Fred C. Tenover
John E. McGowan Jr.
A Brief History of Resistance Mechanisms and the Discovery of Gene Transfer
The development of antimicrobial agents active against a wide array of microbial pathogens in the 1940s and 1950s enabled physicians to begin to turn the tide against a variety of infectious scourges. However, since the early 1970s, the development and spread of bacterial strains that are resistant to these drugs has emerged as a global problem [1,2,3,4,5]. The development of resistance to antimicrobial agents was not anticipated as a serious problem in the beginning of the antimicrobial era because it was assumed that the only mechanism of resistance was likely to be random mutations leading to altered target sites that would prevent the binding of the drug [6]. However, beginning with the discovery in 1940 of penicillinase (an enzyme that hydrolyzes penicillin, destroying its antibacterial properties), first in Escherichia coli [7] and then in Staphylococcus aureus [8], it became clear that other mechanisms of resistance were likely to be found. Subsequently, active efflux of drugs [9,10], modification of drug target sites [11], and other mechanisms to inactivate drugs chemically [12] have all been found to mediate resistance to various antimicrobial agents.
In the 1950s, the remarkable finding that multiply-resistant strains of Shigella were able to transfer the resistance phenotype to other bacterial strains during cell-to-cell mating experiments [13] dramatically changed our understanding of the molecular basis of antimicrobial resistance. Resistance was linked to the presence of extrachromosomal DNA calledplasmids. Subsequently, plasmids have been found in most clinically important bacterial pathogens [14,15].
In the 1960s, bacterial viruses were noted to move antimicrobial resistance genes between strains of staphylococci in a process called transduction [16]. In some instances, it appeared that entire plasmids could be moved from one strain of S. aureus to another, enhancing the ability of resistance genes to disseminate. In the 1970s, the recognition of transposable elements, often containing antimicrobial resistance genes, that could move from plasmids to bacteriophage [17] or from plasmids to chromosomal locations independent of the usual DNA recombination mechanisms [18] added another dimension to the ability of resistance genes to move among bacteria. This novel mechanism of mobilizing genes among bacterial cells indicated that the likelihood of widespread gene dissemination was high, particularly in environments where antibiotics were present in high concentrations to offer an advantage to those organisms that possessed a resistance mechanism.
The 1980s saw the discovery of another unique genetic element, the integron [19]. Integrons are mobile DNA elements that encode enzymes capable of inserting resistance gene cassettes into a stable DNA backbone that can be located either on a plasmid or the chromosome of an organism. The pool of gene cassettes in nature contains a variety of resistance determinants, ranging from aminoglycoside resistance genes to β-lactamase genes [19]. Integrons allow the accumulation and transmission of constellations of resistance determinants within and among diverse species of bacteria. As many as five different resistance genes may be accumulated together in a single integron. Integrons have played a major role in the development and spread of
P.224
multidrug resistance among isolates of Enterobacteriaceae in Europe [21] and the United States [22].
In the 1990s, another unique element, the chromosomal cassette, was described [23]. Similar to integrons, these large elements, common in staphylococci, are the repository of themecA element that mediates oxacillin resistance [24]. Five variants of the mecA cassette, called the staphylococcal cassette chromosome mec (SCCmec), have been described [25,26]. The most recently described cassettes, types IV and V, are widely disseminated among community-associated strains of methicillin-resistant Staphylococcus aureus (MRSA) [25,27].
Our understanding of the various mechanisms of antimicrobial resistance has grown tremendously as has our knowledge of the ways by which genes move among bacteria in a variety of healthcare environments. Both the mechanisms of antimicrobial resistance and the modes of resistance gene dissemination in bacteria are reviewed next.
Mechanisms of Resistance
Intrinsic Resistance
Some bacteria are intrinsically resistant to antimicrobial agents because they either lack the target site for that drug or the drug is unable to transit through the organism's cell wall or membrane to reach its site of action [28]. For example, most enterococci are resistant to low levels of aminoglycosides because the drug cannot penetrate the organism's peptidoglycan layer to reach the ribosomes. Only in the presence of cell wall–active agents, such as penicillin or vancomycin, can aminoglycosides reach their site of action [29]. Intrinsic resistance also includes chromosomally encoded enzymes, such as β-lactamases, that are characteristic of some bacterial species. The AmpC enzyme of Enterobacter cloacae, which when induced can mediate resistance to extended-spectrum cephalosporins and cephamycins (such as cefoxitin and cefotetan) is an example of such an enzyme [30].
