33 Attacking the enemy

antimicrobial agents and chemotherapy

The interactions between host, microbial pathogen and antimicrobial agent can be considered as a triangle, and any alteration in one side will inevitably affect the other two sides (Fig. 33.1). In this chapter, two sides of the triangle will be examined in greater detail:

• the interactions between antimicrobial agents and microorganisms

• the interactions between antimicrobial agents and the human host.

Figure 33.1 The interactions between antimicrobial agents, microorganisms and the human host can be viewed as a triangle. Any effect on one side of the triangle will have effects on the other two sides.

Laboratory aspects of antimicrobial susceptibility tests and assays will also be outlined. The third side of the triangle, the interactions between microorganisms and the human host, has been considered in detail in the preceding chapters. The concluding part of the present chapter will draw together the three sides of the triangle.

Selective toxicity

The term ‘selective toxicity’ was proposed by the immunochemist Paul Ehrlich (Box 33.1, Fig. 33.2). Selective toxicity is achieved by exploiting differences in the structure and metabolism of microorganisms and host cells; ideally, the antimicrobial agent should act at a target site present in the infecting organism, but absent from host cells. This is more likely to be achievable in microorganisms that are prokaryotes than in those that are eukaryotes, as the former are structurally more distinct from the host cells. (A comparison of the cellular organization of prokaryotic and eukaryotic cells is given in Ch. 1.) At the other end of the spectrum, viruses are difficult to attack because of their obligate intracellular lifestyle. A successful antiviral agent must be able to enter the host cell, but inhibit and damage only a virus-specific target. The desirable features of ideal antimicrobial agents are summarized in Box 33.2.

Box 33.1 Lessons in Microbiology

Paul Ehrlich (1854–1915)

Just as Pasteur towers over immunomicrobiology, Ehrlich (Fig. 33.2) is the father figure of immunochemistry. His contributions to the science of medicine at all levels are quite extraordinary. He was the first to propose that foreign antigens were recognized by ‘side-chains’ on cells (1890), a brilliant insight that took 70 years to confirm. He also discovered the mast cell, invented the acid-fast stain for the tubercle bacillus, and devised a method to manufacture and commercialize a strong diphtheria antitoxin. He pioneered the development of antibiotics with his work on ‘606’ (or ‘Salvarsan’), a treatment for syphilis, for which he was denounced by the church for interfering with God’s punishment for sin.

While working on the treatment of infections caused by trypanosomes he set forth the concept of ‘selective toxicity’, as illustrated by the following quote: ‘But, gentlemen, it should be made clear that in general this task is much more complicated than that using serum therapy. These chemical agents, in contrast to the antibodies, may be harmful to the body. When such an agent is given to a sick organism, a difference must exist between the toxicity of this agent to the parasite and its toxicity to the host. We must always be aware of the fact that these agents are able to act on other parts of the body as well as on the parasites.’

Like Pasteur, he had a grasp of the continuum from the whole body to the cell and the three-dimensional structure of molecules, and throughout his life, he stressed the importance of molecular interaction as the basis of all biologic function; this is summed up in his famous maxim corpora non agunt nisi fixata or ‘things do not interact unless they make contact’. A Nobel Prize winner in 1908, his name was systematically eliminated from the records by the Nazi regime on account of his Jewish birth, but he was restored to honour by a reconstruction of his laboratory at the Seventh International Congress of Immunology in Berlin in 1989.

Figure 33.2 Paul Ehrlich (1854–1915).

Box 33.2 Desired Properties of a New Antimicrobial Agent

In the design of new antimicrobial agents, both antimicrobial activity and pharmacologic properties of the antibiotic for the host have to be considered.

Antimicrobial properties

• selectivity for microbial rather than mammalian targets

• cidal activity (antibacterial and antifungal agents)

• slow emergence of resistance

• narrow spectrum of activity.^a

Pharmacologic activities

• non-toxic to the host

• long plasma half-life (once-a-day dosing)

• good tissue distribution including CSF

• low plasma-protein binding

• oral and parenteral dosing forms

• no interference with other drugs.

^a The desired attribute depends on drug usage. Narrow-spectrum drugs cause less disturbance to normal flora and may contribute less to emergence of antibiotic resistance, whereas broad-spectrum compounds are more useful for empiric therapy and treatment of polymicrobial infections. CSF, cerebrospinal fluid.

