Antibiotic resistance is a natural phenomenon, and is a result of the overlap of medical technology with the microorganisms of the biosphere. Humans did not invent antibiotics. They were natural products made by microorganisms to defend their niche in the ecosystem against the next microorganism trying to occupy that niche. The genes for antibiotic resistance evolved from heavy metal (eg, mercury or cadmium) resistance genes. These genes and their products were important to the survival of microorganisms in hostile environments, such as undersea volcanoes. As the biosphere evolved, these genes continued to evolve for interspecies competition. The collective "invention" of antibiotics by human medical technology has led to the use and abuse of these antibiotics in both human and veterinary medicine. The microorganisms, with several billion years of past experience, have evolved "new" resistance mechanisms to overcome the activity of our antibiotics. This has resulted in treatment failures, in our hospitals evolving into nosocomial ecosystems, and in the proclamation of a crisis in antibiotic therapy.1"3
Antibiotics act by inhibition of one of five basic functional mechanisms: (1) bacterial cell wall synthesis; (2) cell membrane function; (3) protein synthesis; (4) nucleic acid synthesis; and (5) a metabolic pathway. It is important to understand the biochemical mechanisms of antibiotic action, the microorganism's resistance mechanism, and ihe transmission of these resistance mechanisms among bacterial species. This aids in the understanding of how the hospital laboratory detects in vitro resistance to antibiotics, as a guide to patient therapy, as well as the concept of cross-resistance among antibiotics with closely related chemical structures.
EPIDEMIOLOGY OF ANTIBIOTIC RESISTANCE
Clinicians are familiar with the seemingly inexorable march of antibiotic resistance among the organisms they encounter in their patients. At the start of the antibiotic era, most clinically common organisms (staphylococci, gonococo, haemophilli, pneumococci, and enteric bacteria) were very sensitive to penicillin, tetracycline, streptomycin, chloramphenicol, and sulfonamides. Bacteria isolated from the feces of rats on remote volcanic islands, not previously visited by humans, have few plasmids and rarely exhibit any significant antibiotic resistance. Now, wherever antibiotics are used, it is easy to demonstrate increased prevalence of plasmids and transposons, and a parallel antibiotic resistance in common soil microorganisms. This is apparently due to transmission of resistance from human and animal fecal flora back to the normal soil microorganisms.
Evolution of Antibiotic Resistance In Haemophllus Influenza*?
As soon as antibiotics were used in significant amounts, resistance began to emerge. In fact, based on retrospective studies of bacterial culture collections, some organisms were resistant to certain antibiotics before the antibiotics were manufactured, a consequence of natural resistance. A pattern soon appeared of individual strains of antibiotic-resistant bacteria emerging at scattered locations around the world and then spreading from country to country with travel. This happened as follows. Once the bacteria were introduced into hospitals and nursing homes, there would be intense nosocomial transmission. Strains with limited resistance would be rapidly passed from patient to patient while being exposed to a variety of antibiotic pressures. Genetic recombination mechanisms would produce strains resistant to several antibiotics, especially in areas of intense selection, such as intensive care units. From there, these strains would spread to the community.
The pattern varied from antibiotic to antibiotic and from organism to organism. Intense use of a particular antibiotic in community medicine could also drive the development of resistance. For example, the emerging resistance of Streptococcus pneumoniae to penicillin is due to the complex interactions of children in day care centers, working parents, and pediatricians pressured by parents to treat the child's earache with an antibiotic so that they can go back to work quickly, and the failure of the parents to finish a full course of therapy. Here, inadequate therapy results in a relapse and this leads to re-treatment with penicillin or another antibiotic. Pneumococcal strains are therefore selected for penicillin resistance by cyclic exposure to sub-inhibitory doses of penicillin.
The epidemiology of resistance to an antibiotic may vary from country to country. For example, vancomycin-resistant enterococci (VRE) found in the United States represent a limited number of strains that are resistant to multiple antibiotics as well as to vancomycin.4 In contrast, European countries have more polydonal VRE strains with limited prevalence of multiple antibiotic resistance. Why should this difference exist? The answer is different selection pressures. In the United States, the selective pressure is vancomycin use in hospital environments where organisms are likely to be resistant to other antibiotics. This results in selection for the multiple antibiotic-resistant VRE strain (the intensive care unit strain). The selective pressure in Europe is explained by the use of the glycopeptide antibiotic avoparcin as a growth promoter in animal feeds. Avoparcin produces cross-resistance to vancomycin, but is selecting for resistance in a population of animal strains of enterococci that are usually susceptible to many other antibiotics.
