**4. Antimicrobial resistance mechanisms**

multidrug-resistant organisms), with more still in development. Other β-lactam drugs developed during this time frame were the carbapenems (broad-spectrum activity) and monopen-

Differences in structure, metabolism, virulence factors, etc., between gram-negative and gram-positive bacteria predict which antimicrobial drug groups may be effective. Fewer of the drug groups have good activity against gram-negative bacteria. Those groups include some of the β-lactam drugs (especially second-, third-, and fourth-generation cephalosporins): aminoglycosides, fluoroquinolones, trimethoprim/sulfamethoxazole (TMP/SXT), and nitrofurans (for UTIs) [33]. Intestinal infections with *E. coli* are most commonly self-limiting and require supportive therapy (antiemetics, antidiarrhetics, rehydration) only. Severe or recurring infections (e.g., traveler's diarrhea) may be treated with fluoroquinolone drugs. Acute dysentery caused by EIEC strains may be treated with fluoroquinolones or appropriate cephalosporins. For infections caused by EHEC, antimicrobial therapy is contraindicated as it greatly increases the risk for development of HUS. UTIs caused by UPEC strains are usually treated with antimicrobial drugs and uncomplicated UTIs with nitrofurantoin or TMP/SXT; complicated UTIs may also be treated with fluoroquinolones [34–37]. Infections with ESBL or CRE strains severely limit treatment options. For ESBL strains, carbapenems may still be an option or newer β-lactam/β-lactamase inhibitor drug combinations; for CRE strains, gentamicin, amikacin, colistin, tigecycline, and fosfomycin may be options. Unfortunately, some ESBL

At the beginning of antimicrobial drug resistance, physicians did not realize how the various drugs affected the bacteria. In addition, in an effort to begin antimicrobial therapy as quickly as possible, physicians often ordered a broad-spectrum drug before knowing the causative agent of the infection. These issues (among others) have led to a large amount of resistance to β-lactam drugs, especially among the gram-negative bacteria. Bacteria that produced β-lactamases (enzymes that inactivate β-lactam drugs) were identified as early as the 1940s (around the same time as penicillin was discovered), and the number of different β-lactamases

The issue of resistance is not just with the β-lactam drugs. Over the 70 plus years that antimicrobial drugs have been in existence, resistance mechanisms have been seen for most of these drugs. It does not seem to take the bacteria very long from initial use of a drug to development of resistance to that drug. Important resistant milestones include resistance to aminoglycosides and tetracycline in the 1960s, vancomycin in the 1980s, fluoroquinolones in the 1990s, and linezolid in the 2000s [42]. In addition, some of the bacteria have become resistant to multiple antimicrobial agents from many of the drug classes. These multidrug-resistant bacteria, such as methicillin-resistant *Staphylococcus aureus* (MRSA), are currently a major cause

Similar to the threat of MRSA, members of the gram-negative Enterobacteriaceae, in particular *E. coli* and *Klebsiella pneumoniae*, include strains that are ESBLs and CREs. These organisms can be resistant to most commonly used antimicrobials, which makes the infections they cause extremely difficult to treat, leading to increased morbidity and mortality (and healthcare costs)

ems (aztreonam—activity against gram-negative aerobic bacteria) [30–32].

86 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

and CRE strains may be resistant to even some of these drugs [38–40].

produced has increased over the years to around 1000 [2–4, 41].

of morbidity and mortality [43].

[44, 45].

There are four general antimicrobial resistance mechanisms that bacteria use. These are limiting uptake of the drug, modifying the target of the drug, inactivating the drug, and active efflux of the drug. These mechanisms may be located on the bacterial chromosome and occur naturally in all members of a species (intrinsic) or come from other bacteria, usually via a plasmid (acquired). Intrinsic resistance genes may be expressed constitutively (usually at a low level) or be induced by the presence of antimicrobial drugs. Gram-negative bacteria widely use all four of these mechanisms and are very capable of horizontal transfer of resistance elements. **Table 2** shows which resistance mechanisms and genes are associated with resistance to various antimicrobial drugs [43, 46, 47].


