Antimicrobial Chemotherapy, 4th Edition
General properties of antimicrobial agents
Synthetic antibacterial agents and miscellaneous antibiotics
Various targets other than the cell wall and ribosome are open to attack by chemotherapeutic agents. This chapter describes the properties of inhibitors of bacterial nucleic acid synthesis, membrane-active compounds, and agents used solely for the treatment of mycobacterial disease. Many, but not all, of these compounds are synthetic chemicals rather than antibiotics in the strict sense.
Inhibitors of nucleic acid synthesis
Given the universality of nucleic acid as the basis of life, it is surprising that so many antimicrobial agents have been discovered that selectively interfere with the functions of DNA and RNA. Some, like the sulphonamides and diamino-pyrimidines, achieve their effect indirectly by interrupting metabolic pathways that lead to the manufacture of nucleic acids; others, of which the quinolones and nitroimidazoles are prime examples, exert a more direct action.
The discovery of prontosil in the 1930s was a major breakthrough in the chemotherapy of bacterial infections (see Historical Introduction). Activity of the dye is due to the liberation in the body of sulphanilamide, an analogue of para-aminobenzoic acid (Fig. 3.1), which is essential for bacterial folate synthesis. Most bacteria synthesize folic acid and cannot take it up preformed from the environment. Mammalian cells, in contrast, use preformed folate and cannot make their own. Sulphonamides block an early stage in folate synthesis: the condensation of p-aminobenzoic acid with dihydropteridine to form dihy-dropteroic acid. Depletion of folate leads to various effects, including a failure to synthesize purine nucleotides and thymidine. Chemical modification of the
sulphanilamide molecule has resulted in the production of hundreds of different sulphonamides which differ chiefly in their pharmacological properties.
Fig. 3.1 Structures of prontosil, sulphanilimide, and p-aminobenzoic acid.
Sulphonamides are broad spectrum, though the activity against enterococci, Pseudomonas aeruginosa, and anaerobes is poor. They are predominantly bacter-istatic, and relatively slow to act: several generations of bacterial growth are needed to deplete the folate pool before inhibition of growth occurs. Resistance emerges readily, and bacteria resistant to one sulphonamide are cross-resistant to the others. Sensitivity tests present problems in the laboratory since results depend critically on the composition of the culture medium and the inoculum size (see p. 108).
The emergence of resistant strains and the appearance of safer and more potent agents has relegated the sulphonamides to a minor place in therapy. Even in their traditional role, uncomplicated urinary infection, they are now seldom used. They are still found in combination products with trimethoprim, pyrimethamine, and other diaminopyrimidines (see below).
Topical silver sulphadiazine is used in burns, but the activity probably owes as much to the silver as to the sulphonamide. In curious extension of the good fortune that attended the discovery of the sulphonamides, sulphasalazine, an agent used in ulcerative colitis and rheumatoid arthritis, probably owes its effi-cacy to a breakdown product, 5-aminosalicylic acid.
Most sulphonamides are well absorbed when given orally and are chiefly excreted in the urine, partly in an antibacterially inactive acetylated form. The compounds diffuse relatively well into cerebrospinal fluid (CSF) and were successfully used for treating meningitis before resistance became common. Sulphadiazine is effective in eradicating Neisseria meningitidis from the throat, provided the strain is not sulphonamide resistant.
Less soluble sulphonamides (e.g. sulphathiazole and sulphadiazine) are prone to cause renal damage due to the deposition of crystals in the urinary collecting system. Sulphafurazole (known as sulfisoxazole in the US), sulphadimidine (sulfamethazine), sulphasomidine (sulfisomidine), and sulphamethizole lack this
side-effect and are preferred in the treatment of urinary infection. Some sulphonamides are distinguished by long plasma half-lives (T½) and are known as long-acting sulphonamides. Sulfametopyrazine (T½ = 60 h) and sulfadoxine (T½ = 120 h) are excreted so slowly that they are given at weekly intervals.
Diaminopyrimidines inhibit dihydrofolate reductase, the enzyme that generates tetrahydrofolate (the active form of the vitamin) from metabolically inactive dihydrofolate. Trimethoprim (Fig. 3.2), the most important antibacterial agent of this type, exhibits far greater affinity for the dihydrofolate reductase of bacteria than for the corresponding mammalian enzyme; this is the basis of the selective toxicity of the compound.