How common is the emergence of bacterial resistance? In fact, we all have experienced this evolution. Although penicillin-resistant S. pneumoniae5 and vancomycin-resistant Staphyïococcus aureus6 are examples of recent interest, the development of ampicillin-resistant and chloramphenicol-resistant Haemophilus influenzae during the past 30 years is a practical lesson to all physicians caring for children. As shown in Table 1, ampicillin was the treatment of choice for H. inßuenzae type B infections in the 1960s and early 1970s. As ampicillin was more widely used, the organism acquired a ß-lactamase that rendered it resistant. Selection of resistant populations occurred rapidly. Chloramphenicol, although thought to be associated with a higher risk for serious adverse reactions, was then added to the routine treatment regimen because of increasing treatment failures. By the early 1980s, chloramphenicol resistance was spreading among H. influenzae type B populations, presumably selected by the increase in chloramphenicol use. Fortunately, the third-generation cephalosporins were beginning to be marketed, providing yet another solution to a rapidly emerging crisis in treatment options.
How long this cycle of emerging resistance would have continued before we ran out of alternatives is mere speculation (there certainly are other species that could provide the genetic material for high-level cephalosporin resistance); however, as Table 1 shows, the solution finally came not from the creation of a new antibiotic, but from a vaccine that prevented infection altogether. The lesson with H. influenzae type B may be more generally applied to the entire bacterial (and now viral) ecosystem - these are living organisms fully capable of developing and spreading resistance to any treatment. The more recent experience with S. pneumoniae is another example. In the final analysis, prevention via the development of vaccines is a much more reliable and permanent solution.
GENETICS OF ANTIMICROBIAL RESISTANCE
A microorganism's resistance to antibiotics can be intrinsic or acquired.1 Intrinsic resistance is common to all members of a genus or species. For example, Escherichia coli, Klebsiella species, and Enterobacter species are resistant to methicillin, clindamycin, and vancomycin due to insensitive targets, lack of penetration, or some other mechanism. An organism producing antibiotics in the ecosystem, such as a soil streptomycete, must be resistant to its own antibiotic (streptomycin). If the gene for streptomycin resistance is transferred and gets into another organism, that organism becomes resistant to streptomycin. This acquired resistance is usually found in only a few members of a genus or species.
Expression of antibiotic resistance can be (1) constitutive, meaning expressed with or without exposure to an antibiotic stimulus; (2) inducible, or produced only after exposure to an antibiotic stimulus; or (3) constitutive-inducible, produced at a low level without stimulus and produced at a high level with antibiotic Stimulation. These subspecies "strains" acquire antibiotic resistance as a result of a rare chromosomal mutation or by the acquisition of genetic information (plasmid [Rfactor] or transposon)7 by means of transduction, transformation, or conjugation.
Plasmids are small, circular DNA elements that reside in bacterial cells and replicate independently of the host chromosome. Some plasmids carry genes for antibiotic resistance as well as genes controlling conjugation. Conjugation involves cell-to-cell contact (mating) via a sex pilus, with DNA being transferred from a donor cell to a recipient cell. These plasmids are called R-plasmids (R for resistance), and they may carry multiple genes for different types of antibiotic resistance (called resistance transfer factors or RTFs). They are able to spread drug resistance to other strains and different species as a single unit.
Incorporation of conjugal transmissible plasmids into a broad host range is important for horizontal transmission of antibiotic resistance genes both within a species and among genera. Once the resistance gene cassette is integrated into a plasmid or a chromosome, it is selected for by continued antibiotic pressure. Removal of selective pressure by discontinuing the antibiotic does not automatically result in loss of resistance. Additionally, the loss of resistance genes is not selectable, except by a long process involving many generations of bacterial reproduction in the absence of antibiotic pressure. It takes slightly more energy to maintain a resistant organism, and in the absence of selection, many generations are required for this energy difference to give a selective advantage to a sensitive strain. Nevertheless, it appears that if antibiotic pressure is discontinued, sensitive strains will ultimately replace resistant strains - evidence mat supports the current call for more conservative use of antimicrobial agents.