Discovery and design of antimicrobial agents

The term ‘antibiotic’ has traditionally referred to natural metabolic products of fungi, actinomycetes and bacteria that kill or inhibit the growth of microorganisms. Antibiotic production has been particularly associated with soil microorganisms and, in the natural environment, is thought to provide a selective advantage for organisms in their competition for space and nutrients. Although the majority of antibacterial agents in clinical use today are derived from natural products of fermentation, most are then chemically modified (i.e. semi-synthetic) to improve their antibacterial or pharmacologic properties. However, some agents are totally synthetic (e.g. sulphonamides, quinolones). Therefore, the term ‘antibacterial’ or ‘antimicrobial’ agent is often used in preference to ‘antibiotic’. Agents used against fungi, parasites, and viruses can also be included under antimicrobials, but the terms antifungals, antiprotozoans, anthelmintics, and antivirals are more often used.

The discovery of new antimicrobial agents used to be entirely a matter of chance. Pharmaceutical companies undertook massive screening programmes searching for new soil microorganisms that produced antibiotic activity. In the light of our greater understanding of the mechanisms of action of existing antimicrobials, the processes have become rationalized, searching either for new natural products by target-site-directed screening or synthesizing molecules predicted to interact with a microbial target. Genomic approaches to the identification of novel targets have revolutionized this approach. In addition, knowledge of the crystal structure of the key enzymes involved in viral replication such as protease, reverse transcriptase and helicase leads to the design of new drugs. The steps in a rational design programme are summarized in Box 33.3.

Box 33.3 Rational Design of an Antimicrobial Agent

The discovery process of new antimicrobial agents has moved away from the random screening of soil microorganisms towards a rational design programme. From discovery to development and marketing can take up to 15 years and cost US$800 million. This list identifies different steps in this programme (average 10 years).

• Select an appropriate target.

• Identify a chemical lead (i.e. a new molecule with inhibitory activity on the target).

• Modify the lead compound to enhance potency.

• Evaluate in vitro activity.

• Evaluate in vivo activity and toxicity.

• Test in clinical trials and develop.

Classification of antibacterial agents

There are three ways of classifying antibacterial agents:

1. according to whether they are bactericidal or bacteriostatic

2. by target site

3. by chemical structure.

Some antibacterial agents are bactericidal, others are bacteriostatic

Some antibacterial agents kill bacteria (bactericidal), while others only inhibit their growth (bacteriostatic). Thus, the bactericidal process is irreversible, while bacteriostasis is reversible. Nevertheless, bacteriostatic agents are successful in the treatment of some infections because they prevent the bacterial population from increasing and host defence mechanisms can consequently cope with the static population. However, in immunocompromised patients, bacteriostatic drugs may be less efficacious, and certain infections (e.g., endocarditis) require a bactericidal drug even in an immunocompetent patient.

As a means of classification, the distinction between bactericidal and bacteriostatic agents is blurred because some agents are capable of killing some species, but are only bacteriostatic for others, e.g. chloramphenicol inhibits growth of Escherichia coli, but kills Haemophilus influenzae.

There are five main target sites for antibacterial action

A convenient way of classifying antibacterials is on the basis of their site of action. This classification does not allow an accurate prediction of which antibacterials will be active against which bacterial species, but it does help in the understanding of the molecular basis of antibacterial action, and conversely in the elucidation of many of the synthetic processes in bacterial cells. The five main target sites for antibacterial action are:

• cell wall synthesis

• protein synthesis

• nucleic acid synthesis

• metabolic pathways

• cell membrane function.

These targets differ to a greater or lesser degree from those in the host (human) cells and so allow inhibition of the bacterial cell without concomitant inhibition of the equivalent mammalian cell targets (selective toxicity).

Each target site encompasses a multitude of synthetic reactions (enzymes and substrates), each of which may be specifically inhibited by an antibacterial agent. A range of chemically diverse molecules may inhibit different reactions at the same target site (e.g. protein synthesis inhibitors).

Antibacterial agents have diverse chemical structures

Classification based on chemical structure alone is not of practical use, because there is such diversity. However, a combination of target site and chemical structure provides a useful working classification to organize antibacterial agents into specific families which will be discussed later in this chapter.

Resistance to antibacterial agents

Resistance to antibacterial agents is a matter of degree. In the medical setting, we define a resistant organism as one that will not be inhibited or killed by an antibacterial agent at concentrations of the drug achievable in the body after normal dosage. ‘Some men are born great, some achieve greatness, and some have greatness thrust upon them’ (William Shakespeare, Twelfth Night). Likewise, some bacteria are born resistant, others have resistance thrust upon them. In other words, some species are innately resistant to some families of antibiotics because they lack a susceptible target, are impermeable to or enzymatically inactivate the antibacterial agent. The Gram-negative rods with their outer membrane layer exterior to the cell wall peptidoglycan are less permeable to large molecules than Gram-positive cells. However, within species that are innately susceptible, there are also strains that develop or acquire resistance.

