Appl Microbiol Biotechnol (2003) 61:385392

DOI 10.1007/s00253-003-1274-y
MI NI - RE VI E W
R. P. Elander
Industrial production of b-lactam antibiotics
Received: 27 November 2002 / Revised: 28 January 2003 / Accepted: 31 January 2003 / Published online: 3 April 2003
Springer-Verlag 2003
Abstract The industrial production of b-lactam antibi-
otics by fermentation over the past 50 years is one of the
outstanding examples of biotechnology. Today, the b-
lactam antibiotics, particularly penicillins and cephalos-
porins, represent the worlds major biotechnology prod-
ucts with worldwide dosage form sales of ~US$ 15 billion
or ~65% of the total world market for antibiotics. Over
the past five decades, major improvements in the
productivity of the producer organisms, Penicillium
chrysogenum and Acremonium chrysogenum (syn. Ceph-
alosporium acremonium) and improved fermentation
technology have culminated in enhanced productivity
and substantial cost reduction. Major fermentation pro-
ducers are now estimated to record harvest titers of
4050 g/l for penicillin and 2025 g/l for cephalosporin
C. Recovery yields for penicillin G or penicillin V
are now >90%. Chemical and enzymatic hydrolysis
process technology for 6-aminopenicillanic acid or
7-aminocephalosporanic acid is also highly efficient
(~8090%) with new enzyme technology leading to
major cost reductions over the past decade. Europe
remains the dominant manufacturing area for both
penicillins and cephalosporins. However, due to ever
increasing labor, energy and raw material costs, more
bulk manufacturing is moving to the Far East, with China,
Korea and India becoming major production countries
with dosage form filling becoming more dominant in
Puerto Rico and in Ireland.
Introduction
The total world market for b-lactam antibiotics is now
estimated to be ~US$ (hereafter $) 15 billion with
cephalosporin dosage form sales at ~$9.9 billion and
penicillin dosage form sales at ~$5 billion (Barber 1996,
2000). In 1996, the total world antibiotic market at the
dosage form level was estimated to be ~$23 billion
(Demain and Elander 1999). The United States antibac-
terial market was >$8 billion with cephalosporins
($3.6 billion), penicillins ($1.2 billion), fluoroquinolones
($0.9 billion), tetracyclines ($0.5 billion) and macrolides
($0.4 billion) reported recently by Strohl (1999). In
contrast, the total world sales for antifungal products was
~$3 billion and growing and the antiviral market was
~$2.6 billion with a market forecast of >$5 billion for
2000. The b-lactam antibiotics now account for over 65%
of the world antibiotic market. There are now more than
50 marketed cephalosporins. Many of these are listed in
Table 1.
The biosynthetic pathways for the penicillins, cepha-
losporins and cephamycins are well characterized both
genetically and biochemically (Demain et al. 1998). A
generalized pathway is shown in Fig. 1. Most of the
important genes have now been cloned, with the excep-
tion of the isopenicillin N epimerase (cef D) gene
(Brakhage 1998). Amplification of certain of the genes,
in particular the deacetoxycephalosporin C synthase (cef
E) gene, has been reported to increase cephalosporin C
production and to decrease the levels of deacetoxy-
cephalosporin C (DAOC) in production fermentation
broths (Skatrud et al.1989; Basch and Chiang 1998). The
transfer to and expression of the cef E gene from
Streptomyces clavuligerus in Penicillium chrysogenum
resulted in the formation of adipyl-7-aminodeacetoxy-
cephalosporanic acid (adipyl-7-ADCA) when the altered
P. chrysogenum strain was fed adipic acid (Crawford et
al. 1995). In this manner, the inherent greater biosynthetic
capacity of P. chrysogenum may be used to produce
DAOCs, moleculeswith expanding marketsthat are
far more expensive to manufacture than penicillins
R.P. Elander is a former employee of Bristol-Myers Squibb
(retired)
R. P. Elander (
)
)
Biotechnology Consultant,
318 Gravilla Street, La Jolla, CA 920376006, USA
e-mail: [email protected]
Tel.: +1-858-5514146
Fax: +1-858-5514146
(Cantwell et al. 1990, 1992). This interesting technology
has not been scaled to production levels at this time.
Commercial production of penicillin
The fermentation production of penicillin-G or -V is a
fed-batch process carried out aseptically in stainless steel
tank reactors of 30,000100,000 gallon capacity. The
fermentation usually involves two to three initial seed
growth phases followed by a fermentation production
phase having a time cycle ranging from 120 to 200 h.
