Antimicrobial Chemotherapy
Harold C. Neu
Thomas D. Gootz
GENERAL CONCEPTS
Basis of Antimicrobial Action
Various antimicrobial agents act by interfering with (1) cell wall synthesis, (2) plasma membrane integrity, (3) nucleic acid synthesis, (4) ribosomal function, and (5) folate synthesis.
Action of Specific Agents
Cell wall synthesis is inhibited by ß-lactams, such as penicillins and cephalosporins, which inhibit peptidoglycan polymerization, and by vancomycin, which combines with cell wall substrates. Polymyxins disrupt the plasma membrane, causing leakage. The plasma membrane sterols of fungi are attacked by polyenes (amphotericin) and imidazoles. Quinolones bind to a bacterial complex of DNA and DNA gyrase, blocking DNA replication. Nitroimidazoles damage DNA. Rifampin blocks RNA synthesis by binding to DNA directed RNA polymerase. Aminoglycosides, tetracycline, chloramphenicol, erythromycin, and clindamycin all interfere with ribosome function. Sulfonamides and trimethoprim block the synthesis of the folate needed for DNA replication
Bacterial Resistance
Bacteria can evolve resistance to antibiotics. Resistance factors can be encoded on plasmids or on the chromosome. Resistance may involve decreased entry of the drug, changes in the receptor (target) of the drug, or metabolic inactivation of the drug.
Effects of Combination Therapy
Combinations of antibiotics may act synergistically-producing an effect stronger than the sum of the effects of the two drugs alone or antagonistically, if one agent inhibits the effect of the other.
Adverse Effects of Antimicrobial Agents
Many antibiotics are toxic to the host. Alterations of the normal intestinal flora caused by antibiotics may result in diarrhea or in superinfection with opportunistic pathogens.
INTRODUCTION
The earliest evidence of successful chemotherapy is from ancient Peru, where the Indians used bark from the cinchona tree to treat malaria. Other substances were used in ancient China, and we now know that many of the poultices used by primitive peoples contained antibacterial and antifungal substances. Modern chemotherapy has been dated to the work of Paul Ehrlich in Germany, who sought systematically to discover effective agents to treat trypanosomiasis and syphilis. He discovered p-rosaniline, which has antitrypanosomal effects, and arsphenamine, which is effective against syphilis. Ehrlich postulated that it would be possible to find chemicals that were selectively toxic for parasites but not toxic to humans. This idea has been called the "magic bullet" concept. It had little success until the 1930s, when Gerhard Domagk discovered the protective effects of prontosil, the forerunner of sulfonamide. Ironically, penicillin G was discovered fortuitously in 1929 by Fleming, who did not initially appreciate the magnitude of his discovery. In 1939 Florey and colleagues at Oxford University again isolated penicillin. In 1944 Waksman isolated streptomycin and subsequently found agents such as chloramphenicol, tetracyclines, and erythromycin in soil samples. By the 1960s, improvements in fermentation techniques and advances in medicinal chemistry permitted the synthesis of many new chemotherapeutic agents by molecular modification of existing compounds. Progress in the development of novel antibacterial agents has been great, but the development of effective, nontoxic antifungal and antiviral agents has been slow. Amphotericin B, isolated in the 1950s, remains an effective antifungal agent, although newer agents such as fluconazole are now widely used. Nucleoside analogs such as acyclovir have proved effective in the chemotherapy of selected viral infections.
Biochemical Basis of Antimicrobial Action
Bacterial cells grow and divide, replicating repeatedly to reach the large numbers present during an infection or on the surfaces of the body. To grow and divide, organisms must synthesize or take up many types of biomolecules. Antimicrobial agents interfere with specific processes that are essential for growth and/or division (Fig. 11-1). They can be separated into groups such as inhibitors of bacterial and fungal cell walls, inhibitors of cytoplasmic membranes, inhibitors of nucleic acid synthesis, and inhibitors of ribosome function (Table 11-1). Antimicrobial agents may be either bactericidal, killing the target bacterium or fungus, or bacteriostatic, inhibiting its growth. Bactericidal agents are more effective, but bacteriostatic agents can be extremely beneficial since they permit the normal defenses of the host to destroy the microorganisms.
FIGURE 11-1 Sites of action of different antimicrobial agents. PABA, paraminobenzoic acid; DHFA, dihydrofolic acid; THFA, tetrahydrofolic acid.
Inhibition of Bacterial Cell Wall Synthesis
As noted in earlier chapters, bacteria are classified as Gram-positive and Gram-negative organisms on the basis of staining characteristics. Gram-positive bacterial cell walls contain peptidoglycan and teichoic or teichuronic acid, and the bacterium may or may not be surrounded by a protein or polysaccharide envelope. Gram-negative bacterial cell walls contain peptidoglycan, lipopolysaccharide, lipoprotein, phospholipid, and protein (Fig. 11-2). The critical attack site of anti-cell-wall agents is the peptidoglycan layer. This layer is essential for the survival of bacteria in hypotonic environments; loss or damage of this layer destroys the rigidity of the bacterial cell wall, resulting in death.
FIGURE 11-2 Outer wall of Gram-positive and Gram-negative species and deail of porin channels of Gram-negative bacteria. Antimicrobial agents diffuse easily through the loose outer wall of Gram-positive bacteria, but must go through the narrow channels of the Gram-negative species.