**Table 2.** Common antimicrobial resistance genes and mechanisms in *Escherichia coli*.

### **4.1. Limiting drug uptake**

Gram-negative bacteria have an advantage in combating drugs because of the structure and functions of the LPS cell wall, which provides a natural barrier to certain molecules. The LPS is generally hydrophobic which limits access to small hydrophilic drugs, such as the β-lactams. These hydrophilic drugs gain access by traveling through the OMPs. The main OMPs in *E. coli* are OmpF and OmpC. In addition to the β-lactams, other drugs that may use the porin channels are chloramphenicol, fluoroquinolones, and tetracycline. Hydrophobic drugs such as the aminoglycosides and the macrolides gain access by permeating through the LPS layer. There are two main mechanisms that are used to limit access to drugs via porins: a decrease in the number of porins or a change in charge, within the porin channel, which reduces its function or binding properties. In *E. coli* porin production may be reduced dramatically or even stopped, or a different porin may be produced instead [48, 49].

**4.3. Inactivating the drug**

genes) [43, 49, 61].

*4.3.1. β-lactamases*

[46, 62].

Drug inactivation is accomplished in one of two ways: by actual degradation of the drug or by transfer of a chemical group to the drug. Gram-negative bacteria use drug inactivation against β-lactams and aminoglycosides. The β-lactam drugs are universally inactivated by β-lactamase enzymes, which degrade the drugs, and *E. coli* produces several of these. The aminoglycoside drugs are inactivated fairly universally by enzymes that transfer one of three small chemical groups to the drug. These enzymes include the acetyltransferases (AACs, *aac* genes), nucleotidyltransferases (ANTs, *ant* genes), and the phosphotransferases (APHs, *aph*

Antimicrobial Mechanisms of *Escherichia coli*

http://dx.doi.org/10.5772/67363

89

The β-lactam drugs all share a specific core structure, which consists of a four-sided β-lactam ring. The β-lactamases (also originally called penicillinases and cephalosporinases) are capable of inactivating β-lactam drugs via hydrolyzation of a specific site in the β-lactam ring structure causing the ring to open. The drugs are then not able to bind to their target proteins, the PBPs. Within the large number of β-lactamases which have been identified, there are enzymes which can inactivate any of the current β-lactam drugs. The production of β-lactamases is the most common resistance mechanism used by gram-negative bacteria against β-lactam drugs

The β-lactamase enzymes can be classified based on their primary structure or functional characteristics. Structurally they are placed into four main categories (A, B, C, or D). There are three functional groupings: the cephalosporinases, the serine β-lactamases, and the metallo-βlactamases. These enzymes are also commonly referred to by their enzyme family, for example, the TEM (named after the first patient) family, the sulphydryl variable (SHV) family, and

The first β-lactamase to be characterized was from *E. coli* and is chromosomally encoded by the *ampC* gene (so named for ampicillin resistance). This gene is constitutively expressed at a low level, but mutations may result in overexpression of the gene. The AmpC β-lactamases are most effective against the penicillins and some first-generation cephalosporins. There are also many plasmid-borne β-lactamases, which carry a variety of *bla* genes (β-lactamase genes). Because these β-lactamases confer resistance to later generation cephalosporins, they were designated as ESBLs and include the TEM, SHV, and CTX-M enzyme families. The most commonly seen of these in *E. coli* are the CTX-Ms. The ESBLs may also be resistant to multiple drug classes but are generally sensitive to β-lactamase inhibitors. The β-lactamase inhibitors are structurally similar to β-lactamases and have weak antimicrobial ability alone but work

Recently, there has been emergence of β-lactamases that are active against the carbapenems (carbapenemases), found primarily in the *Enterobacteriaceae*. Bacterial strains that carry these are known as CRE strains. The carbapenemases are all metallo-β-lactamases (MBLs), and the most widely distributed are the IMP-1 (for imipenem resistance) and VIM-1 (Verona integronencoded MBL) types. A new MBL has recently been identified, mainly in strains of *E. coli*. It has been designated as New Delhi MBL (NDM-1). The CRE strains are usually resistance to

the CTX (preferentially hydrolyze cefotaxime) family [56, 63].

synergistically in combination with a β-lactam drug [56, 64–67].