Fig. 3.2 Structure of trimethoprim.
Since sulphonamides and trimethoprim act at different points in the same metabolic pathway they interact synergically: bacteria are inhibited by much lower concentrations of the combination than by either agent alone. For this reason trimethoprim and sulphonamides are often combined in therapeutic formulations, although trimethoprim alone is probably as effective and less toxic. The most commonly used combination is trimethoprim and sulphamethoxazole (co-trimoxazole), but combinations of trimethoprim with sulphadiazine (co-trimazine) and sulphamoxole (co-trifamole) are also available in some countries.
Trimethoprim is active in low concentration against many common pathogenic bacteria, although Ps. aeruginosa is a notable exception. Resistance is on the increase; over 30 per cent of urinary isolates have been reported to be resistant in some series, but this is not a universal experience. The drug is rapidly absorbed from the gut and excreted almost exclusively by the kidneys with a plasma half-life of about 10 h. Toxic side-effects are uncommon, but residual effects on folate metabolism may cause haematological changes in folate-depleted patients.
The chief use for trimethoprim is in urinary tract infection. The combination with sulphamethoxazole has also been extensively used in many other clinical situations, including respiratory tract infections, typhoid fever, and brucellosis.
Other diaminopyrimidines include tetroxoprim, which is less active than trimethoprim, but is marketed in combination with sulphadiazine as co-tetroxazine in some countries; brodimoprim and metioprim, which are very similar to trimetho-prim; the antimalarial agents pyrimethamine and proguanil (Chapter 5); the anti-pneumocystis agent trimetrexate; and the antineoplastic agent methotrexate.
Nalidixic acid (Fig. 3.3) was the first representative to appear of a family of compounds that share close similarities of structure. These agents are generically known as quinolones, although they embrace a variety of molecular types, depending on the arrangement of nitrogen atoms within the heterocyclic structure.
Fig. 3.3 Structures of nalidixic acid and ciprofloxacin.
Among the earlier quinolones are two with modestly improved antibacterial activity: flumequine (which bears a fluorine atom at the C-6 position) and pipemidic acid (with a piperazine substituent at C-7). During the 1980s a new series of quinolones were synthesized in which the 6-fluoro and 7-piperazinyl features were combined. These compounds, of which ciprofloxacin (Fig. 3.3) is a typical example, exhibit considerably enhanced activity. In some subsequent derivatives, piperazine was replaced by other substituents, and members of this family of compounds are now generally referred to as fluoroquinolones. Although flumequine is strictly a fluoroquinolone, the term is usually restricted to compounds that exhibit superior intrinsic activity.
Further attempts to improve the pharmacological and antimicrobial properties of these compounds have led to the appearance of a new group of fluoroquinolones, so that these agents can now be loosely categorized into three types (Table 3.1).
Table 3.1 Categorization of quinolone antibacterial agents
All antibacterial quinolones act against the remarkable enzymes that are involved in maintaining the integrity of the supercoiled DNA helix during replication and transcription. Two enzymes are affected, DNA gyrase and topoiso-merase IV, so that these drugs have a dual site of action. In Gram-negative bacilli the main target is DNA gyrase, with topoisomerase IV as a secondary site, but in Staphylococcus aureus and some other Gram-positive cocci, the
situation is reversed. Relative affinity for the two sites has led to claims of differential activity, especially among some of the newer quinolone derivatives.
Quinolones are generally well tolerated, but rashes and gastrointestinal disturbances may occur; photophobia and various non-specific neurological complaints are also sometimes encountered. These compounds affect the deposition of cartilage in experimental animals, and licensing authorities have cautioned against their use in children and pregnant women. Several promising fluoroquinolones have had to be withdrawn (temafloxacin, grepafloxacin) or have had their use restricted (trovafloxacin) because of unexpected toxicity.
Nalidixic acid and its early congeners
Several quinolone derivatives, including cinoxacin, oxolinic acid, and pipemidic acid, were introduced into clinical use in various countries following the marketing of nalidixic acid in the 1960s (Table 3.1, left-hand column). They are all well absorbed when taken by mouth and are more or less extensively metabolized in the body before being excreted into the urine. Nalidixic acid itself is largely converted to hydroxynalidixic acid (which retains antibacterial activity) and glucuronide conjugates (which do not).
