Hugo W.B., Russel A.D.(ed). Pharmaceutical Microbiology

Подождите немного. Документ загружается.

support the desired amount of early growth. The principal nitrogen source is corn steep

liquor (CSL), a by-product of the maize starch-producing industry. This material was

originally found to be specifically useful for the penicillin fermentation, but it is

recognized as valuable in many fungal antibiotic media. Apart from its primary purpose

in supplying cheap and readily available nitrogen, CSL also contains a useful range of

carbon compounds, such as acids and sugars, inorganic ions and growth factors—in

short, it is virtually a complete growth medium in itself. However, like some of the fed

nutrients, CSL is a complex nutrient, not chemically defined, derived from natural

products and with significant batch-to-batch variation. It is therefore a source of variation

and one of the reasons why no two fermentations are ever absolutely identical.

The medium contains subsidiary nitrogen sources and additional essential nutrients

such as calcium (added in the form of chalk to counter the natural acidity of the CSL),

magnesium, sulphate, phosphate, potassium and trace metals. The medium is sterilized

with steam at 120°C either in the fermenter itself or in ancilliary plant, which may be

worked continuously.

3.4.2 Fed nutrients

The sterile medium is stirred and aerated and its pH and temperature are set to the

correct values on the process control monitors. It is then inoculated and the growth

phase begins. The initial carbon source is sufficient in quantity to maintain early growth

but not sufficient to provide the energy that penicillin production and maintenance

of the cell population need during the rest of the fermentation. Carbon for these

subsequent stages is 'fed' continuously in such a way as to limit net growth. Either

sucrose or glucose is used, possibly as cheaper, impure forms, such as molasses or

starch hydrolysate. As the concentration of residual sugar in the broth is too low to

measure, the rate of feeding has to be learnt by experience and modified on the basis

of systematic observation. An alternative way of attaining carbon limitation without

the complication of a carefully monitored carbon feed rate is to supply all the

carbohydrate at the outset as lactose. The rate-limiting hydrolysis of lactose to hexose

is then relied upon to give a steady, slow feed of assimilable carbohydrate. Originally,

all benzylpenicillin was manufactured using lactose in this way and some manufacturers

still prefer this technique.

Calcium, magnesium, phosphate and trace metals added initially are usually

sufficient to last throughout the fermentation, but the microorganisms need further

supplies of nitrogen and sulphur to balance the carbon feed. Nitrogen is often supplied

as ammonia gas. The word 'balance' is used quite deliberately; the whole system is a

balanced one. Thus, the carbon and nitrogen feeds not only satisfy the organisms'

requirements for these elements in the correct molar ratio, they also maintain an adequate

reserve of ammonium ion and contribute to pH control, the carbon metabolism being

acidogenic and balanced by the alkalinity of the ammonia. Sulphate is usually supplied

in common with the sugar feed and, by obtaining the correct ratio, there is a balanced

presentation of sulphate with an adequate pool of intermediates.

All feeds are sterilized before they are metered into the fermenter. Contaminants

resistant to the antibiotic rarely find their way into the fermenter, but when they do,

their effects are so damaging that prevention is of paramount importance. A resistant,

Manufacture of antibiotics 155

/3-lactamase-producing, fast-growing bacterial contaminant can destroy the penicillin

already made, as well as consuming nutrients intended for the fungus, causing loss of

pH control and interfering with the subsequent extraction process.

The growth phase passes rapidly into the antibiotic-production phase. The optimum

pH and temperature for growth are not those for penicillin production and there may

be changes in the control of these parameters. The only other event that marks the

onset of the production phase is the addition of phenylacetic acid (PAA) by continuous

feed.

3.4.3 Stimulation by PAA

Phenylacetic acid (PAA) supplies the side-chain of benzylpenicillin (see also Chapter

5); without PAA, the organisms synthesize only small quantities of this penicillin. Indeed,

it was the chance presence of phenylacetyl compounds in CSL (formed from phenylalanine

in the grain by the natural bacterial flora during processing) that caused it to be

established in early experiments as the best of the cheap complex nitrogen sources and

led to the use of PAA. Not only does PAA stimulate benzylpenicillin biosynthesis but it

also suppresses the formation of other (unwanted) penicillins. High levels of PAA are,

however, toxic to the organism and so it cannot be added indiscriminately. PAA is

expensive. The feed provides an adequate standing level of PAA without approaching

the toxic limit; the feed is reduced just before the end of the process so that the amount

of unused (irrecoverable) precursor in the final culture is not excessive.