### **4.2. Modifying drug target**

Gram-negative bacteria make use of the modifying of drug targets against several of the antimicrobial groups including β-lactams, aminoglycosides, fluoroquinolones, and the combination drug TMP/SXT. Even though not as widely used as in gram-positive bacteria, the gram-negative bacteria are able to produce penicillin-binding proteins (PBPs) that are resistant to some β-lactam drugs. PBPs are actually peptidases that are involved in the making of the peptidoglycan cell wall. Penicillin drugs that are able to bind to PBPs inhibit the assembly process. There are several different native PBPs produced by *E. coli*, some of which have reduced binding affinity for some of the β-lactam drugs. No acquired modified form of PBP has been shown to be significant in β-lactam resistance in *E. coli* [50, 51].

The aminoglycoside drugs inhibit protein synthesis by binding to the bacterial 30S ribosomal subunit at the A-site of the 16S rRNA. Bacteria are able to modify the ribosomal subunit via acquisition of plasmids carrying 16S rRNA methyltransferases. The methyltransferases are able to modify the structure of the 16S rRNA, which decreases the ability of the drug to bind to it. Several of these methyltransferases have been identified and characterized. The genes involved include *armA* (for aminoglycoside resistance methylase) and several *rmt* (for ribosomal methyltransferase) genes, with *rmtB* being the most common. The bacteria quite often possess several of these genes simultaneously. These genes most often confer clinically significant resistance to amikacin, gentamicin, and tobramycin, among other aminoglycosides [52–56].

The fluoroquinolone drugs interfere with nucleic acid synthesis during DNA replication by inhibiting either DNA gyrase or topoisomerase IV. Resistance to these drugs occurs commonly from mutations in either the chromosomally encoded GyrA subunit of gyrase (*gyrA* gene) or the ParC subunit of topoisomerase IV (*parC* gene). These mutations decrease the binding ability of the drugs, most commonly ciprofloxacin and norfloxacin. There is also some evidence that lowlevel resistance may be acquired via plasmids carrying quinolone resistance (*qnr*) genes [56–58].

The combination drug TMP/SXT is currently a common choice for treatment of UTIs. Both of these drugs target enzymes in the bacterial folate biosynthesis pathway via competitive inhibition. Trimethoprim is an analog of the natural substrate of the dihydrofolate reductase (DHFR) enzyme, and SXT is an analog of *p*-amino-benzoic acid, the natural substrate of the dihydropteroate synthase (DHPS) enzyme. This competitive binding blocks the binding of the natural substrate and stops the pathway at that point. Since TMP and SXT affect two different enzymes on the same pathway, the combination drug makes an effective treatment. Chromosomal mutations (often single point mutations) in the *dhfr* or *dhps* genes are commonly the cause of resistance to these drugs [59, 60].

### **4.3. Inactivating the drug**

OMPs in *E. coli* are OmpF and OmpC. In addition to the β-lactams, other drugs that may use the porin channels are chloramphenicol, fluoroquinolones, and tetracycline. Hydrophobic drugs such as the aminoglycosides and the macrolides gain access by permeating through the LPS layer. There are two main mechanisms that are used to limit access to drugs via porins: a decrease in the number of porins or a change in charge, within the porin channel, which reduces its function or binding properties. In *E. coli* porin production may be reduced dra-