Peptidoglycan synthesis occurs in three stages. The first stage takes place in the cytoplasm, where the low-molecular-weight precursors UDP-GlcAc and UDP-MurNAc-L-Ala-D-Glu-meso-Dap-D-Ala-D-Ala are synthesized. A number of antimicrobial agents interfere with these early steps in cell wall biosynthesis. UTP and N-acetylglucosamine alpha-1-P are converted to UDP-N-acetylglucosamine, which is subsequently converted by the enzyme phosphoenolpyruvate: UDP-GIcNAc-3-enol-pyruvyltransferase. Fosfomycins block this transfer by a direct nucleophilic attack on the enzyme. Because mammalian enzymes such as enolase, pyruvate kinase, carboxykinases, and the shikimate enlaces are not inhibited by these compounds, fosfomycins have no effect on the host. Three amino acids are added to the muramyl peptide to yield a tripeptide to which two more amino acids will be linked. The dipeptide D-alanyl-D-alanine is synthesized from two molecules of D-alanine by the enzyme D-alanyl-D-alanine synthetase. D-Alanine is produced from L-alanine by an alanine racemase. Cycloserine inhibits both alanine racemase and D-alanyl-D-alanine synthetase owing to the structural similarity of cycloserine and D-alanine and to the fact that cycloserine actually binds to the enzymes better than the D-alanine.
The second stage of cell wall synthesis is catalyzed by membrane-bound enzymes. The nonnucleotide portion of the precursor molecules previously made are transferred sequentially to a carrier in the cytoplasmic membrane. This carrier is a phosphorylated undecaprenyl alcohol. The lipid carrier functions as a point of attachment to the membrane for the precursors and allows for transport of the subunits across the hydrophobic interior of the cytoplasmic membrane to the outside surface. Bacitracin is a peptide antibiotic that specifically interacts with the pyrophosphate derivate of the undecaprenyl alcohol, preventing further transfer of the muramylpentapeptide from the precursor nucleotide to the nascent peptidoglycan.
The third stage of cell wall synthesis involves polymerization of the subunits and the attachment of nascent peptidoglycan to the cell wall. Polymerization occurs by transfer of the new peptidoglycan chain from its carrier in the membrane to the nonreducing N-acetylglucosamine of the new saccharide-peptide that is attached to the membrane. The new peptidoglycan is attached to preexisting cell wall peptidoglycan by a transpeptidase reaction that involves peptide chains in both polymers, one of which must possess a D-alanyl-D-alanine terminus. It is believed that the transpeptidase enzyme cleaves the peptide bond between two D-alanyl residues in the pentapeptide and become acylated via the carbonyl group of the penultimate D-alanine residue. This final reaction is inhibited by ß-lactam antibiotics. These antibiotics contain a critical four-membered ring, which undergoes an acylation reaction with the transpeptidases that cross-link the polymers mentioned above. The ß-lactam antibiotics are the penicillins (penams), cephalosporins (including oxacephems and cephamycins), penems, thienamycins (carbapenems), and aztreonam (monobactams) (Fig. 11-3). The enzymes involved in this final process of cell wall formation are called penicillin-binding proteins since they were discovered by labeling with radioactive penicillin G. The enzymes are different in Gram-positive and Gram-negative bacteria and in anaerobic species. Differences in the penicillin-binding proteins explain, to some extent, differences in antibacterial activity of the ß-lactam antibiotics. The penicillin-binding protein, to which a particular ß-lactam antibiotic binds, affects the morphologic response of the bacterium to the agent. For example, some antibiotics bind to a penicillin-binding protein that is involved in forming the septum between dividing cells; as a result, the bacteria continue to grow into long filaments, which eventually die. Binding to another penicillin-binding protein results in rapid lysis of a bacterium because the wall bulges and the bacterium bursts. ß-Lactams such as mecillinam (an amidino penicillin) do not bind to the penicillin-binding proteins of Gram-positive bacteria and therefore do not affect these bacteria. Aztreonam binds only to Gram-negative penicillin-binding proteins and does not inhibit Gram-positive or anaerobic species.
FIGURE 11-3 Basic structures of ß-lactam antibiotics. Penicillins and cephalosporins/cephamycins are widely used to inhibit both Gram-positive and Gram-negative bacilli. Monobactams inhibit only aerobic Gram-negative bacilli, clavulanic acid acts as a ß-lactamase inhibitor, and thienamycin inhibits a wide range of aerobic and anaerobic species. R and R' represent various carbon groups. X can be either hydrogen or a mehoxy group.
Vancomycin interrupts cell wall synthesis by forming a complex with the C-terminal D-alanine residues of peptidoglycan precursors. Complex formation at the outer surface of the cytoplasmic membrane prevents the transfer of the precursors from a lipid carrier to the growing peptidoglycan wall by transglycosidases. Biochemical reactions in the cell wall catalyzed by transpeptidases and D,D-carboxypeptidases are also inhibited by vancomycin and other glycopeptide antimicrobials. Because of its large size and complex structure (Fig. 11-4), vancomycin does not penetrate the outer membrane of gram-negative organisms. With resistance to beta-lactams increasing in frequency among staphylococci and enterococci, glycopeptides such as vancomycin remain important therapeutic agents against such bacteria.
FIGURE 11-4 Structure of vancomycin.
Antibiotics that Affect the Function of Cytoplasmic Membranes
Bacterial Cytoplasmic Membranes
Biologic membranes are composed basically of lipid, protein, and lipoprotein. The cytoplasmic membrane acts as a diffusion barrier for water, ions, nutrients, and transport systems. Most workers now believe that membranes are a lipid matrix with globular proteins randomly distributed to penetrate through the lipid bilayer. A number of antimicrobial agents can cause disorganization of the membrane. These agents can be divided into cationic, anionic, and neutral agents. The best-known compounds are polymyxin B and colistemethate (polymyxin E). These high-molecular-weight octapeptides inhibit Gram-negative bacteria that have negatively charged lipids at the surface. Since the activity of the polymyxins is antagonized by Mg2+ and Ca2+, they probably competitively displace Mg2+ or Ca2+ from the negatively charged phosphate groups on membrane lipids. Basically, polymyxins disorganize membrane permeability so that nucleic acids and cations leak out and the cell dies. The polymyxins are of virtually no use as systemic agents since they bind to various ligands in body tissues and are potent toxins for the kidney and nervous system. Gramicidins are also membrane-active antibiotics that appear to act by producing aqueous pores in the membranes. They also are used only topically.