The genetics of resistance

In parallel with the rapid development of a wide range of antibacterial agents since the 1940s, bacteria have proved extremely adept at developing resistance to each new agent that comes along. This is illustrated for Staphylococcus aureus by the timeline shown in Figure 33.3. The rapidly increasing incidence of resistance associated with slowing down in the discovery of novel antibacterial agents to combat resistant strains is now recognized worldwide as a serious threat to the treatment of life-threatening infections.

Figure 33.3 ‘Time line’ illustrating the chronological emergence of antibiotic resistance in Gram-positive cocci.

Chromosomal mutation may result in resistance to a class of antimicrobial agents (cross-resistance)

Resistance may arise from:

• a single chromosomal mutation in one bacterial cell resulting in the synthesis of an altered protein: for example, streptomycin resistance via alteration in a ribosomal protein, or the single amino acid change in the enzyme dihydropteroate synthetase resulting in a lowered affinity for sulphonamides. A mutational event could also alter (i.e., increase or decrease) the production of a protein resulting in increased resistance.

• a series of mutations, for example changes in penicillin-binding proteins (PBPs) in penicillin-resistant pneumococci.

In the presence of antibiotic, these spontaneous mutants have a selective advantage to survive and outgrow the susceptible population (Fig. 33.4A). They can also spread to other sites in the same patient or by cross-infection to other patients and therefore become disseminated. Chromosomal mutations are relatively rare events (i.e. usually found once in a population of 10⁶–10⁸ organisms) and generally provide resistance to a single class of antimicrobials (i.e. ‘cross-resistance’ to structurally related compounds).

Figure 33.4 A chromosomal mutation (A) can produce a drug-resistant target, which confers resistance on the bacterial cell and allows it to multiply in the presence of antibiotic. Resistance genes carried on plasmids (B) can spread from one cell to another more rapidly than cells themselves divide and spread. Resistance genes on transposable elements (C) move between plasmids and the chromosome and from one plasmid to another, thereby allowing greater stability or greater dissemination of the resistance gene.

Genes on transmissible plasmids may result in resistance to different classes of antimicrobial agents (multiple resistance)

Not content with surviving the antibacterial onslaught by relying on random chromosomal mutation, bacteria are also able to acquire resistance genes on transmissible plasmids (Fig. 33.4B; see also Ch. 2). Such plasmids often code for resistance determinants to several unrelated families of antibacterial agents. Therefore a cell may acquire ‘multiple’ resistance to many different drugs (i.e. in different classes) at once, a process much more efficient than chromosomal mutation. This so-called ‘infectious resistance’ was first described by Japanese workers studying enteric bacteria, but is now recognized to be widespread throughout the bacterial world. Some plasmids are promiscuous, crossing species barriers, and the same resistance gene is therefore found in widely different species. For example, TEM-1, the most common plasmid-mediated beta-lactamase in Gram-negative bacteria, is widespread in E. coli and other enterobacteria and also accounts for penicillin resistance in Neisseria gonorrhoeae and ampicillin resistance in H. influenzae.

Resistance may be acquired from transposons and other mobile elements

Resistance genes may also occur on transposons; the so-called ‘jumping genes’, which by a replicative process are capable of generating copies which may integrate into the chromosome or into plasmids (see Ch. 2). The chromosome provides a more stable location for the genes, but they will be disseminated only as rapidly as the bacteria divide. Transposon copies moving from the chromosome to plasmids are disseminated more rapidly. Transposition can also occur between plasmids, for example, from a non-transmissible to a transmissible plasmid, again accelerating dissemination (Fig. 33.4C).

‘Cassettes’ of resistance genes may be organized into genetic elements called integrons

As discussed previously, antibiotic-resistance genes may individually reside on plasmids, the chromosome, or on transposons found in both locations. However, in some instances multiple resistance genes may come together in a structure known as an integron. As shown in Figure 33.5A, the integron encodes a site-specific recombination enzyme (int gene; integrase), which allows insertion (and also excision) of antibiotic-resistance gene ‘cassettes’ (resistance gene plus additional sequences including an ‘attachment’ region) into the integron attachment site (att). In classic operon fashion, a strong integron promoter controls transcription of the inserted genes. Based on their integration mechanism (integrase, etc.), integrons have been organized into different classes found in both Gram-negative and Gram-positive organisms. Whether acting as independent mobile genetic elements or inserted into transposons, integrons are capable of moving into a variety of DNA molecules, the overall hierarchy of which is depicted in Figure 33.5B. With their ability to capture, organize and rearrange different antibiotic-resistance genes, integrons represent an important mechanism for the spread of multiple antibiotic resistance in clinically important microorganisms.

Figure 33.5 (A) Basic integron structure and (B) overall interrelationship between integrons and other DNA elements. att, integron attachment site; int, integrase.

