelic acid; DNA,deoxyribonucleic acid; EDTA, ethylenediamine-tetraacetic acid; ESBL, extended spectrumb-lactamase; frd, fumerate reductase operon;GlcNAc, N-acetylglucosamine; HPLC, high-performance liquid chromatography; kb, kilo-base; kDa, kiloDalton; MIC, minimuminhibitory concentration; MurNAc, N-acety-lmuramic acid; NIAID, National Institutes ofAllergy and Infectious Diseases; ORF, openreading frame; PampR, ampR promoter; PBC,penicillin binding component; PBP, penicillin-binding protein; PEP, phosphoenolpyruvic acid;PSE, Pseudomonas-specific b-lactamase; SDS-PAGE, sodium dodecyl sulfate-polyacrylamidegel electrophoresis; UDP-GlcNAc, Uridine diphos-phate-N-acetylglucosamine; UDP-MurNAc,Uridine diphosphate-N-acetylmuramic acid.

The authors thank members of the Mathee labora-tory for thoughtful comments on the manuscript. Thework was supported by NIH S06 GM08205 and

5SC1AI081376 (K.M.) and a Florida InternationalUniversity Teaching Assistantship (K.-F.K.).

REFERENCES

1. Bottcher HM. A History of Antibiotics. Phila-delphia: JB Lippincott Company, 1964.

2. Burdon-Sanderson J. Memoirs: The origin anddistribution of microzymes (bacteria) in water,and the circumstances which determine their exis-tence in the tissues and liquids of the living body.Quarterly JMicroscop Sci 1871;11:323–52.

3. Lister J. A contribution to the germ theory ofputrefaction and other fermentative changes,and to the natural history of torulae and bacte-ria. Trans R Soc Edinb 1875;27:313.

4. Roberts W. Studies of biogenesis. Philos TransR Soc Lond. 1874;164:457–77.

5. Tyndall J. The optical deportment of the atmo-sphere in relation to the henomena of putrefac-tion and infection. Philos TransR Soc Lond 1876;166:27–63.

6. Selwyn S. Pioneer work on the ‘‘penicillin phe-nomenon,’’ 1870–1876. J Antimicrob Chemo-ther 1979;5:248–55.

7. Pasteur L, Joubert J. Charbon et septicemia. CRAcad Sci Paris 1877;85:101–15.

8. Tyndall J. Essays on the Floating Matter of theAir in Relation to Putrefactin and Infection.New York: Appleton, 1881.

9. Cornil CV, Babes V. Concurrence vitale des bac-teries: attenuation de leurs proprietes dan demilieux nutritifs modifies par d’autres bacteries:Tentative de therapeautique. J Conaiss Med PratPharmacol 1885;7:321–3.

10. Garre C. Uber Antagonisten unter denBacterien. Corresp Bl Schweiz Aerzte 1887;17:384–92.

11. Vuillemin P. Antibiose et symbiose. CR AssocFr Av Sci 1889;2:525–43.

12. Duchesne E. Contribution a l’etude de la con-currence vitale chez les micro-organismes: an-tagonisme entre les moisissures et les microbes.Lyon: l’Ecole du Service de Sante Militaire deLyon, 1897.

13. Wainwright M. Andre Gratia (1893–1950): for-gotten pioneer of research into antimicrobialagents. J Med Biogr 2000;8:39–42.

14. Steffee CH. Alexander Fleming and penicillin.The chance of a lifetime? N C Med J 1992;53:308–10.

15. Fleming A. On the antibacterial action of cul-tures of a penicillium, with special reference totheir use in the isolation of B. influenzae. Br JExp Pathol 1929;10:226–36.

16. Hare R. New light on the history of penicillin.Med His 1982;26:1–24.

KONG et al.


17. Chain E. The early years of the penicillin discov-ery. Trends Pharmacol Sci J 1979;1:6–11.

18. Abraham EP, Chain E. An enzyme from bacteriaable todestroypenicillin.Nature 1940;146:837.

19. Abraham EP, Chain E, Fletcher CM, FloreyHW, Gardner AD, Heatley NG, et al. Furtherobservations on penicillin. Eur J Clin Pharmacol1941;42:3–9.

20. Henderson JW. The yellow brick road to peni-cillin: a story of serendipity. Mayo Clin Proc1997;72:683–7.

21. Richards AN. Statement released by the Com-mittee on Medical Research. J Am Med Assoc1943;122:235–9.

22. Fleming A. Streptococcal meningitis treatedwith penicillin. Measurement of bacteriostaticpower of blood and cerebrospinal fluid. Lancet1943;2:434–8.

23. Florey HW, Florey ME. General and local ad-ministration of penicillin. JAMA1943;1:387–97.

24. Hobby GL, Meyer K, Chaffee E. Chemothera-peutic activity of penicillin. Proc Soc Exp BiolMed 1942;50:285–8.

25. Keefer CS, Blake FG, Marshall ER, LockwoodJS, Wood WB. Penicillin in the treatment ofinfections. A report of 500 cases. Statement bythe Committee on Chemotherapeutic and otheragents, Division of Medical Sciences, NationalResearch Council. JAMA 1943;122:1217–44.

26. Rammelkamp CH, Keefer CS. The absorption,excretion, and distribution of penicillin. J ClinInvest 1943;22:425–37.

27. Rammelkamp CH, Keefer CS. Penicillin: itsantibacterial effect in whole blood and serum forthe hemolytic Streptococcus and Staphylococ-cus aureus. J Clin Invest 1943;22:649–57.

28. Queener SF. History and origins of beta-lactamantibiotics. In: Queener SF,Webber JA, QueenerSW, editors. Beta-Lactam Antibiotics for Clini-cal Use. New York: Marcel Dekker, Inc, 1986.pp. 3–16.

29. Abraham EP. Historical introduction. In:Florey HW, Heatley NG, Jennings MA, Sand-ers AG, Abraham EP, Florey ME, editors.Antibiotics. Vol. 2. London: Oxford UniversityPress, 1949. pp. 768.

30. Behrens OK, Corse J, HuffDE, Jones RG, SoperQF, Whitehead CW. Biosynthesis of penicillinsIII. Preparation and evaluation of precursors ofnew penicillins. J Biol Chem 1948;175:771–92.