Fungal Membranes
Fungal membranes contain sterols, whereas bacterial membranes do not. The polyene antibiotics, which apparently act by binding to membrane sterols, contain a rigid hydrophobic center and a flexible hydrophilic section. Structurally, polyenes are tightly packed rods held in rigid extension by the polyene portion. They interact with fungal cells to produce a membrane-polyene complex that alters the membrane permeability, resulting in internal acidification of the fungus with exchange of K+ and sugars; loss of phosphate esters, organic acids, nucleotides; and eventual leakage of cell protein. In effect, the polyene makes a pore in the fungal membrane and the contents of the fungus leak out. Prokaryotic cells neither bind to nor are inhibited by polyenes. Although numerous polyene antibiotics have been isolated, only amphotericin B is used systemically (Fig. 11-5). Nystatin is used as a topical agent and primaricin as an ophthalmic preparation.
A number of other agents interfere with the synthesis of fungal lipid membranes. These agents belong to a class of compounds referred to as imidazoles: miconazole, ketoconazole, clotrimazole, and fluconazole. These compounds inhibit the incorporation of subunits into ergosterol and may also directly damage the membrane.
FIGURE 11-5 Structure of amphotericin B.
Antibiotics that Inhibit Nucleic Acid Synthesis
Antimicrobial agents can interfere with nucleic acid synthesis at several different levels. They can inhibit nucleotide synthesis or interconversion; they can prevent DNA from functioning as a proper template; and they can interfere with the polymerases involved in the replication and transcription of DNA.
Interference with Nucleotide Synthesis
A large number of agents interfere with purine and pyrimidine synthesis or with the interconversion or utilization of nucleotides. Other agents act as nucleotide analogs that are incorporated into polynucleotides.
Flucytosine (5-fluorocytosine) is an antifungal agent that inhibits yeast species. It is converted in the fungal cell to 5-fluorouracfl, which inhibits thymidylate synthetase resulting in a deficit of thymine nucleotides and impaired DNA synthesis. Adenosine arabinoside inhibits viruses. It is phosphorylated in virus-infected cells and acts as a competitive analog of DATP, inhibiting the incorporation of DATP into DNA. Acyclovir is a nucleoside analog that, after being converted to a triphosphate, inhibits the thymidine kinase and DNA polymerase of herpes viruses. Zidovudine (AZT) inhibits human immunodeficiency virus (HM replication by interfering with viral RNA-dependent DNA polymerase (reverse transcriptase).
Agents That Impair the Template Function of DNA
A number of substances bind to DNA by intercalation. None of them is useful as an antibacterial agent; however, chloroquine and miracil D (lucanthone) inhibit plasmodia and schistosomes, respectively. These agents are thought to intercalate into the DNA and thereby to inhibit further nucleic acid synthesis. Acridine dyes such as proflavine act by this intercalation mechanism, but because they are toxic and carcinogenic in mammals they cannot be used as antibacterial agents.
Inhibition of DNA-Directed DNA Polymerase
Rifamycins are a class of antibiotics that inhibit DNA-directed RNA polymerase (Fig. 11-6). Polypeptide chains in RNA polymerase attach to a factor that confers specificity for the recognition of promoter sites that initiate transcription of the DNA. Rifampin binds noncovalently but strongly to a subunit of RNA polymerase and interferes specifically with the initiation process. However, it has no effect once polymerization has begun.
FIGURE 11-6 Structure of rifamphin, which inhibits the DNA-directed RNA polymerase.
Inhibition of DNA Replication
DNA gyrase and topoisomerase I act in concert to maintain an optimum supercoiling state of DNA in the cell. In this capacity, DNA gyrase is essential for relieving torsional strain during replication of circular chromosomes in bacteria. The enzyme is a tetrameric protein composed of two A and two B subunits. A transient, covalent bond between the A subunit and DNA occurs during the double strand passage reaction catalyzed by gyrase. Quinolones such as nalidixic acid (Fig. 11-7), bind to the cleavage complex composed of DNA and gyrase during this strand passage. This interaction of quinolone acts to stabilize the cleavage intermediate which has a detrimental effect on the normal DNA replication process. The effects of this inhibition result in the death of the bacterial cell. The newer fluoroquinolones such as ciprofloxacin, norfloxacin, and ofloxacin also interact with DNA gyrase and possess a broad spectrum of antimicrobial activity.
FIGURE 11-7 Structure of quinolone antibiotica. Nalidixic inhibits only aerobic Gram-negative species. In ciprofloxicin, the flourine provides Gram-positive activity, the piperazine group increases activity against members of the Enterovacteriaceae, and the piperazine and cylopropyl groups give activity against Pseudomonas species.
Nitroimidazoles such as metronidazole inhibit anaerobic bacteria and protozoa. The nitro group of the nitrosohydroxyl amino moiety is reduced by an electron transport protein in anaerobic bacteria (Fig. 11-8). The reduced drug causes strand breaks in the DNA. Mammalian cells are unharmed because they lack enzymes to reduce the nitro group of these agents.
FIGURE 11-8 Structure of metronidazole and its mechanism of action. Metronidazole enters an aerobic bacterium where, via the electron transport protein ferrodoxin, it is reduced. The drug then binds to DNA, and DNA breakage occurs.
Antimicrobial Inhibitors of Ribosome Function.
The basic structure and function of ribosomes are presented in Fig. 11-1. A number of antibacterial agents act by inhibiting ribosome function. Bacterial ribosomes contain two subunits, the 50S and 30S subunits, and it is possible to localize the action of antibiotics to one or both subunits. It is also possible to isolate the specific ribosomal proteins to which an agent binds and to isolate bacterial mutants that lack a specific ribosomal protein and therefore show resistance to a particular agent.