MECHANISMS OF RESISTANCE
There are likely many different biochemical mechanisms whereby microorganisms may resist antibiotics.1 As illustrated in the figure, the welldescribed biochemical mechanisms are:
1. The microorganism produces an enzyme that destroys the antibiotic molecule.8 Examples of this are the penicillinase of S. aureus and the aminoglycoside-modifying enzymes of gramnegative and gram-positive bacteria.
2. The microorganism changes cell wall permeability to the antibiotic.9 Examples of this are the tetracycline and erythromycin efflux pumps and altered outer membrane proteins (porins) in gram-negative bacteria. Some gram-negative bacteria have reduced proton motive force, which decreases the influx of aminoglycosides.
3. The microorganism develops an altered structural target for the antibiotic.10 Examples of this are resistance to aminoglycosides by an alteration of a protein in the 3OS ribosomal subunit of the bacterial ribosome, altered penicillin binding proteins (PBFs; eg, methicillin-resistant S. aureus [MRSA] or pneumococcus), and altered DNA gyrase in quinolone resistance. The microorganism may also develop an altered metabolic enzyme that is less affected by the antibiotic than is the enzyme in susceptible organisms. An example of this mechanism is trimethoprim and sulfamethoxazole resistance, where the dihydrofolic acid reducíase is inhibited less efficiently than is the enzyme in susceptible organisms.
4. The microorganism develops an altered metabolic pathway. An example of this mechanism is sulfonamide resistance, where resistant bacteria use preformed folie acid rather than the antibiotic analogue para-aminobenzoic acid.
Why bother to understand these different mechanisms of resistance? Again, our experiences with H. influenzile and S. pneumoniae are perfect examples. H. influenzae resistance to ampicillin is mediated by a ß-lactamase. It engendered the development and marketing of the third-generation cephalosporins (eg, cefotaxime, ceftriaxone, and cefixime) that are unaffected by the enzyme. When ampicillin-resistant S. pneumoniae subsequently began to emerge, many physicians assumed that the same drugs would be effective. Unfortunately, the mechanism of S. pneumoniae resistance is different - caused by an alteration of PBPs and not a ß-lactamase. Failures of the thirdgeneration cephalosporins in the treatment of S. pneumoniae meningitis were reported and different therapeutic strategies had to be developed.
Figure. Examples of common mechanisms of antibiotic resistance in bacteria.
Some microorganisms are able to produce enzymes that destroy the active antibiotic. The classic example of this mechanism of resistance is the ß-lactamases. These enzymes (more than 170 known) degrade molecules that contain the ß-lactam ring (penicillins, carbapenems, and cephalosporins). They may have differing specificities such that one enzyme may degrade a penicillin but not a cephalosporin and vice versa. The ß-lactamases are excreted into the environment (gram-positive bacteria) and degrade the ßlactam before it reaches the target site (growing peptidoglycan) and binds to PBPs to cause an inhibition of transpeptidation (cross-linking) of the murein or rigid structure of the bacterium. In the sensitive organism, this inhibition causes the cell wall to be weakened, resulting in the osmotic rupture of the bacterium. Most gram-negative bacteria make low levels of ß-lactamases; however, some organisms such as Klebsiella species have inducible extended spectrum ß-lactamases. This ß-lactamase is overproduced on induction, extending the resistance to normally susceptible third-generation cephalosporins and imipenem.
The ß-lactamases that mediate resistance are found on plasmids (gram-positive organisms) or on plasmids and host chromosomes (gram-negative organisms). Another example of enzymatic degradation is by the aminoglycosidases, which have been described in Enterococcus species. These enzymes promote high-level aminoglycoside resistance to kanamycin, gentamicin, streptomycin, and neomycin. The aminoglycosides are inactivated by phosphorylation, acetylation, and adenylation enzymes that are found in all types of bacteria but are widespread among gram-negative organisms. Erythromycin, tetracycline, and chloramphenicol are also enzymatically inactivated.