Acquired Resistance
Bacteria that are by nature susceptible to an antimicrobial agent may become resistant by chromosomal mutation or by the acquisition of new genetic material [28]. For example, point mutations that occur in the chromosomal rpsL gene that encodes a ribosomal protein found in E. coli [31] and Mycobacterium tuberculosis [32] may result in an amino acid change that prevents the binding of streptomycin to the ribosome, resulting in resistance.
The acquisition of new resistance genes carried on plasmids, transposable elements, integrons, and other cassettes, usually results in the synthesis of new proteins in the cell. The known modes of plasmid- and transposon-encoded resistance mechanisms include enzymatically inactivating the antimicrobial agent, altering the target sites for the antimicrobial agent, blocking the transport of the agent into the bacterial cell, enhancing the efflux of the antimicrobial agent out of the cell, and bypassing the metabolic steps inhibited by the antimicrobial agent [28] (Table 15-1). Multiple resistance mechanisms may be encoded on a single plasmid, transposon, integron, or cassette. In addition, organisms may harbor >1 mechanism of resistance for a single class of drugs. Examples of resistance mechanisms for several classes of antimicrobial agents are described in the following sections.
Resistance to β-Lactam Drugs
β-lactam drugs include penicillins, cephalosporins, mono-bactams, carbapenems, and a variety of other related compounds [30]. Plasmid-mediated resistance to β-lactam agents most often is a result of β-lactamases, that is, enzymes that hydrolyze the β-lactam ring, deactivating the drug. The number of different β-lactamases described from gram-negative and gram-positive organisms exceeds 400 [33] and can be divided into several different classes based on chemical structure, substrate profile, isoelectric point, and amino acid sequence [34]. Point mutations in the structural genes of β-lactamases, such as in blaTEM, blaSHV, and blaCTX - M may result in amino acid alterations that extend the spectrum of drugs that can be inactivated [35]. Such enzymes are deemed extended-spectrum β-lactamases, or ESBLs. In addition, spontaneous mutations that result in changes in the porins of the organism's outer membrane can enhance resistance to ceftazidime [36], or loss of porins can result in resistance to carbapenems [37]. In the early 2000s, carbapenem resistance mediated by a Class A β-lactamase, designated KPC-1, emerged in the United States [38]. Klebsiella pneumoniae isolates harboring KPC-1 and the related β-lactamases KPC-2 and KPC-3 emerged and spread in several geographic areas, particularly in New York City [39,40]. Similar β-lactamases have been recognized elsewhere in the United States and abroad [41]. Carbapenem resistance also can be mediated by the Class B, metallo-beta-lactamases, which includes the enzymes SPM-1, VIM-1, IMP-1, and other related enzymes that are widely disseminated among Pseudomonas aeruginosa, Acinetobacter species, and other gram-negative organisms [42].
Resistance to β-lactam drugs among bacterial isolates also can be mediated by changes in the organism's penicillin binding proteins (PBPs) (i.e., those proteins involved in cell wall synthesis). Mutations in the genes that encode the PBPs, or the creation of mosaic PBP genes through the acquisition of DNA fragments from related organisms, result in PBPs with reduced affinity for β-lactam agents. These mechanisms of PBP remodeling, together with changes in several other non-PBP loci, are responsible for the dramatic increase in penicillin resistance seen in Streptococcus pneumoniae in the early 1990s [43]. Additional mutations in PBP genes result in higher levels of cefotaxime and ceftriaxone resistance [44]. The formation of mosaic
P.225
PBP genes through acquisition of DNA via transformation has been reported in both S. pneumoniae and Neisseria meningitidis [45].
