Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The development and spread of resistance to antibiotics have been a continual problem since the discovery of antibiotics (Davies and Davies 2010). In the late 1980s, the appearance of resistance to vancomycin in Gram-positive bacteria was especially disturbing (see Chap. 2). This event energized a prolonged search for new agents having activity against resistant bacteria, both Gram-positive and (more recently) Gram-negative species. Many antibiotics (see Chaps. 10, 12–15) inhibit protein synthesis as their mechanism of action (MOA), making it one of the most common and important antibacterial mechanisms (Lange et al. 2007). Despite an increasing prevalence of pathogens that show multiple patterns of resistance, this MOA remains as important today as it has been for the past several decades of antibiotic usage. Furthermore, new antimicrobial agents that inhibit some part of protein synthesis continue to be developed.

Macrolide antibiotics are one of the foremost classes that exert their antibacterial activity through this MOA (Hermann 2005; McCoy et al. 2011; Wilson 2004; Yonath 2005). The parent macrolides are produced by fermentation of soil microorganisms and some of these older macrolides are still useful therapeutic agents today (Demain 2009; Demain and Sanchez 2009; Omura 2011). They also represent an invaluable resource of chemical starting materials that have spawned many important semi-synthetic derivatives which possess various improved features responsible for increased efficacy and safety. This chapter will summarize the macrolide antibiotics, both older agents that are still important and newer agents that are in some stage of the preclinical or clinical development pipeline.

2 Erythromycin and Its Semi-synthetic Derivatives

Macrolide antibiotics constitute a large class that is organized by the size and substitution patterns of their highly substituted macrolactones to which particular saccharide moieties are attached (Kaneko et al. 2007; Kirst 2005; Mitscher 2010). Fermentation-derived macrolide antibiotics have a 14- or 16-membered lactone, while 15-membered macrolides are created by chemical ring expansion of a 14-membered ring. Many other macrolide compounds are known whose macrolactones have different substitution patterns or have fewer than 14 or more than 16 members, but the antibacterial activity of those compounds is generally too weak and/or too limited for useful clinical applications (Shiomi and Ōmura 2002).

Erythromycin A is the prototype of 14-membered macrolide antibiotics (Fig. 11.1). It is the major component of a complex produced by fermentation of a soil actinomycete now classified as Saccharopolyspora erythraea. First generation derivatives of erythromycin were synthesized soon after its discovery that included many acid-addition salts, esters, and salt-ester combinations designed to increase stability under acidic conditions (e.g., stomach) and to improve oral bioavailability. Acid-addition salts also improved water solubility for intravenous administration, but intramuscular administration was too painful upon injection to be used. All of these derivatives revert to erythromycin free base, which is the active entity of these first-generation derivatives. Early studies were conducted to learn the cause for the acid instability of erythromycin. These studies discovered a facile intramolecular cyclization by the C-6 hydroxyl group with the C-9 ketone to form a 6,9-hemiketal followed by 8,9-dehydration to initially yield the 8,9-anhydro-6,9-hemiketal intermediate (Fig. 11.1), which then underwent further degradation (Kurath et al. 1971). This insight provided a mechanistic rationale for structural modifications of erythromycin that later produced the desired greater stability and oral bioavailability.

Fig. 11.1
figure 1

Structures of erythromycin A and its intramolecular cyclization product

Full size image

Second generation semi-synthetic derivatives of erythromycin were prepared by chemical transformations that modified certain of those functional groups that contribute to the intramolecular cyclization of erythromycin. Early members of this group were erythromycin-11,12-cyclic carbonate, 9-(S)-erythromycylamine, and roxithromycin (Fig. 11.2). The presence of either the exocyclic 5-membered ring in the 11,12-cyclic carbonate or the C-9 oxime in roxithromycin made these derivatives less prone than erythromycin to undergo the irreversible intramolecular cyclization sequence. In erythromycylamine, the C-9 ketone was replaced by an amino group which rendered the derivative incapable of forming the 6,9-hemiketal. Erythromycylamine was later re-examined as the active component in the pro-drug, dirithromycin. However, all of these earlier derivatives were superseded by clarithromycin and azithromycin, both of which became the more widely used second generation macrolides (Fig. 11.2) (Sivapalasingam and Steigbigel 2010; Zuckerman et al. 2011).