Staphylococcal genes for methicillin resistance are organized into a unique cassette structure

Staphylococcal genes responsible for resistance to the antibiotic methicillin (discussed below) are found in a specialized cassette arrangement termed staphylococcal chromosomal casette mec (SCCmec). SCCmec inserts into a unique target site on the staphylococcal chromosome. The cassette represents a highly recombinogenic region which may not only rearrange internally but also serve as a target for the insertion of other resistance elements (e.g, transposons and plasmids).

Mechanisms of resistance

Resistance mechanisms can be broadly classified into three main types. These are summarized below, in Table 33.1 and described in more detail where relevant for each antibiotic in later parts of this chapter. Where bacterial mechanisms of antimicrobial resistance have been elucidated, they appear to involve the synthesis of new or altered proteins. As mentioned above, the genes encoding these proteins may be found on plasmids or the chromosome.

Table 33.1 Mechanisms of resistance can be classified into three main types

The target site may be altered

The target may be altered so that it has a lowered affinity for the antibacterial, but still functions adequately for normal metabolism to proceed. Alternatively, an additional target (e.g. enzyme) may be synthesized.

Access to the target site may be altered (altered uptake or increased exit)

This mechanism involves decreasing the amount of drug that reaches the target by either:

• altering entry, for example by decreasing the permeability of the cell wall

• pumping the drug out of the cell (known as an efflux mechanism).

Enzymes that modify or destroy the antibacterial agent may be produced (drug inactivation)

There are many examples of such enzymes, the most important being:

• beta-lactamases

• aminoglycoside-modifying enzymes

• chloramphenicol acetyl transferases.

These will be described in the relevant parts on these antibiotics.

Classes of antibacterial agents

The following parts of this chapter deal with groups of antibacterial agents based on their target site and chemical structure. In each case, the discussion attempts to summarize the answers to the questions set out in Table 33.2, reviewing the interactions between antibacterial agent and bacteria and between the antibacterial and the host (i.e. two sides of the triangle in Fig. 33.1).

Table 33.2 In order to understand the nature and optimum use of an antibacterial agent, the questions listed here must be answered

What is it?	Chemical structure: natural or synthetic product
What does it do?	Target site, mechanism of action
Where does it go? (and therefore preferred route of administration)	Absorption, distribution, metabolism and excretion of the drug in the body of the host
When is it used?	Spectrum of activity and important clinical uses
What are the limitations to its use?	Toxicity to the human host; lack of toxicity, i.e. resistance of the bacteria
How much does it cost?	Great variation between agents but cost is a serious limitation on availability of some agents in resource-poor countries

Inhibitors of cell wall synthesis

Peptidoglycan, a vital component of the bacterial cell wall (see Ch. 2), is a compound unique to bacteria and therefore provides an optimum target for selective toxicity. Synthesis of peptidoglycan precursors starts in the cytoplasm; wall subunits are then transported across the cytoplasmic membrane and finally inserted into the growing peptidoglycan molecule. Several different stages are therefore potential targets for inhibition (Fig. 33.6). The antibacterials that inhibit cell wall synthesis are varied in chemical structure. The most important of these agents are the beta-lactams, the largest group, and the glycopeptides which are active only against Gram-positive organisms. Bacitracin (primarily used topically) and cycloserine (mainly used as a ‘second-line’ medication for treatment of tuberculosis, discussed later in this chapter) have many fewer clinical applications.

Figure 33.6 The synthesis of peptidoglycan is a complex process that begins in the cytoplasm, proceeds across the cytoplasmic membrane and leads to the attachment of new wall units to the growing peptidoglycan chain. This synthetic pathway can be inhibited at a variety of points by antibacterial agents. The precise mechanism of inhibition caused by glycopeptides such as vancomycin is unknown, but the mechanism of action of beta-lactams has now been fully elucidated (see text). NAG, N-acetyl glucosamine; NAM, N-acetyl muramic acid; UDP, uridine diphosphate.

Beta-lactams

Beta-lactams contain a beta-lactam ring and inhibit cell wall synthesis by binding to penicillin-binding proteins (PBPs)

Beta-lactams comprise a very large family of different groups of bactericidal compounds, all containing the beta-lactam ring. The different groups within the family are distinguished by the structure of the ring attached to the beta-lactam ring – in penicillins this is a five-membered ring, in cephalosporins a six-membered ring – and by the side chains attached to these rings (Fig. 33.7).

Figure 33.7 The beta-lactam family. The ring structure is common to all beta-lactams and must be intact for antibacterial action. Enzymes (beta-lactamases) that catalyse the hydrolysis of the beta-lactam bond render the agents inactive. The penicillins and cephalosporins are the major classes of beta-lactam antibiotics, but other members of the family, particularly the carbapenems and monobactams, are the focus of new developments.