CHANGES IN CELL WALL PERMEABILITY
There are three ways that bacteria modify the cell's permeability as a mechanism of resistance to antibiotics. First, the porin molecules of gramnegative bacteria may be altered so that influx of the antibiotic is limited. Transport across the porin is dependent on size, charge, and hydrophobicity of the antibiotic molecule. Hydrophilic molecules move across the porin more efficiently than hydrophobic molecules. Negatively charged antibiotics also move more slowly across the porin. Zwitterionic (balanced charge molecules) antibiotics move efficiently. For example, imipenem (hydrophilic and zwitterionic) is the most efficient ß-lactam at crossing the porin of enteric bacteria. This may in part explain the imipenem susceptibility of enteric bacteria resistant to third-generation cephalosporins. In contrast, altered porins in part explain resistance to aminoglycosides, chloramphenicol, ß-lactams, and quinolones. These mechanisms are found in pseudomonads and enteric bacteria, but are not necessarily die main mechanism of resistance.
Second, for some bacteria, such as staphylococd, pseudomonads, and enterics, the cytoplasmic membrane may act as a barrier to influx of the antibiotic. For the antibiotic to cross this membrane, an expenditure of energy via proton motive force is required. Antibiotics such as ß-lactams and aminoglycosides have decreased influx when this energy-dependent transport is shut down.
Third, there are efflux pumps that actively transport the antibiotic out of the cell as fast as it enters. Thus, the concentration of antibiotic in the cell never gets high enough to have an antibiotic effect. The transporter molecule requires energy to function. This type of mechanism is important in the resistance of bacteria to tetracycline and the macrolides (erythromycin, clarithromycin, and azithromycin). Staphylococci use this mechanism to export quinolones. In Pseudomonas aeruginosa, a special outer membrane protein (OprK) that normally secretes the siderophore pyoverdin also has the ability to transport (secrete) multiple antibiotics and thus be resistant. Mutants that have lost the OprK protein are susceptible to the same antibiotics.
ALTERED STRUCTURAL TARGETS
The PBPs of bacteria are the structural targets of ß-lactams. Bacteria have evolved a mechanism of altering the structure of the PBPs so that the ßlactam no longer binds. The altered proteins are then associated with resistance. The classic mechanism is that of MRSA. In this case the organism makes a new PBP2, designated PBP2a, that has lowered affinity for ß-lactams. This binding protein modification is controlled by a gene known as the MecA gene. Acquisition of MecA results in the organism's being resistant to all ß-lactams. Spread of this gene among S. aureus strains has been a major infection control problem for hospitals around the world. Similar alterations in PBPs are responsible for increased resistance of S. pneumoniae, S. vindans, Pseudomonas species, ß-lactamase-negative H. influenzae and Neìsseria gonorrhoeae, N. meningitidis, E. coli, Enterobacter species, and Adnetobacter baumannii.
Vancomycin works by complexing to the Dalanyl-D-alanine terminus of peptidoglycan precursors, where it prevents die transglycosylation and transpeptidation steps of peptidoglycan synthesis. It also interferes with RNA synthesis and damages the bacterial cytoplasmic membrane. Having three mechanisms of action makes vancomycin an unusual and important antibiotic. All gram-negative organisms are intrinsically resistant to vancomycin because the molecule cannot penetrate the porins due to the large size of the vancomycin molecule. Enterococci, which are gram positive, have been the only bacteria to develop true resistance to vancomycin. The major mechanisms by which enterococci resist vancomycin are (1) the VanA phenotype, which exhibits high-level vancomycin and teicoplanin resistance; and (2) the VanB phenotype, with moderate to high-level resistance and susceptibility to teicoplanin. The genes required for the expression of the VanA phenotype are carried on a transposon designated TN1546. It contains seven vancomycin resistance genes, five of which are required for expression of VanA.
Strategies to Deal With Emerging Antimicrobial Resistance
Since the emergence of VRE, the medical community has been afraid that the plasmid-mediated transfer of the vanA or vanB genes would extend to S. aureus. This transfer was easily accomplished in vitro and in a mouse burn model. Recently, several isolates of MRSA were found to have increased resistance to vancomycin (8 /¿g/ml) as well. This would be scored as intermediate resistance and named vancomycin intermediate Staphylococcus aureus (VISA). These isolates raised alarm because vancomycin is the best antibiotic for treating MRSA infections. When tested, these isolates had not acquired the vanA or vanB (or vanC) genes. The mechanism of resistance appears to be due to an augmentation of cell wall synthesis by overproduction of peptidoglycan monomer precursors and increased production of PBP2a. It is likely that more of these VISA strains will be produced as vancomycin is used to treat MRSA infections. Eventually, the in vitro experiments will be replicated in vivo and a true vancomycin-resistant S. aureus will become problematic.