31. Sheehan JC, Henery-Logan KR. A general syn-thesis of the penicillins. J Am Chem Soc1959;81:5838–9.

32. Fairbrother RW, Taylor G. Sodium methicillinin routine therapy. Lancet 1961;1:473–6.

33. Acred P, Brown DM, Knudsen ET, RolinsonGN, Sutherland R. New semi-synthetic penicil-lin against Pseudomonas pyocyanea. Nature1967;215:25–30.

34. Abraham EP, Newton GGF. The structure ofcephalosporin C. Biochem J 1961;79:377–93.

35. Nagarajan R, Boeck LD, Gorman M, HamillRL, Higgens CE, Hoehn MM, et al. Beta-lac-tam antibiotics from Streptomyces. J Am ChemSoc 1971;93:2308–10.

36. Imada A, Kitano K, Kintaka K, Muroi M, AsaiM. Sulfazecin and isosulfazecin, novel beta-lactam antibiotics of bacterial origin. Nature1981;289:590–1.

37. Sykes RB, Cimarusti CM, Bonner DP, Bush K,Floyd DM, Georgopapadakou NH, et al.Monocyclic beta-lactam antibiotics produced bybacteria. Nature 1981;291:489–91.

38. Well JS, Trejo WH, Bush K, GeorpapadakouNH, Bonner DP, Sykes RB. EM5400, a familyof monobactam antibiotics produced by Agro-bacterium radiobacter. I. Taxonomy, fermenta-tion and biological properties. J Antibiotics1982;35:295–9.

39. Jacoby GA. AmpC beta-lactamases. ClinMicrobiol Rev 2009;22:161–82.

40. Park JT, Uehara T. How bacteria consume theirown exoskeletons (turnover and recycling of cellwall peptidoglycan). Microbiol Mol Biol Rev2008;72:211–27.

41. Zapun A, Contreras-Martel C, Vernet T.Penicillin-binding proteins and beta-lactam resis-tance. FEMSMicrobiol Rev 2008;32:361–85.

42. Gardner AD. Morphological effects of penicillinon bacteria. Nature 1940;146:837–8.

43. Duguid JP. The sensitivity of bacteria to theaction of penicillin. EdinbMed J 1946;53:401–12.

44. Salton MRJ. Specific detection of glucose onpaper chromatograms. Nature 1960;186:966–7.

45. Bartholomew JW, Mittwer T. The Gram stain.Bacteriol Rev 1952;16:1–29.

46. Salton MRJ. The relationship between the nat-ure of the cell wall and the Gram stain. J GenMicrobiol 1963;30:223–35.

47. Chapman GB, Hiller J. Electron microscopy ofultra-thin sections of bacteria. I. Cellular divisionin Bacillus cereus. J Bacteriol 1953;66:362–73.

48. Amako K, Murata K, Umeda A. Structure ofthe envelope of Escherichia coli observed by therapid freezing and substitution fixation method.Microbiol Immunol 1983;27:95–9.

49. Hobot JA, Carlemalm E, Villiger W, Kellenber-ger E. Periplasmic gel: new concept resultingfrom the reinvestigation of bacterial cell enve-lope ultra-structure by new methods. J Bacteriol1984;160:143–52.

50. Park JT, Johnson MJ. Accumulation of labilephosphate in Staphylococcus aureus grown inthe presence of penicillin. J Biol Chem 1949;179:585–92.

51. Park JT. Uridine-5¢-pyrophosphate derivatives.III. Amino acid-containing derivatives. J BiolChem 1952;194:897–904.



52. Park JT, Strominger JL. Mode of action of peni-cillin. Science 1957;125:99–101.

53. Salton MRJ, Horne RW. Studies of the bacte-rial cell wall. II. Methods of preparation andsome properties of the cell wall. Biochim Bio-phys Acta 1951;7:177–97.

54. Strange RE, Dark FA. An unidentified amino-sugar present in cell walls and spores of variousbacteria. Nature 1956;177:186–8.

55. Work E. Biochemistry of the bacterial cell wall.Nature 1957;179:841–7.

56. Cummins C. The chemical composition of thebacteria cell wall. Int Rev Cytol 1956;5:25–50.

57. Brumfitt W, Wardlaw AC, Park JT. Develop-ment of lysozyme-resistance in Micrococcus lys-odiekticus and its association with an increasedO-acetyl content of the cell wall. Nature 1958;181:1783–4.

58. Salton MRJ. Bacterial cell wall. IV. Composi-tion of the cell wall of Gram-positive and Gram-negative bacteria. Biochim Biophys Acta1953;10:512–23.

59. Rietschel ET, Westphal O. History of endotox-ins. In: Brade SMOH, Vogel SN, Morrison DC,editors. Endotoxins in Health and Disease. NewYork: Marcel Dekker, 1999: 1–30.

60. Repaske R. Lysis of Gram-negative organismsand the role of versene. Biochim Biophys Acta1958;30:225–32.

61. Weidel W, Primosigh J. Biochemical parallelsbetween lysis by virulent phage and lysisby penicillin. J Gen Microbiol 1958;18:513–7.

62. Weidel W, Frank H, Martin HH. The rigid layerof the cell wall of Escherichia coli strain B.J Gen Microbiol 1960;22:158–66.

63. Mandelson J. Isolation of lysozyme-soluble mu-copeptides from the cells wall of Escherichiacoli. Nature 1961;189:855–6.

64. van Heijenoort J. Formation of the glycanchains in the synthesis of bacterial peptidogly-can. Glycobiology 2001;11:25R–36R.

65. Siewert G, Strominger JL. Biosynthesis of thepeptidoglycan of bacterial cell walls. XI. Forma-tion of the isoglutamine amide group in the cellwalls of Staphylococcus aureus. J Biol Chem1968;243:783–90.

66. Araki Y, Shimada A, Ito E. Effect of penicillinon cell wall mucopeptide synthesis in a Escheri-chia coli particulate system. Biochim BiophysRes Commun 1966;23:518–25.

67. Izaki K, Matsuhashi M, Strominger JL. Bio-synthesis of the peptidoglycan of bacterialcell walls. VIII. Peptidoglycan transpeptidaseand D-alanine carboxypeptidase: penicillin-sensitive enzymatic reaction in strains ofEscherichia coli. J Biol Chem 1968;243:3180–92.