|
TABLE 15-1 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Resistance to Erythromycin and Other Macrolides
Erythromycin resistance in bacteria may be mediated by several different mechanisms. These include efflux mechanisms, such as the msr(A) pump of staphylococci and the mef genes of S, pneumoniae and streptococci, and modification of the drug binding sites in rRNA [46]. Methylation of the 23S RNA of the 50S ribosome unit typically occurs at a specific adenine residue in multiple gram-positive organisms, leading to resistance not only to macrolides but also to lincosamides (e.g., clindamycin), and streptogramin B-type drugs. The erm genes (for erythromycin rRNA methylase) constitute a family of resistance determinants that have been isolated from many gram-positive species, including staphylococci, streptococci, actinomycetes, and a variety of anaerobes; gram-negative organisms, including Haemophilus spp., members of the Enterobacteriaeceae, and several genera of gram-negative anaerobes; and Mycobacterium spp. [46]. Erythromycin resistance also can be the result of esterification or phosphorylation of the drug.
Aminoglycoside Resistance
Aminoglycoside resistance is common in both gram-positive and gram-negative organisms and usually is the result of phosphorylation, acetylation, or adenylylation of the antimicrobial agent by plasmid- or transposon-encoded enzymes [12]. The nucleotide sequence diversity among the genes that encode the three subclasses of enzymes is higher than would be expected for genes that diverged in recent times, suggesting that although most of the genes share a common ancestral sequence, divergence occurred long ago, well before the clinical use of antimicrobial agents, making these genes a particularly old group. The enzymes are located at the inner membrane of bacterial cells, and the modified aminoglycoside blocks the transport of the antibiotic into the cell, keeping it from its site of action on the ribosome. Recently, a novel mutation of the aac6'Ib enzyme, designatedaac6'Ib-cr was reported [47]. This subtle change in amino acid sequence enable the enzyme to acetylate ciprofloxacin and several other fluoroquinolones leading to low-level resistance to these agents. The ability of one enzyme to modify two highly divergent classes of antimicrobial agents leading to resistance is very unusual.
P.226
Reduced permeability of the cell envelope, which can occur by spontaneous mutation of the genes that encode the cells' porins, changes in the ribosomal target site, or active efflux of drugs out of the cell also contributes to resistance, particularly in gram-negative organisms [9,10,11].
Tetracycline Resistance
Tetracycline resistance is widespread among gram-positive and gram-negative bacterial species and can be the result of drug efflux, ribosomal protection, or permeability changes [48,49]. Several tetracycline resistance determinants, such as the tet(M) gene, which mediates resistance to tetracycline, doxycycline, and minocycline, are widely distributed in species as diverse as Enterococcus faecalis, Neisseria gonorrhoeae, Mycoplasma pneumoniae, and B. fragilis [49]. This resistance gene and a number of other tetracycline resistance genes are commonly found on transposable elements, often linked together with determinants for resistance to chloramphenicol or macrolides in gram-positive organisms.
Glycopeptide Resistance
Vancomycin is the glycopeptide most commonly used to treat gram-positive healthcare-associated bacterial infections, such as MRSA, ampicillin-resistant enterococcal infections, andClostridium difficile infections. Resistance in enterococci is mediated by a series of plasmid- and transposon-encoded genes that produce altered DNA ligases that result in a cell wall structure that does not bind the drug [50]. The novel ligases, designated as vanA, vanB, vanD, and vanG, are acquired genes, while vanC and vanE are intrinsic to several species of enterococci. The latter loci typically mediate only low-level resistance among Enterococcus gallinarum and Enterococcus casseliflavus isolates and are chromosomally mediated traits.
In 2002, the transfer in nature of the vanA operon from a vancomycin-resistant E. faecalis donor to an MRSA recipient was reported from the State of Michigan [51]. The vancomycin mimimum inhibitory concentration (MIC) for the resulting vancomycin-resistant S. aureus (VRSA) isolate was 1024 µg/ml, which is 1000 times the normal vancomycin MIC for S. aureus(i.e., 1 µg/ml). Molecular studies by Weigel et al. indicated that the vanA operon entered the recipient MRSA strain on an enterococcal plasmid (likely by conjugation), and the plasmid was maintained in the MRSA cell just long enough for the vanA-containing transposon Tn1546 to transfer to a smaller staphylococcal plasmid already present in the MRSA strain [52]. The transfer of the vanA operon on Tn1546 to the MRSA plasmid resulted in the new VRSA isolate. Subsequently, six other VRSA isolates have been documented in the United States, and each contained the vanA resistance gene [53, 54, CDC unpublished data]. However, the latter VRSA isolates contained a variety of plasmids, and not all of the isolates demonstrated high vancomycin MICs. The lower vancomycin MIC of the Pennsylvania VRSA (32 µ/ml) strain was attributed to plasmid instability [55].
