What causes antibiotic resistance in bacteria?

Bacteria have evolved numerous strategies for resisting the action of antibiotics and antibacterial agents. This is particularly true of those bacteria that are antibiotic producers. Bacteria that produce antibiotics do so to gain a selective advantage over other, competing microbes in their natural environment. If they were sensitive to their own metabolic products, such a selective advantage would be lost.

In many hospital units, exploitation of antibiotics is very intensive and this generates an enormous selective pressure for bacteria to acquire the means by which they may become antibiotic-resistant. Under such circumstances, it is not unusual to find that bacteria exhibit resistance to more than one group of antibiotics. Resistance to a particular agent may be accomplished by more than one resistance mechanism. Bacteria may display antibiotic resistance by one or more of the following mechanisms:-

Chlamydia do not have peptidoglycan and are not susceptible to the action of penicillins.

Peptidoglycan in Gram-negative bacteria is inaccessible to penicillins that cannot penetrate the Gram-negative outer membrane. Efflux pumps can actively pump out antibiotics from cells. Gram-negative bacteria resist the activity of tetracyclines by this important mechanism.

Trimethoprim resistance is manifest by alterations in the DHFR target enzyme; quinolone resistance is effected by point mutations in the DNA gyrase, which prevent binding of the drug to its target.

Important examples include the huge range of b-lactamases and the various aminoglycoside-modifying enzymes. Chloramphenicol resistance is most often manifest by acetylation by the chloramphenicol acetyl transferase enzyme.

Meticillin resistance in meticillin-resistant Staphylococcus aureus results from the production of an additional penicillin binding protein: PBP2', which is not susceptible to inhibition by penicillins.

Not all bacteria have peptidoglycan in their cell wall. Rickettsias and chlamydia, for example, lack peptidoglycan. Such bacteria are intrinsically resistant to the action of cell wall inhibitors such as the penicillins and cephalosporins. Interestingly, although chlamydia do not make peptidoglycan they do possess penicillin-binding proteins. This raises the fascinating question, why? What role do these enzymes fulfil in a bacterium that does not have a cell wall?

Having a target that is inaccessible to antibiotics may be achieved in a variety of ways. The outer membrane of Gram-negative bacteria may act as a permeability barrier for antibiotics. Many Gram-negative bacteria are intrinsically resistant to antibiotics like benzyl penicillin because such drugs cannot penetrate the outer membrane and so cannot reach their target.

Alterations to the side chain attached to the penicillin nucleus may overcome the problem of membrane penetration and semi-synthetic penicillins such as ampicillin have a broad-spectrum of activity, encompassing both Gram-positive and Gram-negative bacteria. Another way in which bacteria deny access of an antibiotic to its target is actively to pump the drug out of the cell. Gram-negative bacteria may resist the activity of tetracyclines through an energy-dependent active efflux of the drug.

Modification of the antibiotic target is often seen in laboratory generated mutants. For example, bacteria resistant to trimethoprim produce an alternative dihydrofolate reductase. Resistance to the quinolone antimicrobials results from point mutations in the gene encoding DNA gyrase. Aminoglycoside resistance may result from modifications of the ribosome structure. Indeed, in the laboratory ribosomes may be further altered so that they only function in the presence of aminoglycosides. The drug acts to stabilise the functional ribosome in aminoglycoside-dependent bacteria.

Many clinically important bacteria produce enzymes that are capable of chemically modifying or destroying antibiotics. Chloramphenicol may be acetylated by the action of chloramphenicol acetyltransferases. Aminoglycosides may be acetylated by aminoglycoside acetyltransferases, phosphorylated by aminoglycoside phosphotransferases or conjugated with nucleotides. Such modifications render the antibiotic inactive.

Antibiotics may also be enzymatically degraded to an inactive form. The b-lactam bond can be hydrolysed by a large family of enzymes known as the b-lactamases. Some b-lactamases have a preferential activity against penicillins and these are referred to as penicillinases. Cephalosporinases are more active against cephalosporins.

Recently broad-spectrum b-lactamases have evolved that have activity against both penicillins and cephalosporins. There are families of such enzymes that have arisen as the result of point mutations accumulating in the genes that code for penicillinases. Many of these new enzymes are encoded by self-transmissible plasmids and these new resistance determinants can spread with great ease.

Not all b-lactamase activity is associated with bacterial cells. Human kidney cells produce an analogous enzyme that, although it does not readily attack penicillins and cephalosporins, rapidly destroys carbapenems such as imipenem. Because of this, imipenem is administered together with cilastatin, an inhibitor of the human kidney enzyme. This delays the breakdown of imipenem sufficiently to permit it to be active in treating bacterial infection.

Staphylococci have been associated with the production of b-lactamase for many years. Early in the history of the development of semi-synthetic penicillins, compounds were manufactured that were able to resist the activity of staphylococcal penicillinase. These drugs had side-chains that prevented the staphylococcal b-lactamase from binding to the antibiotic and hydrolysing it. Meticillin, a penicillin that is stable in the presence of staphylococcal b-lactamase, was introduced into clinical practice during the 1960's. Until recently, meticillin was known as "methicillin".

Shortly after the introduction of meticillin into medical practice, resistant strains of Staphylococcus aureus were isolated from hospital units where the drug was in regular use. Meticillin-resistance is greater at 30^oC than at 37^oC. Resistance is due to the temperature-sensitive production of an extra penicillin binding protein, PBP 2', that is not susceptible to inhibition by meticillin. Meticillin-resistant Staphylococcus aureus also produce a b-lactamase and they are generally resistant to a very wide range of antimicrobials. Infections caused by meticillin-resistant Staphylococcus aureus can thus be difficult to treat.