68. Tipper DJ, Strominger JL. Biosynthesis of thepeptidoglycan of bacterial cell walls. XII. Inhi-bition of cross-linking by penicillins and cepha-losporins: studies in Staphylococcus aureus invivo. J Biol Chem 1968;243:3169–79.

69. Tipper DJ, Strominger JL. Mechanism of actionof penicillins: a proposal based on their struc-tural similarity to acyl-D-alanyl-D-alanine. ProcNatl Acad Sci USA 1965;54: 1133–41.

70. Wise EM, Park JT. Penicillin: its basic site ofaction as an inhibitor of a peptide cross-linkingreaction in cell wall mucopeptide synthesis. ProcNatl Acad Sci USA 1965;54:75–81.

71. Izaki K, Strominger JL. Biosynthesis of thepeptidoglycan of bacterial cell walls. XIV.Purification and properties of two D-alaninecarboxypeptidases from Escherichia coli. J BiolChem 1968;243:3193–201.

72. Goffin C, Ghuysen JM. Multimodular penicil-lin-binding proteins: an enigmatic family oforthologs and paralogs. Microbiol Mol BiolRev 1998;62:1079–93.

73. Cooper PD, Rowley D. Investigations withradioactive penicillin. Nature 1949;163:480–1.

74. Cooper PD, Rowley D, Dawson IM. Locationof radioactive penicillin in Staphylococcus aur-eus after contact with the drug. Nature 1949;164:842.

75. Maass EA, Johnson MJ. Penicillin uptake bybacterial cells. J Bacteriol 1949;57:415–22.

76. Maass EA, Johnson MJ. The relations betweenbound penicillin and growth in Staphylococcusaureus. J Bacteriol 1949;58:361–6.

77. Schepartz SA, Johnson MJ. The nature of thebinding of penicillin by bacterial cells. J Bacte-riol 1956;71:84–90.

78. Cooper PD. The site of action of penicillin:some properties of the penicillin-bindingcomponent of Staphylococcus aureus. J GenMicrobiol 1955;12:100–6.

79. Spratt BG, Pardee AB. Penicillin-binding pro-teins and cell shape in Escherichia coli. Nature1975;254:516–7.

80. Georgopapadakou NH, Liu FY. Penicillin-binding proteins in bacteria. Antimicrob AgentsChemother 1980;18:148–57.

81. Suginaka H, Blumberg PM, Strominger JL.Multiple penicillin-binding components in Bacil-lus subtilis, Bacillus cereus, Staphylococcusaureus and Escherichia coli. J Biol Chem 1972;247:5279–88.

82. Sauvage E, Kerff F, Terrak M, Ayala JA, Char-lier P. The penicillin-binding proteins: structureand role in peptidoglycan biosynthesis. FEMSMicrobiol Rev 2008;32:234–58.

83. Ghosh AS, Chowdhury C, Nelson DE. Physio-logical functions of D-alanine carboxypeptidasesin Escherichia coli. Trends Microbiol 2008;16:309–17.

KONG et al.


84. Vollmer W, Bertsche U. Murein (peptidoglycan)structure, architecture and biosynthesis inEscherichia coli. Biochim Biophys Acta 2008;1778:1714–34.

85. Vollmer W, Joris B, Charlier P, Foster S. Bacte-rial peptidoglycan (murein) hydrolases. FEMSMicrobiol Rev 2008;32:259–86.

86. Spratt BG. Distinct penicillin binding proteinsinvolved in the division, elongation, andshape of Escherichia coli K12. Proc Natl AcadSci USA 1975;72:2999–3003.

87. Spratt BG, Jobanputra V, Schwarz U. Mutantsof Escherichia coli which lack a component ofpenicillin-binding protein 1 are viable. FEBSLett 1977;79:374–8.

88. Suzuki H, Nishimura Y, Hirota Y. On theprocess of cellular division in Escherichia coli: aseries of mutants of E. coli altered in the penicil-lin-binding proteins. Proc Natl Acad Sci USA1978;75:664–8.

89. Nakagawa J, Tamaki S, Tomioka S, Matsuh-ashi M. Functional biosynthesis of cell wallpeptidoglycan by polymorphic bifunctionalpolypeptides. Penicillin-binding protein 1Bs ofEscherichia coli with activities of transglycosy-lase and transpeptidase. J Biol Chem 1984;259:13937–46.

90. Tamaki S, Nakajima S, Matsuhashi M. Ther-mosensitive mutation in Escherichia coli simul-taneously causing defects in penicillin-bindingprotein-1Bs and in enzyme activity for peptido-glycan synthesis in vitro. Proc Natl Acad SciUSA 1977;74:5472–6.

91. Tamura T, Suzuki H, Nishimura Y, Mizogu-chi J, Hirota Y. On the process of cellulardivision in Escherichia coli: isolation andcharacterization of penicillin-binding proteins1a, 1b, and 3. Proc Natl Acad Sci USA 1980;77:4499–503.

92. Schiffer G, Holtje JV. Cloning and characteriza-tion of PBP 1C, a third member of the multi-modular class A penicillin-binding proteins ofEscherichia coli. J Biol Chem1999;274:32031–9.

93. Park JT, Burman L. FL-1600: a new penicillinwith a unique mode of action. J Bacteriol 1973;51:863–8.

94. Iwaya M, Goldman R, Tipper DJ, Feingold B,Strominger JL. Morphology of an Escherichiacoli mutant with a temperature-dependentround cell shape. J Bacteriol 1978;136:1143–58.

95. de la Rosa EJ, de Pedro MA, Vazquez D. Peni-cillin binding proteins: role in initiation of mur-ein synthesis in Escherichia coli. J Bacteriol1985;82:5632–5.

96. Matsuhashi S, Kamiryo T, Blumberg PM, Lin-nett P, Willoughby E, Strominger JL. Mecha-nism of action and development of resistanceto a new amidino penicillin. J Bacteriol 1974;117:578–87.

97. Essig P, Martin HH, Gmeiner J. Murein andlipopolysaccharide biosynthesis in synchronizedcells of Escherichia coli K 12 and the effect ofpenicillin G, mecillinam and nalidixic acid. ArchMicrobiol 1982;132:245–50.