Fig. 11.2
figure 2

Structures of second generation derivatives of erythromycin

Full size image

Clarithromycin is the 6-O-methyl ether of erythromycin, in which the 6-hydroxyl group is substituted and can no longer engage in intramolecular cyclization. Azithromycin is a ring-expanded 15-membered derivative in which the C-9 ketone is replaced via Beckmann rearrangement and N-methylation with a ring-embedded N-methylamino-methylene unit, a change that eliminates the C-9 ketone from participation in intramolecular cyclization. The collective group of compounds having an amino group incorporated within the macrolactone framework has been named azalides. Dirithromycin and flurithromycin were later entries into second generation derivatives (Fig. 11.2). Dirithromycin is an oxazine pro-drug of 9-(S)-erythromycylamine. Flurithromycin contains an 8-fluorosubstituent that prevents irreversible dehydration of the 6,9-hemiketal. Each of these diverse modifications provided a unique approach to circumventing the propensity of erythromycin for intramolecular cyclization and thereby achieved greater stability in each individual way.

Ketolides constitute the third generation, so-named due to their 3-keto functionality that replaces the 3-O-cladinosyl moiety of erythromycin, as exemplified by the first commercial ketolide, telithromycin (Ketek®) (Fig. 11.3) (Bryskier and Denis 2002; Sivapalasingam and Steigbigel 2010; Van Bambeke et al. 2008; Zhanel and Neuhauser 2005; Zuckerman et al. 2011). More recent ketolides that have entered the antibiotic pipeline include cethromycin, modithromycin, and solithromycin (Fig. 11.3) (Butler and Cooper 2011; Donadio et al. 2010; Kirst 2010). Cethromycin (Restanza™) originated from the antibiotic discovery programs at Abbott Laboratories (ABT-773) (Hammerschlag and Sharma 2008; Rafie et al. 2010). It was in late stage development for treating community acquired bacterial pneumonia (CABP) and was also being investigated for biodefense applications. However, its future development is currently uncertain because Advanced Life Sciences suspended operations in May, 2011 (Advanced Life Sciences 2012; Bush and Pucci 2011; Sutcliffe 2011).

Fig. 11.3
figure 3

Structures of representative ketolides

Full size image

Several series of macrolides have been prepared at Enanta Pharmaceuticals that contain additional rings created by different bridging linkages between various hydroxyl groups of erythromycin. These series have thus been given the general names of bicyclolides, tricyclolides, etc. Bicyclolides are being investigated for both oral and intravenous administration against bacterial pathogens in hospital and community settings and are also being tested for biodefense applications against several pathogenic bacteria (Enanta 2012). The initial clinical candidate, modithromycin (formerly EDP-420 and S-013420) (Fig. 11.3), contains an additional internal ring that bridges the C-6 and C-11 hydroxyl groups. It has been jointly investigated in clinical trials with Shionogi & Company (Furuie et al. 2010; Jiang et al. 2009).

Solithromycin (formerly CEM-101) is the first 2-fluoro-ketolide in clinical development. Its structure also has a four-carbon linker between the 11,12-carbamate and a bis-heterocyclic side chain composed of a relatively stable 1,2,3-triazole and a 2-aminophenyl group (Fig. 11.3) (Pereira and Fernandes 2011). It was licensed by Cempra Pharmaceuticals from Optimer Pharmaceuticals (Cempra 2012). Solithromycin has completed a phase 2 clinical trial for CABP by oral administration while an intravenous formulation is in a phase 1 trial (Cempra 2012; Fernandes et al. 2011; Sutcliffe 2011).