The building blocks for the biosynthesis of benzylpenicillin are three amino acids,

a-aminoadipic acid, cysteine and valine, and PAA. The amino acids condense to a

tripeptide, ring closure of which gives the penicillin ring structure with an cu-aminoadipyl

side-chain, isopenicillin N. The side-chain is then displayed by a phenylacetyl group

from PAA to give benzylpenicillin.

There comes a time when sequential improvements in penicillin productivity

obtained by standard strain improvement techniques (physical and chemical mutagenesis

in conjunction with a variety of selection techniques that apply pressure for high-yielding

variants) become subject to rate-limiting returns. At first, it is easy to double the 'titre'

with each campaign; later in the genealogy even a 5% improvement would be regarded

as excellent.

Recent developments by academic and industrial geneticists may well prove to

have transformed this situation. Tremendous progress has been made since the mid-

1980s both in the isolation and manipulation of the biosynthetic genes in this pathway

and in the related routes to the cephalosporins (via the cephalosporin C-producing

156 Chapter 7

fungus Acremonium chrysogenum) and the cephamycins (via the cephamycin C-

producing bacterium Streptomyces clavuligerus). Antibiotic manufacturers can now

apply recombinant DNA technology to the industrial strains of filamentous micro-

organisms used to produce jS-lactams and there are exciting prospects of making genetic

changes that will very significantly increase productivity. These are discussed further,

later in this chapter. There is plenty of scope for improvement, because the best current

industrial strains and processes convert little more than 10% of all elemental carbon

into penicillin.

3.4.4 Termination

When to stop a fermentation is a very complex decision and several factors have to be

taken into account. Quite often a manufacturer will find it appropriate to harvest shortly

after the first signs of a faltering in the efficiency of conversion of the most costly raw

material (the carbon source, e.g. glucose) into penicillin.

3.5 Extraction

3.5.1 Removal of cells

At harvest, the benzylpenicilhn is in solution extracellularly, together with a range of

other metabolites and medium constituents. The first step in downstream processing

is to remove the cells by filtration or centrifugation. This stage is carried out under

conditions that avoid contamination with (3-lactamase-producing microorganisms which

could lead to serious or total loss of product.

3.5.2 Isolation of benzylpenicillin

The next stage is to isolate the benzylpenicillin. Solvent extraction is the generally

accepted process although other methods are available including ion-exchange

chromatography and precipitation. In aqueous solution at pH 2-2.5 there is a high

partition coefficient in favour of certain organic solvents such as amyl acetate, butyl

acetate and methyl isobutyl ketone. The extraction has to be carried out quickly, as

benzylpenicillin is very unstable at these low pH values. The penicillin is then extracted

back into an aqueous buffer at pH 7.5, the partition coefficient now being strongly in

favour of the aqueous phase. The solvent is recovered by distillation for re-use.

3.5.3 Treatment of crude extract

Benzylpenicillin is produced as various salts according to its intended use, whether as

an input to semisynthetic /3-lactam antibiotics manufacture or for clinical use in its

own right.

The treatment of the crude penicillin extract varies according to the objective but

involves formation of an appropriate salt, probably followed by treatment to remove

pyrogens, and by sterilization. This last is usually achieved by filtration but pure metal

salts of benzylpenicillin can be safely sterilized by dry heat if desired.

Manufacture of antibiotics 157

For parenteral use, the antibiotic is packed in sterile vials as a powder (reconstituted

before use) or suspension. For oral use it is prepared in any of the standard presen-

tations, such as film-coated tablets. Searching tests are carried out on an appreciable

number of random samples of the finished product to ensure that it satisfies the

stringent quality control requirements for potency, purity, freedom from pyrogens and

sterility.