Most Gram-negative bacteria, with the exception of Ps. aeruginosa, are susceptible to nalidixic acid and its early congeners, but Gram-positive organisms are usually resistant (Table 3.2). Susceptible bacteria readily develop resistance in the laboratory and the emergence of resistance sometimes occurs during treatment. These drugs are best suited to the undemanding role of the treatment of cystitis, though they are also used in gastrointestinal disease in some parts of the world. Acrosoxacin exhibits good activity against Neisseria and has been used for the treatment of gonorrhoea.
Table 3.2 Summary of spectrum of activity of quinolones available in the UK (1999)
After the appearance, over a 20 year period, of a series of compounds that offered little improvement over nalidixic acid, the discovery of a family of quinolones that exhibited greatly improved properties came as a surprise. These compounds, listed in the central column of Table 3.1, are much more active than earlier derivatives against enterobacteria, Ps. aeruginosa, and many Gram-positive cocci (Table 3.2), though the activity is somewhat reduced in acidic conditions and in the presence of divalent cations such as magnesium. The spectrum also includes certain problem organisms such as chlamydiae, legionellae, and some mycobac-teria. Similar compounds, including enrofloxacin, danofloxacin, and sarafloxacin, have been introduced into veterinary practice and there has been considerable debate about the impact this may have had on the development of resistance.
The success of the first fluoroquinolones led to an intensive search for derivatives with further improved properties and this has borne fruit with several new agents now on the world market, or at an advanced stage of development (Table 3.1, right-hand column). These compounds are characterized by enhanced activity against Gram-positive cocci, including Staph. aureus and Streptococcus pneumoniae as well as chlamydiae and mycoplasmas; clinafloxacin, gatifloxacin, moxifloxacin, and trovafloxacin also have sufficient activity against anaerobes of the Bacteroides fragilis group to make treatment of infections with those organisms feasible.
Fluoroquinolones are usually administered orally, although some, including ciprofloxacin, ofloxacin, levofloxacin (the L-isomer of ofloxacin), and trova-floxacin (in the form of its pro-drug, alatrofloxacin), can also be given by injection. Therapeutic dosages achieve relatively low concentrations in plasma, but the compounds are well distributed in tissues and are concentrated within mammalian cells. The major route of excretion is usually renal, in the form of native compound or glucuronide and other metabolites, some of which retain antibacterial activity. Some fluoroquinolones, notably rufloxacin and sparfloxacin, but also including moxifloxacin, and trovafloxacin, exhibit long terminal half-lives; these compounds are partly excreted by the biliary route and this may help to explain the long half-life.
Among currently available fluoroquinolones, norfloxacin, enoxacin, and lome-floxacin have found their greatest use as more reliable replacements for earlier derivatives in the treatment of urinary tract infection, though they may also have a place in gonococcal and gastrointestinal infections. Others are used for systemic infection and have become the subject of conflicting marketing claims. Cipro-floxacin is the most widely used; among other indications, it is now the drug of choice for typhoid fever and other serious enteric diseases. The ease of administration of fluoroquinolones makes them attractive candidates for ‘blind’ therapy in hospital and domiciliary practice. However, they are not universal panaceas and should not be used indiscriminately. The newer derivatives are being targeted at the treatment of respiratory infections in the community and, although they
are undoubtedly effective, it is not clear that such use is necessary or wise. Indeed, there is already evidence that indiscriminate use of fluoroquinolones is undermining the effectiveness of these valuable drugs by encouraging resistance.
Novobiocin is a coumarin-like antibiotic first described in 1955. The heterocyclic structure is unrelated to those of other antibiotics, and cross-resistance is not a problem. Like nalidixic acid and other quinolones, novobiocin inhibits DNA gyrase, but binds to a different subunit. Curiously, the spectrum of activity is the mirror image of that of nalidixic acid: good activity against Gram-positive organisms, but no useful activity against enteric Gram-negative rods. Side-effects are common, and bacterial resistance develops readily. The drug is now seldom used.