Research is still very actively in progress to discover new derivatives of erythromycin (Kirst 2010; Ma and Ma 2011; Ying and Tang 2010). In addition to new ketolides, other modifications around the 2,3-position of the core macrolactone include the so-called acylides, alkylides and anhydrolides. Several additional modifications are also being investigated around the 11,12-position (Kirst 2010). BAL19403 possesses a heterocyclic substituent linked to a 11,12-lactone rather than a 11,12-cyclic carbamate. It demonstrated good activity against resistant propionibacteria (Heller et al. 2007). Changes in the two saccharide moieties are also being explored with the synthesis of 3′-N- or 4′′-O-modified derivatives of erythromycin. These efforts indicate that the search will continue for additional new derivatives of erythromycin having improved clinical efficacy and activity against resistant pathogens.

Macrolide antibiotics also have important applications in veterinary medicine. Two of the more recent azalides, tulathromycin (Draxxin®) and gamithromycin (Zactran®) are used exclusively for veterinary purposes such as treatment of respiratory infections in animals (Forbes et al. 2011; Shryock and Richwine 2010).

3 16-Membered macrolide antibiotics

16-Membered macrolide antibiotics are divided into two large families, tylosin and leucomycin-spiramycin, based on different substitution patterns of their macrolactones (Fig. 11.4). Tylosin is produced by fermentation of Streptomyces fradiae and is its family prototype. It is an important veterinary antibiotic, but it has not been developed for use in human medicine (Elanco 2012). A few other members of the tylosin family have also been developed exclusively for applications in veterinary medicine, including two semi-synthetic derivatives of tylosin, tilmicosin (Micotil®, Pulmotil®) and tildipirosin (Zuprevo®), which are being used to treat respiratory infections in animals (Buret 2010; Menge et al. 2012). Some clinical investigations have occurred in the past with a few members of the tylosin family, but none of these compounds appear to have yet been successfully developed for human medicine.

Fig. 11.4
figure 4

Structures of representative 16-membered macrolides

Full size image

The leucomycin family is more numerous and more complicated because many members have been obtained from fermentation of different microorganisms by different research groups and given different names or corporate code numbers (Kirst 2005). Leucomycin was initially isolated as a complex of ten components from culture broths of Streptomyces kitasatoensis (Fig. 11.4) (Omura 2011). Midecamycin and spiramycin were also isolated as multi-component complexes from culture broths of Streptomyces mycarofaciens and Streptomyces ambofaciens, respectively. Some members of this family are used in human medicine, such as josamycin (leucomycin A3), midecamycin, and spiramycin. Although none of these 16-membered macrolides have been registered for the U.S. market, spiramycin is used to treat certain infections caused by Toxoplasma gondii in pregnant women (Montoya and Remington 2008).

The most important semi-synthetic derivatives in the leucomycin family are miokamycin and rokitamycin (Alvarez-Elcoro and Yao 2002). Chemical acylation of leucomycin-type macrolides, especially of their 3′′-hydroxyl group, increased the half-life of antibiotic activity while retaining good in vitro potency. This discovery was applied to prepare the semi-synthetic derivatives miokamycin (9,3′′-di-O-acetyl derivative of midecamycin A1) and rokitamycin (3′′-O-propionyl derivative of leucomycin A5) (Fig. 11.4). As mentioned above for the 16-membered parent macrolides, neither of these two derivatives has been registered for the U.S. market. As with the 14-membered family, new research on 16-membered macrolides also continues in the effort to discover new antibiotics (Cui and Ma 2011; Przybylski 2010). However, the amount of effort has been significantly less than that devoted to 14- and 15-membered agents. One reason may be that 16-membered ketolides synthesized thus far have not demonstrated activity comparable to 14-membered ketolides (Creemer et al. 2002; Mutak et al. 2004; Terui et al. 2006). Analogous to 14-membered ketolides, the attachment of additional substituents to the 16-membered ring may be required to achieve the necessary stronger ribosomal binding and greater activity.