Aminoglycosides act by binding to specific ribosomal subunits. Aminoglycosides are complex sugars connected in glycosidic linkage (Fig. 11-9). They differ both in the molecular nucleus, which can be streptidine or 2-deoxystreptidine, and in the aminohexoses linked to the nucleus. Essential to the activity of these agents are free NH, and OH groups by which aminoglycosides bind to specific ribosomal proteins. Streptomycin, the first aminoglycoside studied, was a useful too] in elucidating protein synthesis. However, it is rarely used clinically today except to treat tuberculosis, and its mode of action differs to some extent from that of the other clinically useful aminoglycosides, which are 2-deoxystreptidine derivatives such as gentamicin, tobramycin, and amikacin. Streptomycin binds to a specific S12 protein in the 30S ribosomal subunit (Fig. 11-10) and causes the ribosome to misread the genetic code. Other aminoglycosides bind not only to the S12 protein of the 30S ribosome, but also to some extent to the L6 protein of the 50S ribosome. This latter binding is quite important in terms of the resistance of bacteria to aminoglycosides. Indeed, the aminoglycoside-type drugs can combine with other binding sites on 30S ribosomes, and they kill bacteria by inducing the formation of aberrant, nonfunctional complexes as well as by causing misreading (Fig. 11-11).
FIGURE 11-9 Structures of three aminoglycoside antibiotics used clinically. Critical aspectos of the molecules are the amino and hydroxy groups that bind to proteins in the ribosomes.
FIGURE 11-10 Diagrammatic representation of inhibition sites of protein biosynthesis by various antibiotics that bind to the 30S and 50S ribosomes.
FIGURE 11-11 Inhibition of protein biosynthesis by aminoglycosides.
Spectinomycin is an aminocylitol antibiotic that is closely related to the aminoglycosides. It binds to a different protein in the ribosome and is bacteriostatic but not bactericidal. It is used to treat penicillin-resistant gonorrhea.
Other agents that bind to 30S ribosomes are the tetracyclines (Fig. 11-12). These agents appear to inhibit the binding of aminoacyl-tRNA into the A site of the bacteria] ribosome. Tetracycline binding is transient, so these agents are bacteriostatic. Nonetheless, they inhibit a wide variety of bacteria, chlamydias, and mycoplasmas and are extremely useful antibiotics.
FIGURE 11-12 Structure of tetracycline showing the area critical for activity and major and minor points of modification.
There are three important classes of drugs that inhibit the 50S ribosomal subunit. Chloramphenicol (Fig. 11-13) is a bacteriostatic agent that inhibits both Gram-positive and Gram-negative bacteria. It inhibits peptide bond formation by binding to a peptidyltransferase enzyme on the 50S ribosome. Macrolides are large lactone ring compounds that bind to 50S ribosomes and appear to impair a peptidyltransferase reaction or translocation, or both. The most important macrolide is erythromycin, which inhibits Gram-positive species and a few Gram-negative species such as Haemophilus, Mycoplasma, Chlamydia, and Legionella.
FIGURE 11-13 Structure of chloranphenicol.
New molecules such as azithromycin and clarithromycin have greater activity than erythromycin against many of these pathogens. Lincinoids, of which the most important is clindamycin, have a similar site of activity (Fig. 11-14). Both macrolides and lincinoids are generally bacteriostatic. inhibiting only the formation of new peptide chains.
FIGURE 11-14 Structure of erythromycin (prototype or macrolide) and clindamycin. Although extremely different in structure, both compounds inhibit protein synthesis by binding to 50S ribosome.
Drugs that Inhibit Other Biochemical Targets
Both trimethoprim and the sulfonamides interfere with folate metabolism in the bacterial cell by competitively blocking the biosynthesis of tetrahydrofolate, which acts as a carrier of one-carbon fragments and is necessary for the ultimate synthesis of DNA, RNA and bacterial cell wall proteins (Fig. 11-15). Unlike mammals, bacteria and protozoan parasites usually lack a transport system to take up preformed folic acid from their environment. Most of these organisms must synthesize folates, although some are capable of using exogenous thymidine, circumventing the need for folate metabolism.
FIGURE 11-15 Structure of sulfonamide and trimethoprim with sites of inhibition of folic metabolism.
Sulfonamides competitively block the conversion of pteridine and p-aminobenzoic acid (PABA) to dihydrofolic acid by the enzyme pteridine synthetase. Sulfonamides have a greater affinity than p-aminobenzoic acid for pteridine synthetase. Trimethoprim has a tremendous affinity for bacterial dihydrofolate reductase (10,000 to 100,000 times higher than for the mammalian enzyme); when bound to this enzyme, it inhibits the synthesis of tetrahydrofolate.
Antibacterial Agents that Affect Mvcobacteria
Isoniazid is a nicotinamide derivative that inhibits mycobacteria. Its precise mode of action is not known, but it affects the synthesis of lipids, nucleic acids, and the mycolic acid of the cell walls of these species. Ethambutol is also an antimycobacterial agent whose mechanism of action is unknown. It is mycostatic, whereas isoniazid is mycocidal. The other antituberculosis drugs, rifampin and streptomycin, affect mycobacteria in the same manner that they inhibit bacteria. Pyrazinamide is a synthetic analog of nicotinamide. It is bactericidal, but its exact mechanism is unknown
Bacterial Resistance
Bacteria have proved adept at developing resistance to new antimicrobial agents. There are a number of ways in which bacteria can become resistant (Table 11-2). Most of the early studies of bacterial resistance focused on single-step mutational events of chromosomal origin. Resistance to the early sulfonamides, for example, was the result of a single amino acid change in the enzyme pteridine synthetase that caused sulfonamides to bind less well than p-aminobenzoic acid. Similarly, a single step mutation that altered a ribosomal protein conferred resistance to streptomycin. In the late 1950s, Japanese workers found that enteric bacteria such as Shigella dysenteriae had become resistant not only to sulfonamides but also to the tetracyclines and chloramphenicol. This resistance was due not to a chromosomal change, but rather to the presence of extrachromosomal DNA that was transmissible. This type of resistance is called plasmid-mediated resistance.
