Design Of Antimicrobial Chemotherapeutic Agents

EDWARD G.M.POWER and A.DENVER RUSSELL

CONTENTS

11.1 INTRODUCTION 435

11.2 PRODUCTION OF CHEMOTHERAPEUTIC AGENTS 436

11.3 MECHANISM OF ACTION OF CHEMOTHERAPEUTIC AGENTS 438

11.3.1 Inhibitors of cell wall synthesis 438

11.3.2 Membrane-active agents 442

11.3.3 Inhibitors of protein synthesis 442

11.3.4 Inhibitors of nucleic acid synthesis 442

11.3.5 Antibacterial folate inhibitors 445

11.3.6 Conclusions and comments 447

11.4 BACTERIAL RESISTANCE TO CHEMOTHERAPEUTIC AGENTS 447

11.4.1 Enzyme-mediated resistance 447

11.4.1.1 p-lactamases 447

11.4.1.2 Aminoglycoside-modifying enzymes 449

11.4.1.3 Chloramphenicol-inactivating enzymes 451

11.4.2 Outer membrane barrier 452

11.4.3 Transferable resistance 452

11.4.4 Conclusions and comments 453

11.5 DESIGN OF p-LACTAM ANTIBIOTICS 454

11.5.1 Penicillins 454

11.5.2 Cephalosporins 457

11.5.2.1 Structure-activity relationships 457

11.5.2.2 Pharmacokinetic properties 459

11.5.3 ß-Lactamase stability 461

11.5.4 ß-Lactamase inhibitors 462

11.5.4.1 ß-Lactams as inhibitors 462

11.5.4.2 Naturally occurring ß-lactamase inhibitors 462

11.5.4.3 Synthetic ß-lactamase inhibitors 464

11.5.4.4 Structure-activity relationships in ß-lactamase inhibitors 464

11.5.4.5 ß-lactamase inducers 464

11.5.4.6 Mutual pro-drugs 465

11.5.5 Other ß-lactam ring systems 465

11.5.5.1 1-Oxacephems 465

11.5.5.2 Penems 465

11.5.5.3 Nocarcidins 466

11.5.5.4 Monobactams 466

11.5.5.5 Carbacephems 466

11.6 DESIGN OF OTHER ANTIBACTERIAL AGENTS 467

11.6.1 Aminoglycoside-aminocyclitol antibiotics 467

11.6.2 Tetracyclines 469

11.6.3 Macrolides 471

11.6.4 Chloramphenicol 472

11.6.5 Folate inhibitors 473

11.6.6 Quinolones 474

11.7 DESIGN OF ANTIFUNGAL AGENTS 476

11.7.1 Polyene antibiotics 476

11.7.2 Imidazole derivatives 478

11.7.3 Griseofulvin 479

11.7.4 Flucytosine 480

11.7.5 Other membrane-active compounds 480

11.7.6 Cell wall-active compounds 481

11.7.7 Novel antifungal agents 483

11.7.8 Conclusions and comments 483

11.8 DESIGN OF ANTIVIRAL AGENTS 483

11.8.1 Mechanisms of inhibition 484

11.8.1.1 Amantadines 484

11.8.1.2 Nucleoside analogues 484

11.8.1.3 Foscarnet 487

11.8.1.4 Interferons 487

11.8.2 Mechanisms of resistance 488

11.9 OVERALL CONCLUSIONS AND

COMMENTS 489

FURTHER READING 490

11.1 INTRODUCTION

The design of a new chemotherapeutic agent, suitable for clinical use in the treatment of human infections, must take into account two aspects above all others. First, the drug must possess high antimicrobial activity and secondly it must be non-toxic to human tissue. Paul Erlich's early concept of a selectively toxic 'magic bullet' is thus just as true in the modern world as when it was first propounded. A host of antimicrobial agents has been examined and many shown to be effective inhibitors of micro-organisms in vitro; unfortunately, several of these have been found to be harmful to human tissues and consequently have no role to play in chemotherapy. This chapter will thus concentrate on several of those chemotherapeutic agents that have proved their worth when employed internally (usually orally or parenterally). Antimicrobial compounds that are used for their disinfectant, antiseptic or preservative qualities will not be dealt with here.