4 Antimicrobial Features

Macrolide antibiotics possess a moderately wide range of antimicrobial activity in which they inhibit susceptible strains of many Gram-positive bacteria, certain Gram-negative bacteria, and a variety of other pathogenic organisms (Dang et al. 2007; Roberts 2008; Sivapalasingam and Steigbigel 2010; Zuckerman et al. 2011). They penetrate well into many cells and tissues and exhibit activity against many microbes that dwell in an intracellular environment (Mulazimoglu et al. 2005). However, they generally lack useful activity against enterococci and most enteric and coliform Gram-negative bacteria. Interestingly, a recent study reported that in vitro activity of macrolides against Pseudomonas aeruginosa was highly dependent on the type of growth medium being used for the MIC test. MIC values were much lower when eukaryotic cell growth media were used compared to higher MICs when cation-adjusted Mueller–Hinton broth was employed (Buyck et al. 2011). This MIC differential may further help to explain the positive clinical effects of macrolides that are observed in cases involving P. aeruginosa, such as diffuse panbronchiolitis and cystic fibrosis (Crosbie and Woodhead 2009; Friedlander and Albert 2010).

The activity of macrolides may be bactericidal or bacteriostatic, depending on the particular microorganism, antibiotic concentration, contact time, and other experimental conditions. Most macrolides contain an amino group in their structures and thus they are basic substances that form acid-addition salts with increased water solubility. However, the un-ionized free base is the active form, so microbial penetration and antimicrobial activity is increased at higher pH values.

Although second generation macrolides had been more focused toward solving the earlier problems involving stability and oral bioavailability, some of those derivatives also showed greater potency against certain microorganisms that partially expanded the antimicrobial spectrum compared to erythromycin and 16-membered macrolides (Ali et al. 2002; Blondeau et al. 2002; Sivapalasingam and Steigbigel 2010; Zuckerman et al. 2011). Clarithromycin and azithromycin emerged as the more widely used macrolides due to some favorable clinical features, including somewhat broadened spectrum of activity and greater efficacy, improved pharmacokinetics, less frequent dosing schedule, and better gastrointestinal tolerance. Among some prominent traits of this group, azithromycin was more effective in lowering MIC values against many Gram-negative bacteria while clarithromycin was more active against Gram-positive bacteria. The efficacy of clarithromycin against Haemophilus influenzae was aided by its in vivo conversion to its more active 14-hydroxy metabolite whereas azithromycin had a lower MIC against H. influenzae. The enhanced activity of these macrolides has been widely useful against pathogens that are responsible for many respiratory tract infections, skin and soft tissue infections, and sexually transmitted diseases. Among other applications, they are used to treat gastrointestinal (GI) problems caused by Helicobacter pylori. They exhibit activity against many non-tuberculous mycobacteria, especially against the Mycobacterium avium complex (MAC) that has aided treatment of MAC infections in AIDS patients (Young and Bermudez 2002).

With problems of stability and oral bioavailability substantially addressed by several second generation derivatives, the alarming rise in microbial resistance to antibiotics started to draw more attention, especially during the late 1980s. In response to this disturbing development, macrolide research began to shift in order to search for new agents that would combat this dangerous trend. The third generation of macrolides was thus intended to address the growing problems of microbial resistance to antibiotics, resulting in the emergence of the first ketolides in the mid-1990s (Bryskier and Denis 2002; Van Bambeke et al. 2008).

5 Mechanism of Action

The bacterial ribosome is a large and complex structure composed predominantly of RNA and protein that performs the vital task of bacterial protein biosynthesis. Thus, disruption of ribosome function by antibiotics causes serious deleterious effects to the microorganism, including death. The highly complex nature of protein biosynthesis on the ribosome makes for multiple ways in which the overall process can be disrupted. In addition, a second MOA involving inhibition of ribosome assembly by macrolides has been proposed (Champney 2006; Siibak et al. 2009).