High dissolved oxygen levels are critical, especially
during peak growth periods that often occur at the 4050 h
time-period of the cycle. The fermentation mode is fed-
batch and crude sugar and precursor are fed throughout
the cycle. Current penicillin fermentations are highly
computerized and automated. Temperature, pH, dissolved
oxygen, carbon dioxide, sugar, precursor, ammonia, etc.
are closely monitored and controlled for optimal antibi-
otic production (Waites et al. 2001).
Various carbon sources have been adopted for the
fermentation including glucose, sucrose and other crude
sugars. About 65% of the carbon is metabolized for
cellular maintenance, 2025% for growth and 1012% for
penicillin production (Van Nistelrooij et al. 1998). Sugar
and precursor are fed continuously and the sugar is also
used to help regulate the pH of the fermentation to
between 6.46.8 during the active penicillin production
phase.
Corn steep liquor and cottonseed or soybean meal,
ammonia and ammonium sulfate represent major nitrogen
sources. The essential precursor substances are phenyl-
acetic acid (for penicillin G) or phenoxyacetic acid (for
penicillin V) that are either fed or batched.
Mini-harvest protocols are often used in penicillin
fermentations. This "batch-fill and withdraw" system
involves the removal of 2040% of the fermentor contents
with replacement with fresh sterile medium. This proce-
dure can be repeated several times during the fermenta-
tion without yield reduction and, in reality, can enhance
the total penicillin yield per fermentor. Penicillin G titers
of ~100,000 U/ml have been reported by researchers at
Panlabs (Rowlands 1991).
Penicillin is excreted into the medium and is recovered
at the end of the fermentation. Whole broth extraction is
usually performed at acidic pH by most manufacturers
and has resulted in a 25% improvement in overall
extraction efficiency by the elimination of the rotary
vacuum filtration step. Solvent extraction of chilled
acidified broth is carried out with amyl, butyl or isobutyl
acetate. Multiple back-extractions into buffer and solvent
at varying pH using countercurrent contactors has led to
considerable penicillin concentration in the early recovery
stages of the purification process. Pigments and other
broth impurities are removed by the use of activated
charcoal. The penicillin is crystallized upon the addition
Table 1 Marketed and experimental b-lactam antibiotics. Antibiotics italicized are major commercial antibiotics
Subclass Marketed b-lactam antibiotics
Penicillins Ampicillin
a
, amoxicillin
a
, bacampicillin, cloxacillin, floxacillin, mezlocillin, nafcillin, oxacillin,
penicillin G
a
, penicillin V
a
Penicillin-resistant penicillins Methicillin, dicloxacillin
Antipseudomonal penicillins Carbenicillin, indanyl piperacillin, ticarcillin
First-generation cephalosporins Cefalothin, cephradine
a
, cefadroxyl
a
, cefazolin, cephalexin
a
Second-generation cephalosporins Cefuroxine, cefaclor
a
, cefotetam, cefmetazole, cefonicid
Third-generation cephalosporins Cefixime
a
, ceftibuten, cefizoxime, ceftriaxone, cefamandol cefoperazone, cefotaxime, proxetil,
cefprozil
a
, ceftazidime, cefuroxime axetil, cefpodexime
Fourth-generation cephalosporins Cefepime
Oxycephams Flomoxef, latamoxef
Cefam Cefoxitin
Carbapenems Loracarbef
a
, imipenem, meropenem, panipenem
Monobactams Aztreonam, carumonam
Clavams (b-lactamase-inhibitors) Clavulanate, sulbactam, tazobactam
Penicillins/b-lactamase inhibitors Amoxicillin/clavulanate, ampicillin/sulbactam, pipericillin/tazobactam, ticarcillin/clavulanate,
cefoperazone/sulbactam
a
Orally administered b-lactams
Table 2 Changes in penicillin manufacturing technology
Fermentation 1950 2000
Carbon source Lactose Glucose/sucrose
Operational mode Batch Fed-batch
Medium sterilization Batch Continuous
Air filtration Depth filters Membrane filters
Feeds None Many
Morphology Filamentous Pelleted
Cycle time 120 h 120200 h
Tank volume
(1,000 gallons)
1020 2060
Assay Bio-assay HPLC
Control Temperature only Computerized
Titer (g/l) 0.51.0 >40
Recovery and purification
Mycelium removal Filtration Whole broth
Operational mode Batch Semi-continuous
Extraction stages Many Single to few
Precursor recovery
and re-use
Discarded Recovered
and re-used
Efficiency (%) 7080 >90
Environmental issues Few Many
Bulk cost ~US$275350/kg ~US$1520/kg
386
of potassium acetate and is isolated as a crystalline
potassium salt. Additional carbon treatments and solvent
washes results in a highly purified final product. Usually,
conversion grade product used for 6-aminopenicillanic
acid (6-APA) has lower purity and lower final product
cost. Table 2 shows a number of the major production
changes that have occurred in both upstream and down-
stream processing over the past five decades.