Resistance-conferring plasmids are present in virtually all bacteria (Table 11-3). For example, resistance to ampicillin appeared in Haemophilus influenzae in 1974 and in Neisseria gonorrhoeae in 1976. In the last several years, organisms such as enterococci have been shown to contain plasmids that confer resistance to drugs such as ampicillin and aminoglycosides.
Bacteria also contain transposons, which can insert into plasmids and also into the chromosome (see Ch. 5). Transposon-mediated resistance to most of the major antibiotics has been found in the past few years.
Antimicrobial agents exert a strong selective pressure on the development of both chromosomal and plasmid-mediated resistance, as discussed below. Administration of an antibiotic destroys the susceptible bacteria in a population, but may permit resistant ones to proliferate. From an epidemiologic viewpoint, plasmid-mediated resistance is the most important type, since it is transmissible, is usually highly stable, confers resistance to many different classes of antibiotics simultaneously, and often is associated with other characteristics that enable a microorganism to colonize and invade a susceptible host.
Mechanisms of Resistance
The basic mechanisms by which a microorganism can resist an antimicrobial agent are (1) to alter the receptor for the drug (the molecule on which it exerts its effect); (2) to decrease the amount of drug that reaches the receptor by altering entry or increasing removal of the drug; (3) to destroy or inactivate the drug; and (4) to develop resistant metabolic pathways. Bacteria can possess one or all of these mechanisms simultaneously.
Resistance Due to Altered Receptors
ß-Lactam Resistance. The ability to analyze changes in receptors for ß-lactams by competition experiments in which [ 14C]penicillin is inhibited from binding to penicillin-binding proteins has explained a number of cases of bacterial resistance to penicillins and cephalosporins. In 1977, Streptococcus pneumoniae strains resistant to penicillin G were encountered in South Africa. Plasmids were not the cause of the resistance. Penicillin-resistant S pneumoniae cells have altered penicillin-binding proteins, which bind penicillin less well. Resistance of S pneumoniae to penicillin has been increasing, and there are now relatively resistant isolates (minimal inhibitory concentration [MIC], 0.1 to 1 mg/ml) in many parts of the world.
Altered penicillin-binding proteins also explain the resistance of some Staphylococcus aureus strains to ß-lactamase-stable penicillins (the so-called methicillin-resistant strains). The ß-lactams induce synthesis of a new penicillin-binding protein, PBP2a, which does not bind any ß-lactam. The ß-lactam resistance of coagulase-negative staphylococci is also the result of altered penicillin-binding proteins. Staphylococcal organisms resistant to methicillin are resistant to all penicillins, cephalosporins, and carbapenems.
The resistance of group D streptococci to ß-lactam antibiotics appears to be the result of lower affinity of the penicillin-binding proteins for the penicillins. Enterococci are resistant to all cephalosporins because of failure to bind to the penicillin-binding proteins. One Gram-negative species for which resistance to ß-lactam antibiotics can be correlated with diminished affinity of the target enzymes is N gonorrhoeae.
Vancomycin Resistance. Certain transposable genetic elements encode special cell wall-synthesizing enzymes which change the structure of the normal D-Ala-D-Ala side chain in the peptidoglycan assembly pathway. The altered side chain (D-Ala-D-Lac) does not bind vancomycin and allows normal peptidoglycan polymerization to occur in the presence of the drug. Depending upon the nature of the vancomycin resistance gene, high-level resistance can occur to glycopeptides. Thus far, this type of resistance has been found in enterococci but not in multi-resistant isolates of Staphylococcus aureus.
Macrolide-lincomycin Resistance. Macrolide-lincomycin resistance in clinical isolates of staphylococci and streptococci has been recognized for several decades. The resistance is due to methylation of two adenine nucleotides in the 23S component of 50S RNA. This resistance is plasmid mediated, and the resistance is encoded on transposons. Resistance results from induction of an enzyme that is normally repressed. The methylated RNA binds macrolidelincomycin-type drugs less well than unmethylated RNA does. Induction of resistance varies by species, and in most Gram-positive species erythromycin is a more effective inducer of resistance than is clindamycin. The plasmids that mediate macrolide-lincomycin resistance in streptococci and staphylococci have extensive structural similarity, indicating that these plasmids readily pass between these species.
Rifampin Resistance. The resistance of bacteria to rifampin is caused by an alternation of one amino acid in DNA-directed RNA polymerase, which results in reduced binding of rifampin. The degree of resistance is related to the degree to which the enzyme is changed, but does not correlate strictly with enzyme inhibition. This form of resistance occurs at a low level in any population of bacteria so that resistance develops by natural selection during a course of therapy. Naturally resistant organisms are more common among members of the Enterobacteriaceae, explaining why agents of urinary tract infections rapidly became resistant to rifampin. The resistance of Neisseria meningitides to rifampin appeared in closed military settings in which rifampin has been used for prophylaxis.
Sulfonamide-trimethoprim Resistance. Sulfonamide can be rendered ineffective by altered or new dihydropteroic synthetase that has poor affinity for sulfonamides and preferentially binds p-aminobenzoic acid. Sulfonamide resistance of this type can result from a point mutation or from acquisition of a plasmid that causes synthesis of the new enzyme. A most serious resistance problem is an increase in resistance to trimethoprim. This plasmid- and transposon-mediated resistance is due to production of an altered dihydrofolate reductase that has markedly reduced affinity for trimethoprim.