98. Ogura T, Bouloc P, Niki H, D’Ari R, Hirag-a S, Jaffe A. Penicillin-binding protein 2 isessential in wild-type Escherichia coli but notin lov or cya mutants. J Bacteriol 1989;171:3025–30.

99. Normark S. Mutation in Escherichia coli K-12mediating spherelike envelopes and changedtolerance to ultraviolet irradiation and someantibiotics. J Bacteriol 1969;98:1274–7.

100. Matsuzawa H, Hayakawa K, Sato T, ImahoriK. Characterization and genetic analysis of amutant of Escherichia coli K-12 with roundedmorphology. J Bacteriol 1973;115:436–42.

101. Iwaya M, Jones CW, Khorana J, StromingerJL. Mapping of the mecillinam-resistant, roundmorphological mutants of Escherichia coli.J Bacteriol 1978;133:196–202.

102. Spratt BG. Temperature-sensitive cell divisionmutants of Escherichia coli with thermolabilepenicillin-binding proteins. J Bacteriol 1977;131:293–305.

103. Spratt BG, Boyd A, Stoker N. Defective andplaque-forming lambda transducing bacterio-phage carrying penicillin-binding protein-cellshape genes: genetic and physical mappingand identification of gene products from thelip-dacA-rodA-pbpA-leuS region of the Escheri-chia coli chromosome. J Bacteriol 1980;143:569–81.

104. Ishino F, Park W, Tomioka S, Tamaki S, Tak-ase I, Kunugita K, et al. Peptidoglycan syn-thetic activities in membranes of Escherichia colicaused by overproduction of penicillin-bindingprotein 2 and RodA protein. J Biol Chem1986;261:7024–31.

105. Vinella D, D’Ari R, Jaffe A, Bouloc P. Penicillinbinding protein 2 is dispensable in Escherichiacoli when ppGpp synthesis is induced. EMBO J1992;11:1493–501.

106. Vinella D, Joseleau-Petit D, Thevenet D, BoulocP, D’Ari R. Penicillin-binding protein 2 inacti-vation in Escherichia coli results in cell divisioninhibition, which is relieved by FtsZ overexpres-sion. J Bacteriol 1993;175:6704–10.

107. Schmidt LS, Botta G, Park JT. Effects of fur-azlocillin, a beta-lactam antibiotic which bindsselectively to penicillin-binding protein 3, onEscherichia coli mutants deficient in other peni-cillin-binding proteins. J Bacteriol 1981;145:632–7.

108. Botta GA, Park JT. Evidence for involvement ofpenicillin-binding protein 3 in murein synthesisduring septation but not during cell elongation.J Bacteriol 1981;145:333–40.



109. Spratt BG. Distinct penicillin binding proteinsinvolved in the division, elongation, and shapeof Escherichia coli K12. Proc Natl Acad SciUSA 1975;72:2999–3003.

110. Tamura T, Imae Y, Strominger JL. Purifica-tion to homogeneity and properties of twoD-alanine carboxypeptidases I from Escherichi-a coli. J Biol Chem 1976;251:414–23.

111. Iwaya M, Strominger JL. Simultaneous deletionof D-alanine carboxypeptidase IB-C and penicil-lin-binding component IV in a mutant ofEscherichia coli K12. Proc Natl Acad Sci USA1977;74:2980–4.

112. Matsuhashi M, Takagaki Y, Maruyama IN,Tamaki S, Nishimura Y, Suzuki H, et al.Mutants of Escherichia coli lacking in highlypenicillin-sensitive D-alanine carboxypeptidaseactivity. Proc Natl Acad Sci USA 1977;74:2976–9.

113. Korat B, Mottl H, Keck W. Penicillin-bindingprotein 4 of Escherichia coli: molecular cloningof the dacB gene, controlled overexpression, andalterations in murein composition. Mol Micro-biol 1991;5:675–84.

114. de Pedro MA, Schwarz U. Heterogeneity ofnewly inserted and preexisting murein in the sac-culus of Escherichia coli. Proc Natl Acad SciUSA 1981;78:5856–60.

115. Spratt BG, Strominger JL. Identification of themajor penicillin-binding proteins of Escherichiacoli as D-alanine carboxypeptidase IA. J Bacte-riol 1976;127:660–3.

116. Spratt BG. Deletion of the penicillin-bindingprotein 5 gene of Escherichia coli. J Bacteriol1980;144:1190–2.

117. Matsuhashi M, Tamaki S, Curtis SJ, StromingerJL. Mutational evidence for identity of penicil-lin-binding protein 5 in Escherichia coli withthe major D-alanine carboxypeptidase IA activ-ity. J Bacteriol 1979;137:644–7.

118. Broome-Smith JK, Ioannidis I, Edelman A,Spratt BG. Nucleotide sequences of the penicil-lin-binding protein 5 and 6 genes of Escherichiacoli. Nucleic Acid Res 1988;16:1617.

119. Amanuma H, Strominger JL. Purification andproperties of penicillin-binding proteins 5 and 6from Escherichia coli membranes. J Biol Chem1980;255:11173–80.

120. Broome-Smith JK, Spratt BG. Deletion of thepenicillin-binding protein 6 gene of Escherichiacoli. J Bacteriol 1982;152:904–6.

121. Buchanan CE, Sowell MO. Synthesis of penicil-lin-binding protein 6 by stationary-phaseEscherichia coli. J Bacteriol 1982;151:491–4.

122. Baquero MR, Bouzon M, Quintela JC, AyalaJA, Moreno F. dacD, an Escherichia coli geneencoding a novel penicillin-binding protein(PBP6b) with DD-carboxypeptidase activity.J Bacteriol 1996;178:7106–11.

123. Henderson TA, Dombrosky PM, Young KD.Artifactual processing of penicillin-binding pro-teins 7 and 1b by the OmpT protease of Escheri-chia coli. J Bacteriol 1994;176:256–9.

124. Romeis T, Holtje JV. Penicillin-binding protein7 ⁄ 8 of Escherichia coli is a DD-endopeptidase.Eur J Biochem 1994;224:597–604.

125. Malouin F, Chamberland S, Brochu N, ParrTR Jr. Influence of growth media on Escherichi-a coli cell composition and ceftazidime suscepti-bility. Antimicrob Agents Chemother 1991;35:477–83.