In 1949, the United States was the major manufactur-
ing country, with penicillin production amounting to
~83 tons of sodium penicillin G. In 1982, penicillin
production was carried out throughout the world, with a
total production of >12,000 tons and Europe being the
major manufacturing sector (Van der Beck and Roels
1984). In 1995, the total world production was reported to
be ~33,000 tons, a five-fold increase since the late 1960s
(Barber 1996).
The tremendous increase in penicillin fermentation
productivity (Elander and Chiang 1991) and correspond-
ing high (>90%) recovery yield has led to significant cost
reduction despite increasing labor, energy and raw
material costs. In 1953, the bulk cost for penicillin G
was ~$300/kg (Sylvester and Coghill 1954). In 1980, the
bulk price for penicillin was ~$35/kg. In the late 1990s,
bulk penicillin cost ranged from $10 to $20/kg and bulk
marketed costs for 6-APA have been estimated to range
Fig. 1 Biosynthetic pathway
for penicillins, cephalosporins
and cephamycins
387
from $35 to $40/kg. Table 3 lists the worlds major
producers of penicillin and important bulk b-lactam
intermediates.
Semi-synthetic b-lactams
Approximately 75% of the total bulk penicillin volume
produced in 1995, ~33,000 tons, was used for the
production of semi-synthetic penicillins and cephalospor-
ins (Barber 1996). A number of these important antibi-
otics are listed in Table 1. The penicillin nucleus (6-APA)
has enabled researchers to develop many excellent semi-
synthetic penicillins. 6-APA can also be chemically ring-
expanded to 7-ADCA to generate a number of important
orally-active cephalosporins (cephalexin, cephradine, ce-
fadroxyl, etc.). 6-APA has now grown to be the worlds
largest selling b-lactam bulk intermediate.
6-APA was discovered in the late 1950s and although
microbial enzymes were soon discovered that hydrolyzed
penicillins to 6-APA, their use in production processes
was abandoned in the late 1960s. Efficient chemical
splitting technology was developed in Holland by Weis-
senburger and van der Hoeven (1970) and was soon
adopted by many manufacturers. Environmental problems
and high solvent and energy costs prompted researchers to
re-investigate enzymatic routes for 6-APA. When the
reaction product is soluble in water, enzyme reuse is
difficult since the enzyme was lost or degraded during the
isolation of 6-APA. This problem was largely obviated in
the chemical splitting technology. Enzyme reuse and its
associated loss problems were soon eliminated by the
development of improved enzyme immobilization meth-
odology enabling long-term enzyme reuse. The immobi-
lized enzyme can now be used in modified fluidized-bed
modules (Matsumoto 1993). Recombinant Escherichia
coli strains are used as extremely efficient producers of
penicillin G acylase (Savidge 1984) and recombinant
strains of Fusarium oxysporum can be used as highly
efficient producers of penicillin V acylase (Chiang and
Basch 1999).
The chemical relationship of the penicillin (thiazoli-
dine) and the cephalosporin (dihydrothiazine) skeletons
was established in the early 1960s, when it was demon-
strated that penicillin sulfoxides could be rearranged to
form a variety of important cephalosporin derivatives.
When esters of penicillin sulfoxides are heated under
acidic conditions, an acid-catalyzed chemical ring-expan-
sion takes place (Morin et al. 1963). Eventually, a more
efficient process was developed using silyl protection
during the ring expansion rearrangement. Silyl protection
chemistry has led to efficient chemical production of
7-ADCA and has led to highly efficient production of
the oral cephalosporins, cephalexin and cephradine.
Cephadroxyl is synthesized after silylation of 7-amino-
cephalosporanic acid (7-ACA) followed by acylation with
a mixed anhydride prepared from a salt of r-hydrox-
yphenylglycine and ethylchloroformate. Amoxicillin is
synthesized using a similar process.
An important Lilly product, the cefaclor, involves a
ring enlargement of a penicillin V ester to an expanded
cephalosporin-S oxide with an exocyclic double bond.
The product is a useful intermediate in that it can be
converted into 3-substituted cephalosporins and into
cefaclor, a highly prescribed oral cephalosporin with
chlorine on the C-3 position.
Cephalosporin C and important semi-synthetic
cephalosporins
Cephalosporin C fermentation
High-yielding strains of A. chrysogenum are used in large-
scale, fed-batch fermentations. Major fermentation pro-
ducers of cephalosporin C obtain harvest titers in the
range of 2025 g/l. Production-scale fermentations are
fed-batch with carbon supplied as simple or complex
carbohydrate feeds during the growth phase of the
fermentation. As the fermentation progresses, sugar feeds
are reduced and are usually replaced by higher energy oils
such as soybean oil or peanut oil. Energy conservation
from oil as a substrate is considerably less efficient and
leads to slower growth, with the vegetative mycelium
becoming largely transformed into multicellular arthros-
pores. The arthrospore stage leads to greater oxygen
availability to the organism and results in rapid cephalo-
sporin production.