Low-level resistance to vancomycin also has been reported in S. aureus mediated by thickened cell walls that contain thousands of copies of the D-ala-D-ala dipeptide, which is the binding site for vancomycin, and changes in cellular metabolic pathways [56,57]. These organisms, designated as vancomycin-intermediate S. aureus (VISA) strains because the vancomycin MICs were in the intermediate range according to Clinical and Laboratory Standards documents [58], were first reported in 1997 [59] and have been reported subsequently from the United States and around the world [60,61,62]. In some isolates, the changes in cell wall structure and metabolic pathways led to decreased susceptibility to vancomycin, but the changes could not be detected by routine susceptibility testing methods [63]. Population analysis revealed subpopulations that are resistant to vancomycin, and these subpopulations likely lead to clinical failure. Such strains are referred to as “vancomycin-heteroresistant” VISA or hVISA for short [63].
Fluoroquinolone Resistance
Resistance to the fluoroquinolones frequently is mediated by point mutations in either the gyrA and gyrB genes and/or the analogous parC and parE loci, which decrease binding of the fluoroquinolones to the target enzymes, DNA gyrase, and topoisomerase IV [64]. Fluoroquinolone resistance is increasing in many species worldwide [65,66]. Plasmid-mediated resistance to the fluoroquinolones, which tends to be low-level, was first reported in 1998 [67]. The gene, designated qnrA, is one of a family of small proteins that mediates fluoroquinolone resistance in gram-negative bacteria. Other qnr genes, including qnrB, have been detected in bacteria in association with aac6'Ib-cr aminoglycoside/ciprofloxacin resistance determinant [47,68].
Trimethoprim and Sulfonamide Resistance
Bacteria can recruit and use novel enzymes to bypass inhibited metabolic pathways to avoid killing by several antimicrobial agents. This is the mechanism of resistance used by bacteria to circumvent the action of trimethoprim [69] and sulfonamides [70]. This mechanism is based on plasmid- or transposon-encoded enzymes that substitute for the chromosomal enzymes normally inhibited by these drugs. Sulfonamides work through competitive inhibition of the enzyme dihydropteroate synthetase. The plasmid-coded enzyme bypassing this inhibition is smaller and more heat sensitive and a thousand times as much sulfonamide is required to inhibit it as is needed to inhibit the chromosomal enzyme. Similarly, some bacterial strains resistant to high levels of trimethoprim, an agent that inhibits dihydrofolate reductase, contain a plasmid-mediated gene coding for a new, trimethoprim-resistant dihydrofolate reductase. In each of these instances, the plasmids provide a mechanism whereby products vital to the bacterial cell
P.227
can be synthesized and the inhibiting effect of the drug bypassed.
Resistance to Multiple Antimicrobial Agents
Resistance to multiple classes of antimicrobial agents, such as chloramphenicol, tetracycline, and fluoroquinolones, in both gram-negative and gram-positive bacterial organisms often is due to efflux of the drugs out of the bacterial cell [9,10]. There are several genetic determinants that mediate such resistance, including marA, mexAB-oprD, and norA [71]. Many multiresistant organisms previously were thought to have permeability barriers that limited the access of drugs to their sites of action. These are now known to contain effective efflux pumps that direct the drug out of the cell before it reaches its internal site of action [9,10]. Of course, solitary resistance to chloramphenicol, tetracycline, or the fluoroquinolones also is seen, although this is often mediated by mechanisms other than efflux.
Antimicrobial Resistance in M. tuberculosis
Resistance to anti-mycobacterial agents, such as isoniaizid, rifampin, streptomycin, and ethambutol, often is associated with point mutations in critical genes, including katG, rpoB, rpsL, and embB [72,73,74]. However, a sizeable percentage of resistant strains do not contain mutations in these loci, indicating that there is still much to learn about resistance mechanisms, particularly in Mycobacterium tuberculosis. Novel DNA sequencing strategies, such as pyrosequencing, allow detection of mutations associated with resistance in organisms directly from sputum samples or from positive mycobacterial broth culture vials [73,74].