126. Henderson TA, Templin M, Young KD. Identi-fication and cloning of the gene encoding peni-cillin-binding protein 7 of Escherichia coli.J Bacteriol 1995;177:2074–9.

127. Armstrong GL, Conn LA, Pinner RW. Trendsin infectious disease mortality in the UnitedStates during the 20th century. JAMA 1999;281:61–6.

128. Cohen M. Changing patterns of infectious dis-ease. Nature 2000;406:762–7.

129. Kirby W. Extraction of a highly potent penicil-lin inactivator from penicillin-resistant staphylo-cocci. Science 1944;99:452–3.

130. Rammelkamp CH, Maxon T. Resistance ofStaphylococcus aureus to action of penicillin.Proc Soc Exp Biol Med 1942;51:386–9.

131. Gallardo E. Sensitivity of bacteria from infec-ted wounds to penicillin. II. Results in onehundred and twelve cases. War Med 1945;7:100–3.

132. Rantz LA, Kirby WMM. Action of penicillin onStaplylococcus in vitro. J Immunol 1944;48:335–43.

133. Selbie FR, Simon RD, McIntosh J. Bacteriolog-ical aspects of penicillin therapy. J Pathol Bacte-riol 1942;57:47–58.

134. Finland M. Emergence of antibiotic-resistantbacteria. N Engl J Med 1955;253:909–22.

135. Finland M, Frank PF, Wilcox C. In vitro sus-ceptibility of pathogenic staphylococci to sevenantibiotics. With a note on the changing resis-tance of staphylococci to penicillin. Am J ClinPathol 1950;20:325–34.

136. Barber M. Staphylococcal infection due to peni-cillin-resistant strains. Br J Med 1947;2:863–5.

137. Barber M. Coagulase-positive staphylococciresistant to penicillin. J Pathol Bacteriol 1947;59:373–84.

138. Barber M, Rozwadowska-Dowzenko M. Infec-tion by penicillin-resistant staphylococci. Lancet1948;2:641–4.

139. Barber M, Whitehead JEM. Bacteriophagetypes in penicillin-resistant staphylococcal infec-tion. Br Med J 1949;2:565–9.

140. Finland M. Changing ecology of bacterial infec-tions as related to antibacterial therapy. J InfectDis 1970;122:419–31.

KONG et al.


141. Rogers DE. The changing pattern of life-threat-ening microbial disease. N Engl J Med 1959;261:677–83.

142. Sneader W. Drug Discovery: A History. WestSussex: John Wiley & Sons, Ltd., 2005.

143. Travis J. Reviving the antibiotic miracle? Sci-ence 1994;264:360–2.

144. Greenwood D. Tarnished gold: sixty years ofantimicrobial drug use and misuse. J MedMicrobiol 1995;43:395–6.

145. Culotta E. Funding crunch hobbles antibioticresistance research. Science 1994;264:362–3.

146. Majiduddin FK, Materon IC, Palzkill TG.Molecular analysis of beta-lactamase structureand function. Int J Med Microbiol 2002;292:127–37.

147. Ambler RP. The structure of beta-lactamases.Philos Trans R Soc Lond B Biol Sci 1980;289:321–31.

148. Bush K, Jacoby GA, Medeiros AA. A func-tional classification scheme for beta-lactamasesand its correlation with molecular structure.Antimicrob Agents Chemother 1995;39:1211–33.

149. Mitsuhashi S, Inoue M. Mechanisms of Resis-tance to Beta-Lactam Antibiotics. New York:Springer-Verlag, 1981.

150. Richmond MH, Sykes RB. The beta-lactamasesof Gram-negative bacteria and their possiblephysiological role. Adv Microb Physiol 1973;9:31–88.

151. Couture F, Lachapelle J, Levesque RC. Phylog-eny of LCR-1 and OXA-5 with class A and classD beta-lactamases. Mol Microbiol 1992;6:1693–703.

152. Matagne A, Misselyn-Bauduin AM, Joris B,Erpicum T, Granier B, Frere JM. The diversityof the catalytic properties of class A beta-lacta-mases. Biochem J 1990;265:131–46.

153. Boissinot M, Levesque RC. Nucleotidesequence of the PSE-4 carbenicillinase gene andcorrelations with the Staphylococcus aureusPC1 beta-lactamase crystal structure. J BiolChem 1990;265:1225–30.

154. Lachapelle J, Dufresne J, Levesque RC.Characterization of the blaCARB-3 gene encod-ing the carbenicillinase-3 beta-lactamase ofPseudomonas aeruginosa. Gene 1991;102:7–12.

155. Naas T, Philippon L, Poirel L, Ronco E,Nordmann P. An SHV-derived extended-spec-trum beta-lactamase in Pseudomonas aerugin-osa. Antimicrob Agents Chemother 1999;43:1281–4.

156. Dufresne J, Vezina G, Levesque RC. Cloningand expression of the imipenem-hydrolyzingbeta-lactamase operon from Pseudomonasmaltophilia in Escherichia coli. AntimicrobAgents Chemother 1988;32:819–26.

157. Watanabe M, Iyobe S, Inoue M, Mitsuhashi S.Transferable imipenem resistance in Pseudomo-nas aeruginosa. Antimicrob Agents Chemother1991;35:147–51.

158. Senda K, Arakawa Y, Ichiyama S, NakashimaK, Ito H, Ohsuka S, et al. PCR detection ofmetallo-beta-lactamase gene (blaIMP) in Gram-negative rods resistant to broad-spectrum beta-lactams. J Clin Microbiol 1996;34:2909–13.

159. Lauretti L, Riccio ML, Mazzariol A, CornagliaG, Amicosante G, Fontana R, et al. Cloningand characterization of blaVIM, a new integron-borne metallo-beta-lactamase gene from a Pseu-domonas aeruginosa clinical isolate. AntimicrobAgents Chemother 1999;43:1584–90.

160. Tsakris A, Pournaras S, Woodford N, PalepouMF, Babini GS, Douboyas J, et al. Outbreak ofinfections caused by Pseudomonas aeruginosaproducing VIM-1 carbapenemase in Greece.J Clin Microbiol 2000;38:1290–2.