PBPs are membrane proteins (e.g. carboxypeptidases, transglycosylases and transpeptidases) capable of binding to penicillin (hence the name PBP) and are responsible for the final stages of cross-linking of the bacterial cell wall structure. Inhibition of one or more of these essential enzymes results in an accumulation of precursor cell wall units, leading to activation of the cell’s autolytic system and cell lysis (Fig. 33.8).

Figure 33.8 Penicillin-binding proteins (PBPs) play a key role in the final stages of peptidoglycan synthesis. They catalyse the cross-linkage of wall subunits, which are then incorporated into the cell wall. Beta-lactams are able to enter the cell (e.g. through pores in the outer membrane of Gram-negatives) and bind to the PBP. This prevents it from catalysing the cross-linkage of subunits, leading to their accumulation in the cell and the release of autolytic enzymes, which causes cell lysis. Within the periplasmic space of Gram-negatives (b1) beta-lactamases can inactivate beta-lactams before they reach their target PBPs, thereby protecting the cell from antibiotic action. Alternatively, mutant PBPs fail to bind beta-lactase, thus allowing peptidoglycan synthesis to occur. In Gram-positive bacteria (b2) beta-lactams may be extracellularly destroyed by beta-lactamases or rendered ineffective, as in Gram-negatives, by mutant PBPs.

Most beta-lactams have to be administered parenterally

Although the majority of beta-lactams have to be administered intramuscularly or intravenously, there are some orally active agents. Most achieve clinically useful concentrations in the cerebrospinal fluid (CSF) when the meninges are inflamed (as in meningitis) and the blood–brain barrier becomes more permeable. In general, they are not effective against intracellular organisms.

A few of the cephalosporins, notably cefotaxime, are metabolized to compounds with less microbiologic activity. All beta-lactams are excreted in the urine, and for some, such as benzylpenicillin, this is very rapid; hence the need for frequent doses. Probenecid can be administered concurrently to slow down excretion and maintain higher blood and tissue concentrations for a longer period of time.

Different beta-lactams have different clinical uses, but are not active against species that lack a cell wall

A vast array of beta-lactam antibiotics are currently registered for clinical use. Some, such as penicillin, are active mainly against Gram-positive organisms, whereas others (e.g. semi-synthetic penicillins, carboxypenems, monobactams, second-, third- and fourth-generation cephalosporins) have been developed for their activity against Gram-negative rods. Only the more recent beta-lactams are active against innately more resistant organisms such as Pseudomonas aeruginosa (Table 33.3).

Table 33.3 Characteristics of representative beta-lactams

Drug class	Category	General spectrum of activity
Penicillins
Penicillin G, V^a	Natural penicillin	Gram-positive bacteria
Cloxacillin^a Dicloxacillin^a Nafcillin^a Oxacillin^a	Semisynthetic (beta-lactamase resistant) penicillin	Gram-positive bacteria (incl. beta-lactamase producers)
Amoxicillin^a^,^b Ampicillin^a^,^b	Semisynthetic (amino) penicillin	Gram-positive bacteria Gram-negative bacteria, including spirochetes, Listeria monocytogenes, Proteus mirabilis and some Escherichia coli
Carbenicillin^a Ticarcillin^b Mezlocillin Piperacillin^b	Semisynthetic (carboxy) penicillin Semisynthetic (ureido) penicillin	Gram-positive bacteria Enhanced coverage of Gram-negatives, including Pseudomonas and Klebsiella
Cephalosporins
Cefadroxil^a Cefazolin Cephalexin^a Cephalothin Cephradine^a	First generation	Gram-positive bacteria
Cefaclor^a Cefamandole Cefonicid Cefprozil^a Cefuroxime^a	Second generation
Cefdinir^a Cefditoren^a Cefoperazone Cefpodoxime^a Cefotaxime Ceftazidime Ceftibuten^a Ceftizoxime Ceftriaxone	Third generation
Cefepime Cefpirome	Fourth generation	Improved activity against Gram-negative bacteria
Ceftaroline	Anti-MRSA	Improved activity, especially against MRSA
Cephamycin^c
Cefmetazole Cefotetan Cefoxitin		Gram-positive bacteria Improved activity against Bacillus fragilis
Carbapenems
Ertapenem Imipenem Meropenem Doripenem		Gram-positive and Gram-negative bacteria
Monobactams
Aztreonam		Gram-negative bacteria including Haemophilus influenza and Pseudomonas aeruginosa

Although there are many beta-lactam agents available, the most commonly used ones are listed, together with their main indications.

a Oral formulation available.

b Can be formulated in combination with beta-lactamase inhibitors (see Fig. 33.9).

c Often classified with second generation cephalospoxins.