Two important properties of any chemotherapeutic agent were mentioned briefly above. Additionally, any drug must ideally have a broad spectrum of activity, with a rapid bactericidal or other microbicidal action. Some bacteria produce enzymes that can inactivate or modify antibiotics, and insusceptibility of a drug to such degradation or modification could result in its playing an important part in therapy. Likewise, some bacteria possess an outer membrane that acts as a permeability barrier to the entry of some, but not all, antibiotics. Drugs that can readily penetrate this barrier might again be expected to be of possible clinical importance. These two aspects are considered in greater detail later (Sections 11.4.1 and 11.4.2).

In addition to being non-toxic, an antibiotic should not cause any hypersensitive reactions, such as those induced in a minority of patients by the penicillins and, to a lesser extent, the cephalosporins. It should, however, be readily absorbed to give high blood and tissue levels and it should be stable to gastric acid. Binding to serum proteins should be of a low order so that high concentrations of the drug are freely available in the plasma. In urinary infection, high urine levels are desirable but some antibiotics are excreted so rapidly that, on occasion, it may be necessary to delay the rate of excretion. These and other pharmacological properties are considered further where appropriate.

Finally, the design of any chemotherapeutic agent must involve a consideration of its chemical and physical properties, since these will be of paramount importance to the pharmacist responsible for formulating a suitable product. Such properties include its aqueous solubility, and its stability in solution at different pHs and temperatures. These aspects are considered to be outside the scope of the present chapter.

11.2 PRODUCTION OF CHEMOTHERAPEUTIC AGENTS

The demonstration in the early 1920s that lysozyme possessed antibacterial activity, the accidental discovery of 'penicillin' by Fleming in the late 1920s and the finding that the azo dye prontosil owed its antibacterial properties to the release in vivo of sulphanilamide, all stimulated research towards the development of chemotherapeutic agents.

Fleming, in 1929, published the results of his chance finding that a Penicillium mould caused lysis of staphylococcal colonies on an agar plate. He also showed that the filtrate of a culture of the mould, growing in a liquid medium, possessed significant activity against Gram-positive bacteria and Gram-negative cocci, although most other types of Gram-negative organisms were resistant. In retrospect, it was indeed fortunate that the original plates of staphylococci did not contain organisms that produced an inactivating enzyme (P-lactamase: Section 11.4.1) otherwise the mould would not have shown any activity and could well have been discarded! It is, of course, highly unlikely with today's knowledge and techniques that benzylpenicillin (the product obtained from the Penicillium mould) would have been lost to mankind, but rather that its introduction into medicine would have been delayed.

It is now known that benzylpenicillin is thermolabile. At the time of Fleming's discovery, however, great difficulty was experienced in extracting the antibiotic from the culture medium, as this property was not appreciated. Later studies by Florey,

Chain and their colleagues at Oxford University, and in the USA, used a cold solvent extraction method that succeeded in extracting and purifying the elusive active principle. Extensive research on the composition of the culture media and on induction of mutants of the mould resulted in conditions that gave enhanced antibiotic yields.

These facts are all relevant to the subsequent development of antibiotics, since the commercial production of benzylpenicillin stimulated world-wide efforts into examining soil samples with a view to obtaining other antibiotics from moulds residing there. Studies at Rutgers (by Waksman and colleagues) and elsewhere were responsible for the development of streptomycin, the tetracyclines and chloramphenicol. These and many other antibiotics—some of them clinically useful— were obtained from Streptomyces species, which comprise the most prolific source of antibiotics. In the 1950s, a Cephalosporium mould isolated off the coast of Sardinia was found to produce significant antibiotic activity and was sent to Oxford University where, as a result of the efforts of Abraham and his colleagues, the birth of the cephalosporins resulted.

Important though these early findings are, they do not in themselves present any coherent pattern in the deliberate design of an antibiotic. This development had to await further discoveries in antibiotic production and an improved knowledge of mechanism of action, bacterial resistance and pharmacokinetics. Towards the end of the 1950s Chain, in collaboration with scientists at Beecham Research Laboratories, observed the production of 6-aminopenicillanic acid (6-APA (11.1)) in media in which Pénicillium chrysogenum was growing when phenylacetic acid (C6H5-CH2-COOH; phenylethanoic acid) was omitted. Phenylacetic acid is the precursor of the side-chain of benzylpenicillin ((11.2) R =C6H5CH2CO) and 6-APA is the nucleus to which a side-chain is attached. This finding has had far-reaching effects since it has become the starting point in the deliberate design and synthesis of a family of penicillins having different or improved properties from existing members. This aspect is considered in more detail in Section 11.5.1.