Gram-negative bacteria make use of the modifying of drug targets against several of the antimicrobial groups including β-lactams, aminoglycosides, fluoroquinolones, and the combination drug TMP/SXT. Even though not as widely used as in gram-positive bacteria, the gram-negative bacteria are able to produce penicillin-binding proteins (PBPs) that are resistant to some β-lactam drugs. PBPs are actually peptidases that are involved in the making of the peptidoglycan cell wall. Penicillin drugs that are able to bind to PBPs inhibit the assembly process. There are several different native PBPs produced by *E. coli*, some of which have reduced binding affinity for some of the β-lactam drugs. No acquired modified form of PBP

The aminoglycoside drugs inhibit protein synthesis by binding to the bacterial 30S ribosomal subunit at the A-site of the 16S rRNA. Bacteria are able to modify the ribosomal subunit via acquisition of plasmids carrying 16S rRNA methyltransferases. The methyltransferases are able to modify the structure of the 16S rRNA, which decreases the ability of the drug to bind to it. Several of these methyltransferases have been identified and characterized. The genes involved include *armA* (for aminoglycoside resistance methylase) and several *rmt* (for ribosomal methyltransferase) genes, with *rmtB* being the most common. The bacteria quite often possess several of these genes simultaneously. These genes most often confer clinically significant resistance to amikacin, gentamicin, and tobramycin, among other aminoglycosides

The fluoroquinolone drugs interfere with nucleic acid synthesis during DNA replication by inhibiting either DNA gyrase or topoisomerase IV. Resistance to these drugs occurs commonly from mutations in either the chromosomally encoded GyrA subunit of gyrase (*gyrA* gene) or the ParC subunit of topoisomerase IV (*parC* gene). These mutations decrease the binding ability of the drugs, most commonly ciprofloxacin and norfloxacin. There is also some evidence that lowlevel resistance may be acquired via plasmids carrying quinolone resistance (*qnr*) genes [56–58]. The combination drug TMP/SXT is currently a common choice for treatment of UTIs. Both of these drugs target enzymes in the bacterial folate biosynthesis pathway via competitive inhibition. Trimethoprim is an analog of the natural substrate of the dihydrofolate reductase (DHFR) enzyme, and SXT is an analog of *p*-amino-benzoic acid, the natural substrate of the dihydropteroate synthase (DHPS) enzyme. This competitive binding blocks the binding of the natural substrate and stops the pathway at that point. Since TMP and SXT affect two different enzymes on the same pathway, the combination drug makes an effective treatment. Chromosomal mutations (often single point mutations) in the *dhfr* or *dhps* genes are com-

matically or even stopped, or a different porin may be produced instead [48, 49].

88 *Escherichia coli* Escherichia coli - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications - Recent Advances on Physiology, Pathogenesis and Biotechnological Applications

has been shown to be significant in β-lactam resistance in *E. coli* [50, 51].

monly the cause of resistance to these drugs [59, 60].

**4.2. Modifying drug target**

[52–56].

Drug inactivation is accomplished in one of two ways: by actual degradation of the drug or by transfer of a chemical group to the drug. Gram-negative bacteria use drug inactivation against β-lactams and aminoglycosides. The β-lactam drugs are universally inactivated by β-lactamase enzymes, which degrade the drugs, and *E. coli* produces several of these. The aminoglycoside drugs are inactivated fairly universally by enzymes that transfer one of three small chemical groups to the drug. These enzymes include the acetyltransferases (AACs, *aac* genes), nucleotidyltransferases (ANTs, *ant* genes), and the phosphotransferases (APHs, *aph* genes) [43, 49, 61].

## *4.3.1. β-lactamases*

The β-lactam drugs all share a specific core structure, which consists of a four-sided β-lactam ring. The β-lactamases (also originally called penicillinases and cephalosporinases) are capable of inactivating β-lactam drugs via hydrolyzation of a specific site in the β-lactam ring structure causing the ring to open. The drugs are then not able to bind to their target proteins, the PBPs. Within the large number of β-lactamases which have been identified, there are enzymes which can inactivate any of the current β-lactam drugs. The production of β-lactamases is the most common resistance mechanism used by gram-negative bacteria against β-lactam drugs [46, 62].