Table 3 Ranking of major b-
lactam producers and world
bulk production volumes
a
. 6-
APA 6-aminopenicillanic acid,
7-ACA 7-aminocephalosporanic
acid, 7-ADCA 7-aminodeace-
toxycephalosporanic acid
Penicillins(1995) Cephalosporins (1999)
1. Gist-Brocades 1. Antibioticos SpA
2. Antibioticos SpA 2. Biochemie/Hoechst
3. Biochemie 3. Glaxo/Wellcome
4. North China Pharma Works 4. Fujisawa
5. Glaxo/SmithKline 5. Cheil Jedang (Korea)
6. Bristol-Myers Squibb 6. Bristol-Myers Squibb
World bulk production volumes (metric tons)
Penicillins ~33,000 Cephalosporin C ~4,300
6-APA ~8,800 7-ACA ~2,140
7-ADCA ~1,950
Intermediates ~2,130
a
Adapted from Barber (1996, 2000)
388
dl-Methionine addition, which also results in the onset
of arthrospore formation, is often added to the medium
during the early growth phase of the fermentation. The
formation of arthrospores is also correlated with improved
dissolved oxygen concentration in the broth and is critical
for maximal expression of the important biosynthetic
cyclase and expandase enzymes.
Organic nitrogen is often supplied as a combination of
soybean and cottonseed meals supplemented with ammo-
nium sulfate and ammonia that is also used to help control
the pH throughout the fermentation. Corn steep liquor is
also supplied as a cheap nitrogen source and is rich in
amino acids, vitamins, organic acids and trace elements.
The pH of the fermentation is maintained between 6.2 and
7.0 and the temperature range is controlled between 24
and 28C.
A major problem associated with cephalosporin C
fermentation is the inherent chemical instability of the
cephalosporin C molecule. This is probably one of the
major reasons why long-cycle cephalosporin C fermen-
tations often result in reduced cephalosporin production
compared to typical long-cycle penicillin fermentations.
Cephalosporin C is readily degraded to compound X
(2-(d-4-amino-4-carboxybutyl)-thiazole-4-carboxylic acid),
which can account for as much as a ~40% loss of the
cephalosporin C produced (Usher et al. 1988). The
biosynthetic precursor molecules of cephalosporin C,
deacetylcephalosporin C and DAOC have much more
chemical stability. Strains of the yeast, Rhodosporidium
toruloides possess a potent acetyl esterase and, when the
organism is added to active cephalosporin C fermenta-
tions, result in increased levels of deacetylcephalosporin
C with an increase in total cephalosporin nucleus levels of
~40% (Chiang and Basch 1999).
Over the past decade, the cloning of many of the genes
involved in the biosynthetic pathway of cephalosporins
has resulted in more productive strains. Researchers at
Lilly demonstrated that the expandase/hydroxylase activ-
ity could be rate-limiting in certain production strains
(Skatrud et al. 1989). Transformants with an extra copy of
the cef EF gene had twice the expandase/hydroxylase
activity of the non-transformed strain, a reduction of
penicillin N levels and an increase in cephalosporin C
titer. The increase in cephalosporin C titer varied from
47% in laboratory scale fermentations to 15% in pilot
plant fermentors (Skatrud et al. 1989). There have been
no further reports from Lilly regarding the productivity of
the engineered strains in large production fermentors.
Basch and Chiang (1998) reported on the use of
genetic engineering strategies to reduce the levels of
undesirable byproducts in cephalosporin C fermentations
at Bristol-Myers Squibb. They showed that using a
recombinant strain of A. chrysogenum with an increased
copy number of the bifunctional expandase/hydroxylase
(cef EF) gene resulted in a reduced level of DAOC in
large production fermentors. The recovery and purifica-
tion of these broths and subsequent chemical conversion
to 7-ACA resulted in significant reduction of contami-
nating 7-ADCA.
Cephalosporin C recovery and purification
The purification and recovery of harvest cephalosporin C
broth begins with the rapid chilling of the active broth to
35C followed by removal of the mycelial solids either
by filtration or by centrifugation. The active broth
contains not only the desired cephalosporin C component,
but also small quantities of the biosynthetic precursors,
penicillin N, DAOC, deacetylcephalosporin C and the
degraded cephalosporin C product, compound X.
Two major strategies can be used for the recovery and
purification of cephalosporin C. One strategy involves the
use of activated carbon or the use of a non-ionic resin.