Detailed knowledge has been rapidly expanding about the ribosome’s structure, its mechanisms for functioning, and its interactions with antibiotic substances (Allen 2002; Blanchard et al. 2010; Bogdanov et al. 2010; Dunkle et al. 2010; Garrett et al. 2000; Kannan and Mankin 2011; Mankin 2008; McCoy et al. 2011; Wilson 2011). To briefly summarize, the programmed sequential addition of individual amino acids onto a growing peptide chain occurs at the peptidyl transferase center (PTC) located in the large (50S) subunit of the ribosome. The PTC catalyzes the sequential formation of the growing peptide’s amide bonds. As the peptide chain becomes extended upon the addition of each new amino acid, the lengthening peptide moves outward through the exit tunnel of the ribosome. Macrolide antibiotics bind in the region of the exit tunnel near the PTC where their presence either completely blocks or partially hinders progression of the nascent peptide out through this tunnel. Different macrolides may bind in different arrangements, but the overall result is inhibition of protein synthesis by preventing the proper elongation of the peptide. Depending on the macrolide, various types of prematurely terminated peptides may be released. Some macrolides such as 16-membered ones containing the 5-O-mycaminosyl-mycarosyl disaccharide have sufficient length to reach the PTC and disrupt formation of amide bonds.

X-ray crystallographic studies using co-crystals of macrolides bound in the large ribosomal subunit have now been performed using many different macrolides and ribosomes from several different microbes (Wilson 2011). Those results are consistent with the overall MOA and provide valuable visual evidence of how different macrolides bind to particular ribosomes in each individual manner. Although ribosomes are generally considered to have conserved structures, it is now recognized that antibiotic-ribosome interaction and binding may differ between ribosomes from different microbial species, so overly generalized interpretations of results may not be valid (Kannan and Mankin 2011; Wilson 2011). Additional studies are likely forthcoming that will greatly expand our detailed knowledge about this MOA. They will also suggest specific ways by which macrolide binding might be strengthened and thus will guide medicinal chemistry research in synthesizing new more potent derivatives (Sutcliffe 2005; Wimberly 2009). The importance of this technology and its significant impact on new drug discovery was celebrated by the award of the 2009 Nobel Prize in Chemistry to Profs. Ramakrishnan, Steitz, and Yonath for their pioneering work on the structure and function of ribosomes.

6 Microbial Resistance to Macrolides

The isolation of erythromycin from fermentation cultures and the first reports of clinical studies were both published in 1952 (Haight and Finland 1952; Heilman et al. 1952; McGuire et al. 1952). Unfortunately, microbial resistance to erythromycin was also observed soon after its clinical appearance (Leclercq and Courvalin 1991a). In addition, it was early recognized that the level of resistance could be correlated with the amount of antibiotic usage which had placed selective pressures on the microbial population and thereby selected resistant strains (Westh 1996). The clinical significance of resistance to macrolides was initially considered as low, but as years of antibiotic usage increased, so also did serious concerns steadily increase about the continuous rise in resistance to not just macrolides, but to all antibiotics (Boucher et al. 2009; Leclercq and Courvalin 1991b; Mulazimoglu et al. 2005).

The more common mechanism of resistance to macrolide antibiotics is modification of the target site responsible for activity, which is the ribosome. Other resistance mechanisms include antibiotic efflux systems, decreased uptake or permeability into the cell, various mutations to ribosomal RNA and proteins, and modification of the antibiotic structure by inactivating enzymes (Dang et al. 2007; Douthwaite and Vester 2000; Mlynarczyk et al. 2010; Roberts 2008; Sutcliffe and Leclercq 2002).