The β-lactamase enzymes can be classified based on their primary structure or functional characteristics. Structurally they are placed into four main categories (A, B, C, or D). There are three functional groupings: the cephalosporinases, the serine β-lactamases, and the metallo-βlactamases. These enzymes are also commonly referred to by their enzyme family, for example, the TEM (named after the first patient) family, the sulphydryl variable (SHV) family, and the CTX (preferentially hydrolyze cefotaxime) family [56, 63].

The first β-lactamase to be characterized was from *E. coli* and is chromosomally encoded by the *ampC* gene (so named for ampicillin resistance). This gene is constitutively expressed at a low level, but mutations may result in overexpression of the gene. The AmpC β-lactamases are most effective against the penicillins and some first-generation cephalosporins. There are also many plasmid-borne β-lactamases, which carry a variety of *bla* genes (β-lactamase genes). Because these β-lactamases confer resistance to later generation cephalosporins, they were designated as ESBLs and include the TEM, SHV, and CTX-M enzyme families. The most commonly seen of these in *E. coli* are the CTX-Ms. The ESBLs may also be resistant to multiple drug classes but are generally sensitive to β-lactamase inhibitors. The β-lactamase inhibitors are structurally similar to β-lactamases and have weak antimicrobial ability alone but work synergistically in combination with a β-lactam drug [56, 64–67].

Recently, there has been emergence of β-lactamases that are active against the carbapenems (carbapenemases), found primarily in the *Enterobacteriaceae*. Bacterial strains that carry these are known as CRE strains. The carbapenemases are all metallo-β-lactamases (MBLs), and the most widely distributed are the IMP-1 (for imipenem resistance) and VIM-1 (Verona integronencoded MBL) types. A new MBL has recently been identified, mainly in strains of *E. coli*. It has been designated as New Delhi MBL (NDM-1). The CRE strains are usually resistance to all the β-lactam drugs and are not inactivated by the standard β-lactam/β-lactamase inhibitor combination drugs. There is a newer β-lactamase inhibitor, avibactam, which has been approved for use with ceftazidime against gram-negative bacteria. In addition, avibactam is being tested for use with aztreonam against CREs [62, 66–68].

**5. Conclusion**

better options become available.

Address all correspondence to: reygaert@oakland.edu

**Author details**

Wanda C. Reygaert

**References**

32832d52e0.

10.1586/eri.11.153.

Microbiol Rev. 1982;**46**:241-280.

1232. DOI: 10.1086/507962.

For many strains of pathogenic *E. coli*, the most common course of therapy is supportive and does not require the use of antimicrobial drugs, or in the case of EHEC, antimicrobial therapy is not recommended. For severe intestinal infections and UTIs, antimicrobial therapy may be necessary. Unfortunately with the issue of ever increasing antimicrobial resistance, the antimicrobial options are becoming fewer. With the emergence of ESBL and CRE *E. coli* strains, the options have gotten extremely limited, and antimicrobial development has not been able to keep up with the demand. Hopefully the newer carbapenem/β-lactamase inhibitor combination drugs and other drugs being developed under the tetracycline and aminoglycoside drug classes will prove to be equal to the task or at least keep the bacteria under control until

Antimicrobial Mechanisms of *Escherichia coli*

http://dx.doi.org/10.5772/67363

91

Oakland University William Beaumont School of Medicine, Rochester, Michigan, USA

negative sepsis. BMC Infect Dis. 2012;**12**:56. DOI: 10.1186/1471-2334-12-56.

2009;**15**(Suppl 3):12-15. DOI: 10.1111/j.1469-0691.2009.02725.x.

[1] Micek S, Johnson MT, Reichley R, Kollef MH. An institutional perspective on the impact of recent antibiotic exposure on length of stay and hospital costs for patients with gram-

[2] Goossens H. Antibiotic consumption and link to resistance. Clin Microbiol Infect.