6-APA is a naturally occuring antibiotic. In contrast, the nucleus (7-aminocephalosporanic acid, 7-ACA (11.3)) of the cephalosporin group does not occur

H,N <1

COO H

CHj-O-CO-CH

COO H

COOII

(11,4} cephalosporin C

COOII

CHfQ-CO-CHy

COOII

(11,4} cephalosporin C

COOII

naturally. Studies at Oxford revealed that C. acremonium produced more than one antibiotic: 'cephalosporin P' (active against Gram-positive bacteria, but subsequently shown not to be a cephalosporin) and 'cephalosporin N' (with activity against Gramnegative bacteria and later found to be a penicillin). Further studies disclosed that 'cephalosporin N' was actually contaminated with a true cephalosporin, cephalosporin C (11.4) which showed a high degree of stability to staphylococcal P-lactamase. Cephalosporin C is converted to 7-ACA by appropriate chemical means. Like 6-APA, 7-ACA can be considered as being the starting point in the development of newer antibiotics. The cephalosporins are discussed in Section 11.5.2.

The production of semi-synthetic P-lactam antibiotics has formed part of an exciting era in chemotherapy. In the meantime, other semisynthetic antibiotics, e.g. some tetracyclines, have been described and some antibiotics, notably chloramphenicol and the quinolones, have been totally synthesized chemically. These are considered in later Sections.

11.3 MECHANISM OF ACTION OF CHEMOTHERAPEUTIC

AGENTS

The great majority of studies of the mechanism of action of chemotherapeutic agents have involved investigations of individual compounds or of drugs within a particular group, e.g. tetracyclines. A considerable amount of information is now available (Table 11.1) and will be summarized in this section, since in the space available only a bare outline can be presented. Where possible, ways in which this knowledge can be used to further the design of new antibiotics will be discussed.

The bacterial cell wall is a complex structure (see also Section 11.4.2 and Figure 11.1). Differences occur between the walls of Gram-positive and Gram-negative bacteria, but they all contain a basal peptidoglycan (murein, mucopeptide). This consists of the amino sugars A-acetylglucosamine (GlcNAc) and A-acetylmuramic acid (MurNAc) to which are attached amino acids, some in the unnatural D-configuration. In brief, peptidoglycan synthesis involves the stepwise addition of amino acids to MurNAc, the linking to GlcNAc to form a linear polymer and finally a cross-linking (transpeptidation, via a transpeptidase enzyme) of the linear polymers to form a rigid structure, in which the degree of cross-linking varies. A simplified example of cross-linked peptidoglycan is given in Figure 11.2(a). D-cycloserine (11.5), a structural analogue of D-alanine (11.6) inhibits two enzymes (a racemase and

11.3.1 Inhibitors of cell wall synthesis a synthetase) involved in the synthesis of the D-alanyl-D-alanine dipeptide. P-lactam antibiotics inhibit the cross-linking (transpeptidation) reaction (see Figure 11.2(b)).

Table 11.1 Mechanisms of action of antibacterial agents.

Effect

Example(s)_Comments

Inhibition of cell wall synthesis

Effect on the cytoplasmic membrane

D-cycloserine

ß-lactams

Glycopeptides Polymixins

Ionophores

Polyenic antibiotics

Inhibition of protein Streptomycin synthesis Tetracyclines

Competitive inhibition of alanine racemase and synthetase Inhibition of transpeptidases Binding to PBPs (specific) Binding to peptidoglycan precursor

Affect outer membrane of Gram-negative bacteria also

Specific cation conductors: non-selective Bind to membrane sterols in fungi (bacteria unaffected)

Inhibits initiation stage Inhibits binding of aminoacyl-tRNA to 30S ribosomal subunit

Chloramphenicol Inhibits peptidyl Erythromycin transferase

Inhibition of RNA synthesis

Puromycin

Actinomycin D Rifampicin

Inhibition of DNA Mitomycin C

synthesis

Inhibition of tetrahydrofolate synthesis

Quinolones Novobiocin Sulphonamides Trimethoprim

Inhibits translocation Binds to peptidyl transferase: nonselective Binds to double stranded DNA

Inhibits DNA polymerase Covalent linking to DNA Effect on DNA gyrase Effect on DNA gyrase Competitive inhibitors of dihydropteroate synthetase (see also text) Inhibits dihydrofolate reductase

Figure 11.1 Outer layers of (a) Gram-positive and (b) Gram-negative bacteria.
Figure 11.2 Cross-linked peptidoglycan (a) and role of transpeptidase (b) in Staphylococcus aureus. MurNAc, N-acetylmuramic acid; GlcNAc, N-

acetylglucosamine; AGLA, respectively L-alanine, D-glutamine, L-lysine, D-alanine; (Gly)5. 5 molecules of glycine.