Because of the high selectivity of the resin, cephalosporin
C is preferentially adsorbed over penicillin N or the
contaminating biosynthetic precursor molecules. Most of
the penicillin N is removed in the pH 2.0 acidification
step. An additional anion- and cation-exchange step
usually results in high quality cephalosporin C. A large
fraction of the cephalosporin C is converted to 7-ACA
and derivatized to semi-synthetic cephalosporins.
A second purification strategy involves the substitution
of the amine moiety on the a-aminoadipyl side-chain at
C-7. Two substituted derivatives, N-2,4-dichlorobenzoyl
cephalosporin C and tetrabromocarboxybenzoyl cephalo-
sporin C, can be crystallized from acidic aqueous
solution. Alternatively, salts can be formed between the
N-substituted derivatives and an organic base such as
dicyclohexylamine or dimethylbenzylamine results in
cephalosporin salts that are solvent extractable. Bristol-
Myers Squibb uses a solvent-extractable process resulting
in the isochlorobutylformate (ICBF) ester of cephalo-
sporin C, termed cephalosporin D. Several extraction
steps are usually necessary to achieve the final desired
purity. N-Substituted cephalosporin C salts containing
small amounts of contaminants can be effectively
converted to 7-ACA.
Efficient enzymatic processes are now utilized for the
conversion of cephalosporins to 7-ACA, which has
resulted in dramatic cost reduction for this important
bulk intermediate. Two key genetically engineered
enzymes are involved. The initial step is reaction of the
a-aminoadipyl group with d-amino acid oxidase to
produce glutaryl-7-ACA. This reaction proceeds through
a keto-7-ACA intermediate that undergoes an oxidative
decarboxylation in the presence of hydrogen peroxide. A
glutaryl acylase is used to remove the glutaryl side-chain
to produce 7-ACA.
About one-third of commercial cephalosporins are
derived from 7-ADCA. Due to the lower cost of
penicillin, 7-ADCA is usually produced from penicillin
G by ring expansion of a penicillin sulfoxide ester to yield
a cephalosporin ester. The ester group is removed,
followed by removal of the phenylacetyl side-chain to
give 7-ADCA.
Two-thirds of the commercial cephalosporins are
derived from 7-ACA that is produced from cephalosporin
C by either chemical or enzymatic deacylation. In the
chemical process, after protection of the amino and
389
carboxyl groups, reaction with potassium pentachloride in
the presence of base forms an iminochloride derivative.
The iminoether is formed on the addition of alcohol. The
iminoether is hydrolyzed to form 7-ACA.
Enzymatic processes are now used by the major
producers of 7-ACA. Recent bulk market costs for 7-ACA
ranges from $115 to $130/kg. Antibioticos SpA and
Biochemie/Hoechst produced nearly 50% of the worlds
market needs in 1998 (Barber 2000). Other major
producers of 7-ACA are Glaxo-Wellcome, Fujisawa and
Bristol-Myers Squibb (Table 3).
Cephamycins
The cephamycins are derived from the cephalosporins by
methoxylation at the C-7 position. An enzymatic system
was identified in S. clavuligerus that converted cephalo-
sporin C or O-carbamoyl-DAOC to the a-methoxy
derivatives. The discovery of cephamycin C at Merck in
the early 1970s led to considerable research and devel-
opment on prokaryotic cephalosporins, since the presence
of the methoxy group on the b-lactam ring made the
molecule more active against Gram-negative and anaer-
obic pathogens and more resistant to Gram-negative b-
lactamase (Stapley et al. 1979). Thus, for the first time,
methoxylated cephalosporins were available that showed
a high degree of stability to b-lactamase enzymes.
Cephamycin C was never used clinically, but was
employed for the semi-synthesis of many medically
useful compounds.
The early fermentation studies of 7-methoxy-
cephalosporin antibiotics by Streptomyces lipmanii and
S. clavuligerus were reported by Nagarajan (1972) and by
Stapley et al. (1972). Extremely low titers of 130170 mg/
ml were reported.
The cephamycin broths are recovered and purified by
inactivating penicillin N by adjusting the pH of the
clarified broth to 2.02.5. The cephamycins are isolated
by a combination of active carbon adsorption followed by
adsorption on an anion-exchange resin. Small amounts of
DAOC can be removed by passage through silica gel
columns with 30% aqueous acetonitrile as the eluant
(Nagarajan 1972).
Cefoxitin is a marketed cephamycin-type cephalospor-
in that is now manufactured by chemical synthesis. Other
marketed molecules in this class are cefmetazole,
temocillin and cefotetan (Table 1).