Quinolone Resistance. Resistance to quinolones can be caused by mutations in DNA gyrase subunits A or B, reduced outer membrane permeability in gram-negative cells, or to active efflux transporters found in many bacteria. The highest level of resistance to the newer fluoroquinolones is most frequently associated with chromosomal mutations, causing amino acid substitutions in a highly conserved region in the A subunit of DNA gyrase. Multiple-mechanisms of resistance can occur in a single isolate of bacteria, leading to a higher level of resistance to many fluoroauinolones.
Resistance Due to Decreased Entry of a Drug
Tetracycline Resistance. The uptake of tetracycline by members of the Enterobacteriaceae is biphasic. In an initial energy-independent rapid phase, tetracycline binds to cell surface layers and passes by diffusion through the outer layers of the cell. In the second, energy-dependent phase, tetracycline crosses the cytoplasmic membrane, probably by means of a proton-motive force. The precise transport system has not been identified.
Tetracycline resistance is common in both Gram-positive and Gram-negative bacteria. In most cases it is plasmid encoded and inducible; however, chromosomal, constitutive resistance is found in some organisms such as Proteus species. Many plasmid-encoded specified tetracycline resistance determinants have been found in enteric bacteria. The most common of these determinants, TetB, is also present in H influenzae. Tetracycline resistance in Staphylococcus aureus is due primarily to small multicopy plasmids; chromosomal resistance is rare. Tetracycline resistance is found on nonconjugative plasmids in Streptococcus faecalis and on the chromosome of S pneumoniae, S agalactiae (group B streptococci), and oral streptococci. Clostridium species such as C difficile harbor chromosomal genes for tetracycline resistance.
Basically, tetracycline resistance is due to a decrease in the levels of drug accumulation. Decreased uptake and increased efflux both probably participate. Resistant bacteria bind less tetracycline, and the tetracycline they do accumulate is lost by an energy-dependent process when they are in a drug-free milieu.
Plasmid-mediated resistance to tetracyclines can be partially overcome in Gram-positive species by modifying the tetracycline nucleus. Hence, achievable concentrations of minocycline and doxycycline, in particular, will inhibit some tetracycline-resistant streptococci such as S pneumoniae, and some S aureus strains. Molecular modification has not been successful in overcoming the tetracycline resistance of members of the Enterobacteriaceae or Pseudomonas or most Bacteroides species.
Tetracycline resistance is a major concern because it is located on plasmids near insertion sites, and these plasmids readily acquire other genetic information to enlarge the spectrum of resistance. The widespread use of tetracycline in animal feeds may be a factor in the extensive, worldwide resistance of members of the Enterobacteriaceae, particularly enteric species such as Salmonella, to tetracyclines and subsequently to many other drugs. Not only can tetracycline resistance move among members of the Enterobacteriaceae on plasmids, but plasmids mediating tetracycline resistance have moved between S aureus, S epidermidis, S pyogenes, S pneumoniae, and S faecalis.
Fosfomycin Resistance. Fosfomycin and fosmidomycin, which inhibit cell wall synthesis, enter bacteria by means of a glycerol-phosphate or glucose-6-phosphate transport system. Gram-positive bacteria in which the glucose-6-phosphate transport system is poorly developed do not take up these drugs in concentrations adequate to inhibit the cell wall synthesis. This resistance usually is chromosomal. The resistance of Gram-negative bacteria to these agents is related primarily to the presence in the population of some bacteria that can function without the transport system. Plasmids and transposons that transfer resistance to fosfomycin have been found in bacteria such as Serratia marcescens.
Aminoglycoside Resistance. In the most important form of aminoglycoside resistance, the compound is modified outside the cell and resistance is due partly to poor uptake of the altered compound. Also, all aminoglycosides have free amino and hydroxy groups that are essential for binding to ribosomal proteins. A number of enzymes can acetylate the amino groups and phosphorylate or adenylate the hydroxyl groups (Fig. 11-9). Other forms of resistance, such as altered binding site on 30S ribosomes, are much less common.
In members of the Enterobacteriaceae and in Pseudomonas species, the aminoglycosides pass through the cell wall via channels designed to admit cationic molecules to the periplasmic space. These channels, called porin channels, are lined by the porin protein. Aminoglycosides are then translocated across the cell membrane by an energy-dependent proton-motive force and, in the cytoplasm, bind to ribosomes located just below the membrane. Aminoglycosides bind only to ribosomes actively engaged in protein synthesis. Binding to the ribosomes induces a protein involved in the uptake of the aminoglycosides.
Bacteria may contain in the periplasmic space enzymes that acetylate, phosphorylate, or adenylate aminoglycosides to various degrees. It is not clear whether the enzymes are free in the periplasmic space or bound to the cytoplasmic membrane. The modified aminoglycosides do not bind well to ribosomes, and hence uptake is poor or absent (Fig. 11-16).
FIGURE 11-16 Diagrammatic representation of transfer and transfer reduction of aminoglycoside across the bacterial cell wall. If it is modified by acetylation, adenylation, or phosphorylation (see box), the drug will not bind to ribosomes and will leave the bacterial cell.
Aminoglycoside-modifying enzymes have been found in Gram-positive species such as S aureus, S faecalis, S pyogenes, and enzymes are particularly prevalent in members of the Enterobacteriaceae and P aeruginosa. S pneumoniae. These Many of the genes for aminoglycoside-modifying enzymes are carried on transposons
Anaerobic organisms such as Bacteroides species are resistant to aminoglycosides because they lack an oxygen-dependent transport system to move the drugs across the cytoplasmic membrane. Although most resistance of S aureus to aminoglycosides is due to aminoglycoside-modifying enzymes, small-colony variants of staphylococci also show resistance, which may be due to a defect in adenylate cyclase or in cyclic adenosine 5'-monophosphate (cAMP)binding proteins such that cells with a reduced growth rate do not transport aminoglycosides into the cytoplasm. Some members of the Enterobacteriaceae and P aeruginosa appear to be resistant because of altered porin channels, since these bacteria do not take up any drug and do not have aminoglycoside-inactivating enzymes.