Target site modification in bacteria disrupts macrolide ribosomal binding which thereby prevents or hinders the antibiotic from accomplishing its objective of inhibiting protein synthesis. In one common manifestation, the ribosomal binding sites overlap between the macrolide, lincosaminide, and streptogramin B antibiotics (see Chap. 14) resulting in cross-resistance between these three structurally unrelated classes and producing a phenotype named MLSB resistance (Leclercq and Courvalin 1991a; Weisblum 1995a). MLSB resistance is caused by enzymatic N6-methylation of an adenine residue located in the overlapping binding region of ribosomal RNA. That N-methylation produces a conformational change in the ribosome that significantly weakens bonding by the antibiotic. This enzymatic methylation is genetically controlled by numerous readily transferable erm (erythromycin ribosome methylase) genes that are now found in a wide host of bacteria (Roberts 2008, 2011). MLSB resistance can be either inducible or constitutive and 16-membered macrolides are generally non-inducers (Allen 1977, 1995; Weisblum 1995b). Two old fermentation-derived 3-keto-14-membered macrolides (pikromycin and narbomycin) were also shown to be non-inducers that were nevertheless active against macrolide-inducibly-resistant staphylococci (Allen 1977). Ketolides show a similar pattern in their response to inducibility (Bonnefoy et al. 1997). However, like other macrolides, they are not active against constitutively resistant strains (Sivapalasingam and Steigbigel 2010; Van Bambeke et al. 2008).

Ketolides possess several important structural changes compared to traditional derivatives of erythromycin that lead to significant advantages in antimicrobial activity and resistance patterns. These structural changes include replacement of the 3-O-mycarosyl substituent with a 3-keto group, addition of a rigid ring system across either the 11,12- or 6,11-positions, and attachment of a bis-heterocyclic moiety via a short carbon linker to various positions within the C-6 to C-12 region (Fig. 11.3). Among additional changes, modithromycin also contains a C-9 acyl-imine in place of the C-9 ketone. As a result from these structural changes, ketolides acquired a second ribosomal binding site to accommodate the heterocyclic chain in addition to the single ribosomal binding site used by older macrolides (Dang et al. 2007; Wilson 2011; Zhanel and Neuhauser 2005). X-ray structures of telithromycin-ribosomal complexes depict the binding patterns in these ribosomes (Dunkle et al. 2010; Tu et al. 2005). The second binding site strengthens ketolide-ribosomal binding affinity which increases antimicrobial potency. For macrolide-resistant strains, extension of ketolide binding into a second domain provides a new mechanism to overcome or circumvent the ribosomal N-methylation resistance mechanism and thus gives rise to activity against those resistant bacteria (Zhanel and Neuhauser 2005; Zuckerman et al. 2011).

Analysis of a crystal study of Escherichia coli ribosomes complexed with solithromycin proposed the presence of three binding sites with the third site coming from the positioning of the 2-fluoro substituent (Fernandes et al. 2011; Llano-Sotelo et al. 2010; Sutcliffe 2011). Such a result could further strengthen ketolide-ribosomal binding and increase potency relative to non-fluorinated analogs. However, the exact mechanism and in vitro activity resulting from a 2-fluoro substituent may depend on the specific ketolide structure rather than follow a generalized SAR rule for all ketolides (Hwang et al. 2008; Keyes et al. 2003; Llano-Sotelo et al. 2010). Hydrogen bonding from the 2-aminophenyl group also contributes to overall ribosomal binding of solithromycin. The proposal of three binding sites for a single ketolide structure would provide a valuable new mechanism for overcoming the N-methylation and other macrolide-resistance mechanisms in macrolide resistant strains (McGhee et al. 2010). It will be interesting to watch the results of future SAR studies focused in this direction.