Clavulanic acid
Clavulanic acid has relatively weak antibacterial activity,
but has been shown to be a potent inhibitor of the b-
lactamases produced by staphylococci and plasmid-
mediated b-lactamases of Escherichia coli, Klebsiella,
Proteus, Shigella, Pseudomonas and Hemophilus (Brown
et al. 1976). The molecule is a naturally produced
compound consisting of a b-lactam ring fused to an
oxazolidine ring (Howarth et al. 1976). The antibiotic is
produced by strains of S. clavuligerus (Reading and Cole
1977).
Clavulanic acid shows broad affinity for a number of
penicillin-binding proteins and is currently used in
combination with amoxicillin and is marketed under the
trade name Augumentin for the treatment of infections
caused by b-lactamase producing pathogenic bacteria.
Strains of S. clavuligerus are propagated on a fermen-
tation medium containing soy bean meal, soluble starch,
glycerol and potassium phosphate at a pH of 6.5 at 26C
for 100 h (Lawrence and Lilly 1980). There is no recently
reported production data for the clavulanic acid fermen-
tation.
The bulk of the clavulanic acid is found in culture
filtrates. Adsorption of the antibiotic on active carbon
followed by elution with aqueous acetone or solvent
extraction at pH 2.0 using butanol with back-extraction
into water at pH 7.0 proved to be effective primary
purification protocols. Secondary purification has been
achieved by conversion of the purified clavulanate to the
benzyl ester, which was then dissolved in ethyl acetate
and subjected to two chromatographic steps using
Sephadex LH20 and silica gel. The purified benzyl ester
was then hydrogenated over 10% Pd/C in the presence of
sodium bicarbonate to yield sodium clavulanate tetrahy-
drate.
Clavulanic acid possesses only weak antibiotic activity
against a variety of Gram-positive and Gram-negative
bacteria, but is an excellent inhibitor of a variety of b-
lactamases. The molecule has been coformulated with a
variety of broad-spectrum semi-synthetic penicillins with
amoxicillin being one of the best known formulations.
Augmentin had a world sales value of ~$1.3 billion in
1995 and was the second largest selling antibacterial that
year.
Carbapenems
The carbapenems resemble the penicillins, having a b-
lactam ring fused to a five-membered ring that does not
contain sulfur. Sulfur is present in the molecule outside
the ring in all carbapenems produced by streptomycetes.
A large number of carbapenems have been discovered,
but thienamycin produced by a unique strain of Strepto-
myces cattleya is the only carbapenem that is medically
and industrially important at this time.
Thienamycin is one of the most potent, broad-
spectrum, non-toxic antibacterial compounds ever dis-
covered. It was discovered at Merck by Kahan et al.
(1979) by a highly sensitive screening protocol based on
the inhibition of peptidoglycan synthesis. The chemical
structure of thienamycin was reported by Albers-Schoen-
berg et al. (1978).
The multi-component nature of carbapenem fermenta-
tions, its extremely low titer and high chemical instability
made recovery and purification of the antibiotic extreme-
ly difficult. These many difficulties led Mercks chemists
390
to employ organic synthesis rather than fermentation as
the preferred route for its commercial production (Miller
1981).
Thienamycin decomposes in dilute aqueous solution
and its decomposition accelerates as its concentration is
increased. A more stable and more potent derivative, N-
formimidoylthienamycin was developed and marketed as
imipenem. The molecule has high resistance to bacterial
b-lactamases, but is readily degraded by human renal
peptidase. A renal peptidase inhibitor, cilastatin, was
synthesized and formulated with the product. Imipenem is
a highly successful antibiotic and its world market, which
had grown to <$500 million by 1995, is among the top ten
marketed b-lactam antibiotics.
Future prospects regarding the industrial production
of b-lactam antibiotics
The production efficiencies of the two major marketed b-
lactam antibiotic classes, penicillins and cephalosporins,
have now been under major development for over four
decades. As a consequence, the bulk production costs for
penicillin and its important bulk intermediate, 6-APA, can
now almost be considered as commodity chemicals when
compared to other major marketed pharmaceuticals.
However, with increasing higher labor, energy and raw
material costs, especially in Europe, the United States and
Japan, more bulk penicillin manufacturing is moving to
the Far East and other developing countries (Barber
1996). Bristol-Myers Squibb has the only major manu-
facturing plant producing bulk penicillin V and 6-APA in
the United States.
Cephalosporins are continuing to gain market share,
and their final product cost still justifies continued
development and manufacture in Europe, Japan and the
United States. New and effective cephalosporins continue
to be introduced and there is still a major need to find
more effective cephalosporins for methicillin-resistant
(MRSA) and other resistant pathogens.
The development of new, fifth-generation, cephalos-
porins will depend more on an increased understanding of
b-lactam biosynthetic regulation and sophisticated meta-
bolic engineering of the producing organisms.