Dissemination of Resistance Genes in Nature
Bacteria can exchange genetic information by means of transformation, transduction, and conjugation [75,76]. Resistance determinants can be encoded on an organism's chromosome or on the plasmids, bacteriophages, transposons, or integrons harbored in the cell, all of which may be moved from cell to cell by transformation, transduction, or conjugation. Thus, there is a multitude of pathways by which resistance genes move from one organism to another. Each of the various genetic elements that harbor resistance determinants has unique features.
Plasmids
Plasmids are self-replicating, extrachromosomal segments of DNA that can be found in many species of bacteria and yeast. Plasmids usually are circular and range in size from 2 to 400 kilobases [15]. The number of proteins encoded on plasmids can be substantial, with larger plasmids (~300 kilobases) encoding 50 to 75 proteins. The proteins may include enzymes involved in antimicrobial resistance, virulence, or the molecular machinery for plasmid transfer, in addition to those proteins required for plasmid maintenance. Plasmids usually can be acquired or lost by bacteria without affecting basic cellular functions because most of the genetic information necessary for metabolism and growth of the bacterial cell is located on the chromosome. For some plasmids, no phenotypic properties are known, and they are deemed “cryptic.”
Plasmids are categorized by their ability to transfer themselves to other organisms. Those that are self-transmissible from one bacterial cell to another are termed “conjugative” [15]. When a gram-negative bacterial cell contains a conjugative plasmid, a proteinaceous appendage called a “pilus” is synthesized on the outside of the cell; this, together with other plasmid-mediated proteins, enables cell-to-cell transmission of the plasmid DNA. Conjugative plasmids of gram-positive bacteria do not use pili for conjugation; rather, direct cell-to-cell contact is required. In some instances, this is facilitated by plasmid-coded proteins (pheromones) that enhance clumping of donor and recipient cells [77]. Both staphylococci and enterococci have been shown to contain conjugal plasmids.
Nonconjugative plasmids often are smaller in size than conjugative plasmids. Some nonconjugative plasmids can still be transferred to recipient organisms by a process called mobilization in which a co-resident, nonconjugative plasmid takes advantage of the transfer of a conjugative one present in the same cell. In addition, nonconjugative and conjugative plasmids also may be transmitted by bacterial virus vectors through transduction. In this process, plasmid DNA instead of phage DNA is packaged in the viral protein coat and, on infection of a suitable recipient cell, the plasmid DNA is released and begins replication in the new host. Because transduction requires that the plasmid DNA be packaged in the protein coat of a bacteriophage, the amount of plasmid DNA that can be transduced is limited to approximately the size of the phage genome. This process operates much more efficiently for the smaller nonconjugative plasmids and appears to be an important mechanism of plasmid exchange in S. aureus. In gram-negative bacilli and streptococci, conjugation and mobilization seem to be the most common means of transfer. Direct uptake of both conjugative and nonconjugative plasmid DNA by a recipient cell also can occur through transformation. However, even in those naturally transformable species, this mode of plasmid transfer probably is uncommon.
Transposons and Insertion Sequences
A transposon is a segment of DNA that can move as an intact unit from one replicating DNA unit (replicon) to another by mechanisms other than those used in generalized recombination [75]. These sequences can “jump” from plasmid to plasmid, plasmid to bacteriophage, plasmid to chromosome, or the reverse using a “transposase” enzyme that is
P.228
encoded in the element. Transposons encoding–resistance genes to a variety of antimicrobial agents have been described. In theory, any plasmid could gain a transposon if DNA containing the transposon were to coexist in a cell long enough for transposition to occur. Transposons also promote a variety of rearrangements of DNA in adjoining regions, including deletions, inversions, and duplications. Some transposons, particularly those in gram-positive organism, such as enterococci, are conjugative and promote their own transfer from the chromosome of the donor cell to that of the recipient [75,76]. Such elements also have been recognized in S. pneumoniae [78]. Simpler genetic structures, called insertion sequences, do not contain antimicrobial resistance determinants but promote gene rearrangements and can modify the expression of resistance genes by inserting strong promoter elements upstream of open reading frames.