All stages of antibiotic manufacture from fermentation through to finished product

are governed by the code of good manufacturing practice (GMP), of which quality

control is one aspect. GMP requires that 'there should be a comprehensive system, so

designed, documented, implemented and controlled, and so furnished with personnel,

equipment and other resources as to provide assurance that products will consistently

be of a quality appropriate to their intended use'.

4 The production of penicillin V

By the addition of different acyl donors to the medium, different penicillins can be

biologically synthesized. For example, penicillin V is made by a similar process to

benzylpenicillin, but with phenoxyacetic acid as the precursor instead of PAA. In the

biosynthetic pathway, the a-aminoadipyl side-chain of isopeniciUin N is replaced by a

phenoxyacetyl group.

The microorganism is again P. chrysogenum. A manufacturer may use the

same mutant strain to make both products or may have different mutants for the two

penicillins. Parallel situations of a single organism producing more than one natural

product occur with other types of antibiotics, for example strains of Streptomyces

aureofaciens are used for both chlortetracycline and demethylchlortetracycline

fermentations.

Like benzylpenicillin, penicillin V is still widely used in its own right but can also

be used as a starting material for the manufacture of the semisynthetic penicillins,

none of which can be made by direct fermentation.

5 The production of cephalosporin C

It is possible to convert penicillin V or benzylpenicillin to a cephalosporin by chemical

ring expansion. The first-generation cephalosporin cephalexin, for example, can be

made in this way. Most cephalosporins used in clinical practice, however, are semi-

synthetics produced from the fermentation product cephalosporin C.

The ancestral strain of Acremonium chrysogenum (at that time called

Cephalosporium acremonium) was isolated on the Sardinian coast in 1945 following

an observation that the local sewage outlet into the sea cleared at a quite remarkable

rate. Advances were slow because the activity was associated with a number of different

types of compound. Cephalosporin C was first isolated in 1952, but it was a further

decade before clinically useful semisynthetic cephalosporins became available.

The biosynthetic route to cephalosporin C is identical to that of the penicillins as

far as isopeniciUin N (section 3.4.3). The further route to cephalosporin C is shown on

p. 160. Note the branch into a third series of /3-lactam drugs, the cephamycins (see

Chapter 5).

158 Chapter 7

Isopenicillin N

Penicillin N

Desacetoxycephalosporin C

Desacetylcephalosporin C —> Cephamycin C (in certain Streptomyces)

Cephalosporin C

The similarities in the routes to the three classes of antibiotics have facilitated

progress in the understanding of the underlying molecular genetics. Most of the

genes coding for the relevant enzymes have been isolated. Modern DNA techniques

are being targeted at rate-limiting biosynthetic steps. Amplification of gene copy

numbers, improving gene expression efficiencies, transferring genes to bacterial host

organisms and manipulation of pathways of antibiotic synthesis all have potential in

strain development. However, in the production of antibiotics, economic benefits from

the application of recombinant DNA technology have thus far been limited.

Manufacturing processes for cephalosporin C and benzylpenicillin are broadly

similar. In common with many other antibiotic fermentations, no specific precursor

feed is necessary for cephalosporin C. There is sufficient acetyl group substrate for the

terminal acetyltransferase reaction available from the organism's metabolic pool.

The product is extracted from the culture fluid by adsorption onto carbon or resins

rather than by solvent. This illustrates an important general point that antibiotic

manufacturing processes differ from one another much more in their product recovery

stages than in their fermentation stages. Figure 7.4 illustrates a typical production route

from inoculum to bulk antibiotic.

6 Further reading

Bu'Lock J.D., Nisbet L.J. & Winstanley D.J. (eds) (1983) Bioactive Microbial Products, vol. II,

Development and Production. London: Academic Press.

Calam C.T. (1987) Process Development in Antibiotic Fermentations, Cambridge Studies in

Biotechnology, 4 (eds Sir James Baddiley, N.H. Carey, J.F. Davidson, I.J. Higgins & W.G. Potter).

Cambridge: Cambridge University Press.