Resistance Due to Destruction or Inactivation of a Drug
Chloramphenicol Resistance. Many Gram-positive and Gram-negative bacteria, including some recently discovered H influenzae strains, are resistant to chloramphenicol because they possess the enzyme chloramphenicol transacetylate, which acetylates hydroxyl groups on the chloramphenicol structure. This enzyme, unlike the aminoglycoside-inactivating enzymes and ß-lactamases, is an intracellular enzyme of higher molecular weight and subunit structure. Acetylated chloramphenicol binds less well to the 50S ribosome.
ß-Lactam Resistance. The best-known mechanism of bacterial resistance is the resistance to ß-lactams, which is mediated by penicillinase enzymes. Resistance of E coli to penicillin was recognized in 1940, before sufficient penicillin was made to be clinically useful. In the 1940s, resistance of staphylococci was shown to be due to a penicillinase. As these enzymes also attack other ß-lactam compounds such as cephalosporins, carbapenems, and monobactams, they would be more appropriately designated ß-lactamases. The most important activity of these enzymes is alteration of the ß-lactam nucleus (Fig. 11-17). ß-lactamases are widely distributed in nature and are usually classified on the basis of the principal compounds they destroy (e.g., as penicillinases or cephalosporinases) (Fig. 11-18). ß-Lactamases may be chromosomally or plasmid mediated, and they may be constitutive or inducible.
FIGURE 11-17 Site of ß-lactamase attack in penicillins and cephalosporins.
FIGURE 11-18 ß-Lactamase found in bacteria and their classification and synthesis, whether chromosomally or plasmid mediated.
In Gram-positive species, ß-lactamases are primarily exoenzymes; that is, they are excreted into the milieu around the bacteria. Virtually all hospital isolates of staphylococci, both S aureus and S epidermidis, have ß-lactamases, and 50 to 80 percent of community-acquired staphylococcal isolates produce ß-lactamases. In Gram-negative species, both aerobic and anaerobic, ß-lactamases are contained in the periplasmic space, thus effectively protecting the penicillin-binding proteins.
Resistance of staphylococci to ß-lactams was soon overcome with the antistaphylococcal penicillins and the cephalosporins. Some strains of S aureus produce more ß-lactamase constitutively and can destroy some of the cephalosporins.
In 1974, H influenzae was shown to possess a plasmid mediated ß-lactamase. At present 10 to 35 percent of H influenzae strains in the United States produce ß-lactamases. The TnA transposon has become more widespread, and the resistance of Haemophilus species to penicillin G and ampicillin seems to be increasing yearly. The Haemophilus ß-lactamase is the same structurally as the enzyme found in E coli, Salmonella, Shigella, and N gonorrhoeae. The enzyme has generally been called the TEM enzyme after the initials of the Greek girl from whom an E coli strain containing a plasmid ß-lactamase was first isolated. These enzymes are also called Richmond-Sykes class IIIa enzymes from a classification proposed by Richmond and Sykes in 1973. By far the most common plasmid ß-lactamase found in nature is TEM-1, which accounts for 75 to 80 percent of plasmid-mediated ß-lactamase resistance worldwide. Recently new ß-lactamases have been found that hydrolyze compounds such as inomethoxy cephalosporin, which were not destroyed by other plasmid-encoded ß-lactamases. The new ß-lactamases have an altered amino acid composition, which permits binding to the cephalosporin and subsequent hydrolysis. How common these new enzymes will become is unknown.
Chromosomally mediated ß-lactamases are present in many Enterobacter, Citrobacter, Proteus-Providencia, and Pseudomonas species. All Klebsiella species possess a ß-lactamase, which acts primarily as a penicillinase and is chromosomally mediated. Constitutively produced ß-lactamases are also present in many anaerobic species.
Table 11-3 lists the major ß-lactamases of clinical importance. ß-Lactamases vary in their ability to destroy penicillins and cephalosporins. ß-Lactamase activity is only one component of the ß-lactam resistance of Gram-negative bacteria, since resistance to ß-lactams is a combination of decreased entry, ß-lactamase stability and affinity of the compounds for penicillin-binding proteins.
Synthesis of Resistant Metabolic Pathway
No synthesis of a new type of cell wall resistant to ß-lactams has occurred, but some bacteria, particularly some streptococci, lack the hydrolytic enzymes necessary for forming a new cell wall, and so ß-lactams do not lyse these bacteria. An altered hydrolytic system thus converts a bactericidal antibiotic into a bacteriostatic agent. Whether such resistance occurs in Gram-negative species is not clear.
Some thymidine-requiring streptococci are not inhibited by trimethoprim and sulfonamides and so are not killed by these agents. These organisms are a rare cause of urinary tract infections. Other bacteria produce adequate deoxyribosylthymine 5'-monophosphate (DTMP) by alternative methods and, as a result. survive exposure to these folate inhibitors.
Certain Candida or Cryptococcus yeasts are resistant to flucytosine because they cannot convert it to its active component, fluorouracil. Other fungi can resist the polyenes and imidazoles because they synthesize membrane components by different metabolic mechanisms.
Combinations of Antimicrobial Agents
Antibiotics are frequently used in combination for the following reasons: (1) to treat a life-threatening infection; (2) to prevent emergence of bacterial resistance; (3) to treat mixed infections of aerobic and anaerobic bacteria; (4) to enhance antibacterial activity (synergy); and (5) to use lower doses of a toxic drug. Combined treatment is reasonable when the precise agents of a serious infection are unknown. Use of two or more drugs to prevent the emergence of resistance is effective for tuberculosis and for therapy of some chronic infections. The use of combinations to achieve synergy is more complicated. Synergy occurs when a combination of two drugs causes inhibition or killing when used at a fourfold-lower concentration than that of either component drug used separately (Fig. 11-19). However, indifference or antagonism may occur instead. Indifference means that the combined action is the same as with either component; antagonism refers to a reduction in the activity of one or both components in the presence of the other (Fig. 11-19).