As a group, the imidazoles are remarkable in that derivatives are known which between them cover bacteria, fungi, viruses, protozoa, and helminths—in fact, the whole antimicrobial spectrum. Outside the antimicrobial field certain imida-zoles have been shown to exhibit radio-sensitizing properties and have attracted attention as adjuncts to radiation therapy for some tumours.
The members of this family of compounds used as antibacterial agents are 5-nitroimidazoles, of which metronidazole (Fig. 3.4) is best known. Related 5-nitroimidazoles include tinidazole, carnidazole, ornidazole, secnidazole, and nimorazole; they share the properties of metronidazole but have longer plasma half-lives.
Fig. 3.4 Structure of metronidazole.
Metronidazole was originally used for the treatment of trichomoniasis, and subsequently for two other protozoal infections—amoebiasis and giardiasis. The antibacterial activity of the compound was first recognized when a patient suffering from acute ulcerative gingivitis responded spontaneously while receiving metronidazole for a Trichomonas vaginalis infection. Anaerobic bacteria are commonly incriminated in gingivitis, and it was subsequently shown that metronida-zole possesses potent antibacterial activity that was at first thought to be confined to strict anaerobes (the protozoa against which the drug is effective also exhibit anaerobic metabolism) since even oxygen-tolerant species such as Actinomyces
and Propionibacterium are resistant. However, some micro-aerophilic bacteria, including Gardnerella vaginalis and Helicobacter pylori, are commonly susceptible to nitroimidazoles, and metronidazole features in drug regimens for the treatment of infections with these organisms.
Metronidazole is so effective against anaerobic bacteria and resistance is so uncommon that it is now the drug of choice for the treatment of anaerobic infections. It is also commonly used for prophylaxis in some surgical procedures in which post-operative anaerobic infection is a frequent complication. It is the preferred alternative to vancomycin in the treatment of antibiotic-associated colitis caused by Clostridium difficile toxins (p. 251).
The basis of the selective activity against anaerobes resides in the fact that a reduction product is produced intracellularly at the low redox values attainable by anaerobes, but not by aerobes. The reduced form of metronidazole is thought to induce strand breakage in DNA by a mechanism that has not been precisely determined.
The 5-nitroimidazoles are generally free from serious side-effects, though gastrointestinal upset is common and ingestion of alcohol induces a disulfiram-like reaction. Since these drugs act on DNA they are potentially genotoxic and tumorigenic, but there is no evidence that these problems have arisen despite widespread clinical use. None the less, these compounds are best avoided in pregnancy.
A number of nitrofuran derivatives have attracted attention over the years, among which nitrofurantoin (Fig. 3.5) is much the most important. Others include fura-zolidone, nifurzide, and nifuroxazide (orally non-absorbed compounds sometimes used for intestinal infections); nitrofurazone (a topical agent also used for bladder irrigation); nifuratel and nifuroxime (used in some countries for T. vaginalisinfection and vaginal thrush); nitrofurtoinol (very similar to nitrofurantoin); and nifurtimox (used in Chagas' disease; see p. 68).
Fig. 3.5 Structure of nitrofurantoin.
Nitrofurantoin is the only nitrofuran derivative available in the UK. Its use is restricted to the treatment of urinary infection since it is rapidly excreted into urine after oral absorption and the small amount that finds its way into tissues is inactivated there. It is active against most urinary tract pathogens, but Proteus spp. and Ps. aeruginosa are usually resistant. The occurrence of resistant strains
among susceptible species is uncommon. The activity is affected by pH, being favoured by acid conditions.
Nausea is fairly common after administration of nitrofurantoin; macrocrys-talline formulations have improved gastrointestinal tolerance and are generally preferred for this reason.
The mode of action of nitrofurantoin (or other nitrofurans) has not been precisely elucidated and is probably complex. The nitro group is reduced intra-cellularly in susceptible bacteria and it is likely that one effect, as with metronidazole, is the interaction of a reduction product with DNA.
The clinically useful rifamycins, of which rifampicin (known in the US as rifampin) is the most important, are semi-synthetic derivatives of rifamycin B, one of a group of structurally complex antibiotics produced by Streptomyces mediterranei. These compounds interfere with mRNA formation by binding to the β-subunit of DNA-dependent RNA polymerase. Resistance readily arises by mutations in the subunit. For this reason, the drugs are normally used in combination with other agents.