161. Sanders CC. Chromosomal cephalosporinasesresponsible for multiple resistance to newerbeta-lactam antibiotics. Annu Rev Microbiol1987;41:573–93.

162. Galleni M, Lindberg F, Normark S, Cole S,Honore N, Joris B, et al. Sequence and compar-ative analysis of three Enterobacter cloacaeampC beta-lactamase genes and their products.Biochem J 1988;250:753–60.

163. Lodge JM, Minchin SD, Piddock LJ, BusbyJW. Cloning, sequencing and analysis of thestructural gene and regulatory region of thePseudomonas aeruginosa chromosomal ampCbeta-lactamase. Biochem J 1990;272:627–31.

164. Dale JW, Smith JT. A direct comparison of twounusual R-factor-mediated beta-lactamases.Biochem J 1972;128:173–4.

165. Medeiros AA, Cohenford M, Jacoby GA. Fivenovel plasmid-determined beta-lactamases. An-timicrob Agents Chemother 1985;27:715–9.

166. Kong KF, Jayawardena SR, Del Puerto A,Wiehlmann L, Laabs U, Tummler B, et al.Characterization of poxB, a chromosomal-enco-ded Pseudomonas aeruginosa oxacillinase. Gene2005;358:82–92.

167. Girlich D, Naas T, Nordmann P. Biochemicalcharacterization of the naturally occurring oxa-cillinase OXA-50 of Pseudomonas aeruginosa.Antimicrob Agents Chemother 2004;48: 2043–8.

168. Poirel L, Weldhagen GF, De Champs C,Nordmann P. A nosocomial outbreak ofPseudomonas aeruginosa isolates expressing theextended-spectrum beta-lactamase GES-2 inSouth Africa. J Antimicrob Chemother 2002;49:561–5.

169. Sanders CC, Sanders WE. Microbial resistanceto newer generation beta-lactam antibiotics:clinical and laboratory implications. J Infect Dis1985;151:399–405.



170. Sabath LD, Jago M, Abraham EP. Cephalospo-rinase and penicillinase activities of a beta-lac-tamase from Pseudomonas pyocyanea. BiochemJ 1965;96:739–52.

171. Sykes RB, Matthew M. The beta-lactamases ofGram-negative bacteria and their role in resis-tance to beta-lactam antibiotics. J AntimicrobChemother 1976;2:115–57.

172. Proenca R, Niu WW, Cacalano G, Prince A.The Pseudomonas cepacia 249 chromosomalpenicillinase is a member of the AmpC familyof chromosomal beta-lactamases. AntimicrobAgents Chemother 1993;37:667–74.

173. Seoane A, Francia MV, Garcia Lobo JM.Nucleotide sequence of the ampC-ampR regionfrom the chromosome of Yersinia enterocoli-tica. Antimicrob Agents Chemother 1992;36:1049–52.

174. Lindberg F, Westman L, Normark S. Regula-tory components in Citrobacter freundii ampCbeta-lactamase induction. Proc Natl Acad SciUSA 1985;82:4620–4.

175. Honore N, Nicolas MH, Cole ST. Induciblecephalosporinase production in clinical isolatesof Enterobacter cloacae is controlled by a regu-latory gene that has been deleted from Escheri-chia coli. EMBO J 1986;5:3709–14.

176. Poirel L, Guibert M, Girlich D, Naas T, Nord-mann P. Cloning, sequence analyses, expression,and distribution of ampC-ampR from Morga-nella morganii clinical isolates. AntimicrobAgents Chemother 1999;43:769–76.

177. Mahlen SD, Morrow SS, Abdalhamid B, Han-son ND. Analyses of ampC gene expression inSerratia marcescens reveal new regulatoryproperties. J Antimicrob Chemother 2003;51:791–802.

178. Lau SK, Ho PL, Li MW, Tsoi HW, Yung RW,Woo PC, et al. Cloning and characterization of achromosomal class C beta-lactamase and its reg-ulatory gene in Laribacter hongkongensis. Anti-microbAgents Chemother 2005;49:1957–64.

179. Nadjar D, Labia R, Cerceau C, Bizet C,Philippon A, Arlet G. Molecular characteriza-tion of chromosomal class C beta-lactamase andits regulatory gene in Ochrobactrum anthropi.AntimicrobAgents Chemother 2001;45:2324–30.

180. Jaurin B, Grundstrom T. ampC cephalosporin-ase of Escherichia coli K-12 has a differentevolutionary origin from that of beta-lactamasesof the penicillinase type. Proc Natl Acad SciUSA 1981;78:4897–901.

181. Jaurin B, Normark S. In vivo regulation ofchromosomal beta-lactamase in Escherichiacoli. J Bacteriol 1979;138:896–902.

182. Jaurin B, Grundstrom T, Edlund T, Normark S.The E. coli beta-lactamase attentuator mediatesgrowth rate-dependent regulation. Nature 1981;290:221–5.

183. Gootz TD, Sanders CC. Characterization ofbeta-lactamase induction in Enterobacter cloa-cae. Antimicrob Agent Chemother 1893;23:91–7.

184. Hennessey TD. Inducible beta-lactamase inEnterobacter. J Gen Microbiol 1967;49:277–85.

185. Seeberg AH, Tolxdorff-Neutzling RM, Wiede-mann B. Chromosomal beta-lactamases ofEnterobacter cloacae are responsible for resis-tance to third-generation cephalosporins. Anti-microb Agents Chemother 1983;23:918–25.

186. Yamamoto T, Murayama SY, Sawai T. Cloningand expression of the gene(s) for cephalosporin-ase production of Citrobacter freundii. Mol GenGenet 1983;190:85–91.

187. Lindberg F, Normark S. Contribution of chro-mosomal beta-lactamases to beta-lactam resis-tance in enterobacteria. Rev Infect Dis 1986;8:S292–304.

188. Giwercman B, Rasmussen JW, Ciofu O, Clem-mentsen I, Schumacher H, Hoiby N. Antibodiesagainst chromosomal beta-lactamase. Antimic-rob Agent Chemother 1994;38:2306–10.

189. Goering RV, Sanders CC, Sanders WEJ, GuayR, Guerin S. Heterogeneity in ampR-ampCgene interaction in Enterobacter cloacae. RevInfect Dis 1988;10:786–92.