Hugo W.B. & Mol H. (1972) Antibiotics and chemotherapeutic agents. In Materials and Technology

(eds L.W. Codd, K. Dijkoff, J.H. Fearon, C.J. van Oss, H.G. Roeberson & E.G. Stanford). London:

Longman & de Bussy.

Peberdy J.F. (ed.) (1987) Penicillin and Acremonium, Biotechnology Handbooks, 1 (Series eds

T. Atkinson & R.F. Sherwood). New York: Plenum Press. (See, in particular, Chapters 2 and

5.)

Office for Official Publications of the European Community (1992) The Rules Governing Medicinal

Products in the European Community, vol. IV, Guide to Good Manufacturing Practice for the

Manufacture of Medicinal Products.

Queener S.W. (1990) Molecular biology of penicillin and cephalosporin biosynthesis. Antimicrob Agents

Chemother, 34, 943-948.

Rohm H.-J., Reed G., Piihler A. & Stadler P. (eds) (1993) Biotechnology, vol. Ill, Bioprocessing. New

York: VCH.

Smith J.E. (1985) Biotechnology Principles. Aspects of Microbiology Series No. 11. Wokinham: Van

Nostrand Reinhold.

160 Chapter 7

Stowell J.D., Bailey P.J. & Winstanley D.J. (eds) (1986) Bioactive Microbial Products, vol. Ill,

Downstream Processing. London: Academic Press.

Van Damme E.J. (1984) Biotechnology of Industrial Antibiotics. New York: Marcel Dekker.

Verrall M.S. (Ed) (1985) Discovery and Isolation of Microbial Products. Society of Chemical Industry

Series in Biological Chemistry and Biotechnology. Chichester: Ellis Horwood.

A good source of articles on individual antibiotics, groups of antibiotics, fermentation plant and related

topics is the series Progress in Industrial Microbiology edited originally by D.J.D. Hockenhull and

published by Heywood Books, London. These articles normally carry extensive references to the original

literature.

Manufacture of antibiotics 161

Mechanisms of action of antibiotics

3 Protein synthesis

3.1 Protein synthesis and selective

inhibition

3.2 Aminoglycoside-aminocyclitol

antibiotics

3.3 Tetracyclines

3.4 Chloramphenicol

3.5 Macrolides and azalides

3.6 Lincomycin and clindamycin

3.7 Fusidic acid

3.8 Mupirocin

4 Chromosome function and replication

mammalian cells

5.2 Sulphonamides

5.3 DHFR inhibitors

6 The cytoplasmic membrane

6.1 Composition and susceptibility

of membranes to selective

disruption

6.2 Polymyxins

6.3 Polyenes

6.4 Imidazoles and triazoles

6.5 Naftidine

7 Further reading

Introduction

The antibiotics described in Chapter 5 are used to treat microbial infections caused by

bacteria, fungi or protozoa. Most exert a highly selective toxic action upon their target

microbial cells but have little or no toxicity towards mammalian cells. They can therefore

be administered at concentrations sufficient to kill infecting organisms (or at least inhibit

their growth) without damaging mammalian cells. By comparison, the disinfectants,

antiseptics and preservatives described in Chapter 10 are too toxic for systemic treatment

of infections. Study of the mechanism of action of the antibiotics reveals the basis of

their selective toxicity. Table 8.1 lists the five broad target areas of action (cell wall,

ribosome, chromosome, folate metabolism, cell membrane) with the major antibiotics

which act upon them and a summary of the basis of selective action. Note that the

majority of the antibiotics are used for treatment of bacterial infections; comparatively

few agents are available for fungal or protozoal infections.

1 Introduction 4.1 The basis for selective inhibition

of chromosome replication and

2 The bacterial cell wall function

2.1 Peptidoglycan biosynthesis and its 4.1.1 Synthesis of precursors

inhibition 4.1.2 Unwinding of the chromosome

2.1.1 D-Cycloserine 4.1.3 Replication of DNA strands

2.1.2 Glycopeptides—vancomycin and 4.1.4 Transcription

teicoplanin 4.2 Quinolones

2.1.3 /3-Lactam antibiotics—penicillins, 4.3 Nitroimidazoles (metronidazole) and

cephalosporins, carbapenems and nitrofurans (nitrofurantoin)

monobactams 4.4 Rifampicin

2.2 Mycolic acid and arabinogalactan 4.5 5-Fluorocytosine

synthesis in mycobacteria

2.2.1 Isoniazid 5 Folate antagonists

2.2.2 Ethambutol 5.1 Folate metabolism in microbial and

t All antibacterial except: *antimycobacterial agent; **antiprotozoal agent; ***antifungal agent.