|
Figure 15-1 A schematic diagram showing the stepwise acquisition of antimicrobial resistance genes by Escherichia coli, beginning with (1) a chromosomal mutation in gyrA that results in ciprofloxacin resistance (star), followed by (2) acquisition of a plasmid containing a blaTEM - 1 ampicillin resistance gene, followed by (3) insertion of a transposon containing a aac[3]-I gentamicin resistance genes into the plasmid, followed by (4) insertion of an integron with sulfa and trimethoprim resistance (sulI and dhfrI genes, respectively) into the chromosome, and finally (5) insertion of a blaCTXM - 2 cefotaxime resistance gene on a cassette into the integron backbone. |
Integrons
Integrons are mobile DNA elements that contain “hot spots” of recombination where a variety of antimicrobial resistance gene cassettes can insert, mediated by an integrase enzyme [19]. The gene cassettes must contain a key 59-base-pair region at their termini to be recognized by the integrase enzyme. The gene cassettes do not contain their own promoters, and thus expression depends on insertion into the integron in the appropriate orientation downstream from a promoter element [19]. The cassettes appear to undergo integration and excision as covalently closed circular molecules. Large, complex integrons with as many as five resistance genes have been detected on plasmids in enteric organisms [20,21]. The proximity of the cassette to the integron's promoter has a major effect on expression, with those genes that are proximal to the promoter showing strong and consistent expression and those that are distal showing low-level or no expression.
Development of Multiresistant Organisms
Multiresistant organism can develop over time by the successive acquisition of mutations and resistance genes using a number of different mechanisms (Figure 15-1). These may include spontaneous mutation, acquisition of plasmids and transposons, and insertion of gene cassettes into integrons. Dissemination of epidemic plasmids in closed populations (sometimes referred to as “plasmid outbreaks”) continues to plague hospitals as demonstrated by Neuwirth et al., who noted dissemination of an ESBL-producing strain ofEnterobacter aerogenes in their hospital followed by transfer of the β-lactamase-encoding plasmid to other species of Enterobacteriaceae [79]. Rubens et al. demonstrate how a gentamicin-resistance determinant originally present on a small nonconjugative plasmid was transposed onto a larger, self-transferable plasmid [18]. Once on the conjugative element, the resistance was efficiently transferred among a variety of genera in the hospital. This process provided a mechanism to assemble new resistance determinants on a preexisting plasmid and appears to be one way that multiresistance plasmids can develop.
Environmental Conditions that Favor the Development of Resistant Organisms—Selective Pressure
Resistant organisms, particularly those that develop through spontaneous mutation, usually will not survive unless there is an advantage to maintaining the resistance phenotype. The environment in many healthcare
P.229
institutions (hospitals, long-term care facilities, dialysis centers, etc.) is rich with antimicrobial agents that are hostile to antimicrobial-susceptible organisms [80]. Thus, resistant organisms have a survival advantage in this environment. In genetic terms, the presence of antimicrobial agents in the environment is the “selective pressure” that favors the development and spread of resistant strains of bacteria. However, selective pressures are not unique to healthcare institutions. To control the spread of resistant organisms, the selective pressures provided by both hospital and community must be considered [80].
The Healthcare Environment
The most frequently encountered healthcare-associated pathogens, staphylococci, enterococci, and gram-negative bacilli [81], all have demonstrated the ability to acquire and disseminate resistance genes. Intensive use of antimicrobial agents exerts a selective pressure favoring those organisms that have acquired resistance determinants through mutation or genetic exchange. Even organisms that develop low-level resistance may survive long enough to acquire additional mutations or resistance genes that facilitate long-term survival [82]. Antimicrobial therapy alters or eliminates the normal bacterial flora found in many niches of the human body, which can enhance colonization of skin or other sites with resistant strains because the innate defense provided by the presence of normal flora is gone. This, in turn, likely contributes to enhanced gene transfer among organisms, particularly in the gastrointestinal tract. Plasmid transfer, in particular, often is observed in aqueous reservoirs peculiar to the hospital environment, such as urinary catheter collection bags. Conditions that enhance transfer of plasmids also favor the transmission of transposons and integrons, further increasing the dissemination of resistance genes [76]. Immunocompromised patients who are receiving antimicrobial agents and who are hospitalized for prolonged periods provide a reservoir for resistant organisms [80]. Antimicrobial agents typically find their way into the general inanimate environment of healthcare facilities, which also contributes to the selective advantage obtained by plasmid- or integron-containing organisms. Thus, the healthcare environment is conducive to the development and spread of resistant bacteria.