[3] Tacconelli E. Antimicrobial use: risk driver of multidrug resistant microorganisms in healthcare settings. Curr Opin Infect Dis. 2009;**22**:352-358. DOI: 10.1097/QCO.0b013e

[4] Griffith M, Postelnick M, Scheetz M. Antimicrobial stewardship programs: methods of operation and suggested outcomes. Expert Rev Anti Infect Ther. 2012;**10**:63-73. DOI:

[5] Bentley R, Meganathan R. Biosynthesis of vitamin K (menaquinone) in bacteria.

[6] Lee SY, Kotapati S, Kuti JL, Nightingale CH, Nicolau DP. Impact of extended-spectrum β-lactamase-producing *Escherichia coli* and *Klebsiella* species on clinical outcomes and hospital costs: a matched cohort study. Infect Control Hosp Epidemiol. 2006;**27**:1226-

### **4.4. Drug efflux**

Bacteria possess methods for disposal of toxic substances to the outside of the cell. The most commonly used mechanism is the efflux pump. Most bacteria have chromosomally encoded efflux pump genes. Some of these pumps are expressed constitutively, and expression of others is induced by various environmental stimuli. Many of these pumps are capable of transporting a variety of substances and are also described as multidrug (MDR) efflux pumps. There are five efflux pump family groups: the ATP-binding cassette (ABC) family, the multidrug and toxic compound extrusion (MATE) family, the major facilitator superfamily (MFS), the small multidrug resistance (SMR) family, and the resistance-nodulationcell division (RND) family. The RND pumps are generally found only in gram-negative bacteria as these pumps are multicomponent pumps that function in association with an OMP [69–72].

There is only one ABC efflux pump in *E. coli* that is known to contribute to antimicrobial resistance. That is the MacAB transporter that confers resistance to some macrolides [73]. There is also only one MATE efflux pump found in *E. coli*, the NorE pump which is able to transport fluoroquinolones. It is still in question if the NorE pump has a clinically significant impact on antimicrobial resistance [74, 75]. There are five known MFS efflux pumps found in *E. coli*. These are capable of transporting macrolides (MefB and MdfA pumps), fluoroquinolones (QepA2, EmrAB-TolC, and MdfA pumps), tetracycline (EmrAB-TolC and MdfA pumps), trimethoprim (Fsr pump), and chloramphenicol (MdfA pump). In addition, there are several MFS pumps that may be acquired by *E. coli* (e.g., via plasmids) that are specific for tetracyclines, with *tetA* and *tetB* being the most common [76, 77]. There are no clinically significant SMR efflux pumps found in *E. coli* [78].

The RND efflux pumps are the most clinically significant pumps found in gram-negative bacteria. These pumps consist of three components (tripartite): an inner membrane transporter, an outer membrane porin, and a periplasmic accessory protein that functions to connect the other two components. In *E. coli*, the OMP that is associated with all of the antimicrobial efflux pumps is TolC. There are five known RND pumps in *E. coli*: AcrAB-TolC, AcrAD-TolC, AcrEF-TolC, MdtABC-TolC, and MdtEF-TolC. AcrAD-TolC has been shown to efflux aminoglycosides and β-lactams. AcrEF-TolC has been shown to efflux quinolones and tigecycline. MdtABE-TolC has been shown to efflux quinolones. MdtEF-TolC has been shown to efflux erythromycin. The level of expression of these four pumps is relatively low, and if operating alone, the amount of antimicrobials effluxed would probably not be significant. Because *E. coli* has five efflux systems plus multiple other types of antimicrobial resistance mechanisms in play, these pumps undoubtedly help out. The other RND efflux pump in *E. coli*, AcrAB-TolC, is the most clinically significant and accounts for major antimicrobial efflux. This pump has been shown to efflux β-lactams, fluoroquinolones, tetracyclines, chloramphenicol, and lincosamides [72, 79, 80].