In some cases the only drugs available effectively to treat infections caused by meticillin-resistant Staphylococcus aureus are the glycopeptides such as vancomycin. Bacterial resistance to this agent was unknown until recently. Vancomycin resistance first appeared in enterococci. These bacteria are resistant to all currently available standard antimicrobial therapies. In an experiment of dubious ethical status, the gene encoding vancomycin resistance was transferred in the laboratory from a vancomycin-resistant enterococcus into a meticillin-resistant Staphylococcus aureus. In 1997, the first naturally occurring vancomycin-resistant, meticillin-resistant Staphylococcus aureus appeared in Japan. New antibiotics used in the treatment of MRSA include synercid - a combination of quinupristin and dalfopristin - and the new broad-spectrum tetracycline tigecycline. It remains to be seen how long these will remain effective.

Some antimicrobial resistance genes have only ever been found located on the bacterial chromosome. Others have been found to lie on plasmids. Plasmids encoding antibiotic resistance are often called resistance factors, R-factors or R-plasmids. R-plasmids may encode resistance to several unrelated antibiotics.

Some R-plasmids are self-transmissible and can move from strain to strain, even between different bacterial genera. Other R-plasmids, although not self-transmissible may be mobilised by other plasmids, that need not necessarily encode antibiotic resistance. Furthermore, antibiotic resistance genes are frequently located within transposons. These can move more or less at random around the bacterial genome.

Genes encoded by transposons may thus spread very easily because many transposable elements may become associated with transmissible plasmids. In this way, antibiotic resistance genes may become rapidly disseminated. Control of antimicrobial resistance in pathogenic microbes is one of the greatest challenges currently facing medical microbiology. If we are unsuccessful, we will surely enter the post-antibiotic era.

Six babies on the SCBU developed systemic infection caused by Serratia marcescens and two died as a result of their infection. All six isolates were resistant to netilmicin; the aminoglycoside then used to treat neonates because of its apparent lack of toxicity. Netilmicin was used as empirical therapy on the SCBU at the time of the incident.

Netilmicin-sensitive Serratia marcescens were isolated during environmental sampling of the SCBU during this incident. Exposure of these bacteria to netilmicin or to kanamycin quickly leads to the selection of a resistance phenotype that is relatively slow to return to a sensitive phenotype.

Resistance in these bacteria was the result of increased expression of the aac6' gene, encoding an aminoglycoside modifying enzyme. The AAC6' acetylates the aminoglycosides at their 6' site, thus inactivating them. Kanamycin and netilmicin are inactivated by this enzyme; other clinically important aminoglycosides do not act as substrates for this enzyme.

Mutations in the gene promoter are responsible for its overproduction in resistant isolates of Serratia marcescens. ALL strains of this bacterium carry a copy of the aac6' gene and thus have the potential to mutate to a resistant phenotype. In most strains, this gene is expressed at very low levels and thus bacteria appear susceptible upon laboratory testing. Exposure of sensitive Serratia marcescens rapidly selects resistant mutants that only very slowly revert to a sensitive phenotype. The AAC6' enzyme probably has an important yet unidentified role in the metabolism of Serratia marcescens.

The tet(M) gene was first observed in streptococci but, in 1985, it first caused resistance in Neisseria gonorrhoeae. Treatment failures appeared more or less simultaneously in Georgia, Pennsylvania and New Hampshire. Within six months, tetracycline-resistant Neisseria gonorrhoeae had appeared in seventeen states in the USA.

In addition, in 1985, tetracycline-resistant isolates of Neisseria gonorrhoeae were found in Holland. This mirrored an event ten years previously, when penicillin resistant Neisseria gonorrhoeae evolved independently in Africa and Asia. Penicillin resistance is caused by the production of a b-lactamase.

In 1990, strains of tetracycline-resistant penicillin-resistant Neisseria gonorrhoeae had evolved. Tetracycline-resistant Neisseria gonorrhoeae had also spread globally. The first tetracycline-resistant Neisseria gonorrhoeae in the United Kingdom appeared in 1987, isolated in Leeds.

The tet(M) gene is often found in a transposon, but not in Neisseria gonorrhoeae, where it is stably inserted into a plasmid. The structure of the "American" and the "Dutch" plasmids are different and close inspection shows that the tet(M) genes on the plasmids are also distinct from one another. Tetracycline-resistant Neisseria gonorrhoeae isolated from Gabon have a "Dutch-type" plasmid but an "American-type" gene.

The following bacterial genera harbour the tet(M) gene on their chromosomes:

• Mycoplasma;
• Ureaplasma;
• Gardnerella;
• Campylobacter;
• Clostridium;
• Listeria;
• Fusobacterium;
• Veillonella;
• Peptostreptococcus;

while in the following genera the tet(M) gene has a plasmid location:

• Kingella;
• Eikenella;
• Gardnerella
• Haemophilus;
• Neisseria;
• Clostridium;
• Listeria;
• Fusobacterium.

Over a ten-week period in the paediatric oncology unit of St James' University Hospital, six children had septicaemia caused by a member of the Enterobacteriaceae that produced an extended-spectrum b-lactamase. In the case of one child, two separate cases of septicaemia were recorded.

• 28 Escherichia coli,
• 28 Klebsiella oxytoca,
• 11 Klebsiella pneumoniae,
• 10 Citrobacter freundii,
• 3 Enterobacter sp. &
• 1 Serratia marcescens.

Resistance to ceftazidime was conferred by b-lactamases of the TEM family. These bacteria all showed multiple resistance and were resistant to antibiotics in classes other than the b-lactams.

• spectrum of antimicrobial activity,
• biochemical examination including enzyme kinetics,
• iso-electric focusing &
• nucleotide sequence determination.