The Community Environment
The use and sometimes misuse [83] of antimicrobial agents in the community contributes to the selective pressure that results in healthcare-associated pathogens because the gene pools for the organisms are the same [84]. Selective pressures outside of the hospital, including the use of antimicrobial agents in humans, particularly in children for respiratory infections [85], in animals both therapeutically and as growth promoters, in fish (aqua-culture), and on vegetables [1,2,3,4], suggests that the development of resistant strains can and does occur virtually everywhere. The development and spread of plasmid-mediated ampicillin-resistant Haemophilus influenzae [86] in the 1970s, multiresistant pneumococci in the 1970s and 1980s [87], and the descriptions of multiply resistant Salmonella [88] and Shigella [89] provided convincing evidence that problems with resistant organisms are not confined to the hospital. The two locations often come together, as when patients from nursing homes who harbor MRSA are transferred to acute care community hospitals where the patients serve as a nidus for the spread of the pathogen [90].
Epidemiologic Studies of Bacterial Resistance
To study the epidemiology of antimicrobial-resistant bacteria, it is imperative to have strain typing methods that are both discriminatory and reproducible [91]. When an outbreak of a resistant strain is suspected in a healthcare system, several techniques can be used to confirm that the isolates are clonal (i.e., derived from a common parent). Biochemical patterns and antimicrobial susceptibility test results can provide initial clues to strain identity, but more definitive information, as provided by pulsed-field gel electrophoresis (PFGE) patterns, repetitive element polymerase chain reaction (rep-PCR) typing, multiple locus variable number tandem repeat assays (MLVA), or multilocus sequence typing (MLST) often are necessary [92].
Pulsed-Field Gel Electrophoresis
Of all of the strain typing methods currently available, PFGE of chromosomal DNA comes the closest to being a universal typing method for bacteria [93]. PFGE uses infrequent cutting restriction endonucleases to cleave the chromosome of an organism into a relatively small number (usually 10 to 25) of fragments (Figure 15-2). The fragments are subjected to electrophoresis through agarose in a special chamber in which the direction of the electrical current is switched frequently according to a preset pattern. This resolves the fragments into discrete patterns that easily can be photographed and analyzed [93]. Interpretive criteria for analyzing restriction fragment banding patterns have been published [93]. This typing method has been validated against epidemiologic data collected during outbreak investigations for a large number of bacterial species.
REP-PCR
REP-PCR uses the repetitive elements present in most bacterial and fungal species to generate banding patterns that can be analyzed in a fashion similar to the banding patterns of PFGE [94,95,96]. Commercial systems that use
P.230
REP-PCR for both strain typing and species identification have been developed [97]. This system yields results that are comparable or slightly less discriminatory than PFGE but with the advantage of much more rapid turnaround time and sophisticated on-line data analysis and reporting [95,96].
|
Figure 15-2 Example of pulsed-field gel electrophoresis showing differentiation of eight of the major strain types of methicillin-resistant Staphylococcus aureus [104]. Molecular size standards are shown in lanes 2 and 9. |
MLVA
Analysis of the genomes of a variety of bacterial species has shown that many organisms have chromosomal loci that contain variable numbers of tandem repeats (VNTRs) [98]. The structure and function of the repeats differ widely, but the value of VNTRs for strain typing, particularly for monomorphic species such as Bacillus anthracis, is significant. VNTR may represent direct or inverted repeats and may be responsible for antigenic variation or gene regulation. PCR amplification products resulting from the use of specific primers can be analyzed directly for variation in primary nucleotide sequence. Alternatively, variation in the sizes of the PCR products can be assessed by agarose gel or capillary electrophoresis coupled with computer software to facilitate fragment analysis [98]. Multiplex PCR assays can be developed that allow products from multiple loci to be assessed simultaneously [99]. However, when fragment size rather than the actual sequence is used for strain differentiation, the datasets have the limitations of laboratory to laboratory reproducibility inherent in any gel-based technique.