The extended and stronger binding that results from the structural changes in the C-6 to C-12 region of ketolides more than compensates for the reduction in activity that occurs upon removal of the 3-O-cladinosyl subunit from erythromycin, an absence that does have the positive effect of removing inducibility of resistance (Allen 1977). Consequently, ketolides tend to show greater activity compared to erythromycin and second generation derivatives against both susceptible and resistant staphylococci, streptococci, and other important pathogens. 2-Fluoro-ketolides appear to increase that activity differential even further although the universality of that trend is still unproven and needs to be more fully investigated. Another caution is that binding of macrolides to ribosomes from different species may yield different results so over-generalizations should be avoided (Kannan and Mankin 2011; Wilson 2011).

Detailed analyses of comparative potencies or resistance patterns among ketolides and older macrolides are beyond the scope and available space of this review and such surveys have been published by many others (Dang et al. 2007; Rafie et al. 2010; Sivapalasingam and Steigbigel 2010; Sutcliffe 2011; Van Bambeke et al. 2008; Zhanel and Neuhauser 2005; Zuckerman et al. 2011). Driven by the medical needs and therapeutic potential that is still available from the ketolide template, it is likely that the creation of novel ketolide structures has not yet reached any limits and additional innovative structures should be revealed in due course.

7 Pharmacology

The two semi-synthetic derivatives clarithromycin and azithromycin are the dominant macrolide antibiotics currently being used in clinical practice. They are prescribed to treat upper and lower respiratory tract infections caused by a range of pathogens, skin and skin structure infections, several sexually transmitted diseases, and a wide spectrum of other infections caused by various bacteria and other pathogenic organisms (Van Bambeke et al. 2008; Sivapalasingam and Steigbigel 2010; Zuckerman et al. 2011). Among the latter uses is treatment of MAC infections in AIDS patients and eradication of gastrointestinal H. pylori often by means of combination therapy. Macrolides also play an important clinical role as an alternative to β-lactam antibiotics for patients who are allergic to the latter agents.

In addition to their overtly bacteriostatic or bactericidal activities against pathogens, macrolides have been long known to display a variety of anti-inflammatory (AIF) and immunomodulatory (IMM) properties in the host that make some significant contributions to the overall efficacy of these agents. Numerous studies, analyses, and reviews of these systems have been made over several decades by many investigators (Altenburg et al. 2011; Buret 2010; Harvey et al. 2009; Kovaleva et al. 2012; Zarogoulidis et al. 2012). However, the complexities of the numerous AIF and IMM networks cause difficulties in separating the component parts and in dissecting primary causes from many secondary effects. The situation is further complicated because different macrolides may show opposite effects, thereby making generalities difficult to establish. Consequently, many of the basic mechanisms by which these effects occur still remain incompletely understood. Some attempts have also been made to create derivatives that dissociate the direct antimicrobial activity from non-antibiotic effects, but this objective has thus far only met with very limited success. The most successful separation of activities has been found with compounds derived from intramolecular cyclization of erythromycin, first as motilin agonists in the GI tract and more recently as lead structures for AIF or IMM applications (Sugawara et al. 2011).

As described above, successive generations of derivatives have steadily improved many clinical attributes of this class, allowing it to remain an important contributor to the therapeutic armamentarium for nearly 60 years. In this therapeutic role, macrolides are generally regarded as among the safest antibiotics, with the majority of side effects involving various disturbances of the GI tract. One advantage of several second generation derivatives was a lower incidence and reduced severity of GI effects compared to erythromycin (Periti et al. 1993).