Rifampicin is one of the most effective weapons against two major mycobacte-rial scourges of mankind: tuberculosis and leprosy. It also exhibits potent bactericidal activity against a range of other bacteria, notably staphylococci and legionellae. Rifampicin has established itself as such a useful antimycobacterial drug that there has been a move to confine its use to tuberculosis and leprosy on the grounds that more widespread use might inadvertently encourage the emergence of resistant strains of mycobacteria. Critics of this view claim that an agent which possesses exceptionally good antistaphylococcal activity and useful activity against other bacteria is being unnecessarily restricted on unproven grounds. In fact, restriction of the use of rifampicin has been steadily eroded: it is now often used in combination with erythromycin in Legionnaires' disease; it is also used to eliminate meningococci from the throats of carriers and for the protection of close contacts of meningococcal and Haemophilus influenzae type b disease.
Rifampicin is well absorbed by the oral route, although it may also be given by intravenous infusion. Serious side-effects are relatively uncommon, but can be more troublesome when the drug is used intermittently, as it may be in antituber-culosis regimens. Hepatotoxicity is well recognized and the antibiotic induces hepatic enzymes, leading to self-potentiation of excretion and antagonism of some other drugs handled by the liver, including oral contraceptives. A potentially alarming side-effect arising from the fact that rifampicin is strongly pigmented is the production of red urine and other bodily secretions; contact lenses may become discoloured. Patients should be warned of these potential problems.
Two other rifamycin derivatives are of particular interest: rifapentine exhibits an extended plasma half-life, but is otherwise similar to rifampicin; rifabutin (ansamycin) also has a prolonged half-life and retains activity against some rifampicin-resistant bacteria. Much interest has focused on the possibility that these agents may be useful in infections caused by organisms of the Myco-bacterium avium complex, which often cause disseminated disease in patients with cancer or acquired immune deficiency syndrome (AIDS). Although these mycobacteria are commonly resistant to standard antituberculosis drugs, rifabutin and rifapentine display good activity in vitro. Clinical success has been modest, but rifabutin has proved of sufficient value to warrant its inclusion in some multidrug regimens for prophylaxis against infection with organisms of the M. avium complex. Rifabutin also inhibits replication of human immunodeficiency virus (HIV) in vitro, but it is unlikely that this translates into clinical benefit.
Rifaximin, rifamide, and rifamycin SV have limited availability. They are poorly absorbed after oral administration and although they are marketed in some countries for gastrointestinal infections and for topical application, they are not recommended.
Agents affecting membrane function
The polymyxins are a family of five compounds (polymyxins A, B, C, D, and E) produced by Bacillus polymyxa and related bacteria. Only polymyxins B and E are used therapeutically. Polymyxin E is usually known by its alternative name, colistin. Structurally, the polymyxins are polypeptides with a long hydrophobic tail. Most of the peptide portion is arranged in a cyclic fashion, reminiscent of bacitracin and the cyclic peptides. The polymyxins act like cationic detergents by binding to the cell membrane and causing the leakage of essential cytoplasmic contents. The effect is not entirely selective, and both polymyxin B and colistin exhibit considerable toxicity.
Derivatives of the polymyxins in which up to five diaminobutyric acid residues are substituted with sulphomethyl groups are better tolerated and more quickly excreted than the parent compounds. These sulphomethylpolymyxins exhibit diminished antibacterial activity, but the precise loss in activity is difficult to estimate because the substituted compounds spontaneously break down to the more active parent.
The antibacterial spectrum of the polymyxins encompasses most Gram-negative bacteria except Proteus spp., but the importance of these antibiotics hinges on their activity against Ps. aeruginosa. With the appearance of antipseudomonal β-lactam agents, aminoglycosides, and fluoroquinolones, the polymyxins have virtually fallen into disuse for systemic therapy, although they are still used in
some topical preparations. They are also included in selective decontamination regimens aimed at preventing endogenous infection in profoundly neutropenic patients (p. 220) and there are advocates of the use of colistin in cystic fibrosis by instillation into the lungs of those suffering exacerbation of pseudomonal infection.