Most of the improvement in productivity of penicillin
and cephalosporin biosynthesis was obtained through
intensive screening of variant strains following mutagen-
esis and through recent applications of genetic engineer-
ing (Chiang et al. 1991; Elander and Chiang 1991). With
an increased knowledge of biosynthetic gene clustering
and possible gene transfer between already efficient b-
lactam organisms, this may lead to more rational
approaches to new, improved b-lactam products and
improved process technologies for more efficient b-
lactam antibiotic manufacture. Researchers at Merck and
Panlabs have reported on the introduction of the ex-
pandase gene from S. clavuligerus to a high-producing
strain of P. chrysogenum (Crawford et al. 1995). The
recombinant penicillin-producing strain, when fed adipic
acid, produced adipyl-7-ADCA, which can be easily
hydrolyzed to 7-ADCA, the major building block for
many oral cephalosporins. In this way, the greater
antibiotic-producing capability of P. chrysogenum could
be used for the manufacture of 7-ADCA. This novel
genetic engineering strategy could result in significant
cost reduction for many oral cephalosporin products.
Gene amplification of the bifunctional expandase/
hydroxylase gene in strains of C. acremonium has led to
improved fermentation of cephalosporin C as well as
decreased production of contaminating DAOC in large
production fermentors at Bristol-Myers Squibb. The
lower DAOC content resulted in more acceptable 7-
ACA with a higher purity (Basch and Chiang 1998).
The future, and need, for new b-lactam antibiotics
remains bright. Major pharmaceutical companies contin-
ue to have research and development efforts for new and
improved b-lactam biotechnology and chemistry because
of the need to discover new molecules due to the ever
increasing, continuing emergence of resistant bacterial
pathogens. New types of b-lactamase enzymes continue
to be discovered and the need for new semi-synthetic or
chemically-synthesized inhibitors is great.
References
Albers-Schoenberg G, Arison BH, Hensens OD, Hirshfield J,
Hoogsteen K, Kaczka EA, Rhodes RE, Kahan JJ, Kahan FM,
Ratcliffe RW, Walton E, Ruswinkle LJ, Morin RB, Chris-
tiansen BG (1978) The structure and absolute configuration of
thienamycin. J Am Chem Soc 100:64916499
Barber M (1996) The penicillins business. Michael Barber and
Associates, 18 Croydon Road, Catterham, Surrey, UK
Barber M (2000) The cephalosporins business: 2000 and beyond.
Michael Barber and Associates, 18 Croydon Road, Catterham,
Surrey, UK
Basch J, Chiang S-JD (1998) Genetic engineering approach to
reduce undesirable products in cephalosporin C fermentation. J
Ind Microbiol Biotechnol 20:344351
Brakhage AA (1998) Molecular regulation of b-lactam biosynthesis
in filamentous fungi. Microbiol Mol Biol Rev 62:547585
Brown AG, Butterworth D, Cole M, Hanscomb G, Hood JD,
Reading C, Rolinson CN (1976) Naturally occurring b-lacta-
mase inhibitors without antibacterial activity. J Antibiot
29:668669
Cantwell CA, Beckman RJ, Dotzlaf JE, Fisher DL, Skatrud PL,
Yeh WH, Queener SW (1990) Cloning and expression of a
hybrid Streptomyces clavuligerus cef E gene in Penicillium
chrysogenum. Curr Genet 17:213221
Cantwell CA, Beckman R, Whiteman P, Queener SW, Abraham E
(1992) Isolation of deacetoxycephalosporin C from fermenta-
tion broths of Penicillium chrysogenum transformants: con-
struction of a new biosynthetic pathway. Proc R Soc London B
238:283289
Chiang SJ, Chang LT, Chen YS, Hou HH, Elander RP (1991) Strain
improvement in Penicillium chrysogenum, from classical
genetics to genetic engineering. In: Kleinkauf H, von Dohren
H (eds) 50 years of penicillin applications: history and trends.
Public, Czech Republic, pp 245257
Chiang S-JD, Basch J (1999) Cephalosporins. In: Flickinger MC,
Drew SW (eds) Encyclopedia of bioprocess technology:
fermentations, biocatalysis and bioseparations, vol 1. Wiley,
New York, pp 560570
Crawford L, Stepan AM, McAda PC, Rambosek JA, Conder MJ,
Vinci VA, Reeves CD (1995) Production of cephalosporin
391
intermediates by feeding adipic acid to recombinant Penicilli-
um chrysogenum strains expressing ring expansion activity.
Biotechnology 13:5862
Demain AL, Elander RP (1999) The b-lactam antibiotics: past,
present and future. Antonie van Leeuwenhoek 75:519
Demain AL, Martin JF, Elander RP (1998) Penicillin biochemistry
and genetics. In: Mateles RI (ed) Penicillin: a paradigm for
biotechnology. Candida, Chicago, pp 98114
Elander RP, Chiang S-J (1991) Genetics and antibiotic process
improvement: from classical genetics to genetic engineering.