It is important to remember that beta-lactams are not active against species that lack a cell wall (e.g. Mycoplasma) or those with very impenetrable walls such as mycobacteria, or intracellular pathogens such as Brucella, Legionella and Chlamydia.

Resistance to beta-lactams may involve one or more of the three possible mechanisms

The beta-lactamases of Gram-positive bacteria are released into the extracellular environment (Fig. 33.8A) and resistance will only be manifest when a large population of cells is present. The beta-lactamases of Gram-negative cells, however, remain within the periplasm (Fig. 33.8B).

To date, hundreds of different beta-lactamase enzymes have been described. All have the same function but with differing amino acid sequences that influence their affinity for different beta-lactam substrates. Some enzymes specifically target penicillins or cephalosporins, while others are especially troublesome in broadly attacking most beta-lactam compounds (i.e. extended-spectrum beta-lactamases, ESBLs). Some beta-lactam antibiotics (e.g. carbapenems) are hydrolysed by very few enzymes (beta-lactamase stable), whereas others (e.g. ampicillin) are much more labile. Beta-lactamase inhibitors such as clavulanic acid (Fig. 33.9) are molecules that contain a beta-lactam ring and act as ‘suicide inhibitors’, binding to beta-lactamases and preventing them from destroying beta-lactams. They have little bactericidal activity of their own.

Figure 33.9 Clavulanic acid, a product of Streptomyces clavuligerus, inhibits the most common beta-lactamases (e.g. TEM enzymes) and allows amoxicillin to inhibit cells producing these enzymes. Augmentin is the most widely used of these combination drugs. Other combinations include ticarcillin and clavulanic acid, and piperacillin and tazobactam.

Side effects

Toxic effects of beta-lactam drugs include mild rashes and immediate hypersensitivity reactions

Statistics regarding allergy to beta-lactam drugs are complicated by the fact that the problem historically involves self-reporting by patients who are often mistaken in their ‘diagnosis’. Nevertheless, serious allergy to beta-lactam drugs in the form of an immediate (type 1) hypersensitivity reaction may occur in ca. 0.5–4% of patients, although anaphylaxis occurs much less frequently (ca. 0.004 to 0.04% of penicillin treatment courses). Mild idiopathic reactions, usually in the form of a rash, are more common (ca. 25% of treatment courses), especially with ampicillin. Patients who are allergic to penicillin are often allergic to cephalosporins (less with third-generation compounds) and vice versa, but aztreonam, a monobactam, shows negligible cross-reactivity.

Neurotoxicity and seizures can occur with all the beta-lactams if improperly dosed for body weight and kidney function, especially in patients with renal impairment. This toxicity is manifest as fits, unconsciousness, myoclonic spasms and hallucinations. Carbenicillin can cause platelet dysfunction and sodium overload (because it is given as a sodium salt), especially in patients with liver failure, renal failure and congestive heart failure.

Glycopeptides

Glycopeptides are large molecules and act at an earlier stage than beta-lactams

Glycopeptides include vancomycin and teicoplanin. Both are very large molecules and therefore have difficulty penetrating into Gram-negative cells. Teicoplanin is a natural complex of five different but closely related molecules.

Glycopeptides are bactericidal and interfere with cell wall synthesis by binding to terminal d-alanine-d-alanine at the end of pentapeptide chains that are part of the growing bacterial cell wall structure (see Fig. 33.6). This binding inhibits the transglycosylation reaction and prevents incorporation of new subunits into the growing cell wall. As glycopeptides act at an earlier stage than beta-lactams, it is not useful to combine glycopeptides and beta-lactams in the treatment of infections.

Vancomycin and teicoplanin must be given by injection for systemic infections

Vancomycin and teicoplanin are not absorbed from the gastrointestinal tract and do not penetrate the CSF in patients without meningitis. However, bactericidal concentrations are achieved in most patients with meningitis because of the increased permeability of the blood–brain barrier. Excretion is via the kidney.

Both vancomycin and teicoplanin are active only against Gram-positive organisms

Vancomycin and teicoplanin are used mainly for:

• the treatment of infections caused by Gram-positive cocci and Gram-positive rods that are resistant to beta-lactam drugs, particularly multiresistant Staphylococcus aureus and Staphylococcus epidermidis

• for patients allergic to beta-lactams

• the treatment of Clostridium difficile in antibiotic-associated colitis, although concerns that this may promote emergence of glycopeptide-resistant enterococci in the gut flora have led to the increasing use of alternative compounds.