190. Walther-Rasmussen J, Johnsen AH, Hoiby N.Terminal truncations in ampC beta-lactamasefrom a clinical isolate of Pseudomonas aerugin-osa. Eur J Biochem 1999;263:478–85.

191. Gates ML, Sanders CC, Goering RV, SandersWE. Evidence for multiple forms of type I chro-mosomal beta-lactamase in Pseudomonas aeru-ginosa. Antimicrob Agents Chemother 1986;30:453–7.

192. Campbell JI, Ciofu O, Høiby N. Pseudomonasaeruginosa isolates from patients with cysticfibrosis have different b-lactamase expressionphenotypes but are homogeneous in the ampC-ampR genetic region. Antimicrob Agents Chem-ther 1997;41:1380–4.

193. Bagge N, Ciofu O, Hentzer M, Campbell JI,Givskov M, Hoiby N. Constitutive highexpression of chromosomal beta-lactamase inPseudomonas aeruginosa caused by a newinsertion sequence (IS1669) located in ampD.Antimicrob Agents Chemother 2002;46:3406–11.

194. Bergstrom S, Lindberg FP, Olsson O, Nor-mark S. Comparison of the overlapping frdand ampC operons of Escherichia coli withthe corresponding DNA sequences in otherGram-negative bacteria. J Bacteriol 1983;155:1297–305.

195. Lindberg F, Normark S. Common mechanismof ampC beta-lactamase induction in enterobac-teria: regulation of the cloned Enterobacter clo-acae P99 beta-lactamase gene. J Bacteriol 1987;169:758–63.

KONG et al.


196. Schell MA. Molecular biology of the LysR fam-ily of transcriptional regulators. Annu RevMicrobiol 1993;47:597–626.

197. Lindquist S, Lindberg F, Normark S. Bindingof the Citrobacter freundii AmpR regulator toa single DNA site provides both autoregulationand activation of the inducible ampC beta-lactamase gene. J Bacteriol 1989;171:3746–53.

198. Jacobs C, Frere JM, Normark S. Cytosolic inter-mediates for cell wall biosynthesis and degrada-tion control inducible beta-lactam resistance inGram-negative bacteria.Cell 1997;88:823–32.

199. Kong KF, Jayawardena SR, Indulkar SD, DelPuerto A, Koh CL, Hoiby N, et al. Pseudomo-nas aeruginosa AmpR is a global transcriptionalfactor that regulates expression of AmpC andPoxB beta-lactamases, proteases, quorum sens-ing, and other virulence factors. AntimicrobAgents Chemother 2005;49:4567–75.

200. Lodge J, Busby S, Piddock L. Investigation ofthe Pseudomonas aeruginosa ampR gene and itsrole at the chromosomal ampC beta-lactamasepromoter. FEMS Microbiol Lett 1993;111:3315–20.

201. Gualerzi CO, Pon CL. Initiation of mRNAtranslation in prokaryotes. Biochemistry 1990;29:5881–9.

202. Preston M, Seed P, Toder D, Iglewski B, Oh-man D, Gustin J, et al. Contribution of prote-ases and LasR to the virulence of Pseudomonasaeruginosa during corneal infections. Infect Im-mun 1997;65:3086–90.

203. Toder DS, Gambello MJ, Iglewski BH. Pseudo-monas aeruginosa LasA: a second elastaseunder the transcriptional control of LasR. MolMicrobiol 1991;5:2003–10.

204. Nicolas MH, Honore N, Jarlier V, Philippon A,Cole ST. Molecular genetics analysis of cephalo-sporinase production and its role in beta-lactamresistance in clinical isolates of Enterobactercloacae. Antimicrob Agents Chemother 1987;31:295–9.

205. Lindberg F, Lindquist S, Normark S. Inactiva-tion of the ampD gene causes semiconstitutiveoverproduction of the inducible Citrobacterfreundii beta-lactamase. J Bacteriol 1987;169:1923–8.

206. Honore N, Nicolas MH, Cole ST. Regulation ofenterobacterial cephalosporinase production:the role of a membrane-bound sensory trans-ducer. Mol Microbiol 1989;3:1121–30.

207. Lindquist S, Galleni M, Lindberg F, NormarkS. Signaling proteins in enterobacterial AmpCbeta-lactamase regulation. Mol Microbiol1989;3:1091–102.

208. Keilty S, Rosenberg M. Constitutive function ofa positively regulated promoter reveals newsequences essential for activity. J Biol Chem1987;262:6389–95.

209. Ponnambalam S, Chan B, Busby S. Functionalanalysis of different sequence elements in theEscherichia coli galactose operon P2 promoter.Mol Microbiol 1988;2:165–72.

210. Spratt BG, Cromie KD. Penicillin-binding pro-teins of Gram-negative bacteria. Rev Infect Dis1988;10:699–711.

211. Korfmann G, Sanders CC, Moland ES. Alteredphenotypes associated with ampD mutations inEnterobacter cloacae. Antimicrob Agents Che-mother 1991;35:358–64.

212. Kopp U, Wiedemann B, Lindquist S, NormarkS. Sequences of wild-type and mutant ampDgenes of Citrobacter freundii and Enterobactercloacae. Antimicrob Agents Chemother 1993;37:224–8.

213. Jacobs C, Huang LJ, Bartowsky E, NormarkS, Park JT. Bacterial cell wall recycling pro-vides cytosolic muropeptides as effectors forbeta-lactamase induction. EMBO J 1994;13:4684–94.

214. Tuomanen E, Lindquist S, Sande S, Galleni M,Light K, Gage D, et al. Coordinate regulationof beta-lactamase induction and peptidoglycancomposition by the amp operon. Science 1991;251:201–4.

215. Holtje JV, Kopp U, Ursinus A, Wiedemann B.The negative regulator of beta-lactamase induc-tion AmpD is a N-acetyl-anhydromuramyl-L-alanine amidase. FEMS Microbiol Lett 1994;122:159–64.

216. Jacobs C, Joris B, Jamin M, Klarsov K, VanBeeumen J, Mengin-Lecreulx D, et al. AmpD,essential for both beta-lactamase regulation andcell wall recycling, is a novel cytosolic N-acety-lmuramyl-L-alanine amidase. Mol Microbiol1995;15:553–9.