Considerable progress has been made in understanding the action of P-lactam antibiotics at the molecular level. Bacterial cell membranes (inner membranes) contain several proteins, known as penicillin binding proteins (PBPs), which are associated with specific enzyme activity (Table 11.2), with which P-lactam antibiotics may combine.

Table 11.2 Function of penicillin-binding proteins (PBPs) in Escherichia coli.

Considerable progress has been made in understanding the action of P-lactam antibiotics at the molecular level. Bacterial cell membranes (inner membranes) contain several proteins, known as penicillin binding proteins (PBPs), which are associated with specific enzyme activity (Table 11.2), with which P-lactam antibiotics may combine.

Table 11.2 Function of penicillin-binding proteins (PBPs) in Escherichia coli.

PBP Enzyme activity

Function

Result of

Example(s) of ß-

inhibiting lactam inhibiting

1A

Transglycosylase,

Cell wall

Lysis

Benzylpenicillin,

transpeptidase

growth

cephalosporins

1B

Transglycosylase,

Cell wall

Lysis

Benzylpenicillin,

(a, b, transpeptidase

growth

most

d)

cephalosporins

2.

Transglycosylase,

Initiation of

Oval cells

Mecillinam,

transpeptidase

cell wall growth

imipenem

3.

Transglycosylase,

Septum

Filaments

Many

transpeptidase

formation, cell division

cephalosporins, piperacillin, aztreonam

4.

Carboxypeptidase

Regulation

?

Benzylpenicillin,

endopeptidase

of cross-linking

ampicillin, imipenem

5.

Carboxypeptidase

Regulation of cross-linking

?

Cefoxitin

B.

Carboxypeptidase

Regulation of cross-linking

?

Cefoxitin

7/8

Unknown

?

?

Penems

Most of these antibiotics bind to only one or two PBPs and, very importantly, the morphological effects induced by various P-lactams are determined by the PBP to which they bind predominantly (examples are provided in Table 11.2). PBPs 1B, 2 and 3 appear to be the most important in E. coli, since binding to PBPs 4, 5 and 6 has no adverse effect on the cells. Thus, a future possibility might well be to design a combination of two P-lactam antibiotics which have different PBP specificity in order

Most of these antibiotics bind to only one or two PBPs and, very importantly, the morphological effects induced by various P-lactams are determined by the PBP to which they bind predominantly (examples are provided in Table 11.2). PBPs 1B, 2 and 3 appear to be the most important in E. coli, since binding to PBPs 4, 5 and 6 has no adverse effect on the cells. Thus, a future possibility might well be to design a combination of two P-lactam antibiotics which have different PBP specificity in order to achieve a synergistic result, P-lactams that bind to PBP 1 induce rapid cell lysis or sphaeroplasts (osmotically fragile forms), whereas those binding predominantly to PBP3 induce filamentation. P-lactams binding to PBP2 induce the formation of spherical, osmotically stable forms.

These PBPs are associated with specific enzyme activity, as depicted in Table 11.2. Changes in PBPs may be associated with bacterial resistance to P-lactam antibiotics, e.g. (a) a laboratory strain of E. coli resistant to mecillinam possessed a PBP2 with lower affinity for this antibiotic; (b) a hospital isolate of Neisseria gonorrhoeae resistant to benzylpenicillin was found to possess PBPs 1 and 2 with a lower affinity for this drug than sensitive cells; (c) penicillin-resistant strains of pneumococci have been found to possess an altered PBP with reduced affinity for the drug.

Interaction of a P-lactam with a 'penicillin-sensitive enzyme' (PSE) involves binding of the antibiotic with a serine group in the enzyme by opening of the P-lactam ring system. The penicilloyl (or cephalosporyl)-enzyme complex is more stable than the complex of enzyme and natural substrate (D-alanyl-D-alanine), and the enzyme is consequently trapped in an inactive form, and is thus unable to fulfil its normal role in peptidoglycan synthesis.