Telithromycin is the most recent commercial macrolide and is currently the only ketolide that has received regulatory approval, which occurred in Europe and some Latin countries in 2001 and the U.S. in 2004. However, during its more extensive use following the clinical trials and approvals, serious problems were reported which included incidents of severe hepatotoxicity, certain visual side effects, and exacerbation of myasthenia gravis. In response to these safety concerns, stronger labeling warnings were written and in 2007, the U.S. FDA restricted use of telithromycin to the treatment of CABP (Van Bambeke et al. 2008; Sivapalasingam and Steigbigel 2010; Zuckerman et al. 2011). One recent study proposed that certain nicotinic acetylcholine receptors that may be associated with those side effects are located in the liver, eye, and muscle. These receptors may be inhibited by telithromycin and may be responsible for these undesirable effects (Bertrand et al. 2010; Fernandes et al. 2011; Sutcliffe 2011). Furthermore, the pyridine component in telithromycin has been suspected of involvement in this activity. The older macrolides, clarithromycin and azithromycin, and the newer ketolide, solithromycin, did not show the same level of inhibition as telithromycin, suggesting that this test could perhaps be used to predict the possibility of these side effects. These developments are encouraging that the side effects of telithromycin may be more structure-specific and not shared by all ketolides.

8 Biosynthesis

Early studies of the biosynthesis of erythromycin and other macrolides revealed the formation of their aglycones by sequential coupling of small organic acids (acetate, propionate, etc.) (Corcoran 1964). Following the addition of each acid, the newly formed subunit was then appropriately modified to give the desired stereospecific sub-structure using the processes of ketone reduction, dehydration, and enoyl reduction as appropriate to produce the final product (Kwan and Schulz 2011). Lastly, cyclization of the resultant 14- or 16-membered acyclic chains yielded the aglycones (Corcoran 1981; Omura and Tanaka 1984).

Later studies discovered a strongly programmed process that assembled the aglycones via large and highly organized modular structures called a polyketide synthase (PKS) (Cortes et al. 1990; Donadio et al. 1991). Following cyclization that cleaves the polyketide chain from the PKS, the resultant aglycone is converted to the macrolide antibiotic by appropriate post-PKS transformations, such as hydroxylation, O-methylation, O-glycosylation, etc. (Rix et al. 2002; Zhao and Liu 2010). Investigations by numerous researchers have revealed many further details about the general biosynthetic pathways and PKS-controlled processes and confirmed the generality of this biosynthetic mechanism for the construction of numerous polyketide structures (Cane 2010; Hertweck 2009; Khosla 2009; McDaniel et al. 2005; van Lanen and Shen 2008). This greatly detailed knowledge about the biochemistry and genetics of biosynthesis now allows more rationale and control for genetic engineering of biosynthetic pathways in microorganisms, including applications for combinatorial biosynthesis to create new molecules and for improvements in the fermentative production of known compounds (Baltz 2006; Khosla et al. 2007). All of these biosynthetic possibilities open additional routes to new structural diversification and nicely complement the chemical synthetic routes to produce new antibiotic structures.

9 Conclusions

Macrolide antibiotics continue to be an important class for treatment of many infectious diseases. Their 2009 sales in the U.S. were $4.8 billion, making them the fourth largest class in sales (after cephalosporins, broad spectrum penicillins, and fluoroquinolones) (Hamad 2010). Approximately 60 years have passed since erythromycin and many other macrolides were discovered and isolated from culture broths of soil microorganisms. During that period, an extremely large number of semi-synthetic 14-, 15-, and 16-membered macrolides have been prepared and evaluated, which can be divided conveniently into three generations of derivatives. Two second generation derivatives of erythromycin, clarithromycin and azithromycin, are currently the more utilized macrolides. Ketolides have emerged as third generation derivatives of erythromycin that show useful activity against many macrolide-resistant bacteria. Even though the first approved ketolide, telithromycin, has encountered some serious problems with side effects, other ketolides are being synthesized and some have entered the clinical development pipeline. Based on both the undeveloped potential still remaining for this class and the medical need for new agents, research efforts within the macrolide class will undoubtedly continue. From these continuing efforts, new members possessing important and useful improvements in antimicrobial spectrum, efficacy, and safety should be discovered and developed. Such future discoveries will ensure that the macrolide antibiotic class will remain an important contributor to the global anti-infective armamentarium.