MLVA has been used successfully to type highly monomorphic organisms, such as B. anthracis and Yersinia pestis, and to differentiate among lineages of common pathogens, such as Shiga-toxin-producing E. coli, Salmonella typhi, and S. aureus [98]. Initially, the MLVA assay for B. anthracis was focused on a single variable repeat region, vrrA, which consisted of a series of 12 base pair tandem repeats. The addition of other chromosomal loci and sites on the two virulence plasmids (pXO1 and pXO2) has provided remarkable discrimination among isolates of this highly monomorphic pathogen [100].
MLST
MLST is a DNA sequence-based typing scheme that is primarily used for studying the population biology and structure of bacterial species [92,101]. The principles of MLST are based on an older technique that examined variations in the electrophoresis patterns of multiple enzymes in starch gels, namely multilocus enzyme electrophoresis (MLEE) [102]. In MLEE, various electromorphs were equated with alleles. Similarly, in MLST, isolates are classified based on polymorphisms in the DNA coding sequences of 7 to10 different metabolic (“housekeeping”) genes, where every nucleotide variant in a sequence is considered a unique allele. The set of loci that is sequenced must be identified and validated for each bacterial species [92]. When used for strain typing of isolates associated with outbreaks, the discriminatory power of MLST is typically less than PFGE. However, many pathogenic clones of bacteria are designated using a combination of strain typing methods that often include MLST types, such as pneumococci [103], MRSA [104], and N. meningitidis [105].
MLST databases have been established for a variety of organisms and a Web site (www.mlst.net) is available to help analyze sequence data generated for a number of bacterial species, ranging from N. meningitidis to S. aureus. MLST may be useful in some epidemiologic studies, particularly if discriminatory power is enhanced by examining additional loci.
Additional Methods for Laboratory Studies of Resistant Microorganisms
The dissemination of resistance genes occurs on three levels in hospitals. The first level is dissemination of a resistant strain, often on the hands of healthcare workers or through patient contact with environmental reservoirs of resistant organisms. The next level occurs when a resistance plasmid transfers from one organism to another or from one species to another, which amplifies the extent of the problem and makes epidemiologic studies more complex and the outbreak more difficult to control. The third level involves the acquisition of novel resistance genes by the outbreak strain through generalized recombination, transposition, or insertion of resistance gene cassettes into integrons. Examples of this include the acquisition of chloramphenicol resistance by strains of MRSA during an outbreak in a hospital in Seattle [106] or the acquisition
P.231
of gentamicin-resistance transposons by an epidemic strain of P. aeruginosa [107]. Furthermore, spontaneous mutation to a resistant phenotype followed by dissemination of the novel strain also occurs as evidenced by the explosive increase in ciprofloxacin-resistant strains of MRSA reported by Blumberg et al. [108], in which the new resistant strains represented >90% of MRSA isolates over the course of a single year after the introduction of ciprofloxacin in the hospital. Thus, spread of a resistant strain, transfer of plasmids, spontaneous mutation, and genetic exchange in the presence of antibiotic selective pressure are the key factors in the development and spread of resistant organisms in the hospital setting.
Conclusions
Antimicrobial resistance in bacteria comprises a wide variety of biochemical mechanisms and processes that allow microorganisms to grow in the presence of antimicrobial agents. Bacteria are able to transmit this acquired “knowledge” to other species through transformation, transduction, and conjugation. Specialized structures, including transposons, integrons, and chromosomal cassettes, allow cells to recruit additional genetic information outside of the normal processes of mutation and generalized recombination. The web of genetic exchange in bacteria is very broad and even crosses between gram-positive and gram-negative organisms. Finally, there is no barrier between hospital and the community when it comes to resistant organisms. All bacteria can draw on the same gene pool to find ways to cope with the presence of antibiotics in their environment.
References
P.232
P.233