LPS, lipopolysaccharide.

Table 8.1 Target sites for antimicrobial action

Target

Bacterial cell wall

Bacterial ribosome

function

Chromosome function

Folate metabolism

Cytoplasmic

membrane

Antibioticst

j3-Lactams

Glycopeptides

Cycloserine

Isoniazid*

Ethambutol*

Aminoglycosides

Tetracyclines

Chloramphenicol

Macrolides, azalides

Fusidic acid

Mupirocin

Quinolones

Metronidazole (also**)

Nitrofurantoin

Rifampicin (also*)

5-Fluorocytosine***

Sulphonamides (also**)

Trimethoprim

Pyrimethamine**

Trimetrexate**/***

Polymyxins

Polyenes***

Imidazoles and triazoles***

Naftidine***

Mechanism of action

Inhibit peptidoglycan synthesis

Inhibits peptidoglycan synthesis

Inhibits mycolic acid synthesis

Inhibits arabinogalactan synthesis

Distort 30S ribosomal subunit

Block 30S ribosomal subunit

Inhibits peptidyl transferase

Block translocation

Inhibits elongation factor

Inhibits isoleucyl-tRNA synthesis

Inhibit DNA gyrase

DNA strand breakage

Inhibits RNA polymerase

Inhibits DNA synthesis

Inhibit folate synthesis

Inhibits dihydrofolate reductase

Disrupt bacterial membranes

Disrupt fungal membranes

Inhibit ergosterol synthesis

Inhibits ergosterol synthesis

Basis of selective toxicity

None in mammalian cells

No action on 40S subunit

Excluded by mammalian cells

No action on mammalian equivalent

Excluded by mammalian cells

No action on mammalian equivalent

Requires anaerobic conditions not

present in mammalian cells

No action on mammalian equivalent

Converted to active form in fungi

Not present in mammalian cells

Mammalian enzyme not inhibited

Toxicity overcome with leucovorin

Bind to LPS and phospholipids

Bind preferentially to ergosterol

Pathway not in mammalian cells

The bacterial cell wall

2.1 Peptidoglycan biosynthesis and its inhibition

Peptidoglycan is a vital component of virtually all bacterial cell walls. It accounts for

approximately 50% of the weight of Gram-positive bacterial walls, around 30% of

mycobacterial cell walls and between 10 and 20% of the Gram-negative envelope. It is

a macromolecule composed of sugar (glycan) chains which are crosslinked by short

peptide bridges (Fig. 8.1).

One characteristic feature of peptidoglycan is the occurrence of D-amino acids,

particularly D-alanine and D-glutamic acid. These D-stereoisomers of amino acids are

not found in proteins. The precise nature of the peptide crosslinks varies among

organisms but the essential structure is the same. The peptidoglycan polymer is

responsible for both the shape of bacterial cells and their mechanical strength and

integrity. If the synthesis of peptidoglycan is blocked selectively by antibiotic action

p-lactams

Fig. 8.1 Biosynthesis of peptidoglycan. The large circles represent A^-acetylglucosamine or N-

acetylmuramic acid; to the latter is linked initially a pentapeptide chain comprising L-alanine, D-

glutamic acid and meso-diaminopiraelic acid (small circles) terminating in two D-alanine residues

(small, darker circles). The lipid molecule is undecaprenyl phosphate. In the initial (cytoplasm) stage

where inhibition by the antibiotic D-cycloserine is shown, two molecules of L-alanine (small circles)

are converted by an isomerase to the D-forms (small, darker circles), after which a ligase joins the

two D-alanines together to produce a D-alanyl-D-alanine dipeptide.

164 Chapter 8