Some antibiotics act on protein synthesis by blocking initiation complexes, peptide translocation, or the addition of new amino acids to the growing protein chain. Resistance to tetracyclines and aminoglycosides is mediated by a mutational alteration of 3OS ribosomal protein binding sites that blocks the attachment of these antibiotics to the ribosome. Protein synthesis therefore can proceed in the presence of the antibiotic. Inhibition of binding to altered 5OS ribosome proteins produces resistance to erythromycin, chloramphenicol, clindamycin, and lincomycin.
Mutations producing altered receptors on DNA gyrase enzymes (enzymes bacteria use to unwind their DNA in preparation for replication) are responsible for quinolone resistance in staphylococci, pseudomonads, and enteric bacteria. Similar mutational alterations in sequential target enzymes essential for bacterial nucleic acid synthesis, dihydrofolate reducíase (inhibited by trimethoprim) and dihydropteroate synthetase (inhibited by sulfonamides), result in resistance to trimethoprim and sulfamethoxazole. This alteration is plasmid mediated.
ALTERED METABOLIC PATHWAY
There are two examples of altered metabolic pathways acting as a bypass of antibiotic mechanisms. Enterococcus species requiring thymine become resistant to trimethoprim and sulfamethoxazole in vitro by developing mutations that bypass biochemical pathways to replace thymine. Enterococci may also use host folinic acid for growth in vivo in the presence of trimethoprim and sulfamethoxazole. The enterococci appear sensitive when tested in the laboratory, but are resistant when the patient is treated.
PRACTICAL SOLUTIONS TO EVOLVING ANTIBIOTIC RESISTANCE
The above review of the mechanisms and the spread of bacterial resistance should engender a healthy respect for the ability of these biologically creative organisms to adapt rapidly to new treatment approaches. Table 2 gives an overview of our potential responses. The most effective, without question, as exemplified by the history of ampicillin resistance in H. influenzai, is the development of an effective vaccine. Despite the high initial cost, this is practical for some common infections (it may soon be so for S. pneumoniae), but not for others that are less prevalent or have less clearly defined virulence factors. So far, the pharmaceutical industry has been successful in modifying existing classes of antibiotics or developing new ones to keep up (sometimes just barely) with emerging resistance. This strategy is expensive and temporary and may become less effective if there are biologic or chemical limits to our scientific creativity. The remaining solution - that of conserving the use of antibiotics for evidence-proven indications - can be effective and can lower costs, if we have the collective discipline to use antibacterial agents only when they are truly necessary.11'12 The Darwinian reality is that populations, especially bacterial ones, respond to the selective pressures of the excess use of antibiotics. It may well be that all four strategies will be necessary to deal with such a formidable opponent.
1. Neu HC. The crisis in antibiotic resistance. Science. 1992257: 1064-1 073.
2. Levy SB. Dawn of the post antibiotic era. Patient Care. 1995;29:84.
3. Levy SB. The challenge of antibiotic resistance. Sd Am. 1998;278:46-53.
4. Goosens H. Spread of vancomycin resistant enterococci: differences between the United States and Europe. Infect Control Hasp Epidemial. 1998;19:546-551.
5. Duchin JS, Breiman RF, Diamond A, et al. High prevalence of multidrug-resistant Streptococcus pneumoniae among children in a rural Kentucky community. Pediatr infect Dis }. 1995;14:745-750.
6. Hiramatsu K. Vancomycin resistance in staphylococci. Drug Resistance Updates. 1998;1:135-150.
7. Bennett PM. Transposable elements. In: Lederberg J, ed. Encyclopedia of Microbiology. New York: Academic Press; 1992:311-325.
8. Davies J. Inactivation of antibiotics and the dissemination of resistance genes. Science. 1994;264:375-381.
9. Nikaido H. Prevention of drug access to bacterial targets: permeability barriers and active efflux. Science. 1994;264:382-387.
10. Spratt B- Resistance to antibiotics mediated by target alterations. Science. 1994;264:388-392.
11. Wellbery C. Are we prescribing too many antibiotics? Am Fam Physician. 1997^5:1535.
12. Treem WR. Resisting the urge to prescribe. / Pediatr. 1996;128:729-730.
Evolution of Antibiotic Resistance In Haemophllus Influenza*?
Strategies to Deal With Emerging Antimicrobial Resistance