Resistance

Some organisms are intrinsically resistant to glycopeptides

As mentioned previously, Gram-negative bacteria are ‘naturally’ resistant to the glycopeptides, since these compounds are too large to efficiently move through the outer membrane to the peptidoglycan. Other organisms have an altered glycopeptide target, such as pentapeptides, terminating in d-alanine-d-lactate (e.g. Erysiplothrix, Leuconostoc, Lactobacillus and Pediococcus) or d-alanine-d-serine (e.g. Enterococcus gallinarum, Enterococcus casseliflavus).

Organisms may acquire resistance to glycopeptides

Historically, the most clinically relevant acquired glycopeptide resistance has been observed in Enterococcus faecium and Enterococcus faecalis (vancomycin-resistant enterococci; VRE), first reported by investigators in the UK in 1986. Since that time, a variety of resistance phenotypes have been described which can be differentiated by transferability (e.g. plasmid association), inducibility and extent of resistance (Table 33.4). The genes associated with the highest levels of glycopeptide resistance are vanA, vanB, and vanD which encode a ligase producing pentapeptides terminating in d-alanine-d-lactate.

Table 33.4 Characteristics of glycopeptide resistance in enterococci

VanA is the best understood mechanism of acquired glycopeptide resistance

VanA-type glycopeptide resistance has been the most extensively studied and is characterized by inducible high-level resistance to both vancomycin and teicoplanin. VanA is associated with transposable elements related to Tn1546 (ca. 11 kb in size) which may be chromosomal or plasmid (transferable) in nature.

VanB is associated with inducible high-level resistance to vancomycin but not teicoplanin (although teicoplanin resistance can be induced by prior exposure to vancomycin). VanB resistance may be chromosomal or plasmid linked and is associated with very large transposable elements such as Tn1549 (34 kb).

VanD is chromosomal in nature and thus non-transferable, resulting in constitutive resistance to high levels of vancomycin but low levels of teicoplanin.

Glycopeptide resistance in the staphylococci occurs by mutation or by acquisition from the enterococci

Within the coagulase-negative staphylococci (CNS), Staphylococcus epidermidis and Staphylococcus haemolyticus are especially prone to development of glycopeptide resistance by mechanisms which remain incompletely understood. Nevertheless, resistant clinical and laboratory-generated isolates have been shown to differ from their susceptible counterparts in a variety of ways including changes in glycopeptide binding capacity, membrane proteins and cell wall synthesis and composition.

Coagulase-positive staphylococci (i.e. Staphylococcus aureus) showing decreased susceptibility to glycopeptides (but not fully resistant) were first described by Japanese investigators in 1996. The reduced susceptibility of these vancomycin-intermediate or glycopeptide-intermediate isolates (VISA or GISA, respectively) may be either homogeneously or heterogeneously expressed. In either case, ‘resistance’ is not associated with VanA, B, or D but, instead, appears to involve other mechanisms affecting cell wall composition (e.g. leading to increased thickness, etc.).

Unfortunately, high-level glycopeptide resistance has also been observed in Staph. aureus. This is due to the vanA gene (apparently acquired from VRE) residing on a staphylococcal plasmid.

Side effects

The glycopeptides are potentially ototoxic and nephrotoxic

Vancomycin is usually given by intravenous infusion, administered slowly to avoid ‘red-man’ syndrome due to histamine release. Particular care must be taken to prevent toxic concentrations accumulating in patients with renal impairment. Oral vancomycin is used for treatment of antibiotic-associated pseudomembranous colitis due to Clostridium difficile. Teicoplanin is less toxic than vancomycin and can be given by intravenous bolus and by intramuscular injection.

Inhibitors of protein synthesis

Although protein synthesis proceeds in an essentially similar manner in prokaryotic and eukaryotic cells, it is possible to exploit the differences (e.g. 70 S vs 80 S ribosome) to achieve selective toxicity. The process of translation of the messenger RNA (mRNA) chain into its corresponding peptide chain is complex, and a range of antibacterial agents act as inhibitors, although the full details of their mechanisms of action are not yet known (Fig. 33.10).

Figure 33.10 The synthetic pathway leading to the production of new protein in bacterial cells is extremely complex and still not fully elucidated. A number of different groups of antibacterial agents act by inhibiting proteins with specific reactions in this synthetic pathway. They can be grouped into those that act on the 30 S subunit of the ribosome (e.g. aminoglycosides and tetracyclines) and those that act on the 50 S subunit (e.g. chloramphenicol, lincosamides, erythromycin and fusidic acid). fmet-tRNA, formylmethionyl-transfer RNA.

Aminoglycosides

The aminoglycosides are a family of related molecules with bactericidal activity

The aminoglycosides contain either streptidine (streptomycin) or 2-deoxystreptamine (e.g. gentamicin; Table 33.5). The original structures have been modified chemically by changing the side chains to produce molecules such as amikacin and netilmicin that are active against organisms that have developed resistance to earlier aminoglycosides.