In: Procop A, Bajpai RK, Ho C (eds) Recombinant DNA
technology and applications. McGraw-Hill, New York, pp 153
170
Howarth TT, Brown AG, King TJ (1976) Clavulanic acid, a novel
b-lactam isolated from Streptomyces clavuligerus: X-ray crys-
tal structure analysis. J Chem Soc Chem Commun 266267
Kahan JS, Kahan FM, Goegelman R, Currie SA, Jackson M,
Stapley EO, Miller TW, Hendlin D, Mochales S, Hernandez S,
Woodrull HB, Birnbaum J (1979) Thienamycin, a new beta-
lactam antibiotic. 1, discovery, taxonomy, isolation and phys-
ical properties. J Antibiot 32:112
Lawrence GC, Lilly G (1980) UK Patent 1,571,888
Matsumoto K (1993) Production of 6-APA, 7-ACA and 7-ADCA
by immobilized cephalosporin amidase. In: Tanaka A, Tosa T,
Kobayashi T (eds) Industrial application of immobilized
biocatalysis. Dekker, New York, pp 6788
Miller TW (1981) Crystalline N-formimidoylthienamycin. US
Patent 4,260,543
Morin RB, Jackson BG, Mueller RA, Lavagnino ER, Scanlon WB,
Andrews SL (1963) Chemistry of cephalosporin antibiotics. 3.
Chemical correlation of penicillin and cephalosporin antibi-
otics. J Am Chem Soc 85:18961897
Nagarajan R (1972) b-Lactam antibiotics from Streptomyces. In:
Flynn EH (ed) Cephalosporins and penicillins: chemistry and
biology. Academic Press, New York, pp 636661
Reading E, Cole M (1977) Clavulanic acid: a beta-lactamase
inhibitor from Streptomyces clavuligerus. Antimicrob Agents
Chemother 11:852857
Rowlands RT (1991) Industrial strain improvement and the Panlabs
penicillin club. In: Kleinkauf H, von Dohren H (eds) 50 years of
penicillin applications: history and trends. Public, Czech
Republic, pp 258266
Savidge TA (1984) Enzymatic conversion used in the production of
penicillins and cephalosporins. In: Vandamme E (ed) Biotech-
nology of industrial antibiotics. Dekker, New York, pp 191224
Skatrud PL, Tietz AJ, Ingolia TD, Cantwell CA, Fisher DC,
Chapman JL, Queener SW (1989) Use of recombinant DNA to
improve the production of cephalosporin C by Cephalosporium
acremonium. Biotechnology 7:477485
Stapley EO, Jackson M, Hernandez S, Zimmerman SB, Currie SA,
Mochalis S, Mahta JM, Woodruff HB, Hendlin D (1972)
Cephamycins, a new family of b-lactam antibiotics. 1.
Production by actinomycetes, including Streptomyces lactam-
durans sp. n. Antimicrob Agents Chemother 2:122131
Stapley EO, Birnbaum J, Miller AK, Wallick H, Hendlin D,
Woodruff HB (1979) Cefoxitin and cephamycins: microbio-
logical studies. Ref Infect Dis 1:7387
Strohl WR (1999) Secondary metabolites; antibiotics. In: Flickinger
MC, Drew SW (eds) Encyclopedia of bioprocess technology:
fermentation, biocatalysis and bioseparations, vol 5. Wiley,
New York, pp 23482365
Sylvester JC, Coghill RD (1954) The penicillin fermentation. In:
Underkofler LA, Hickey RJ (eds) Industrial fermentations, vol
2. Chemical Publishing, New York, pp 219263
Usher JJ, Lewis MA, Hughes DW, Compton BJ (1988) Develop-
ment of the cephalosporin C fermentation taking into account
the instability of cephalosporin C. Biotechnol Lett 10:543548
Van der Beck CP, Roels JA (1984) Penicillin production:
biotechnology at its best. Antonie van Leeuwenhoek 50:625
639
Van Nistelrooij, Krijgsman J, DeVroom E, Oldenhof C (1998)
Penicillin update. In: Mateles RI (ed) Penicillin: a paradigm for
biotechnology. Candida, Chicago, pp 8591
Waites MJ, Morgan NL, Rockey JS, Higton G (2001) Industrial
microbiology: an introduction. Blackwell, Oxford, UK
Weissenberger HWO, van der Hoeven MB (1970) An efficient non-
enzymatic conversion of benzylpenicillin to 6-aminopenicil-
lanic acid. Recl Trav Chim Pays-Bas 89:10811084
392