Other membrane-active agents
Naturally-occurring oligopeptides that destabilize bacterial membranes are widespread in nature, where they play a part in innate defence against microbial infection. Compounds of this type include cecropins (originally described in insects), magainins (from frogs), defensins (from mammalian leucocytes), and lanthionine-containing ‘lantibiotics’ (from bacteria). Whether these, or related synthetic compounds, have any future as therapeutic agents remains to be seen.
Daptomycin, a semi-synthetic lipopeptide antibiotic not unlike the polymyxins in structure, has various effects on bacteria, but the primary mode of action is thought to lie in disruption of the cell membrane. Development of the compound in the 1980s was stopped because of fears of toxicity, but the rise to prominence of multi-resistant Gram-positive cocci has revived commercial interest. Activity is restricted to Gram-positive cocci and is greatly enhanced in vitro by the presence of magnesium ions.
Antibiotics of the tyrothricin complex (gramicidin and tyrocidine), which are used in some topical preparations, are cyclic decapeptides that bind to the cell membrane and interfere with its function. Monensin, a polyether used in animal husbandry as a growth promoter and coccidiostat, and the depsipeptide valino-mycin act in a similar manner by forming transmembrane channels. These agents possess good activity against Gram-positive organisms, but they also bind to mammalian cell membranes and are far too toxic to be used systemically in man.
Toxicity also precludes the systemic use of the many disinfectants, including phenols, quaternary ammonium compounds, biguanides, and others, that achieve their antibacterial effect wholly or in part by interfering with the integrity of the cell membrane.
Compared with the number of agents at the disposal of the prescriber for the therapy of most bacterial infections, the resources available to treat mycobacte-rial disease are precariously meagre. Part of the reason is that mycobacteria are unusual organisms with a relatively impermeable waxy coat, but the fact that they are very slow growing and are able to survive and multiply within macrophages and necrotic tissue also makes them difficult targets.
In general, the development of drugs for the treatment of tuberculosis and leprosy has tended to evolve along specialized lines, but some important
antimycobacterial agents, such as rifampicin (see above) and certain aminogly-cosides (Chapter 2), have wider uses. Agents specifically used for the treatment of tuberculosis include isoniazid (isonicotinic acid hydrazide; INH), pyrazi-namide, ethambutol, and thiacetazone (thioacetazone). Para-Aminosalicylic acid (PAS), which was formerly much used in antituberculosis regimens, is no longer recommended, but this and other compounds with activity against Mycobacterium tuberculosis, such as capreomycin, cycloserine, and viomycin, may be considered if first-line treatment fails. In leprosy, the most important agents (apart from rifampicin) are dapsone (or its pro-drug, acedapsone) and clofazimine. The thioamides ethionamide and prothionamide (protionamide) are sometimes used, but are hepatotoxic.
Some fluoroquinolones and macrolides display quite good activity against mycobacteria, including M. leprae and organisms of the M. aviumcomplex, and these agents might widen the options for treating mycobacterial disease at a time when resistance is emerging as a serious problem.
Because of the difficulties in studying mycobacteria in the laboratory, less is known about the mode of action of antimycobacterial drugs than about other antibacterial agents. Isoniazid is a nicotinic acid derivative that arose from early studies with thiosemicarbazones (of which thiacetazone is an example). Various theories have been put forward to explain its action. The most widely held view is that an oxidized product inhibits the formation of the mycolic acids that are peculiar to the cell walls of acid-fast bacilli. Other derivatives of nicotinic acid, including pyrazinamide, ethionamide, and prothionamide, may act in the same way. Such evidence as is available suggests that the diamine ethambutol inhibits the formation of arabinan, a component of the cell wall arabinogalactan of mycobacteria. Dapsone (diaminodiphenyl sulphone) and PAS are related to the sulphonamides and have been assumed to share the same mode of action, but this is by no means certain. The mode of action of the antileprosy agent clofaz-imine has not been determined, but it may, like rifampicin, inhibit DNA-dependent RNA polymerase. Some puzzling aspects of the idiosyncratic spectrum of antimycobacterial agents may be explained by differences in uptake into susceptible cells.
Further information on these agents is given in the context of their use in Chapter 26.