Table 33.5 Aminoglycoside-aminocyclitol antibiotics classified according to their chemical structure

4,6-distributed 2-deoxystreptamines
Gentamicin^a	Complex of 3 closely related structures; first aminoglycoside with broad spectrum
Tobramycin^b	Activity very similar to gentamicin but slightly better against Pseudomonas aeruginosa
Amikacin	Semi-synthetic derivative of kanamycin; active against many gentamicin- resistant Gram-negative rods
Netilmicin^b	Activity spectrum similar to amikacin: possibly lower toxicity
4,5-disubstituted 2-deoxystreptamines
Neomycin^b	Too toxic for parenteral use but has topical uses in decontaminating mucosal surfaces
Streptidine-containing
Streptomycin^b	Oldest aminoglycoside; now use restricted to treatment of tuberculosis

They are also differentiated by the genus of microorganisms that produces them, and this is reflected in the spelling of the names.

a Micins from Micromonospora species.

b Mycins from Streptomyces species.

Aminoglycosides act by binding to specific proteins in the 30 S ribosomal subunit, where they interfere with the binding of formylmethionyl-transfer RNA (fmet-tRNA) to the ribosome (Fig. 33.10), thereby preventing the formation of initiation complexes from which protein synthesis proceeds. In addition, aminoglycosides cause misreading of mRNA codons and tend to break apart functional polysomes (protein synthesis by multiple ribosomes tandemly attached to a single mRNA molecule) into non-functional monosomes.

Aminoglycosides must be given intravenously or intramuscularly for systemic treatment

Aminoglycosides are not absorbed from the gut, do not penetrate well into tissues and bone, and do not cross the blood–brain barrier. Thus, they are usually administered as an intravenous infusion. Intrathecal administration of streptomycin is used in the treatment of tuberculous meningitis, and gentamicin may be administered by this route in the treatment of Gram-negative meningitis in neonates. Aminoglycosides are excreted via the kidney.

Gentamicin and the newer aminoglycosides are used to treat serious Gram-negative infections

Gentamicin, tobramycin, amikacin and netilmicin are important for the treatment of serious Gram-negative infections, including those caused by P. aeruginosa (Box 33.4). They are not active against streptococci or anaerobes, but are active against staphylococci. Against P. aeruginosa, amikacin is most active. Amikacin and netilmicin may be active against strains resistant to gentamicin and tobramycin (see below). Streptomycin is now reserved almost entirely for the treatment of mycobacterial infections. Neomycin is not used for systemic treatment, but can be used orally in gut decontamination regimens in neutropenic patients.

Box 33.4 Indications for Aminoglycoside Therapy

Aminoglycosides are valuable additions to the clinician’s armamentarium despite their potential toxicity. They are important agents active against Gram-negative facultative bacteria and are often used in combination with beta-lactams to broaden the spectrum to include streptococci and some anaerobes, which are not susceptible to aminoglycosides alone. Resistance to aminoglycosides, particularly among enterobacteria and staphylococci, is mediated by the production of aminoglycoside-modifying enzymes, which react with groups on the aminoglycoside molecule to yield an altered aminoglycoside product. This competes with the unmodified aminoglycoside for uptake into the cell and binding to the ribosome.

Basic rule: use only in severe, life-threatening infections

• Gram-negative septicaemia (including Pseudomonas) usually in combination with beta-lactam

• Septicaemia of unknown aetiology arising from:^a

• hospital-acquired infection

• malignancy

• immunosuppressive therapy

• major trauma, major surgery or major burns

• intravenous catheter

• urinary catheter

• extremes of age

• Bacterial endocarditis for synergy with beta-lactam

• Staphylococcus aureus septicaemia in combination with beta-lactam

• Pyelonephritis for difficult cases

• Post-surgical abdominal sepsis in combination with anti-anaerobe therapy.

^a Every effort should be made to establish aetiology.

Production of aminoglycoside-modifying enzymes is the principal cause of resistance to aminoglycosides

Although relatively uncommon, resistance to aminoglycoside antibiotics may occur by alteration of the 30 S ribosomal target protein (e.g. a single amino acid change in the P12 protein prevents streptomycin binding). Resistance may also arise through alterations in cell wall permeability or in the energy-dependent transport across the cytoplasmic membrane.

Production of aminoglycoside-modifying enzymes is the most important mechanism of acquired resistance (Fig. 33.11). The genes for these enzymes are often plasmid-mediated, located on transposons, and transferable from one bacterial species to another. The enzymes alter the structure of the aminoglycoside molecule, thus inactivating the drug. The type of enzyme determines the spectrum of resistance of the organism containing it.