FIGURE 11-19 Example of how two antibiotics (A and B) may interact with synergy, indifference, or antagonism.
Important examples of bacterial synergy include (1) combinations of anti-cell wall agents with aminoglycosides, (2) use of ß-lactamase inhibitors with ß-lactamase-susceptible antibiotics, and (3) combinations of drugs that act on sequential steps in bacterial metabolic or synthetic pathways. Examples of synergy that have proved clinically important are the combination of penicillin and streptomycin to treat Enterococcus faecalis endocarditis and the combination of carbenicillin and gentamicin to treat P aeruginosa infections. Recently the ß-lactamase inhibitor clavulanic acid or sulbactam has been combined with amino penicillins to inhibit S aureus, Klebsiella pneumoniae, H influenzae, and anaerobic organisms such as Bacteroides species, all of which are resistant to amoxicillin or ampicillin when they contain ß-lactamases. The combination of sulfamethoxazole and trimethoprim attacks two parts of the folic acid cycle and synergistically inhibits many bacteria. Finally, combinations of two penicillins that affect different stages of cell wall synthesis in Gram-negative bacteria are synergistic. This is true for the combination of mecillinam (amdinocillin), which binds to PBP2 of bacteria, and penicillins or cephalosporins, which bind to PBP lb or 3.
Antagonism can occur when a bacteriostatic agent is combined with a bactericidal agent. The classic example has been the combination of chlortetracycline and penicillin in treatment of pneumococcal meningitis. This effect has not been explained from a molecular standpoint, and tetracyclines and penicillins or cephalosporins are used to treat mixed infections such as pelvic inflammatory disease due to N gonorrhoeae and Chlamydia. Some ß-lactam antibiotics can induce ß-lactamases that inactivate other ß-lactams, and antagonism can be shown in the test tube, but the relevance for clinical infections is not established.
Toxicology of Antimicrobial Agents
Antimicrobial agents can be directly toxic, can interact with other drugs to increase their toxicity, or can alter microbial flora to cause infection by organisms that are normally saprophytic. Allergic reactions can be caused by an agent, but penicillins can produce either immediate, IgE-mediated, or delayed hypersensitivity reactions. Cutaneous reactions have been reported with every class of antimicrobial agent. Hematologic reactions can range from the life-threatening blood dyscrasia that occurs in 1 in 60,000 individuals who receive chloramphenicol to hemolytic anemia due to sulfonamides in individuals who lack the enzyme glucose-6-phosphate dehydrogenase. Depression of blood platelet activity has occurred with many agents. By altering the gastrointestinal flora, almost all antibiotics can cause overgrowth of Clostridium difficile, which produces a toxin that causes diarrhea and even pseudomembranous colitis. Alteration of intestinal flora by antibiotics can also result in overgrowth of Candida in the mouth, vagina, or gastrointestinal tract. Since a number of antibiotics are metabolized in the liver, damage to the liver can occur. This has been of particular concern with isoniazid, which is used to treat tuberculosis. Damage to the kidneys can follow the use of aminoglycosides. Neurologic toxicity is fortunately fairly uncommon, but the aminoglycosides can damage the auditory or vestibular apparatus if the dosage is not closely monitored.
Bacteria continue to evolve new mechanisms of resistance to old and to new antimicrobial agents. Some bacteria such as P aeruginosa are particularly adept at utilizing a number of different mechanisms simultaneously to become resistant to agents in virtually every class and those with such diverse sites of action as cell wall, protein biosynthesis, or DNA and RNA synthesis. Progress in medicine will keep patients alive who have nosocomial infections with resistant pathogens.
Mechanism to Reduce Bacterial Resistance
Proper selection of new antibiotics will be a major force in slowing the development of antimicrobial resistance. Proper hygiene practices will reduce plasmid transfer and the establishment of multiple drug-resistant bacteria in the hospital and will delay the appearance of such species in the community. Table 11-4 lists a number of mechanisms to prevent bacterial resistance. The health care provider must be continually alert to the appearance of antibiotic resistance within the hospital and community.
REFERENCES
Arthur M: Genetics and mechanisms of glycopeptide resistance in enterococci. Antimicrob Agents Chemother 37:1563, 1993
Gale EF, Cundliffe E, Reynolds PE et al: The Molecular Basis of Antibiotic Action. 2nd Ed. John Wiley & Sons, New York, 1981
Kucers A, Bennett N: The Use of Antibiotics. 4th Ed. JB Lippincott, Philadelphia, 1985. Lorian V (ed): Antibiotics in Laboratory Medicine. 3rd Ed. Williams & Wilkins, Baltimore, 1991
Murray B: New Aspects of antimicrobial resistance and the resulting therapeutic dilemmas. J Infect Dis 163:1185, 1991
Neu HC (ed): Update on antibiotics. 1. Med Clin N Am 71:1051, 1987
Neu HC (ed): Update on antibiotics. 11. Med Clin N Am 72:555, 1988
Neu H: The crisis in antibiotic resistance. Science 257:1064, 1992
Norrby SR, Bergan T, Holm SE et a] (eds): Evaluation of new beta-lactam antibiotics. Rev Infect Dis 8 (Suppl. 3):S235, 1986
Schaberg D: Resistant gram-positive organisms. Ann Emergency Med 24(3):462, 1994
Waxman DJ, Strominger JL: Beta-lactam antibiotics: biochemical modes of action. p. 210. In Morin RB, Gorman M (eds): Chemistry and Biology of Beta-Lactam Antibiotics. Academic Press, San Diego, 1982
Wolfson JS, Hooper DC (eds): Quinolone Antimicrobial Agents. 2nd Ed. American Society for Microbiology, Washington, 1993