The glycopeptides (vancomycin, teicoplanin, ristocetin) also inhibit peptidoglycan synthesis by binding to the D-alanyl-D-alanine terminus of various petidoglycan precursors, preventing the transglycosylation step by which glycan units are polymerized within the peptidoglycan. Unlike P-lactam action, the transglycosylase enzyme is not inhibited, but the complex of vancomycin with the dipeptide prevents the substrate from interacting with the active site of the enzyme.

11.3.2 Membrane-active agents

The term 'membrane-active agent' is generally taken to mean an agent that affects the cytoplasmic membrane in micro-organisms. Gram-negative bacteria, however, also possess an outer membrane (Section 11.4.2) which may act as a penetration barrier to some drugs. The polymyxins, for example, cause the leakage of intracellular constituents by damaging the cytoplasmic membrane of Gram-negative bacteria, but they also disrupt the outer membrane lipopolysaccharide. They are highly toxic to mammalian cells.

Polyenic antibiotics combine with sterols in the cytoplasmic membrane of yeasts, fungi and mammalian cells. However, nystatin and amphotericin B have some selective action against fungi because they exhibit a greater binding to ergosterol than to cholesterol (Section 11.7.1).

Ionophoric drugs facilitate the passage of specific inorganic cations across the cytoplasmic membrane of Gram-positive bacteria, e.g. valinomycin and monactin are K+-conducting ionophores. Unfortunately, they show a lack of specific toxicity, as they exert the same effect on mammalian membranes also.

11.3.3 Inhibitors of protein synthesis

Bacterial ribosomes have different characteristics (sedimentation coefficient of 70, i.e. are 70S, with 50S and 30S subunits) from those of mammalian ribosomes (80S, with

60S and 40S subunits). Most antibacterial antibiotics that are clinically important inhibitors of protein synthesis have a preferential effect on 70S ribosomes. For example, chloramphenicol affects 70S ribosomes, but not 80S. In contrast, the tetracyclines inhibit protein synthesis on isolated 70S and 80S ribosomes. The tetracyclines bind to the 30S subunit, as does streptomycin. In contrast, erythromycin binds entirely to the 50S subunit. The reason for their selective inhibition of bacterial protein synthesis in vivo resides in the energy-dependent active transport system, present in bacterial but not mammalian cells, that transports these antibacterial agents into bacteria.

Bacterial protein synthesis, i.e. peptide chain extension, is carried out on 70S ribosomes. In brief, an amino acid, activated and attached to its transfer RNA (tRNA) binds to the acceptor site of the ribosome. The peptide in the donor site is transferred, as a result of peptidyl transferase activity, to the new amino acid at the acceptor site. Subsequently, translocation of the extended peptide from the acceptor site to the donor site takes place with loss of the preceding tRNA and movement of the ribosome relative to messenger RNA (mRNA) which specifies the next amino acid.

The action of some antibiotics is shown in Table 11.1.

11.3.4 Inhibitors of nucleic acid synthesis

Inhibitors of nucleic acid synthesis fall into two main categories: (a) those that inhibit the synthesis of purine and pyrimidine nucleotides, e.g. azaserine, a glutamine analogue, although this is not selectively toxic as it is also harmful to mammalian cells; (b) those that ihibit synthesis at the polymerization level. This involves the appropriate polymerase, with nucleic acid as a template, and is the stage where condensation of nucleoside triphosphates into a polynucleotide chain, with joining by 3',5'-phosphodiester linkages takes place.

Some drugs are intercalating agents (e.g. acridines) and inhibit DNA synthesis and DNA-dependent RNA synthesis in whole cells and in cell-free systems. Another intercalating agent, actinomycin D, has a selective action on RNA synthesis. Mitomycin C cross-links to DNA but, like actinomycin D, is also toxic to mammalian cells.

Quinoline (11.7) derivatives based on 4-quinolone (11.8), 8-aza-4-quinolone (11.9), 2-aza-4-quinolone (11.10) and 6,8-diazo-4-quinolone (11.11) have been examined for antimicrobial activity. Ciprofloxacin (11.12) has been shown to be the most active 4-quinolone against most aerobic Gram-negative bacteria. Nalidixic acid (11.13) has no

Qrriinilrrir

4- Qurrmforif o

S-Ajn-4-Quinnlont O

0 0

Post a comment