217. Holtje JV, Tuomanen EI. The murein hydrolas-es of Escherichia coli: properties, functions andimpact on the course of infections in vivo. J GenMicrobiol 1991;137:441–54.

218. van Heijenoort J, Parquet C, Flouret B, vanHeijenoort Y. Envelope-bound N-acetylmura-myl-L-alanine amidase of Escherichia coli K 12.Purification and properties of the enzyme. Eur JBiochem 1975;58:611–9.

219. Langaee TY, Dargis M, Huletsky A. An ampDgene in Pseudomonas aeruginosa encodes anegative regulator of AmpC beta-lactamaseexpression. Antimicrob Agents Chemother1998;42:3296–300.

220. Langaee TY, Gagnon L, Huletsky A. Inactiva-tion of the ampD gene in Pseudomonas aeruginosaleads to moderate-basal-level and hyperinducibleAmpC beta-lactamase expression. AntimicrobAgents Chemother 2000;44:583–9.

221. Juan C, Moya B, Perez JL, Oliver A. Stepwiseupregulation of the Pseudomonas aeruginosachromosomal cephalosporinase conferring high-



level beta-lactam resistance involves threeAmpD homologues. Antimicrob Agents Che-mother 2006;50:1780–7.

222. Moya B, Juan C, Alberti S, Perez JL, Oliver A.Benefit of having multiple ampD genes foracquiring beta-lactam resistance without losingfitness and virulence in Pseudomonas aerugin-osa. Antimicrob Agents Chemother 2008;52:3694–700.

223. Werner V, Sanders CC, Sanders JWE, GoeringRV. Role of beta-lactamases and outer mem-brane proteins in multiple beta-lactam resistanceof Enterobacter cloacae. Antimicrob AgentsChemother 1985;27:455–9.

224. Korfmann G, Sanders CC. ampG is essential forhigh-level expression of AmpC beta-lactamasein Enterobacter cloacae. Antimicrob AgentsChemother 1989;33:1946–51.

225. Lindquist S, Weston-Hafer K, Schmidt H, PulC, Korfmann G, Erickson J, et al. AmpG, a sig-nal transducer in chromosomal beta-lactamaseinduction. Mol Microbiol 1993;9:703–15.

226. Chahboune A, Decaffmeyer M, Brasseur R,Joris B. Membrane topology of the Escherichiacoli AmpG permease required for recycling ofcell wall anhydromuropeptides and AmpC beta-lactamase induction. Antimicrob Agents Che-mother 2005;49:1145–9.

227. Cheng Q, Li H, Merdek K, Park JT. Molecularcharacterization of the beta-N-acetylglucosa-minidase of Escherichia coli and its role in cellwall recycling. J Bacteriol 2000;182:4836–40.

228. Votsch W, Templin MF. Characterization of abeta-N-acetylglucosaminidase of Escherichiacoli and elucidation of its role in muropeptiderecycling and beta-lactamase induction. J BiolChem 2000;275:39032–8.

229. Cheng Q, Park JT. Substrate specificity of theAmpG permease required for recycling of cellwall anhydro-muropeptides. J Bacteriol 2002;184:6434–6.

230. Ghoul M, Pommepuy M, Moillo-Batt A,Cormier M. Effect of carbonyl cyanide m-chlorophenylhydrazone on Escherichia colihalotolerance. Appl Environ Microbiol 1989;55:1040–3.

231. Bartowsky E, Normark S. Purification andmutant analysis of Citrobacter fruendii AmpR,the regulator for chromosomal AmpC beta-lac-tamase. Mol Microbiol 1991;5:1715–25.

232. Edlund T, Normark S. The Escherichia colibeta-lactamase attenuator mediates growthrate-dependent regulation. Nature 1981;292:269–71.

233. Jaurin B, Grundstrom T, Normark S. Sequenceelements determining ampC promoter strengthin Escherichia coli. EMBO J 1982;1:875–81.

234. Jaurin B, Normark S. Insertion of IS2 creates anovel ampC promoter in Escherichia coli. Cell1983;32:809–16.

235. Olsson O, Bergstrom S, Lindberg FP, NormarkS. ampC beta-lactamase hyperproduction inEscherichia coli: natural ampicillin resistancegenerated by horizontal chromosomal DNAtransfer from Shigella. Proc Natl Acad Sci USA1983;80:7556–60.

236. Keseler IM, Bonavides-Martinez C, Collado-Vides J, Gama-Castro S, Gunsalus RP, JohnsonDA, et al. EcoCyc: a comprehensive view ofEscherichia coli biology. Nucleic Acids Res2009;37:D464–70.

237. Romero P, Karp P. PseudoCyc, a pathway-gen-ome database for Pseudomonas aeruginosa.J Mol Microbiol Biotechnol 2003;5:230–9.

238. Winsor GL, Lo R, Sui SJ, Ung KS, HuangS, Cheng D, et al. Pseudomonas aeruginosaGenome Database and PseudoCAP: facilitat-ing community-based, continually updated,genome annotation. Nucleic Acids Res 2005;33:D338–43.

239. Winsor GL, Van Rossum T, Lo R, Khaira B,Whiteside MD, Hancock RE, et al. Pseudo-monas Genome Database: facilitatinguser-friendly, comprehensive comparisons ofmicrobial genomes. Nucleic Acids Res2009;37:D483–8.

240. Sauvage E, Powell AJ, Heilemann J, Jose-phine HR, Charlier P, Davies C, et al. Crystalstructures of complexes of bacterial DD-pep-tidases with peptidoglycan-mimetic ligands:the substrate specificity puzzle. J Mol Biol2008;381:383–93.

241. Fonze E, Vanhove M, Dive G, Sauvage E, FrereJM, Charlier P. Crystal structures of theBacillus licheniformis BS3 class A beta-lactam-ase and of the acyl-enzyme adduct formed withcefoxitin. Biochemistry 2002;41:1877–85.

242. Ghuysen JM. Serine beta-lactamases and peni-cillin-binding proteins. Annu Rev Microbiol1991;45:37–67.

KONG et al.


Beta-lactam antibiotics: from antibiosis to resistance …szabojudit.web.elte.hu/antibiotikum...

Documents

Transcript of Beta-lactam antibiotics: from antibiosis to resistance …szabojudit.web.elte.hu/antibiotikum...